How and Why Wall Street Programmers Earn Top Salaries

msmoriarty writes "Given the level of interest in the recent highest-paid programmers discussion, our reporter decided to do a follow-up looking into the languages and skills needed to work on high-frequency trading systems. There's actually a pretty wide range of languages/tools used, but Linux is the 'default' OS and, not surprisingly, the 'ability to work under pressure when the traders are screaming at you' is a must-have skill."

Comments

Perversion of Capitalism

[NFN_NLN]

"In high-frequency trading (HFT), programmers eke out every last incremental tick in performance to build algorithms that battle other algorithms for computational supremacy and millions in profits -- and earn a lot in the process."

Skimming money off billions of micro-transactions. Ahh, yes... forget investing in ideas and backing well managed companies... this is the way capitalism was envisioned.

Re:Perversion of Capitalism

[zippthorne]

Look, superman III had a lot of lessons to teach. It's really too bad on villains watched it...

Re:Perversion of Capitalism

[Anonymous Coward]

Skimming money off billions of micro-transactions. Ahh, yes... forget investing in ideas and backing well managed companies... this is the way capitalism was envisioned.

Oh, admit it; this has been every computer programmer's dream since they saw Superman III as a child.

Re:Perversion of Capitalism

[jo42]

I call BS. You don't need much skill or talent in writing code that effectively does this in the end:

account.balance *= 10.0;

Re:Perversion of Capitalism

[Anonymous Coward]

Skimming money off billions of micro-transactions. Ahh, yes... forget investing in ideas and backing well managed companies... this is the way capitalism was envisioned.

Skimming is a mischaracterization of enormous magnitude. What they usually do is arbitrage. Something that has been part of various markets both physical and financial for a thousand years or more. It's effectively no different than buying potatoes in location A where they cost 10 cents each, driving them to location B using enough fuel to bring the cost of each potato up to 15 cents each and selling them for 20 cents each because location B has a shortage. They just do it really really fast and before the other guy can and are buying and selling contracts for physical assets rather than the assets themselves (or financial derivatives based on the physical assets in the case of financial trading). It's not rocket science and it's not nefarious in the least but the fastest guy makes millions and the second fastest guy has a bunch of expensive metaphorical potatoes that he can't sell for a profit on this hands.

Re:Perversion of Capitalism

[russotto]

I call BS. You don't need much skill or talent in writing code that effectively does this in the end:

account.balance *= 10.0;

The challenge isn't making the code do that. The challenge is making it look like the code is intended to do something entirely different, and that part is merely an unavoidable side effect.

Re:Perversion of Capitalism

[Anonymous Coward]

Who Is John Galt?

Re:Perversion of Capitalism

[Anonymous Coward]

who fucking cares?

Re:Perversion of Capitalism

[Anonymous Coward]

Do you know the legend of Atlantis?

Re:Perversion of Capitalism

[Anonymous Coward]

Yup, and when the cheap energy runs out, there's going to be some "readjusting" going on in that model. But don't worry, it won't affect your hard-working heroes and moral models, it'll affect you and me. After all, we CHOSE to have this system....

Re:Perversion of Capitalism

[Anonymous Coward]

I didn't incorrectly mod it off-topic, I correctly modded it troll.

http://www.snopes.com/politics/soapbox/proportions.asp

Re:Perversion of Capitalism

[baitisj2]

http://www.snopes.com/politics/soapbox/proportions.asp

Attributing this to David Kaiser is incorrect. "As far as we know, this piece began as a comment posted to Pat Dollard's blog in November 2008 by an author identified only as 'TPS'."

Re:Perversion of Capitalism

[fliptout]

Oh, please, spare me. I see this sentiment everywhere, and it's nonsense. Traditional investing still works. At the end of the day, investors are still going to look at how much money a company makes and assess how much risk that company's stock poses. If HFT causes blips in a stock's price, the the market will eventually correct the price.

Re:Perversion of Capitalism

[PopeRatzo]

Skimming is a mischaracterization of enormous magnitude. What they usually do is arbitrage.

And arbitrage is nothing but skimming of an enormous magnitude.

Let's not bullshit here, OK?

Re:Perversion of Capitalism

[NFN_NLN]

Oh, please, spare me. I see this sentiment everywhere, and it's nonsense. Traditional investing still works. At the end of the day, investors are still going to look at how much money a company makes and assess how much risk that company's stock poses. If HFT causes blips in a stock's price, the the market will eventually correct the price.

Like this blip?
http://www.bloomberg.com/news/2010-05-06/electronic-trading-to-blame-for-stock-market-plunge-nyse-s-leibowitz-says.html

Re:Perversion of Capitalism

[NFN_NLN]

This guy nails it - http://boards.straightdope.com/sdmb/archive/index.php/t-601887.html

"It's not about acting on market information. It is purely arbitrage. A mis pricing allows one to buy and sell simultaneously and lock in the difference minus trading costs.

In the old days, traders used to do this in the trading pit. Now it's computers closest to the exchange feed.

Tied in with this is the automatic trading. In the case of that big intra day fall, a wrong trade was entered. I forget the details but it was big enough to push down the market xx amount, which triggered automatic sell programs from non-arbitrage automated computer selling, which triggered a market sell off, which in turn triggered more selling until the market circuit breakers kicked in.

During the mandatory no trading period, the original bad trade was discovered and reversed. This IIRC also triggered automated buy programs and the whole thing went in reverse. The bad thing is that the market whipsaw really hammered some real end investor trades as collateral damage.

I remember watching the Hang Seng Index the day that Soc Gen announced Jerome Kerviel's fraud and liquidated the positions. It was a full trading day of massive market swings for big losses to big gains several times throughout the day. It was almost all computer generated programmatic trading."

Re:Perversion of Capitalism

[aaarrrgggh]

Dean of architecture at KU. Why do you ask? ;)

Re:Perversion of Capitalism

[MarkvW]

We pay our taxes to these guys as much as we pay them to the government.

Re:Perversion of Capitalism

[Anonymous Coward]

"Perversion of Capitalism"

No it;s exactly what you get in the real world where market values (profit) drive everything. We see it all the time in offshored jobs, destroyed lives, rolling back of the welfare state, the election of extreme right wing conservatives like Stephen harper.

This "there is some pure capitalism we have to get back to" bullshit is just that - bullshit. The left was born from captialism killing workers, it caused two world wars and then then there was the cold war. This idea that is some 'benign' capitalism we have to get back to is just utter american ignorance.

Re:Perversion of Capitalism

[Anonymous Coward]

Much of the HFT volume disappeared as the market plunged. Things had gone far enough that everyone knew that some trades were going to be canceled, but the standards for which trades would be kept and which would be canceled were not known in advance. Normally another firm's mistake has HFTs salivating, but no one wants to end up with a bunch of canceled trades.

In other words, if an HFT causes blips in a stock's price, another HFT will be happy to immediately correct the price, but only if they're confident they'll be able to keep the money they make from it.

Re:Perversion of Capitalism

[Sir_Sri]

Why?

Seriously, what's the problem? If there's a difference in price between two exchanges those prices should balance out. If there's a difference in price between trading in different currencies, they should even out. This is just doing it quickly. The money was always there to be made, it's just in smaller increments now.

If you can, with the right data, specify an algorithm on which to base the buying of companies, then it is a problem best executed by machines rather than people. Your concept of "investing in ideas and well managed companies" is an entirely separate problem. If you want to invest 5 billion dollars in a company you don't go to your HFT stock broker and ask him to buy at right this nanosecond. You call up the CEO of the company, arrange the price, and have a press conference and news release about it. If you want to invest $5000 in a local company you can call a guy up and still do that. But if you notice a differential in price of 1 cent on 5000 shares between a stock exchange in Chicago vs new york, or a sell order and a buy order with a price difference of even 1 cent (or a fraction of a cent if you're into that sort of thing) you can buy up from the seller, and sell to the buyer, and make that fractional price difference. Both the buyer and seller got what they wanted, at the price they wanted, and you made up for the fact that one person was too lazy to check if there was a buy/sell at already on the market at the price they would accept. Building the tools to facilitate the transaction between those two has value, you made money on it. What's the problem?

Trying to move money rapidly carries inherent risk. Trying to move money at all carries the same risk (buy low, sell high, pick an 'well managed company" are all just exercises in guesswork based on public data and personal impressions), but it manifests differently. But the risk is in effect the same. If you buy a car from a manufacturer, to sell to dealerships you're banking on the car company not going bankrupt, if you buy airplanes to sell to airlines you're banking on al qaeda not crashing more planes into buildings tomorrow. Any item you want to sell is a liability as long as you hold it. If that's a share in a company or a bond in a company (both of which have value), or a more durable, or more perishable good. HFT is good in that it normalizes markets and injects short term liquidity into transactions.

It's bad in that just as quickly as it can put liquidity in, it can take it out if the algorithm cannot resolve viable strategies. To a long term (or at least longer term than 24 hours) investor these liquidity shifts don't matter much. No more than it should matter to you if the US government has 76 billion dollars in cash or 74 billion. It does however have the potential to create very odd short term spikes in data. I'm not sure they matter, on the whole, and the reason it is automated, and works well automated is that these are largely automatic things to worry about anyway. There's no reason the price of a share of GE should be different in New York or Chicago, if it is, eventually someone will buy low and sell high. If that happens at high frequency the gains are less, but the losses are less if there's a small price shift, so in the end it's better for everyone.

There are certainly places where HFT could be unfair - for example if HFT's are given access to data before others or the like. That is, to me, and implementation issue, not a concept issue. (Imagine trading where one trader can watch a conference live, and trade during that conference, whereas everyone else watches on TV, on a 30 millisecond delay. The trader watching live has a small advantage, but it shouldn't significantly change the overall outcome of the stock performance in relation to the conference).

Put another way. If you're a long term investor, HFT shouldn't effect you, at all. You can still, quite successfully, read quarterly and yearly reports and guess how valuable ideas are and how good m

Re:Perversion of Capitalism

[Ex-MislTech]

You will care when the system collapses and the dollar is
worth the same amount as the money in the 34 other nations
that went thru hyperinflation.

At that time you will "get it".

http://en.wikipedia.org/wiki/Hyperinflation#Examples_of_hyperinflation

Or maybe you can read and "get it" before it happens.

Re:Perversion of Capitalism

[arth1]

I call BS. You don't need much skill or talent in writing code that effectively does this in the end:

account.balance *= 10.0;

Remember that every cycle counts. Depending on the architecture and compiler, you might want to refactor that code as:

account.balance += account.balance + account.balance << 3;

Re:Perversion of Capitalism

[obarthelemy]

it does feel kinda gratuitous to go for ever faster churning. is it really productive, useful, to invest ever greater efforts and money into accelerating arbitrages from say, once per hour, to once per millisecond ? picoseconds next ?

meanwhile, others are building factories...

Re:Perversion of Capitalism

[Anonymous Coward]

888 didn't read shit

Re:Perversion of Capitalism

[improfane]

It's not genuine value. It produces nothing for society. It's a loophole.

Re:Perversion of Capitalism

[bjourne]

I think you are missing the point completely. It is not that money is being moved around in what basically amounts to a huge zero-sum game. One daytrader has better computers or lower ping to the nyse and beats out another trader who hasn't. That's really not a problem. The problem is the huge amounts of resources that is wasted on this game and the impact we are letting it have on our lives. The worlds brightest minds are spent in the game. You may not see it as a problem that the best mathematicians and programmers are working in the finance industry instead of developing a cure for cancer, affordable space shuttles, electric cars, aids vaccine or whatever because the salaries are much higher there so obviously that is what the market wants and the market is always right. But I do, I think it is a waste. But the worst problem is the importance we are giving to the stock market game. The idea was that the stock price should reflect the progress if its company. Now it's the other way around. It doesn't matter what the company does, if the stock price is high, then that's good otherwise it is bad. Oh and if the price of most stocks are low, and most players in the game have lost, then that is really bad. It's a depression coming and because the game was busted the rest of society will have to clean it up.

Re:Perversion of Capitalism

[Anonymous Coward]

More importantly, they don't really "eke out every last incremental tick in performance".
From what I've seen, they're still quite far from reaching the peak performance of HPC systems.

The real problem is that they need to be very reactive, rather than very fast, which has led them to focus their efforts on networking stuff rather than on pure computational matters.

Re:Perversion of Capitalism

[Anonymous Coward]

Be careful when you read that long post; who knows who made it and with what purpose. Try to determine for yourself which sentences ring true and which are shaped as a straw-man.
This is from the Wikipedia article http://en.wikipedia.org/wiki/Definitions_of_fascism: Umberto Eco defines fascism with the following features:

The features of fascism he lists are as follows:

• * "The Cult of Tradition", combining cultural syncretism with a rejection of modernism (often disguised as a rejection of capitalism).
• * "The Cult of Action for Action's Sake", which dictates that action is of value in itself, and should be taken without intellectual reflection. This, says Eco, is connected with anti-intellectualism and irrationalism, and often manifests in attacks on modern culture and science.
• * "Disagreement Is Treason" - fascism devalues intellectual discourse and critical reasoning as barriers to action.
• * "Fear of Difference", which fascism seeks to exploit and exacerbate, often in the form of racism or an appeal against foreigners and immigrants.
• * "Appeal to a Frustrated Middle Class", fearing economic pressure from the demands and aspirations of lower social groups.
• * "Obsession with a Plot" and the hyping-up of an enemy threat. This often involves an appeal to xenophobia or the identification of an internal security threat. He cites Pat Robertson's book The New World Order as a prominent example of a plot obsession.
• * "Pacifism Is Trafficking with the Enemy" because "Life is Permanent Warfare" - there must always be an enemy to fight.
• * "Contempt for the Weak" - although a fascist society is elitist, everybody in the society is educated to become a hero.
• * "Selective Populism" - the People have a common will, which is not delegated but interpreted by a leader. This may involve doubt being cast upon a democratic institution, because "it no longer represents the Voice of the People".
• * "Newspeak" - fascism employs and promotes an impoverished vocabulary in order to limit critical reasoning.

Re:Perversion of Capitalism

[ozmanjusri]

What makes you think he's American?

Re:Perversion of Capitalism

[Anonymous Coward]

Look, what am I supposed to do.

I'm a chip designer. I used to work on beautiful hardware, you might be using some of my products to play games even today. Turns out, ASIC design is screwed as a business in so many ways, and it's not because of outsourcing. People just don't want the hardware all that much, and they don't pay a lot for it. The chips are priced by cost of manufacturing, not cost of R&D, and so R&D is basically cut to the bone on salaries.

Outside of financial companies (and FPGA vendors, who have been trying to sell this idea unsuccessfully to everyone on the planet for the last 10 years), nobody seems to have recognized that bringing stuff into hardware (FPGAs, for instance) can make money through performance. So that's where I work.

Re:Perversion of Capitalism

[Anonymous Coward]

In medieval times, people had to pay tithes to the church. New we pay tithes to the feudal oligarchs. Progess is great.

Re:Perversion of Capitalism

[Anonymous Coward]

You don't have to be American to shit yourself when thinking about USD hyperinflation

Re:Perversion of Capitalism

[Anonymous Coward]

Also it's recently come to light that it was Goldman Sachs who received the loans at pretty much zero interest. "You'd have to be stupid not to take it."

Re:Perversion of Capitalism

[Anonymous Coward]

IIRC the regulatory agency also forcefully reversed most of the automated trades that contributed to the crash, but upheld trades from actual investors made when the stocks were low, but re-priced them to the non-crash level. Guess who lost out there too.

Re:Perversion of Capitalism

[luis_a_espinal]

I call BS. You don't need much skill or talent in writing code that effectively does this in the end:

account.balance *= 10.0;

Remember that every cycle counts. Depending on the architecture and compiler, you might want to refactor that code as:

account.balance += account.balance + account.balance << 3;

Sadly, I'm afraid the average /. poster (or recently graduate CS'tist for that matter) wouldn't even understand the code refactoring you just proposed. Everything here is "blah, it's easy, I'm awesome, let me show you how", and then they drop a naive turd as an example.

Re:Perversion of Capitalism

[Opportunist]

Careful when reading? Who are you, the Catholic Church, expecting people to "fall" for the slings of the devil if they only as much as read his words?

People are, fortunately, usually quite capable of debunking bull. Those that aren't, well, there's always a church for them.

Re:Perversion of Capitalism

[Opportunist]

Eventually is such a great word. It means "sometimes in the future, and if it didn't happen yet, it certainly will. Some day. Eventually. You can't say I'm wrong just 'cause it didn't happen yet."

Re:Perversion of Capitalism

[Opportunist]

Patience, my friend. Soon we'll pay more of our taxes for these guys. It's all a matter of time and things will get sorted out.

Then we'll finally have the small government everyone dreams of. Because there won't be any money for anything but a small government left after the locusts had their share.

Re:Perversion of Capitalism

[Anonymous Coward]

Yeah... One of the USA's greatest exports - inflation via the petrodollar.

Re:Perversion of Capitalism

[Opportunist]

I am always amused if people consider themselves so much better off today than it was in the old days, with aristocracy dictating and us not having a voice.

I fail to see how that would have changed.

Re:Perversion of Capitalism

[Anonymous Coward]

we're not dying at the ripe old age of 30 in a mud hovel blind and covered in our own feces. Although you probably will. HTH HAND

Re:Perversion of Capitalism

[Rich0]

Uh, if you're going to be pedantic, at least try to be correct... :)

Re:Perversion of Capitalism

[Rich0]

I'm sure everybody working in this industry is doing it for a reason, and it usually has to do with putting food on the table. So are the clerks who mail out RIAA legal threats. That doesn't mean that society doesn't have a problem.

Sure, if there aren't any other jobs, then go ahead and create your ASICs to do HFT. However, society should be looking to:

1. Find some better way to gainfully employ our best and brightest.
2. Eliminate the perverse incentives to optimize our financial markets on the nanosecond timescale. I mean, really, is society really losing something when two markets are out of equilibrium for an entire millisecond, or even an entire hour or even day?

Re:Perversion of Capitalism

[Unoriginal_Nickname]

Under the old campaign finance rules, each political party is funded by the Canadian government based on the number of votes they receive. The NDP is not a "real" party, it's a scam; they ran parachute candidates in as many ridings as possible so Jack Layton could collect a $200,000+ annual salary for leading the NDP in addition to his and his wife's salaries as MPs.

You can call Stephen Harper "extreme right-wing" if you want, but that doesn't change the fact that Canada is politically broken on the left-wing, and a lot of that damage is being caused by the new crop of mid-20s Facebook-addicted special snowflakes with no educations, no work ethics, no marketable skills and a complete dependency on mommy, daddy and the welfare state for personal solvency.

Re:Perversion of Capitalism

[Khashishi]

Bad analogy. In your example, the merchant is adding some value to the potatoes by transporting them to another place, and moving to another geographic area with a different price. In the case of high frequency trading, there is no transportation to different geographic areas going on. The high frequency trader is simply buying up property and selling it for higher in the same place. He isn't taking advantage of a price differential between two places. Really, a better analogy is scalping.

Re:Perversion of Capitalism

[adamofgreyskull]

meanwhile, others are building factories...

Ooh, the power of the ellipsis, suggesting so much, but saying so little! Unfortunately, it leaves you open for:

...building factories producing trans-atlantic cables (and the ships which will lay them) in order to reduce latency between NY and London. Building factories which will produce newer, faster chips which will drive faster network connections and churn through trading algorithms milliseconds faster. This race to be the fastest is benefitting us. It is a by-product, but it benefits us nevertheless.

Re:Perversion of Capitalism

[adamofgreyskull]

There's a grocer down the road. He buys fruit and vegetables from farmers and sells them to people. What does he produce for society? Fuck all. He's acting as a middle-man between a seller and a buyer. Shit, we'd better close this loophole! It's not just the NYSE that it's affecting, it's right on our doorsteps!!!11!

Re:Perversion of Capitalism

[improfane]

Vegetables are tangible and benefits everyone's lives. Having a supply near you is beneficial.

Credit swaps and and financial trades are intangible and worthless to all those that manipulate them. They also destablize society. Vegetables don't.

Re:Perversion of Capitalism

[sgt scrub]

The challenge is making it look like the code is intended to do something entirely different

Ooooh. Now I get it! I thought they used Haskell to protect the code. ie. C++ you just decompile it. With Haskell you get everything, non readable code, no compiler to wait on, and nobody is sure what your code does. Sweet.

Re:Perversion of Capitalism

[fafaforza]

No it isn't. Any of us can take part in arbitrage at your home right now, with 0% cash advance offers from credit cards and savings/CD/MM accounts. (Albeit, currently, it likely wouldn't make any sense due to the low interest rate from the government, and I haven't paid attention into whether those offers are still prevalent, but you can play arbitrage with anything, at any quantity.)

Re:Perversion of Capitalism

[Opportunist]

I thought everyone is so damn opposed to healthcare in the hands of the government, where does this come into play now? I'm a bit confused, please elaborate how advances in medicine and technology have anything to do with a change in government.

Re:Perversion of Capitalism

[guanxi]

Where do you think electric car manufacturers get money to pay researchers, engineers, etc.?

Re:Perversion of Capitalism

[Anonymous Coward]

Now tell us how you really feel.

Re:Perversion of Capitalism

[epine]

You're concerned about Layton making $200,000 annual salary as leader of Canada's official opposition, who publicly subjects himself to the kind of nonsense you just spouted 24/7?

College Football Coaches Salaries - Top 100

Layton misses the $375,000 cut by a Canadian country mile. Obviously, coaching a football team takes real talent; if the players lose faith, they tune you out; you're under incredible public scrutiny to perform now; it's not the easy street of running a country, giving good quotes that don't blow up in your face, or placing well in national elections.

I understand that snowflakes tend to congregate in snow drifts, but how is it that our precious snowflakes are stuffing the ballet box? Our elections are won and lost on the fat demographic bulge of baby boomers nearing retirement with visions of Nortel still in their heads. These were the precious snowflakes of previous generations. Have you seen the footage of previous generations? I bet you've heard the music.

The jobs that used to fall off trees for young people have A) been outsourced to China, B) locked up by older Canadians who still dominate the work force, and C) no longer subsidized by the free-love price of gasoline in 1960.

The precious snowflakes aren't quite as stupid as you make them out to be. They understand that ten years from now when the boomers have mostly retired and are struggling to manage their booming personal health care burdens, there's going to be a sharp correction in the value of compensation available. Someone has to process all the paperwork for the millions of immigrants we are going to require to stretch our underfunded pension programs to the Freedom 95 finish line.

By the way, our pension funds aren't half so "underfunded" as you hear in the telling. When pensions were first established, not many people lived well past retirement age. It has never been economically feasible to support 1/3 of a western country's adult population out of the work force at a three car garage standard of living. Why were those promises made in the first place? The voting power of a grasping demographic bulge. It sold well to the masses, didn't it? All you had to do was convince the Saudi's to play along. Why conserve a precious resource for a rainy day when you can promise it to the flower children in their golden years instead? Besides, it never rains in Saudi Arabia. The pension program "underfunding" is a lot like the Madoff losers, who at first reported losses against what the crook had promised, oblivious to their failure of due diligence. Harry Markopolos wrote a newsletter packed with salient insight, and no-one subscribed. You wonder what these greedy people did to earn their success in the first place, or if the bulk of privilege accrues from hitching yourself to the right crook on the way up and knowing when to severe your ties.

Harper's contribution has mostly been to muzzle the people who work the equations: statisticians and scientists. The feelings of love emanating from the newly retired that get a politician re-elected for multiple terms are mainly threatened by accurate accounting. Nobody ever said Harper was stupid. It's our job, as the electorate, to demand better, should be choose to use it.

Some of us did. Layton will do a fairly competent job of clearing his throat and shaking the carpet as the worst of the Tory blinkering. The whole dynamic is entirely stupid. It's a natural thing when you've promised a large group of people eternal sunshine (that you could never afford) to shoot the messenger when the heroine wears off. Harper would be insane to accelerate this. That's why a democracy also needs a competent opposition party: someone has to be the bearer of bad news if anything is going to change soon enough to make a difference.

I don't see how our precious snowflakes could do much to deflect this iceberg even if they lived o

Re:Perversion of Capitalism

[gbjbaanb]

and can you imagine that this won't happen again. "oh dear, I accidentally pressed 0 a few too many times, whoops. silly me."

and the computers make a killing for their employers, the bad trade gets correctly later and all the small investors who panicked lose out.

I think the only moral story to this is that the small investor needs to take a much longer term view to the stock market - you just can't beat these guys at their own game ('cos they've rigged it against you)

Re:Perversion of Capitalism

[PopeRatzo]

Any of us can take part in arbitrage at your home right now, with 0% cash advance offers from credit cards

Wait, that's your example of something that is NOT skimming and NOT a scam?

Re:Perversion of Capitalism

[skids]

Actually your example of arbitrage is faulty. A person "buying potatoes in location A where they cost 10 cents each, driving them to location B using enough fuel to bring the cost of each potato up to 15 cents each and selling them for 20 cents each because location B has a shortage" is performing a service, namely, they are transporting the potatoes, and they are also assuming the risk that something will happen to the potatoes or that the shortage will be over by the time they arrive at location B. That's not pure arbitrage. There's no service performed in arbitrage -- it is simply a person using a middleman position to skim money off the market.

Re:Perversion of Capitalism

[arth1]

Uh, if you're going to be pedantic, at least try to be correct... :)

Very sound advice. You might want to follow it, because the joke is on you - it is correct.

A likely cause for your confusion is that it's not intuitive for everyone that x *= 10 is the same as x += 9x

Coders commanding six figures should know things like this in their bones, though.

Re:Perversion of Capitalism

[Sir_Sri]

Credit swaps are not HFT. Please don't confuse the two.

HFT lets you execute existing trades faster. That's it. If the trades themselves are of dubiously valuable goods that doesn't make the ability to trade them any less genuine. It's probably faster to drive to macdonalds than to drive to a grocery store.

People used to get on camels and mules and tromp half way across the world to get silk from china to europe. Then they built ships that did the same thing, then trains, then big ships, and aircraft and so on. That's high frequency trading. Silk has virtually no value other than as a luxury (or parachute), but people still paid for it.

Facilitating connections between buyers and sellers of securities, normalizing prices between exchanges etc. are all *good* for the economy.

Re:Perversion of Capitalism

[Sir_Sri]

Again, that's buying into the: we trade 100 dollars 1000x a day, 365 days a year.

It's a specialized knowledge field. It's tricky to get right, and it involves a lot of skills that stock brokers use. I'm sure one could argue John Carmack and his million + a year salary (or whatever he makes now, he said about a month ago he got a 100x pay increase at one point), is wasted on games. I'm a game developer, people complain at me that I should be doing something other than making games on a regular basis because I should be doing medical imaging or something that helps society or whatever. Hell half the people on /. probably run web servers that have relatively little value. Being able to design and implement HFT systems well should pay well if it makes a lot of money. That guy who made minecraft has like 30 million dollars now. See what I'm getting at? Make a product that does what it's supposed to well, and you get paid well for it. Write a program that finds places to make small amounts of money, you should be rewarded for how much money it makes. Make a game that sells well, you should, again, be rewarded for how many copies it sells.

I think HFT only really pays a lot of money because you can't afford to have people who screw up too much. The worst paid doctor still makes more money than I do because you have to actually be a doctor. HFT may not be medicine, but it's still a very specialized set of knowledge.

The guys making 100K-400k a year with PhD's doing medical imaging are doing alright for themselves too. And it's a much lower risk environment.

Re:Perversion of Capitalism

[TheRaven64]

Anyone with a CS degree from somewhere other than the back of a cornflakes box will have seen the algorithm for decomposing multiplication into shifts and adds. They will typically also have studied the idea of pipelining, and will see that the proposed 'optimisation' will require multiple trips through the pipeline, so they'll be able to explain why it's a bad idea. They will probably also know that instruction cache misses are one of the biggest performance killers in code for systems with a typical memory hierarchy, and so bloating the code from one instruction up to 3 is not such a great idea.

Those who have studied compilers (sadly, not all of them - well, not too sadly, since it means there's a skill shortage in that field, which is great for me) will also know that compilers for modern pipelined architectures will actually do the opposite transform, and turn this 'optimised' version into the original, with a single multiply.

The ones who have done an advanced architecture module will also be able to tell you that ARM is the one exception to this, where the fact that you get the shift for free and the latency for add is shorter than for multiply means that this may be marginally faster.

The ones that did any kind of software engineering module (a requirement for accreditation in the UK, not sure about the USA) will know that obfuscating your code for a marginal performance gain based on probably-obsolete assumptions about the underlying architecture is a really good way of writing unmaintainable code that runs slowly and ends up being thrown away and completely rewritten.

Re:Perversion of Capitalism

[russotto]

The problem is the huge amounts of resources that is wasted on this game and the impact we are letting it have on our lives. The worlds brightest minds are spent in the game. You may not see it as a problem that the best mathematicians and programmers are working in the finance industry instead of developing a cure for cancer, affordable space shuttles, electric cars, aids vaccine or whatever because the salaries are much higher there so obviously that is what the market wants and the market is always right.

So what? Are you a slaver? Developing a cure for cancer or an AIDS vaccine? Pays for shit, particularly for anyone lacking that biochemistry Ph.D; the grad students actually programming the models are probably living on ramen noodles, and the postdocs not much better. Affordable space shuttles? Nobody wants them. Electric cars? Not such great pay either, unless you're one of the principals. Probably better than ramen noodles, but still not going to make any mathematician or programmer rich.

So what we have with high frequency trading is a field where a mathematician or a programmer can cut himself a good-sized slice of the pie. Still nothing like what the principals are making, but more than he'll make on the staff of an electric car company and far more than in medical R&D. And you'd take this away, not because HFT is somehow harmful in and of itself, but because you want those mathematicians and programmers working in the fields of YOUR choice, not theirs. You ARE a slaver, whether you realize it or not.

If you really think electric cars, cancer cures, aids vaccines, and the like are being held back for lack of mathematical and programming talent, you should be looking for ways to increase the compensation for mathematicians and programmers in those fields. Not to take away the fields which have better compensation so they have no choice.

(I don't work for an HFT trader; I work for Google. But I also know that the existence of extremely high-paying jobs in the HFT field puts upward pressure on salaries for programmers in other fields, including making electric cars or writing search engines. So I do benefit from their existence)

Re:Perversion of Capitalism

[mikael]

A search for topics on "stochastic calculus" will be enlightening.
Basically it's a form of one-dimension Brownian motion or card-counting with random noise rather than cards. You've got a share price you want to watch because it's constantly moving up and down; it's got an upper and lower bound as well as maximum/minimum deltas, so you know there are good times to buy and good times to sell. The electronic trading system gives you the option of buying and selling and canceling orders at different times in the very short-term future. Like the throw of a coin, the price can go up or down (theoretically, the coin could land on its edge, and the share price could remain static, but there's enough noise in the system to prevent that from happening).

As time goes on, the probability of a lucky streak or a straight run of downs or ups becomes infinitely small. So the trading algorithm has to determine the optimum times of buying and selling. The system can send in an order to buy at time t0, and another order to sell at time t1. If at some time in-between, the price falls, the orders are canceled, otherwise if the price exceeds the target profit, the shares are sold anyway. Because the computer systems are so closed to the actual trading system, all of this can happen thousands of times faster than any remote punter could.

Re:Perversion of Capitalism

[TheLink]

The HFT traders who know what they are doing are NOT doing arbitrage. Because they get a 30 millisecond advantage:

http://www.nytimes.com/imagepages/2009/07/24/business/0724-webBIZ-trading.ready.html

http://www.nytimes.com/2009/07/24/business/24trading.html

When the favoured companies screw up they get the trades rolled back.
When they get beaten by "normal humans", they prosecute the humans: http://www.computerworlduk.com/news/security/3244186/norwegian-traders-convicted-for-outsmarting-us-stock-broker-algorithm/
When they beat the humans, they keep the profit as "rightfully earned".

Re:Perversion of Capitalism

[Your.Master]

You're not paying any attention to what he said.

He's not castigating people for taking these jobs, nor is he even saying that those jobs should just vanish into thin air. He's pointing out a societal problem that leads people to making these choices. Basically, he's agreeing with everything you said about researching the cure for cancer paying less than HFT, and calling that the problem. Surely you can imagine a society where curing cancer is paid like HFT and HFT is paid like curing cancer, even if you can't imagine how to get there from here. The only thing you said that countered the GP rather than reinforcing it was the argument that HFT applies upward salary pressure.

The narrative where people should not be allowed to make their own choices in favour of your choices, is yours alone. At best, it's mistaken, irrational hyperbole. At worst, it is a battle-cry against change and progress and any analysis whatsoever about the status quo.

Re:Perversion of Capitalism

[fishbowl]

The worst-paid doctors are working for no larger reward than student loan forgiveness, with no higher aspirations than being debt free by their early 50s.

Re:Perversion of Capitalism

[fishbowl]

Very few people are at all opposed to healthcare in the hands of the government. Where do you get that idea? Even in the US where "everyone" is opposed to government being involved in healthcare, health programs collectively account for one of the largest parts of the federal budget. I can only assume this is a result of the consensus will of the people. Some government healthcare programs are "good' (e.g., Medicare, the VA, the FDA, the DEA), some are "bad" (e.g., Medicaid and anything approved of by anyone in the current Democratic party). But I don't see any universal opposition to government involvement in healthcare.

The problem I'm seeing is with a perception of health care costs, where there are certain classes of people who are able to take advantage of employer-provided healthcare (people working for mid- to large corporations in salaried positions) and then there are all the rest (people working for those same corporations as hourly employees or contractors, people working for smaller companies, small business owners, and the unemployed). One person might be getting a prescription filled for a $10 co-pay, and another person is expected to pay $400 for the same thing (and ends up not getting it at all). One person might be getting a payroll deduction of $95 for a health plan that another would have to pay $750 a month for (and also ends up not doing it at all).

I'm not making an argument with respect to "government" in health care. I actually think that *employers* should *all* get out of the healthcare business so that the playing field is leveled for everyone, not just for a shrinking privileged class of a certain kind of employee who gets health benefits.

Re:Perversion of Capitalism

[Rich0]

Doh! :)

Re:Perversion of Capitalism

[Setsquare]

account.balance += account.balance + account.balance << 3;

Sadly, I'm afraid the average /. poster (or recently graduate CS'tist for that matter) wouldn't even understand the code refactoring you just proposed. Everything here is "blah, it's easy, I'm awesome, let me show you how", and then they drop a naive turd as an example.

The thing is they're right : shift has lower precedence than adding so what you're really doing is
account.balance += (account.balance + account.balance ) << 3;
which actually multiplies your balance by 17.

Re:Perversion of Capitalism

[Anonymous Coward]

"no educations, no work ethics, no marketable skills and a complete dependency on mommy, daddy and the welfare state for personal solvency."

You mean like corporations that depend on government handouts, er, subsidies, to keep afloat?

Re:Perversion of Capitalism

[Anonymous Coward]

No, the sad part is neither you nor him understand that his little kludge (1) isn't any faster, and (2) won't work with non-integer variables.

Re:Perversion of Capitalism

[Dogtanian]

Skimming money off billions of micro-transactions. Ahh, yes... forget investing in ideas and backing well managed companies... this is the way capitalism was envisioned.

Look, superman III had a lot of lessons to teach. It's really too bad on villains watched it...

Yeah, well if they're getting their ideas from Superman III, I'd be damn worried. Remember what happened *after* they moved on from simple money skimming and designed their own computer?

Re:Perversion of Capitalism

[Tacvek]

Sure, but a such an adjustment is not binding on the actual investor. It cannot be, since no contract at the adjusted price was ever made by the trader. The trader is free to refuse to settle at the adjusted price, and no sane court would compel payment.

Re:Perversion of Capitalism

[The Dawn Of Time]

We aren't aiming for 'benign' capitalism. We simply reject the notion that it is our lot in life to prop up the weak. The great thing about we capitalists is that unlike you idealists, we don't pretend that utopia is achievable. We don't even care about it. We just do our thing.

Re:Perversion of Capitalism

[Anonymous Coward]

And I hope your yeast-filled pussy keeps you vag deep in croutons.

Re:Perversion of Capitalism

[russotto]

The thing is they're right : shift has lower precedence than adding so what you're really doing is
account.balance += (account.balance + account.balance )

So by doing that you've increased your profits by 70% with a perfectly plausible optimization. If it's ever discovered you just claim honest error.

Re:Perversion of Capitalism

[Anonymous Coward]

It's also good for introducing bugs. Depending on the language, that left shift operation may silently truncate to integer, which is bad news if you were using floats. (javaScript does this, for instance.)

Re:Perversion of Capitalism

[arth1]

The thing is they're right : shift has lower precedence than adding so what you're really doing is
account.balance += (account.balance + account.balance )

Thus the "depending on language" part. But yes, it would have been clearer to write (account.balance

Re:Perversion of Capitalism

[Pseudonym]

Many slashdotters are pushing 40 these days, and did their CS degrees back in the day when pipelines were shorter and even Booth multipliers were relatively slow.

The shift/add combination probably won't take "multiple trips through the pipeline" on any modern non-embedded core. If it's an x86 or x86_64, the compiler will probably generate a single LEA instruction (i.e. it won't pollute the I-cache), which is probably compiled to several uops which are issued together. Rather than round-tripping through the entire pipeline, the intermediate values will just get circulated around the CDB, which is fairly quick.

If you're as old as I am, chances are that you may not know this because Tomasulo schedulers weren't in vogue during your CS degree.

But as you are a fellow compiler writer, you must know that even on a modern CPU, the most optimal strategy depends on a lot of context. Using the multiplier isn't always a good idea, for example, since it's usually a more scarce resource than other functional units. Even a relatively recent AMD K10 core only has a single integer multiplier unit (which admittedly only costs a couple of cycles, but there's still only one), and a single floating-point combined multiplier/divider/square root unit (for non-SIMD operations, anyway).

You probably also know that simple integer operations can be effectively "free" if you really know what you're doing, since they can be used to fill pipeline delays from other operations.

But your central point is valid: This is the sort of optimisation that the programmer should not be wasting their time with. Leave the micro-optimisations up to the compiler, because it will almost always do a better job.

Re:Perversion of Capitalism

[Pseudonym]

A likely cause for your confusion is that it's not intuitive for everyone that x *= 10 is the same as x += 9x

Especially people who understand floating point, where it isn't the same.

Re:Perversion of Capitalism

[arth1]

The ones who have done an advanced architecture module will also be able to tell you that ARM is the one exception to this, where the fact that you get the shift for free and the latency for add is shorter than for multiply means that this may be marginally faster.

Careful with that "the one exception". I'm sure there are others too, but just because I remember m68 assembly:

.mul10a
  muls.l #10,d0 ; 42-74 cycles
 
.mul10b
  move.l d0,d1 ; 4 c
  lsl.l #3,d1 ; 14 c
  add.l d0,d0 ; 6 c
  add.l d1,d0 ; 6c, Total: 30 cycles
 
.mul10c
  add.l d0,d0 ; 6 c
  move.l d0,d1 ; 4 c
  add.l d1,d1 ; 6 c
  add.l d1,d1 ; 6 c
  add.l d1,d0 ; 6 c, Total: 28 cycles

But anyhow, yes, cycle shavers should know their environment well. And remember that the best optimization is "don't do it".
Instead of shaving cycles, there will often be better algorithms that will reduce the work at the critical time, but finding and implementing that might mean taking two steps back to look at the bigger picture.

Re:Perversion of Capitalism

[Richard Steiner]

Translation: I already have mine. I don't have to care.

Re:Perversion of Capitalism

[gidyn]

it caused two world wars and then then there was the cold war

Silly me, I thought the second world war was started by fascists, and that the cold war had something to do with communists. If only my history teachers had known that an evil cabal of capitalists was behind it all.

Re:Perversion of Capitalism

[TheRaven64]

Yes, you're right - I was simplifying my response to the level of knowledge that I'd expect from students that I've taught (I did a module on high-performance computing last term, and covered all of the points from my post), so it does contain some oversimplifications.

Re:Perversion of Capitalism

[TheRaven64]

My point is really that I'd expect any CompSci graduate to understand that change, and to know that it's a bad idea. Irrespective of whether it's actually faster on a specific target, you're stealing information from the compiler. If you write a * b, then the compiler can work out the most efficient way of performing a multiplication, which may be via a multiply instruction, via a sequence of adds and shifts, or by something more baroque. If you write a sequence of adds and shifts, then the compiler needs an extra optimisation step to work out that you meant a multiply from this. This kind of pattern matching is harder, and it needs to be done to undo any 'optimisation' that the programmer might do, so it's generally a lower priority than making code fast for programmers that actually say what they mean.

Quite often, this kind of optimisation removes optimisation opportunities from the compiler. You can generally get much more consistent improvements by using things like __builtin_unreachable() (tell the optimiser that a specific case is impossible, so it can optimise assuming that it won't happen).

Re:Perversion of Capitalism

[Anonymous Coward]

Actually, algorithmic trading has been employed for a longtime to avoid large swings cause buy huge buys or sells. The quantity is split and each piece trades at closer to the midpoint. It's better for everyone.
The sells on the way down and buys on the way up are not hft but simple stop limit orders that anyone can put on through a broker.
Hft is not as exposed to large price swings because they hold hedged positions not naked ones.

Re:Perversion of Capitalism

[adamofgreyskull]

Well, actually, to tell the truth, I buy my vegetables from the supermarket. Interestingly, they exploit their enormous buying power and reduced logistical costs and I get a much better price than if I went to the grocer. The grocer's prices are higher because they can't trade in the same volumes and aren't prepared to take the risk of being over-exposed. They only buy small quantities of things, so that if they can't sell them, they haven't lost much. The supermarket, I think, can afford certain losses because they sell higher volumes and can make their money back on pumpkins if the swedes aren't selling well. Anyway, I could get them marginally cheaper if I went directly to the farmers, but that takes too much effort, and I'm prepared to pay the small amount extra at the supermarket for the convenience. See the parallel?

Credit swaps? Who brought that up? (No-one.) Being able to buy and sell stocks at better prices is beneficial to society. So what if it's intangible? My mp3 collection is intangible. My internet connection is intangible. The added value I may get from the supermarket selling me cheaper vegetables is no different from the added value I may get as a result of HFT. In one case, I have bought some potatoes, in the other, I have some stock. In both cases, a third party satisfies the needs of a buyer and a seller and his reward is the difference between the buy price and the sell price. If the seller gets what he wants and the buyer gets what he wants, how is the third party cheating anyone? Fair exchange is no robbery, sport.

Re:Perversion of Capitalism

[Bob the Super Hamste]

If only they had let the market correct then more of the HFTs would have lost their shorts. Some times the invisible hand is trying to bitch slap some actors.

Re:Perversion of Capitalism

[improfane]

What do you think these were caused by? Do you think these 'benefit

You seem quite content that our monetary system is based on the firm foundations of bullshit. That everyone in the world is affected by for survival. None of these trades produce any natural value.

An Mp3 on your PC is entertainment, your internet connection is a communication medium. They have genuine value. A trader produces no genuine value except for himself. People do not invest anymore, they buy and sell on speculation. Nothing is honest. The banks are backstabbers (Goldman Sachs)

This is why you need to have commodity stocks that include the physical delivery of such a product. We're trading on lies. You can read more about this if you do a search...

"The added value I may get from the supermarket selling me cheaper vegetables is no different from the added value I may get as a result of HFT." There is a huge difference. A supermarket's logistics system of lorries and software to plan them benefits everyone. A HFT architecture benefits a bank and makes some rich people richer.

Read Wealth of Nations by Adam Smith. Should be basic economics reading for everybody really. Prosperity comes from mutual trade and mutual value. My wellbeing and job should not be based on a trade of virtual numbers elsewhere. Money reflects a trade in labour. Please do some reading on this...

Re:Perversion of Capitalism

[Bob the Super Hamste]

My understanding is that it is basically they are doing arbitrage since they are playing the spread between what the buyer is asking and seller is willing to sell for.

Re:Perversion of Capitalism

[Bob the Super Hamste]

No in real capitalism all actors would have the exact same access to markets and information as everyone else does. Why do HFTs get to preview orders before everyone else? Why is it that when they make a mistake do their transactions get rolled back? Why is it when the invisible hand of the market attempts to bitch slap a bad actor back to the stone age the bad actor is made whole at the expense of the tax payer. This is much closer to what those on the political left call crony capitalism and I do have to agree.

Re:Perversion of Capitalism

[arth1]

Especially people who understand floating point, where it isn't the same.

Care to back that up with an example in floating point where x *= 10 is not the same as x += 9*x ?
(And we're not talking micro-rounding discrepancies here)

Re:Perversion of Capitalism

[tehcyder]

If a method of making money is currently legal but immoral, you can either choose to make it illegal, or you can just forget about morality.

I don't see any reason to have stock markets that respond in fractions of a second. Investments should be based on long term prospects, not short term profit. It's just gambling, which should be viewed as a hobby rather than an economic model.

Re:Perversion of Capitalism

[tehcyder]

There's a grocer down the road. He buys fruit and vegetables from farmers and sells them to people. What does he produce for society? Fuck all. He's acting as a middle-man between a seller and a buyer. Shit, we'd better close this loophole! It's not just the NYSE that it's affecting, it's right on our doorsteps!!!11!

Bollocks, the grocer stops you having to get up at dawn and visit out of town fruit and veg wholesalers and buying in bulk, so you are paying for a service that saves you time and money.

There is nothing that a stockbroker or stockbroking system does that you couldn't do yourself as long as you were allowed to trade directly over the internet.

Re:Perversion of Capitalism

[TheLink]

Arbitrage is exploiting the difference in different markets. This is happening in the same market.

Seems closer to front running than arbitrage. Only the ones with the 30 millisecond advantage can do it. The differences between this and front-running is:
1) Front running is when a broker buys stuff before buying the same stuff for their customer. This is being able to buy anything before everyone else AFTER you see that they want to buy lots of it. In short they can "front-run" anybody who does not also have the 30 millisecond advantage.
2) This is currently legal, front running is illegal.

Re:Perversion of Capitalism

[Sir_Sri]

geeze what country do you live in? All the people who I went to high school and university with that are doctors, and I'm 31 years old - so that's only 6 or 7 working years, are debt free by now. And we're in canada.

Re:Perversion of Capitalism

[Bob the Super Hamste]

True, but conceptually it is similar.

Re:Perversion of Capitalism

[Anonymous Coward]

I call bullshit on that.

So you are saying that if you want to sell your MSFT stock, there will be a buyer willing and ready to buy that stock. There isn't - not at the price you want to sell. So you sell to the high frequency shop when you sell, and the high frequency shop waits to find the seller who is willing to sell. Some of the time, the high frequency shop can't sell it for as much as they paid. Some of the time, they can. They hope to be able to sell it for more more often than they sell it for less. But they are taking real risk by buying your stock when there isn't anyone else there to buy it at that price.

Remember that the markets have price improvement. So if you go to sell, you'll always sell to the buyer willing to pay the most. If the high frequency shop bids that highest price, then they buy your stock. But they had to take a risk that someone else might be willing to pay more at some later time.

The high frequency player always has to step in front and be willing the pay the MOST to get your trade. Certainly this is good for you, the seller, vs. waiting for another buyer to come up to the price you want.

Since they are providing you this service (you get to trade now, you don't have to wait around to find willing sellers, you get a better price than the current sellers are offering, etc.), there should be payment for that service. There really isn't even payment for these guys though - they have to take real risk, and hope (via quite a bit of math) that they were right about where the stock price was going.

So you get a better price, you get instant liquidity, and all this for a whopping fee of $0.00. Not bad.

Note that in the past, you used to go to your broker, who would do the same thing. Only they were dumber and slower, so in order to make a profit, they would charge you a much bigger spread.

The grocer analogy is pretty good ...

Re:Perversion of Capitalism

[Anonymous Coward]

The fact that this works implies there is no perfect market.
The fact that mistakes are reversable implies that they cannot lose.
I have some trades I would like to reverse, how come I cannot?

Re:Perversion of Capitalism

[Anonymous Coward]

You say rolling back of the welfare state like that's a bad thing.

Re:Perversion of Capitalism

[Pseudonym]

(And we're not talking micro-rounding discrepancies here)

Translation: "Care to back that up with an example of where X and Y are not the same, leaving aside the cases where X and Y are not the same?"

Re:Perversion of Capitalism

[arth1]

Um, no. It means "don't do silly things like comparing two floats for equality". Always take the documented precision of the architecture into account when comparing.

Floats are never going to be accurate - you can't shoehorn a real into a finite format. Heck, even comparing (10 * a) calculated one place with (10 * a) calculated somewhere else for absolute equality isn't guaranteed to be true - the compiler might do different optimizations, perhaps calling a subroutine for the second calculation because there must be a branch anyhow, or too few spare registers to do an inline optimization.

This is rather obvious to anyone who has worked with floats, which is why my comment was in parentheses - only UI for those not aware of the nature of floats.

Re:Perversion of Capitalism

[fafaforza]

How is it a scam, exactly? You were the one to characterize arbitrage as some large scale scam, which it isn't.

Re:Perversion of Capitalism

[PopeRatzo]

You were the one to characterize arbitrage as some large scale scam, which it isn't.

Sure it is. It produces nothing, contributes nothing to the economy, and despite its original intent, only adds instability.

It's a bunch of guys who figured out that if you jigger the Coke machine just right, you get free cans of Coke. They've made so much money now that they've insisted that all Coke machines, by law, have to give free cans of Coke only to them.

Re:Perversion of Capitalism

[Pseudonym]

Um, no. It means "don't do silly things like comparing two floats for equality".

There were no equality comparisons in any code snippets so far. But as an aside, this piece of advice is wrong.

Floats are never going to be accurate - you can't shoehorn a real into a finite format.

And so is this.

You can, for example, store a 32-bit integer in an IEEE-754 double-precision floating point number and do arithmetic and do equality comparisons in perfect safety.

Heck, even comparing (10 * a) calculated one place with (10 * a) calculated somewhere else for absolute equality isn't guaranteed to be true - the compiler might do different optimizations, perhaps calling a subroutine for the second calculation because there must be a branch anyhow, or too few spare registers to do an inline optimization.

If a == a, then 10*a == 10*a. If not, your floating point implementation is non-compliant.

Yes, this does indeed mean that many common "optimisations" used on 8087 ISA derivatives are technically non-compliant, because of the incorrectly rounded 80 bit intermediates. SSE does not have this problem.

Re:Perversion of Capitalism

[fafaforza]

So I guess we'll just keep saying "Yes it is" and "No it's not" ad infinitum.

Traders

[Anonymous Coward]

"the 'ability to work under pressure when the traders are screaming at you' is a must-have skill.""

Yeah, most of the 'traders' I've met were complete assholes, probably because they realised their high salaries were more down to luck than skill.

Fortunately when I screw up air traffic controllers just have to start rerouting airliners with hundreds of people on board so they don't crash into each other, so I clearly couldn't justify being paid a Wall Street salary. I can't imagine the stress level of having a 'trader' screaming at me.

Re:Traders

[artor3]

It's not even luck, either. It's down to your ability to hob-nob with other psychopaths. My friend's brother-in-law was a trader five years back (pre-recession) and lost something like $800 million (of other people's money) in bad speculations. He was fired for it, but hired into a new firm thanks to his "connections" a few weeks later, at a higher salary than his previous job. Oh, and of course he kept the bonuses he made at the first firm before his investment went bust.

Wall street is the world's greatest argument against the notion that capitalism rewards people in proportion to their skill and hard work.

Re:Traders

[Anonymous Coward]

Traders are an interesting breed, yes they can come across as assholes, that's pretty much a requirement (you need an ego the size of Texas for one thing). But having worked around them, there's no way I'd want their job, for any money. The hours, the stress, the never-seeing-your-family. Even if I could do it (and I really don't think I'd be capable) I'd never want to. So as far as I'm concerned they earn their money...doing something other people probably think is easy but is very, very far from it.

Re:Traders

[Jane Q. Public]

The thing is: Wall Street anymore is not very "capitalist", in the historic sense of the term.

"Capitalism", in general, refers to making capital investments in goods that in turn will make a profit. Often, that means indirectly and over time: you spend $1M to buy a machine that makes thousands of widgets a day, and sell the widgets for $5.00 each.

However, with the current obsession with turnover and a quick dollar, what you end up with on Wall Street is really much more pure money speculation, which doesn't have much in the way of capitalist underpinnings. In fact it is a lot more like gambling, which has been around a lot longer than capitalism has.

We simply haven't been seeing the long-term investment in capital goods, which in turn power a robust economy, and at one time made the US the greatest economy in the world.

Until we see a return to something resembling real capitalism (and it still does happen, just not so much on Wall Street anymore), we will continue to have economic problems. A corollary to that is something I have been saying for a long time: Washington needs to stop concentrating on viewing Wall Street as some kind of driver for the economy. That's a false economy. Reasoning: anything that can lose a large percentage of itself overnight is not a stable thing on which to base a real economy. Keynesian "created" fiat dollars can disappear just as fast as they are created, as we saw quite clearly in 2008.

And for similar reasons, I believe HFT should be outlawed. It isn't even a real market anymore; it's simply a contest to see who can spot a discrepancy in perceived value the fastest.

Re:Traders

[timeOday]

I agree HFT is rubbish, but where do you see justification for long-term investment in capital goods in the US? It seems that industry is being moved to where it can be operated more cheaply - there is no economic justification for more investment here. In fact there's a surplus of "capital," at lest in the sense of invested money seeking good returns. (Overall the S&P has returned almost exactly 0% over the last 10 years.) So we keep getting bubbles due to over-investment, first in .com, then housing... what is the "next big thing" where we could invest to bring real growth?

Re:Traders

[Anonymous Coward]

Keynes didn't invent fiat currency. There's nothing Keynesian about credit default swaps.

Keynesianism is the idea that a recession is when the private sector prefers to hold on to cash and financial assets rather than buying stuff and building capital assets. With that basic model of the economy, Keynes argues that in a recession, the right thing for the government to do is to borrow and spend more money to increase demand.

It's simple, fairly obvious, and it works. Because it works, the world's financial elites and power brokers have refused to allow it to be tried this time around.

Just to be clear, Keynes would ask the President to create a TVA, a CCC, and half a dozen other jobs programs, building the infrastructure people would depend on for the next 80 years while revitalizing the economy by putting cash in people's pockets with which they turn around and buy stuff. Sadly, we were stuck with Obama instead of FDR this time around, and not quite 60 Democrats in the Senate.

Re:Traders

[Anonymous Coward]

Having worked on two trading floors (one as a member of a HFT/algorithmic team), this aspect of the job is somewhat exaggerated. Yes, there is a lot of pressure from the traders when something goes wrong, but the traders themselves usually maintain their composure quite admirably. It goes with the territory--you need to be able to remain calm and collected when the unexpected (or undesirable) happens.

I have never witnessed any screaming. There is definitely shouting, but it's not in any way belligerent. The HFT developers sit on the same trading floor (maybe a few seats over or one row back), and when the traders see something is amiss, they'll shout over to us to make sure we're aware of it without leaving their seats or taking their eyes off the screen. We then shout back to tell them that we're looking into it, bouncing some component, etc.

I suspect that if a trader actually did start screaming, the lead trader would tell him to get out and not come back until he can keep his cool.

Another thing to consider is that on an HFT team, the trading is algorithmic. While some traders occasionally place trades manually, much of their job involves pouring over trade logs looking for anomalies: missed opportunities, trades which shouldn't have been executed, periods of especially low or high activity, etc. Luck doesn't really factor in to an HFT trader's performance since they're not the ones doing the trading. Much of that stress gets shifted to the developers--particularly those who develop the strategies.

Re:Traders

[Anonymous Coward]

When people talk about 'return to a capitalistic base'. They are talking about the sort of thing where 500 people build things on assembly lines making things. That sort of thing only exists in very low wage countries these days. In higher wage countries those jobs are automated.

I was watching a 'how its made' show. They showed a factory that was picked up 100% brick by brick and moved to another country from the US. Why? The factory was build for building things using humans. The labor here costs too much. It is that simple. You can buy 2-3 workers in another country for the same as here. Most of the shows they have 20 people working in a small factory that pumps out 100k in items a day. Most of the things they show are nearly automated are so close its not even funny.

Wall street is not a real economy but they have *MAJOR* effects on our economy. To minimize them is not right either. But the fed is made up from people that come from there. So they do what they know...

Also the dollars disappearing was a result of bad loans unfolding. It was also an effect of 15 years of unwinding of the 1930s laws that protected the rest of us from those silly antics. In the name of 'simpler paper work'. It was an effect of low hold rates placed upon the banks to create the ability of banks to loan money on the cheap. In many ways it was an echo of the 2000 crash. Which was an echo of the 1992 crash. Which were a direct result of the 1986 crash and wanting to 'free up cash to drive the economy'. They unhinged the very things reigning in wall street. To keep wall street from creating a 1930s crash. Which once those reigns were unshackled in 1997 they did. The 2008 crash is almost exactly the same as the 1930s one.

HFT is just a game to see who can screw each other faster. Making millions off the backs of others all in the name of 'liquidity'. They could put and end to it by doing 3 things. Making people take delivery of a % of commodity goods they buy. Also by limiting the number of times per day you can put in an order for a good say start with 3. Creating a minimum owned time. You buy something you must keep it for at least 1 day.

HFT is creating demand that doesnt really exist and in effect creating liquidity that doesnt exist. As my econ teacher said on the first day talking about government regulation or lack of it. You can cheat the market but eventually the market fixes itself. 2008 was a major correction. There is another bubble building in the commodities market. Whoever gets caught on the other end of that one is really going to get their shorts burnt.

This 'false' liquidity pushes prices up and creates cash that does not exist. It is why you are seeing 100+ oil, and 1600oz gold. Never mind the printing spree the gov went on. Plus the 'possibility' of default. The last 2 we can not really gauge the effects because of the first one. I read in the WSJ of 1 guy controlling 75-90% of the gold market 2 years ago. With only 1% of the cash needed to do so. Something is very wrong there...

Re:Traders

[MaskedSlacker]

Trading is a job like playing slot machines is a job.

Re:Traders

[sjames]

Bank robbers are an interesting breed, yes they can come across as assholes, that's pretty much a requirement (you need an ego the size of Texas for one thing). But having worked around them, there's no way I'd want their job, for any money. The hours, the stress, the never-seeing-your-family. Even if I could do it (and I really don't think I'd be capable) I'd never want to. So as far as I'm concerned they earn their money...doing something other people probably think is easy but is very, very far from it.

Re:Traders

[Anonymous Coward]

Well, I hate to sound like I defend HFT, but I think the best way to discourage its use is to prove that it doesn't actually do anything valuable. When I read the report of the last flashcrash (of 2010, prior to reading about this year's flashcrash), one thing came to mind: are these algorithms really computing anything like economic value? Anyways, I'm thinking they're not proper algorithms as they're not actually calculating the marginal [best] use of a good they're trading compared to that of a trader who does (in some respects). The price of a good doesn't always reflect its economic value nor its marginal use. I'm not sure if I should pursue this line of thinking as a Master thesis, but I might float it around the CS department at my university to see what they think.

Re:Traders

[LordLimecat]

Wall street is the world's greatest argument against the notion that capitalism rewards people in proportion to their skill and hard work.

And human nature (as well as history) is the greatest argument against any idea that tries to replace capitalism.

Re:Traders

[Anonymous Coward]

Well, you're an idiot. For every HFT out there, there's 2 fundamental investors (probably a lot more than that although I don't have hard numbers). Obivously the HFT do a lot more trading since they're high-frequency, but in terms of dollar capital, there is probably way more fundamental investing.

Re:Traders

[Anonymous Coward]

I believe HFT should be outlawed. It isn't even a real market anymore; it's simply a contest to see who can spot a discrepancy in perceived value the fastest.

You need to get your perspectives straight. Do you really think HFTs are causing enough harm to the global economy to merit outlawing?

The most harm an HFT has ever caused was the flash crash in May 2010, and that lasted about 20 minutes, and a LOT of rational people made a ton of money when they realized a bomb hadn't gone off somewhere and it was just somebody that wanted to sell a lot at market price without a dark pool.

There are two reasons why things are shitty right now. (1) People got greedy to the point of being extremely self-destructive, and (2) Banks got greedy on people's greed and then went under with them when people defaulted across the country.

Wall Street is always going to speculate. Banks will always try to screw you. (Free checking is as "fuck the poor" as it gets.) That's what these entities exist for. I don't see how regulating how long a person/firm holds a stock will do anything but create a race to who can get as close to that limit as possible. Speed is here, it's not going away.

Regular people started speculating housing prices, had a little tulip mania, and here we are.

Normal salaries vs housing price ratios are historically 1:3. During the housing bubble in the time coming up to ~2007, in sand states the rates were more like 1:8 or 1:10. When one investigated those housing loans, you'd often find empty houses. People were outright speculating that the price of housing would continue to rise, "real estate is an investment!" mantra on crack. Then we get banks like AIG FP that are packaging up subprime mortage loans (70% of the population had a better credit rating than the pop that parent company AIG heavily insured and eventually lead to need a gov' bail out to prevent mas).

This all wouldn't have been possible without floating interest rate loans (initially 2% interest for the first three years! Then an effective 12% mortgage on a 30 year loan. Gee, I wonder what's going to happen when middle class Americans have $500,000 mortages at 12%?). Banks made this stuff REALLY popular (researched two dozen credit cards today, and not a single one told me what my interest rate would be after the first 6 months without clicking through and opening their full conditions in fine-print pdf).

Re:Traders

[Jane Q. Public]

"I agree HFT is rubbish, but where do you see justification for long-term investment in capital goods in the US? It seems that industry is being moved to where it can be operated more cheaply - there is no economic justification for more investment here."

Industry WAS being moved to where it can operate more cheaply. And there is plenty of justification. If you're American.

First, let me say this and get it out of the way: China's economy has been slowing. It's not exactly something they advertise, but it's true nevertheless. Further (and I won't claim this is directly related): there is a new Chinese "consumer class" that scarcely existed before, and they are hankering for goods that are not made at home. But having said that, I'm not going to belabor it, and instead I will concentrate on the U.S.

We have learned some things over the last couple of decades. One of the things we have learned is that regardless of whether it results in cheaper goods, outsourcing is bad for the economy. We know for a fact (it is not a guess or just a theory anymore) that it costs jobs at home, and those jobs are more valuable to our economy than the cheaper goods are.

Further, outsourcing is not "fair trade", because it artificially bypasses exchange rates. It is easy enough to say that "U.S. labor can't compete with foreign labor for the same amount of pay". Which is a true statement... but it's only half the story. It's a lie by omission. Because we aren't really competing on an equal basis. Let me construct a simplified example to illustrate the point (numbers just pulled out of the air to make the point):

Company X wants to hire people to manufacture widgets. It can hire unskilled Americans for $10 per hour, or Crotobaltoslavonian workers for $2 per hour. But here's the thing: In America, that $10 buys about 5 pounds of rice. In Crotobaltoslavonia, that $2 buys 10 pounds of rice. So in actual purchasing power, you're paying that Crotobaltoslavonian the equivalent of $20 per hour in America. Americans literally can't compete for that amount because it doesn't pay them enough to eat... in THEIR economy.

And that's the whole point. Economies are different. That's just the reality of the situation. And that's why we have exchange rates... to keep trading FAIR between countries. But hiring labor in a foreign country, in effect, bypasses that exchange rate, and so it is not a "fair trade". Americans are hurt in the process... even if rich Company X gets to keep a few more dollars. We are exporting even more dollars than what they keep as a result... and that doesn't help our economy, either. So it hurts in several different ways, not just one.

Savvy companies actually realize this, but they just don't care. Which is pretty unethical, if you ask me. They are willing to sacrifice your $10 so they can make an extra $1 in eventual profits. And you should be angry about that.

Plain and simple: we need to bring the manufacturing jobs back home. Partly because of the direct monetary imbalance that outsourcing causes, and partly because we need to keep that capital investment (in machines and buildings and other capital goods that can actually MAKE things) here at home, rather than letting it be done overseas at our expense.

"In fact there's a surplus of "capital," at lest in the sense of invested money seeking good returns. ... So we keep getting bubbles due to over-investment, first in .com, then housing... what is the "next big thing" where we could invest to bring real growth?"

The bubbles aren't due to a surplus of capital, per se. I mean they are, but the proximate cause of that surplus is indisputably government monetary policy. They have been keeping interest rates artificially low to "stimulate" the economy (BEFORE the 2001 and 2008 bubble burstings). Easy money means more capital. But it wasn't invested in capital goods, intended to

Re:Traders

[Jane Q. Public]

I meant to add:

Q: How can we bring jobs back home?

A: By directly -- and heavily -- taxing companies that outsource labor and manufacturing. This avoids the pitfalls of tariffs: we would not be blocking trade from coming into or going out of the country. However, we would be demanding compensation for our economy, from the companies that have been so harming it by their practices. That's about as fair as it gets.

(It's also not a subsidy by the government... so it avoids those problems, too. Instead, it is a source of revenue TO the government.)

Re:Traders

[obarthelemy]

X-Lib-of-Cong-ISSN: 1098-7649 -=[ Fun_People ]=-
  X-http://www.langston.com/psl-bin/Fun_People.cgi
  Forwarded-by: Nev Dull
  Forwarded-by: "John P. Kole"
  Forwarded-by: Vince Cavasin
Monkey Trumps Wall Street With 200 Percent Gain
  Dart-Throwing Monkey Returns to Wall Street With New Picks For Year 2000

  Business Wire
  01/12/00, 9:41p
  (Copyright ' 2000, Business Wire)

LOS ANGELES--(BUSINESS WIRE)--Jan. 12, 2000--Raven, the dart-throwing monkey with her own Web site, showed up many of Wall Street's finest with her 213 percent gain for the year.

  "It's all in the wrist action," stated Raven, age six. A Web site and index have been created to monitor her performance. The Web site and the index can both be found at www.monkeydex.com.

  MonkeyDex is the Internet's first index of Internet stocks picked by an actual monkey. MonkeyDex was created in January of 1999 when Raven, a six-year-old female monkey, tossed darts at a dartboard of 133 Internet-related stocks. Raven returned to Wall Street this year with a dart toss at a dartboard of 281 Internet-related stocks.

  "She quadrupled the performance of the Dow and doubled the performance of the Nasdaq composite," stated Roland Perry, editor of the Internet Stock Review and creator of the MonkeyDex. "Not bad considering she wasn't able to participate in any of the hot new issue offerings," he added.

  "And yes, she beat the Internet Stock Review's top 20 picks for 1999, which gained 79 percent last year. We did manage to outperform her with our Original Watch List, which gained 490 percent in 1998 and went on to finish 1999 with a two-year gain of 1,566 percent."

  Had Raven been employed at a Wall Street Mutual Fund, her performance would rank her as the 22nd best money manager in the country outperforming over 6,000 Wall Street pros.

  Raven's picks for 1999 were AudioHighway (Nasdaq:AHWY), down 21%; CMGI Inc. (Nasdaq:CMGI), up 935%; iMall (Nasdaq:IMAL), bought out; Inktomi (Nasdaq:INKT), up 177%; ISS Group (Nasdaq:ISSX), up 158%; Kushner-Locke (Nasdaq:KLOC), down 36%; Lycos (Nasdaq:LCOS), up 194%; Netspeak (Nasdaq:NSPK), up 88%; Onsale (Nasdaq:ONSL), bought out; and Ozemail, which merged with MCI WorldComm (WCOM).

  Raven's picks for year 2000 are Audible.com (Nasdaq:ADBL), Broadcom (Nasdaq:BRCM), eToys (Nasdaq:ETYS), Litronic (Nasdaq:LNTX), Medium4.com (OTCBB:METV), Lycos (Nasdaq:LCOS), N2H2 (Nasdaq:NTWO), Prodigy (Nasdaq:PRGY), Software.com (Nasdaq:SWCM) and StarMedia (Nasdaq:STRM).

  This year's dart-tossing ceremonies were recorded by a film crew from the National Enquirer and will air on National Enquirer TV (http://www.nationalenquirertv.com/):

  Boston, WFXT -- 12:30 p.m., Midnight
  Chicago, WFLD -- 12:30 p.m., Midnight
  Los Angeles, KCAL -- 3 p.m., 12:30 a.m.
  Miami, WBZL -- 1 p.m., 2 a.m.
  New York, WNYW -- 11 a.m.
  Philadelphia, WTXF -- Noon
  San Francisco, KBWB -- 11 p.m., 12:30 p.m.
  Washington, D.C., WTTG -- 12:30 p.m., 5 a.m.

  About the Internet Stock Review Online

  The Internet Stock Review Online is an Online newsletter designed to give broad coverage to Internet-related publicly traded companies. Each issue of the Online newsletter aggregates and reports on New Buys & Sales on Wall Street, coverage initiated or dropped on Wall Street, Big Movers, Earnings (or lack of) reports, news and/or news releases from Business Wire or PR Newswire, news from industry media or trade journals, news from traditional media outlets.

  Each issue includes hyperlinks to the news origination source for full coverage. The newsletter additionally reports on the availability of streaming audio/video profiles, news releases, shareholder conferences and interviews with senior management of the companies covered.

  MonkeyDex
www.monkeydex.com

Re:Traders

[Anonymous Coward]

Hey, here's an idea. Maybe you can't get fucking richer than you already are? The whole world is looking at Wall Street as the panacea, as the example to follow, and you still want more?

Re:Traders

[Anonymous Coward]

Fortunately when I screw up air traffic controllers just have to start rerouting airliners with hundreds of people on board so they don't crash into each other, so I clearly couldn't justify being paid a Wall Street salary. I can't imagine the stress level of having a 'trader' screaming at me.

A quote from Terry Pratchett's "Feet of Clay."

When a human doctor, after much bleeding and cupping, finds that a patient has died out of sheer desperation, he can always say, "dear me, will of the gods, that will be thirty dollars please," and walk away a free man. This is because human beings are not, technically, worth anything. A good racehorse, on the other hand, may be worth twenty thousand dollars. A doctor who lets one hurry off too soon to that great big paddock in the sky may well expect to hear, out of some dark alley, a voice saying something on the lines of "Mr. Chrysoprase is very upset," and find the brief remainder of his life full of incident.

Quite a lot of people don't value other human beings as much as they do dollars. Ask someone out there if they'd let someone die for $100,000, $250,000, $1,000,000. Most people would just argue about what "letting someone die" means instead of rejecting it outright. Which shows that they're willing to assign dollar signs to people's lives. (I included by the way. I'm not immune to this either.)

Is it any wonder that a trader would feel more stressed out than someone with actual human lives on their hands? It just shows how fucked up out society is sometimes.

Re:Traders

[snowgirl]

I love how it's not "real" capitalism because they're turning over stocks for profit instead of objects with physical properties.

Re:Traders

[Jane Q. Public]

"Keynes didn't invent fiat currency. There's nothing Keynesian about credit default swaps."

I agree. I wasn't implying that the concept of fiat currency was Keynesian; the Fed was created before he wrote his treatises on economics.

However, Keynesian monetary policy could not exist without fiat currency. When the value of money is set by a free market of goods, government manipulation of the money supply via interest rates (which is what Keynesian theory is all about) is simply not possible.

"Keynesianism is the idea that a recession is when the private sector prefers to hold on to cash and financial assets rather than buying stuff and building capital assets. With that basic model of the economy, Keynes argues that in a recession, the right thing for the government to do is to borrow and spend more money to increase demand.

No, Keynesianism is not just one idea, but if it could be summed up as a single idea, that idea would be that the government can control the economy by influencing the money supply via interest rates.

It's simple, fairly obvious, and it works. Because it works, the world's financial elites and power brokers have refused to allow it to be tried this time around."

Wrong. It's simple, fairly obvious, and it DOESN'T work. If it did, 80 years of Keynesian theory would not have landed us in this mess.

Re:Traders

[Jane Q. Public]

"Wall street is not a real economy but they have *MAJOR* effects on our economy. To minimize them is not right either. But the fed is made up from people that come from there. So they do what they know..."

I do not agree with your assessment of capital investment, but this is certainly true as far as it goes. On the other hand, there are subtleties you do not account for.

First off, they do have major effects on our economy, but that's simply because we let them. All of Keynesian economic theory is wrapped around the idea that money "created" by the stock market (and fractional-reserve banking, but that's another issue) is "real" money.

And yet we have seen that it is not real money. Many times in history, but most recently in 2008, we have experienced that money going "poof!" back into the thin air from which it was created. And unfortunately (no pun intended), it has a tendency to go "poof" a lot faster than it was created in the first place.

If we did not treat that as real money... if we again had, say, a gold standard... then this could not happen. Markets could still collapse, don't get me wrong. But SOMEBODY would still have the money. It couldn't simply disappear.

"It was also an effect of 15 years of unwinding of the 1930s laws that protected the rest of us from those silly antics."

Agreed 100%. You obviously know your history and have been paying attention. +10.

"HFT is just a game to see who can screw each other faster. Making millions off the backs of others all in the name of 'liquidity'."

Again, agreed. This is basically what I was saying, in different words.

We do not disagree so much. Except about the capital investment thing. On that we will have to agree to disagree.

Re:Traders

[Trogre]

I think Dave Barry said it best (okay it was some time in the 1980s, but still holds true today):

The stock market of the 1920s was very different from the stock market of today. Back then, the market was infested by greed-crazed slimeballs, get-rich-quick speculators with the ethical standards of tapeworms, who shrieked "buy" and "sell" orders into the telephone with no concern whatsoever for the nation's long-term financial well-being. Whereas today they use computers.

Re:Traders

[Jane Q. Public]

"Well, I hate to sound like I defend HFT, but I think the best way to discourage its use is to prove that it doesn't actually do anything valuable."

You have grasped the essence of it, but missed the big picture: Because HFT simply does the same thing, only faster, then in order to prove that it doesn't actually do anything valuable for the real economy, you would have to prove that the stock market, per se, does not do anything valuable for the real economy.

Which is true. But good luck with that. Because people have this idea so entrenched in their minds that they let it effect the real economy, even when it shouldn't.

Re:Traders

[Jane Q. Public]

"You need to get your perspectives straight. Do you really think HFTs are causing enough harm to the global economy to merit outlawing?

The most harm an HFT has ever caused was the flash crash in May 2010, and that lasted about 20 minutes, and a LOT of rational people made a ton of money when they realized a bomb hadn't gone off somewhere and it was just somebody that wanted to sell a lot at market price without a dark pool."

My perspectives are straight enough. And frankly, I don't give a damn about the "global economy", and neither should anybody else. The current popular concept of "global economy" is a myth based on manipulated fiat currency, and is best ignored. Not that our so-called "leaders" will let us actually ignore it, but nevertheless THAT house of cards needs to fall down, and soon, so that people can get back to genuine, rather than artificially and politically created, economic issues.

Saying "the most harm it's ever done" was a little blip is like saying "the bullet only grazed me". It bears no relationship to the damage it could cause.

"(1) People got greedy to the point of being extremely self-destructive,"

Which people? Consumers? You mean the people who were sold loans when they had inadequate income and credit credentials, and were told by the experts that it was okay, and that they could afford it? Oh, yeah. Those greedy people. The ones who believed what they were told by the government, the news, and the finance companies. We simply must protect America from those "greedy" people!!! They wanted homes and were told by all the people in positions of power and knowledge that they could afford it. They must be evil.

"(2) Banks got greedy on people's greed and then went under with them when people defaulted across the country."

Well, the first part is right except for the "people's greed" part, because it was in fact the banks and finance companies who talked them into taking those loans. It wasn't after the fact.

But more to the point is: the banks really didn't go under when the people did. Did they? For the most part, the banks got bailed out, but the people didn't. And the banks knew that would be the case should disaster strike, so they DELIBERATELY made those bad investments. Don't blame it on the consumer. The banks and finance companies deliberately made malinvestments, and they knew full well what they were doing... unlike their victims.

"This all wouldn't have been possible without floating interest rate loans (initially 2% interest for the first three years! Then an effective 12% mortgage on a 30 year loan. Gee, I wonder what's going to happen when middle class Americans have $500,000 mortages at 12%?)"

Nonsense. The government was deliberately keeping interest rates low, and they even said it was to help people buy homes. It was the government's deliberate policy of making interest rates artificially low that caused the malinvestment... that, coupled with the repeal of regulations preventing banks from investing in speculative instruments. We know for indisputable fact that bad loans were being made DELIBERATELY, and then those debts bundled into derivative instruments that were grossly overvalued.

None of this is a secret, and it is ALL due to the banking and finance industries (if you count mortgage companies as finance, which you should), knowingly taking advantage of a regulatory situation which they KNEW would let them get away with murder. Blaming this on the "greed" of consumers, who simply wanted to own a home and were told they could, is grossly distorting the actual picture. The amount of malinvestment and defaults due simply to ARMS had very little to do with the overall picture, OR the dollar value.

Re:Traders

[Jane Q. Public]

Well, it isn't. Turning over stocks for short-term profit is gambling, not capitalism. Please look up the historic meaning of "capitalism". It isn't just money-mongering.

Re:Traders

[Lord_Jeremy]

From my understanding, stronger regulation of the financial sector essentially has the same effect of having a gold-backed currency, with the advantage that there is actually a degree of control over the market fluctuations. Again, this is only my understanding, but the concept of money "disappearing" is essentially someone thinks they have money but they actually don't. That's a gross oversimplification but I think it's an apt way of describing what essentially is a broker thinking his plan for gains has executed properly when he was in reality outpaced (outsmarted) by other "forces" in the market.

The phrase "unwinding of the 1930s laws" reminds me of a clip from Capitalism: A Love Story in which Michael Moore talks about a photograph showing treasury dept. members that were formerly involved with Lehman (I think?) cutting away the red tape of financial regulation. Wall Street is a casino, with gambling fanatics running around with no net gain. But the casino owners, the few who have put themselves in a position to run the game, are coming out farther ahead than can be imagined. There most certainly are people who are fully aware of where this has been heading since the beginning.

Re:Traders

[Jane Q. Public]

To be a bit more specific:

It isn't "physical objects", per se. You are confusing my characterization of fiat dollars (which are, indeed, nothing but smoke with influence) with my description of market trading. They are related but I don't pretend they are the same things.

It is what you are trading for that makes the difference. I was not referring to, say, investing in some commodity (pork bellies is the usual jest) versus stock in some company that you think might go somewhere. There is nothing wrong with investing in the stock of some company... as long as it is a real company and it is doing real things in the economy. Producing some kind of product (whether knowledge, software, paper, or steel), paying employees, and so on.

However, Wall Street has become (has been for some time really) a morass of people investing not in companies with a real future, a "capitalist" investment in the classical sense, but rather in what value their stock is from one day to the next. Which turns it into pretty much money speculation, rather than actual investment in companies that produce real goods over time. (Again, without regard to what those goods are. I don't care if it is software, toilet paper, reports from a consulting engineering company, or adult toys. It's still a real product.)

High-frequency trading has simply taken this idea to an extreme, and exemplified how much Wall Street has become a disease, rather than part of the metabolism.

Re:Traders

[Jane Q. Public]

Perhaps the link I posted earlier will do a better job of explaining what real capital wealth is, than I have. It's late and I am tired. But please note what this man says. He was exactly right about the economy (as Austrian economists have been for quite a while), while all the people following Keynesian theory (official government policy) were 100% wrong.

He explained just exactly what was going to happen, and why.

And as he says: real capital wealth consists of goods and the infrastructure to produce those goods, not whether the stock index is at 12,000 or 13,000. The goods need not necessarily be physical: informational goods are still goods.

Re:Traders

[Jane Q. Public]

"From my understanding, stronger regulation of the financial sector essentially has the same effect of having a gold-backed currency, with the advantage that there is actually a degree of control over the market fluctuations."

We disagree in principle. Historically, the most stable and robust economy has been one of free market, free of outside "control".

We have had 80 years of direct government manipulation of the economy, and it has almost invariably resulted in bad consequences, throughout the entire period.

I do not agree that more government control, even if it is more effective, is the answer. Frankly I believe, and the preponderance of the evidence shows, that if governmental control were more effective, things would be that much more the worse.

I'll pass on that, thanks very much.

As for money disappearing, in order to understand that it actually does disappear, you have to understand how it is created in the first place. When faced with the reality of this, many people are actually horrified and choose not to believe. I will admit, that video is perhaps overly dramatic, but it gets the facts pretty straight. (A few minor details are not exact. For example: "I killed the bank" were not Jackson's actual last words, although he did give that answer to the question on his deathbed, on the same day he died. These tiny details do not affect the essential truths told by the video.)

I agree with you about the casino bit. However all of what I have been saying here should indicate to you that I take issue with Michael Moore's blaming of our economic situation on capitalism. On the contrary: it is our deviation from capitalism that has led us to this mess. Modern government monetary policy could not be said to be "capitalism" by any sane person who knew the definition of the term.

I happen to like Michael Moore as a person. However, he tends to get idealism confused with reality, and sound-bites confused with actual theory. That is the most generous statement I am likely to make about him right now.

Re:Traders

[scamper_22]

and what makes one think there should ever be economic growth?

I ask that as a very serious question. For most of human history, economic growth was essentially 0%. Your parents lived pretty much the same life you did who would live pretty much the same life as your great great great grand children.

This changed briefly during the industrial revolution, which I'll say broadly happened from about the 1700s to now. But now we're post industrialized.

We have our running water, our shopping centers, our communications, our electrical grid... I dare say humanity really isn't that complex. I grew up in the 'developing world' and by in large people want to do the same things as people in North America. Have fun, eat good food, interact with friends and relatives... in short nothing different in the big picture.

Now don't get me wrong, I'm sure well invent crazy things, but can you really get more real economic growth? Our needs are satisfied in the industrialized world. For lack of a better term, walmart is there. Is there anything walmart can't satisfy? Walmart might get replaced, but it will at the end of the day just be another company replacing what walmart already does. For example, amazon might do it, but its not really creating economic growth on a broad basis. How many pairs of jeans are you going to buy each year? How much are people willing to work to get more? Not much more.

Now sure, we'll invent new things and gain efficiencies. NetFlix will replace blockbusters. And we'll continue to invent all kinds of technology... but we're not really doing anything that will generate economic growth. For example, the 'green' industry is not a real new industry generating new wants and needs. It's just a replacement industry.

Now sure, I could be wrong and maybe there will be the 'next big thing' that drives demand. However, I think the industrial revolution was a unique point in human history. It is the exception, not the rule.

It was especially good when fulfilling the new wants and needs required mass labor. For example, at its peak, the telephone industry required telephone switch operators. This has all changed due to advances in computing.

Whatever new fields we do enter will not generate mass employment due to computing.

And so yes, we have generated huge institutions around the industrial revolution that are by in large struggling. Wall Street is one of them, dependent on economic growth that is getting harder and harder to find. But it's not just Wall Street. The public sector as well is dependent on economic growth. The industrial revolution and wall street allowed your teachers, doctors, police officers to be paid more than what society can actually afford. They were paid based on promises of economic growth and the highly paid labor of the industrial revolution.

I think at this point we have to separate 'capitalism' from the 'free market'. Make up whatever words you want.

The free market can adapt to new circumstances as it essentially just says that people choose the goods and services they wish to purchase for a price they can freely agree to.

But 'capitalism' is suffering from the same problems of communism. Capitalism has transformed into a command and control system. It's just bankers in charge instead of social planners. But the results the same. They are unable to adapt to new realities, beholden to special interests or predict the future enough to plan for it.

Our entire society is based around economic growth which simply is not guaranteed to exists. And so yes, we have bubbles. We have cheap money trying to inflate bubbles. We have HFT to try and keep wall street afloat. It's the only thing we can largely do to preserve the current system.

And so I agree we have had really 0% real growth. I just happen to think that's the historical norm. I don't think there's any real growth left in the big market as a whole.

As we look around the world, the only growth left is in countries industrializing. Human population is also stabilizing which also hinders economic growth..

Rather than try to get 'real growth'. I'd suggest we treat growth as a bonus when it happens. But our society should not be dependent on economic growth.

Re:Traders

[Anonymous Coward]

I agree HFT is rubbish, but where do you see justification for long-term investment in capital goods in the US? It seems that industry is being moved to where it can be operated more cheaply - there is no economic justification for more investment here.

Less costly operations usually mean more costly capital. eg: it's cheaper to have a robot than a human bolt your car together. The robot costs a hell of a lot more than a screwdriver, and requires a much greater capital investment that you hope to recover through lower operating costs

Re:Traders

[Anonymous Coward]

what is the next 'big thing'?

Gold - a lot of gold holdings are reprsesented by a piece of (possibly electronic) paper saying the person 'owns' some gold. Sounds just like a fiat currency to me. What happens when sum(value of these claims to gold) > sum(value of physical gold)?

Re:Traders

[AchilleTalon]

"the 'ability to work under pressure when the traders are screaming at you' is a must-have skill." So, being deaf is a plus on your CV.

Re:Traders

[Anonymous Coward]

And the companies we are taxing will just move out of the country entirely. Since we won't be using tariffs, their prices will go back down, and now even more foreign businesses will have a competitive advantage over American businesses, and even more jobs will go overseas, and our economy will get even worse.

Re:Traders

[Bacon Bits]

You realize all that would happen then is that the company itself would relocate overseas. It might have very large offices in the US, but they would all be "branch offices". All this tax would do is punish companies that are assholes enough to outsource labor, but not assholes enough to evade taxes by leaving the country.

What do you do then? Require businesses that do business here to be based here? Ultimately, it's a walled garden approach. Someone will root it and break the system.

No, the best option is to let them outsource. Eventually -- and I mean in probably a century -- local economies will realign to the global economy and outsourcing will not save the amounts of money it does today. This is what is happening in China and India. The problem with businesses screwing people over is that people aren't as stupid as business would like them to be. People figure it out, and start demanding silly things like equality and justice, and those are things that governments tend to take an interest in.

Re:Traders

[Opportunist]

Wall Street can be a driver for economy, but we have to get rid of the locusts. The "investors" that try to squeeze money quickly from companies, bleed them dry, dump them and repeat this bloodletting with the next. To do this, there is very little actually needed: Reward long term investment and the creation of goods and services (and jobs!), and punish short time gambling. For this, all you have to do is to install a hefty tax on short term trading and reward holding stock. One could, e.g. create a tax that is crippling for daytrading, but progressively diminishes the longer the time between buying and selling that stock gets. Make dividends paid according to the actual productive performance (productive! I.e. based on the creation and selling of goods, not shifting money around and dabbling in banking, insurance and real estate) tax exempt or at least lowly taxed.

That way investors would be keenly interested in investing long term in a company that produces goods and services. Which is, essentially, what Wall Street and investment is good for. If it's done for everything else, it's an economical parasite and should be eliminated from the system.

Re:Traders

[Antique Geekmeister]

> We have had 80[0,000] years of direct government manipulation of the economy, and it has almost invariably resulted in bad consequences, throughout the entire period.

I fixed that for you.

The manipulation of local economies by group warehousing of food, building permanent dwellings, and pooling resources to care for children or defend against organized raiders certainly predates writing. It's not clear when money entered the mix, but economies certainly exist in barter based societies. And there is every anthropological and psychological reason to believe that family, tribe, religious, and city leaders manipulated and taxed those economies long before we could develop subtleties like banking and stock markets.

Re:Traders

[Opportunist]

Very wise words. Why do we need growth? Because we have to pay interest, that's why.

Let's imagine a very tiny model economy. Me, as a producer. Some guy, as a consumer and a bank. Also, our economy contains the massive amount of 100 bucks, all of them held by the consumer.

I want to produce something and need 100 bucks for it. That guy I mentioned has 100 bucks, but there's nothing to buy for them because I do not have those 100 bucks to build what I'd want to build. So, lacking anything better to do with his money, he hands it to the bank, for interest. In a year or so, he'll get 104 bucks back from the bank, and since that's better than sitting on those 100 bucks with nothing good to do with them in the meantime, he'll just do that. The bank doesn't have 104 bucks. But it has found someone who is willing to pay back 110 bucks in a year. Me. So I get 100 bucks from the bank. Everyone thinks it's a great deal. The consumer gets 4 bucks interest, the bank pays 4 but gets 10 from me, netting 6 bucks and I finally get to build my item. And I plan to sell it for 110 bucks, which it's worth and which would also allow me to pay my loan back. So, let's do that!

I build my item and want to sell it for 110 bucks. Unfortunately, my customer only has 100, and those 100 are even locked down in the bank. But he really, really wants it. And 10 bucks in debt ain't that much, is it? So he asks his bank for his money back and 10 bucks as a loan. The bank grants it to him, albeit only in book money because it simply doesn't really have 110 bucks. But that's good enough, we all trust that bank to have the money, sometimes in the future. He gets 110 bucks which he gives to me for the item I built. I can pay my loan back and everyone's happy.

Well, not really. I, at least, get out at zero sum. My customer is owing the bank 110 bucks, getting 4 for his 100 bucks that he handed them means that after paying off everything he had, he still owes 6 bucks that he cannot pay back. Of course not. They did never exist. But he now owes the bank 6 bucks.

Sweet, being able to create money from thin air, ain't it?

And this is why we NEED growth. Without, we cannot pay back our loans. And it's either perpetual growth or the collapse of the economy. I.e. what we witness currently.

Re:Traders

[Opportunist]

What about it doesn't work? The core problem we're facing now is that there is no money in the pockets of the consumers and hence demand plummets. Look around you, supply ain't the problem. There's plenty of crap to sell, there's just nobody with the money to buy it. If we need something now, and fast, it's some way to give people money to buy stuff. And preferably this should be people who cannot what's commonly called "handle money", i.e. people who would spend quickly rather than save it.

We might even have missed the point where this could have solved the problem, with people so deeply in debt and nearing defaulting (if they haven't already reached it), where they can't simply take the money they get and spend it on anything but trying to pay off their debts. Which would not result in the economy revival intended since people wouldn't go and consume.

The idea is correct, though. We need more demand. Whether creating government jobs is still an option is another question, though.

Re:Traders

[fafaforza]

The whole TARP program is the epitome of this. Covering the asses of those that played with things they did not understand, but pretended they did. The US taxpayers were those people's rich parents who bailed them out.

Re:Traders

[Anonymous Coward]

And for similar reasons, I believe HFT should be outlawed.

Exactly. It serves NOONE (except HFT) that you can turn around your position within seconds. An SEC rule that requires any security to be held for at least 5-business-days would eliminate that whole issue completely. There would still be speculators, and liquidity in the markets... they'd just do a lot more research than simply looking at the last sale, 20-min moving average, deviation, and current spread.

Re:Traders

[sgt scrub]

By directly -- and heavily -- taxing companies that outsource labor and manufacturing.

Have you leaned nothing from the /. right wing? You need to sugar coat it. For example, if a company employs 10 US citizens they can receive a percentage of the amount of taxes and benefits as a write off and a write off for each employee. If the same company employs 10 US citizens and 10 over seas citizens or H1 workers the number of over seas/H1 workers is subtracted from the number they can write off therefor completely negating the write off. If the company only has 10 over seas/H1 workers they get taxed what they would have received as a write off. This way the goobermonkeys can say they are passing a bill cutting taxes for US employers.

Re:Traders

[Xyrus]

Why do hate America so much? Raising taxes is what Hitler would do. /snark

Re:Traders

[Anonymous Coward]

1. As long as a I can maintain my standard of living with cheap goods and services today (forget about tomorrow - I want it now!)

2. government is ineffiecient and equally short sighted (I am either just concerned about the next election or I am a useless/careless civil servant)

I don't see any solution - it is just the cycle of rise and fall of empires.

Re:Traders

[Lord_Jeremy]

I've seen that video. From what I've read from other sources it seems pretty poignant on some points but gets lost in the dramatic on others. Personally what I see as an issue with the economic ideologies that are thrown around presently is that every group thinks that their plan will be 100% effective. A completely free market has issues, a completely controlled economy and total communism has issues. It has always been my belief that there needs to be a balance. Eliminating greed would help too...

I'd have a long response but I'm not very awake at the moment/p.

Re:Traders

[Anonymous Coward]

If you heavily tax companies that outsource, they'll just move completely.

The software company I used to work for has been expading like crazy in other countries. Here in the US they've been static.

Re:Traders

[TheRaven64]

As the other poster said, it wouldn't work. Companies that outsource everything have no reason to stay in the USA except for their tax status or other regulation (e.g. News Corp - foreign corporations can't own more than a certain percentage of US news organisations). If this goes, then they'll just reincorporate somewhere else. Their US branch will be a wholly owned subsidiary, which will probably make a loss (on paper) every year.

If you want to bring the jobs back home, then you want real free trade, with a level playing field. If you import goods from a country that has a lower minimum wage, or laxer pollution laws, then you pay an import tariff equal to the difference in cost of producing them without the kind of worker and environmental protections that the customers consider to be essential.

Re:Traders

[TheLink]

Q: How can we bring jobs back home?
A: Convince the rest of the world to emulate you - in at least the disadvantageous bits... ;)

Tell them to "harmonize" their laws with yours otherwise you won't buy their stuff.

Re:Traders

[Anonymous Coward]

> So we keep getting bubbles due to over-investment, first in .com, then housing... what is the "next big thing" where we could invest to bring real growth?

Interesting statement. A complaint about over-investment in selected markets followed by a request for a new selected market. You nailed the problem. Everyone is sitting on their money waiting for the next bubble market instead of investing broadly across existing and future potential markets. All of which is rewarded by the special tax treatment bestowed upon capital gains. Tax all investment related capital gains at the same rate as other earned income and investors will shift to long term investments. CEOs will stop chasing the next quarter. Corporations will begin funding long-term R&D projects. And pretty soon America will start looking a lot more like the jobs engine it was before the Ronnie Raygun Revolution.

Re:Traders

[snowgirl]

Perhaps the link I posted earlier will do a better job of explaining what real capital wealth is, than I have. It's late and I am tired. But please note what this man says. He was exactly right about the economy (as Austrian economists have been for quite a while), while all the people following Keynesian theory (official government policy) were 100% wrong.

I will avoid copypasta here, and just link to relevant text. No theory is perfect, and this should be recognized by any valid scientist. You know, Creationists think that the Bible is exactly right about the world, and that evolutionary theory has been 100% wrong. It's easy to make such assertions when your entire theory is unfalsifiable.

And as he says: real capital wealth consists of goods and the infrastructure to produce those goods, not whether the stock index is at 12,000 or 13,000. The goods need not necessarily be physical: informational goods are still goods.

Please provide a rigorous and unambiguous way to differentiate stock trading based on speculation from informational goods. I agree that speculation is a bad thing, it's the entire reason for economic bubbles, and boom-and-bust cycles. But when me and my friends sat down and declared, "we can use magic to enforce rules and regulations, how would we stop trading based on speculation". We realized that there was no way to do it, even with omnipotent magic.

Re:Traders

[snowgirl]

Well, it isn't. Turning over stocks for short-term profit is gambling, not capitalism. Please look up the historic meaning of "capitalism". It isn't just money-mongering.

I like how you can assert out of thin air that current capitalism isn't valid capitalism, because of the "historic meaning of 'capitalism'." How am I supposed to even take this as a valid assertion? You seem to be arguing by vigorous handwaving, and appeal to authority. No less it completely ignores semantic drift, and that words are constantly expanded to apply to newer versions of older concepts.

Should electric powered cars not be "cars" because historically cars have had internal combustion engines? Thus your argument is invalid as demonsted by analogy.

Re:Traders

[snowgirl]

I won't argue that HFT is a disease, and speculation is the cause of nearly all economic misery in modern times, because it's demonstrable fact.

However, you can't just reject it as "capitalism" because it makes the word look bad. You have to have valid and rational reasons why one should reject speculation from the definition of capitalism. And honestly, you haven't made any rigorous argument to that effect. Just a bunch of hand waving, saying "THIS ISN'T CAPITALISM, CAPITALISM WORKS, SINCE THIS DIDN'T WORK, IT'S OBVIOUSLY NOT CAPITALISM!!!"

Re:Traders

[Anonymous Coward]

S&P is -2.9% over the last decade.

Re:Traders

[Hognoxious]

And that's the whole point. Economies are different. That's just the reality of the situation. And that's why we have exchange rates... to keep trading FAIR between countries.

It is? I always thought it was down to relative demand for different currencies, which in turn is down to the relative demand for goods & services made in their countries, with an adjust for psychological factors like stability and confidence..

But enlighten us. What person, committee or organisation decides what exchange rates are "fair", and how? When and where do they meet?

Re:Traders

[Wannabe Code Monkey]

Q: How can we bring jobs back home?

A: By directly -- and heavily -- taxing companies that outsource labor and manufacturing. This avoids the pitfalls of tariffs

Umm.. I'm pretty sure that will only hurt American companies that have some off shoring going on. It won't affect foreign companies at all. What's to stop any company from incorporating their former outsourced department as a completely separate company in another country and then simply buy the output of that new company? They're no longer outsourcing they're manufacturing, they're simply buying manufactured goods from another company, which in your example would not be subject to tariffs.

Re:Traders

[Hognoxious]

By directly -- and heavily -- taxing companies that outsource labor and manufacturing. This avoids the pitfalls of tariffs:

No, it is a tariff, in a laughably thin disguise.

So if foocorp brings in stuff from overseas you'd tax it, but if foocorp buys it from usedtobefoo (India) you wouldn't? What if it came from nouvobar (Brazil)?

You'd have to apply it to all imports, otherwise companies would outflank it by spinning off the overseas part as a separate entity, passing it thorough middlemen etc.

Seriously, having teh gubmint dictate procurement decisions to private businesses? Get real, you've more chance of getting state healthcare passed. In fact you'll have two communist muslim lesbians sharing the white house before either happens.

Re:Traders

[Hognoxious]

Next they'll be telling us that real work doesn't necessarily make your hands dirty.

Re:Traders

[Anonymous Coward]

Overall the S&P has returned almost exactly 0% over the last 10 years.

Yeah, but 10 years ago we were in a bubble, and today we're in a recession. Try looking at the last 15 years, or better yet, waiting 5 years and then looking at the last 20 years. You'll probably see a fairly regular return over those time periods.

S&P price at beginning of 1996 = 616.71 (~850 in 2011 $)
S&P price today = 1292.28

Average return = 2.8%

Not a great rate, historically, but again, we're still in a recession.

Re:Traders

[Hognoxious]

much of their job involves pouring over trade logs looking for anomalies

Pouring what? Coffee? 7 up?

Re:Traders

[Jane Q. Public]

"However, you can't just reject it as "capitalism" because it makes the word look bad."

I didn't. Nowhere did I state any such thing.

What I wrote was: it isn't "capitalism" so much as it is gambling, because it doesn't fit the historic meaning of the word "capitalism".

Capitalism relates to the means of production and how that is financed. If your company does not produce some kind of goods or services (just as I stated earlier), then "capitalism" doesn't apply.

Re:Traders

[retsofaj]

Alternatively why not change the parameters on capital gains taxes?

If the argument is that capital gains should be taxed at lower rate to attract investment and create jobs, then capital gains realized on companies that have outsourced should be taxed at a rate significantly higher than capital gains made on companies that have created jobs.

The devil is in the details (esp. wrt. productivity) but it does try to link the apparent goal of a low capital gains tax with the labour market outcomes of the capital investments.

Re:Traders

[Jane Q. Public]

"No theory is perfect, and this should be recognized by any valid scientist. You know, Creationists think that the Bible is exactly right about the world, and that evolutionary theory has been 100% wrong. It's easy to make such assertions when your entire theory is unfalsifiable."

I'm glad you brought science into this, because that is exactly where Austrian economics shines and Keynesian economics fails. Further, as far as I know nobody (except for you) has tried to make a credible claim that Austrian economics is essentially unfalsifiable. If you think that, then you really don't know much about it. A little history is in order.

In regard to that "main criticism" of Austrian economics, i.e. the assertion that it "lacks scientific rigor": if you know your history of economics, then you know that the Austrian school was around before Keynes, and in fact (you can find some of this in that same Wikipedia article), the present theory and mathematics of mainstream economics owe a great deal to Austrian economists, who first formulated much of it. The supposed "lack of rigor" of the Austrian school refers merely to their assertion that certain elements of the economy cannot be precisely calculated, because there are too many variables. There is nothing at all "unscientific" about that observation.

Any scientist worth his salt will tell you: a theory is only as good as its ability to predict. Keynesians like to point at their equations, charts and graphs (again, many of which were first invented by Austrians), and claim that therefore, their theory is "scientific". However, all the charts and graphs in the world do you no good if your theory can't predict anything. So let's look at some history, shall we?

Again, in that same Wikipedia article, you can see where Austrian-school economists publicly predicted the economic crash of 1929, while proto-Keynesian ("contemporary") economist Fisher, echoing the mainstream view, equally publicly (and quite famously) claimed that the economy was doing fine, right up until the actual crash.

More history: Austrian theory allows (and allowed) for the existence of high inflation and high unemployment at the same time, now commonly called "stagflation". Up until 1974 when it actually happened, Keynesian economics, with its foundations firmly rooted in A href="http://en.wikipedia.org/wiki/File:NAIRU-SR-and-LR.svg">the Phillips curve, not only did not predict stagflation, but denied that it could even exist. It was against fundamental principles of Keynesian theory, which had to be revised when it did in fact happen. I might add that mainstream Keynesians still rely on the Phillips curve, and and teach and preach about this inverse relationship, even though it was disproved by real life in the 70s and early 80s. That's not very "scientific".

I'm not claiming here that Austrian economists actually predicted that stagflation... but their theory at least allowed for it, while Keynesian theory did not... and still does not. The reaction of the Keynesians to stagflation was to scramble to find some kind of explanation for it, and they came up with "supply shocks", which supposedly render the Phillips curve ineffective. However, pretty much by definition it is impossible to predict when supply shocks will occur, and therefore when the Phillips curve will be valid or not. Which renders it -- a staple of Keynesian economic "theory" -- useless for predictions and therefore nothing but an unscientific abstract concept.

Austrians pointed out the existence of, and predicted the crash of, the 2001 bubble. Keynesians did not. In fact, yet again Keyensians were caught with their pants down, proclaiming that the market was doing great and to "Come on in! The water's fine!"

As I have already shown, Austrians (as represented by Peter Schiff) were also right about the 2008 econo

Re:Traders

[Jane Q. Public]

Sheesh. I meant to hit "preview" and hit publish instead. Pardon the typographical slips.

Re:Traders

[Jane Q. Public]

"Please provide a rigorous and unambiguous way to differentiate stock trading based on speculation from informational goods. I agree that speculation is a bad thing, it's the entire reason for economic bubbles, and boom-and-bust cycles. But when me and my friends sat down and declared, "we can use magic to enforce rules and regulations, how would we stop trading based on speculation". We realized that there was no way to do it, even with omnipotent magic."

I do not know of a way to do that. But what you can do is start using actual assets and production capacity as a measure of real wealth, rather than fictitious fiat dollars "created" by speculative trade, which have had this annoying tendency to disappear without warning.

Re:Traders

[Jane Q. Public]

"And there is every anthropological and psychological reason to believe that family, tribe, religious, and city leaders manipulated and taxed those economies long before we could develop subtleties like banking and stock markets."

I felt it was pretty obvious that what I was referring to was the specific government monetary policy we have experienced over the last (approximately) 80 to 100 years, not whether government control of one sort or another has ever existed. That is completely beside the point.

Re:Traders

[Jane Q. Public]

As I pointed out myself, the video is probably overly dramatic. But it is still factual.

I wasn't offering it as an example of any kind of solution to anything; I merely pointed to it as an explanation of how money gets "created".

Re:Traders

[Jane Q. Public]

I guess I should make myself more clear. I do not disagree that better regulation of the "financial industry" is called for. I happen to agree with that. My issue is with government manipulation of the money supply.

Re:Traders

[tqk]

Q: How can we bring jobs back home?

A: By directly -- and heavily -- taxing companies that outsource labor and manufacturing.

And the companies we are taxing will just move out of the country entirely. Since we won't be using tariffs, their prices will go back down, and now even more foreign businesses will have a competitive advantage over American businesses, and even more jobs will go overseas, ...

Interesting. Back in the day, libertarians wondered if it would be moral and ethical to invade the USSR to force the Reds out of power. You've just made an excellent argument in favour of that. Tyranny anywhere hurts everybody everywhere in one way or another eventually, so yes it is our duty to stamp it out everywhere.

The corollary is we should despise any people who allow themselves to be ruled by such regimes. "Grow a fscking backbone already, damnit!"

Where do I sign up? Libya or Syria or Burma/Myanmar I suppose. Anyone want to sponsor my transportation costs?

Re:Traders

[Jane Q. Public]

"I like how you can assert out of thin air that current capitalism isn't valid capitalism, because of the "historic meaning of 'capitalism'." How am I supposed to even take this as a valid assertion? You seem to be arguing by vigorous handwaving, and appeal to authority."

Not at all. As I mentioned in my other reply to you, I did no such thing.

It is YOU who are doing the handwaving and claiming "semantic drift" where it doesn't actually exist. Look up the current, modern definition of the word "capitalism", and you will find that it relates to the means of production of goods (and perhaps services) in a society.

Risky speculation in markets, then (especially things like money markets where no goods are involved), is not a part of "capitalism". We're talking apples and oranges here. It has nothing at all to do with capitalism, by definition. Current definition. I am aware that you seem to disagree with that definition, but if so don't go pointing at me, because I did not make it up.

Re:Traders

[Anonymous Coward]

Company X wants to hire people to manufacture widgets. It can hire unskilled Americans for $10 per hour, or Crotobaltoslavonian workers for $2 per hour. But here's the thing: In America, that $10 buys about 5 pounds of rice. In Crotobaltoslavonia, that $2 buys 10 pounds of rice. So in actual purchasing power, you're paying that Crotobaltoslavonian the equivalent of $20 per hour in America. Americans literally can't compete for that amount because it doesn't pay them enough to eat... in THEIR economy.

How can this be voted +5? I am pretty sure that $2 can only get the Crotobaltoslavonia 1 pounds of rice. You Americans are way overpaid and overfed. You can't compete because you ain't competitive.

Re:Traders

[Lord_Jeremy]

Isn't the current problem that the government doesn't control the money supply? As I recall the private banking sector (i.e. Federal Reserve) has been manipulating the value of US currency for almost 100 years.

Re:Traders

[Jane Q. Public]

Well, that's true. Technically the government does not control it. But as a practical matter, government and the Fed work hand-in-hand and coordinate the monetary policy.

For example, the recent "quantitative easing" required collaboration by the government and the Fed, as did the bailouts and recent government borrowing.

But you make a good point. In the first-ever audit of the Fed (which was restricted by law to investigating events surrounding TARP), the congressional committee concluded that the Fed had made $16 TRILLION in loans to U.S. banks and foreign borrowers over that same period, which they didn't tell anybody about, including the government. JUST during the last couple of years.

And those are newly-created, inflationary dollars that the US government did not even know had been inserted into the economy.

Re:Traders

[snowgirl]

Capitalism is an economic system in which the means of production are privately owned and operated for profit, usually in competitive markets.

Speculation and stock trading are services.

I could very well argue that a supermarket is not real capitalism, because they do not produce anything. Really, they just buy one product, and then sell it at a higher price to other people. Wow, sounds just like what stock traders do!

Re:Traders

[snowgirl]

Capitalism relates to the means of production and how that is financed. If your company does not produce some kind of goods or services (just as I stated earlier), then "capitalism" doesn't apply.

As noted above, then supermarkets are not capitalism.

Re:Traders

[snowgirl]

I'm glad you brought science into this, because that is exactly where Austrian economics shines and Keynesian economics fails. Further, as far as I know nobody (except for you) has tried to make a credible claim that Austrian economics is essentially unfalsifiable. If you think that, then you really don't know much about it. A little history is in order.

Apparently, you can't click links, so here:

For example, Nobel laureate Milton Friedman, after examining the history of business cycles in the US, concluded that "The Hayek-Mises explanation of the business cycle is contradicted by the evidence. It is, I believe, false."[106][107][108] Nobel laureate Paul Krugman has also argued that the Austrian business cycle theory implies that consumption would increase during downturns, and cannot explain the empirical observation that spending declines in all sectors of the economy during a recession.[109]

Oh, and as for falsifiability:

A related criticism[5][100] is applied to Austrian School leaders; these leaders have advocated a rejection of methods which involve directly using empirical data in the development of (falsifiable) theories; application of empirical data is fundamental to the scientific method.[101]

Re:Traders

[Anonymous Coward]

Or simply impose duties on imported goods.

Re:Traders

[Jane Q. Public]

"Apparently, you can't click links, so here:"

An interesting thing to say, since I referred to the same Wikipedia entry in my reply.

"Oh, and as for falsifiability:"

And I showed you verifiable ability to predict economic events. Apparently, you are willing to take the word of some "expert" over actual historical fact. Who is "appealing to authority" now?

Come back when you have real arguments that refute real facts, as opposed to quotes from scholars that contradict actual history.

Sheesh, again. I thought I was having a discussion with someone who actually understood the rules of intellectual argument.

Re:Traders

[Jane Q. Public]

Hint, by the way:

If you want to make any actual refutation, you are going to have to dig a little deeper than just Wikipedia.

"Appeal to authority", indeed.

Re:Traders

[Jane Q. Public]

This isn't even worth arguing with.

The stock market is trading. Capitalism has to do with the production of goods. You aren't even arguing about the same subject you think you are.

Have a nice day.

Re:Traders

[Anonymous Coward]

the problem is that they will simply pass the fee on to the US customers because we all know they afford it.

Re:Traders

[snowgirl]

This isn't even worth arguing with.

The stock market is trading. Capitalism has to do with the production of goods. You aren't even arguing about the same subject you think you are.

Have a nice day.

Perhaps because you can't read? I was saying that hypothetically if you're right, then supermarkets aren't capitalism, because they're trading not producing goods.

Re:Traders

[snowgirl]

"Apparently, you can't click links, so here:"

An interesting thing to say, since I referred to the same Wikipedia entry in my reply.

Oddly, I don't know how you could then make the claim: "Further, as far as I know nobody (except for you) has tried to make a credible claim that Austrian economics is essentially unfalsifiable." when the very article that you referenced yourself contained details and knowledge about someone other than me (and significantly more credible) making the claim that Austrian economics is unfalsifiable.

So, let's cut the bullshit right here. There are credible people (nobel laureates no less) who think that Austrian school lacks rigor, and is fundamentally unfalsifiable. These claims are not my own, and I am not the only one making them. Therefore your statement "Further, as far as I know nobody (except for you) has tried to make a credible claim that Austrian economics is essentially unfalsifiable." is fundamentally inconsistent with facts. But wait, if you knew this statement was false when you made it, then you willfully made a statement that was false with intent for it to be accepted as true. What's the word for that? Oh yeah, LYING.

Wait, I just though of a way that I could be wrong. Perhaps you just happened to not know anything about any of the criticism of Austrian school. That would keep your statement consistent with facts. Sure there are credible claims that Austrian school is unfalsifiable, and the people making those claims aren't me, but you were just unaware of them. Although, now that I think about it, this idea seems a bit foolish, since it requires assuming that you are willfully walking into confirmation bias by ignoring all criticism of your position.

So, crap, it may just be a lack of imagination here, but I only see two possibilities... either you're an idiot, and ignorant of all criticism of your pet economic model, or you're a dishonest liar.

Re:Traders

[snowgirl]

Hint, by the way:

If you want to make any actual refutation, you are going to have to dig a little deeper than just Wikipedia.

"Appeal to authority", indeed.

I don't have to dig deeper than wikipedia to refute your claim that _I_, snowgirl, am the only person that you know about, making a credible claim that Austrian school is unfalsifiable. I'm patently not alone in this assertion (especially considering that the assertion is not my own, but that of nobel laureates).

The details of your Austrian school economics are of little interest to me. It's the details in the simple process of debate that you keep getting wrong that interest me.

Re:Traders

[snowgirl]

Ludwig von Mises wrote of his theory, "its statements and propositions are not derived from experience... They are not subject to verification or falsification on the ground of experience and facts." F.A. Hayek wrote that any theories in the social sciences can "never be verified or falsified by reference to facts."

Shall you argue with your own demi-gods about if Austrian school is falsifiable or not? They appear to have taken a position opposite to your own. (That's fine, Darwin was wrong about a lot of things as well, and I would nitpick a few issues with him if I had the chance.) But none the less, as conceived the historic definition of "Austrian school economics" clearly was intended to eschew falsifiability...

Re:Traders

[Bob the Super Hamste]

Well how about we do the following to correct:
Everyone has the same access to markets (Sorry HFTs no more previewing orders)
If you make a mistake you own it (sorry you don't get to roll back that order you placed to sell at 10% of market price or buy at 10x market price)
Companies listed on exchanges must have open books
No we won't bail you out (sorry Detroit, Sorry investment banks, maybe you shouldn't have piss poor decisions) This will probably never happen because it would do a lot to level the playing field and would limit the amount of gaming going on so of course the big companies would be against it.

Re:Traders

[Bob the Super Hamste]

Well we could "harmonize" our trade policies with theirs. We could also "harmonize" our southern border policy with Mexico's southern border policy but we get condemned when there is even the thought of building a fence, or putting some troops there.

Re:Traders

[Bob the Super Hamste]

It also isn't "real" capitalism as HFTs get to preview orders before to the rest of the market. Real capitalism requires that all actors have the same access to the market and access to information. HFTs have a benefit as they get to preview orders and thus can make trades before the rest of the market even sees the orders. Also HFTs get the benefit from seeing orders in advance of know more info before the rest of the market since they get to look at orders (and see buying and selling prices) before the rest of the market. Here is an analogy for you: it is like getting to know the next set of winning numbers in the lottery and then getting to buy up all the winning tickets before anyone else can even buy a random ticket.

Re:Traders

[Bob the Super Hamste]

I wish it were more like gambling as they don't loose as they basically get to see future information (they get to see the buy and see prices) and act on it (add buy or sell orders in front of the new batch of orders) before the rest of the market gets to see the same info.

Re:Traders

[Bob the Super Hamste]

Yet it we were told it was a failure of the market. To me it looked like the market was trying to get rid of some bad actors.

Re:Traders

[tehcyder]

I didn't understand this, are they trying to say that programmers somehow sit next to the traders, and manipulate the computer in real time?

Re:Traders

[Registered Coward v2]

I meant to add: Q: How can we bring jobs back home? A: By directly -- and heavily -- taxing companies that outsource labor and manufacturing. This avoids the pitfalls of tariffs: we would not be blocking trade from coming into or going out of the country. However, we would be demanding compensation for our economy, from the companies that have been so harming it by their practices. That's about as fair as it gets. (It's also not a subsidy by the government... so it avoids those problems, too. Instead, it is a source of revenue TO the government.)

All that does is add a "tax" on every product made in the US. It would essentially raise the cost of goods produced in the US to an artificially high level since now US labor and production is "competitive" to out sourced labor and production. As a result, the cost of goods increases and less people are able to buy them. You don't magically create more wealth, you just inflate the costs of goods - less people buy and the workers you helped are out of a job again.

The key is productivity - if you can make more, even at a higher labor cost, it's worth producing it here. Trying to keep jobs by tax policy is counterproductive.

Re:Traders

[Jane Q. Public]

"Oddly, I don't know how you could then make the claim: "Further, as far as I know nobody (except for you) has tried to make a credible claim that Austrian economics is essentially unfalsifiable." when the very article that you referenced yourself contained details and knowledge about someone other than me (and significantly more credible) making the claim that Austrian economics is unfalsifiable."

The key word there is "credible".

I don't care if God made the claims, if they contradict verifiable facts. That's not credible.

Re:Traders

[Jane Q. Public]

"The details of your Austrian school economics are of little interest to me."

This is the most hypocritical statement you have made so far. If you have no interest, why are you expending so much effort in the process?

Further, I repeat: I gave you simple facts that you can verify yourself. If you don't understand the role of facts in a debate, as opposed to rhetoric, then it is not I who is getting it wrong.

Re:Traders

[Jane Q. Public]

This whole "conversation" started out with my assertion that speculative money trading was not "capitalism", because it had little or nothing to do with the production of goods.

You took issue with that, because your definition of "capitalism", apparently, was simply trade. Which is FALSE.

Supermarkets have everything to do with capitalism, because they deal in the trade of GOODS.

And I am done here. You have been doing so many contortionist verbal gymnastics to try to support unsupportable arguments, it is a complete waste of my time to be doing this.

Again: have a nice day.

Re:Traders

[snowgirl]

This whole "conversation" started out with my assertion that speculative money trading was not "capitalism", because it had little or nothing to do with the production of goods.

You took issue with that, because your definition of "capitalism", apparently, was simply trade. Which is FALSE.

No, that's not my definition of "capitalism", I have not laid out a definition of capitalism. I've simply objected to your definition being a good definition. Definitions are never ever EVER "false". There are however "good" and "bad" definitions.

Supermarkets have everything to do with capitalism, because they deal in the trade of GOODS.

So, the commodities market has everything to do with capitalism, because they deal in the trade of GOODS? But wait, there has been speculative trading even in the commodities markets as well...

And I am done here. You have been doing so many contortionist verbal gymnastics to try to support unsupportable arguments, it is a complete waste of my time to be doing this.

I'm not the one doing contortionist verbal gymnastics. In fact, it's the Austrian school that holds to the idea that models don't need to be mathematically modeled and empirically validated, but rather than simply through praxeology you can get to the truth, right?

Re:Traders

[snowgirl]

"Oddly, I don't know how you could then make the claim: "Further, as far as I know nobody (except for you) has tried to make a credible claim that Austrian economics is essentially unfalsifiable." when the very article that you referenced yourself contained details and knowledge about someone other than me (and significantly more credible) making the claim that Austrian economics is unfalsifiable."

The key word there is "credible".

I don't care if God made the claims, if they contradict verifiable facts. That's not credible.

Ah... so now we redefine "credible" to get out of the moral quandary of you being either a deceitful liar, or an ignorant moron.

Re:Traders

[snowgirl]

"The details of your Austrian school economics are of little interest to me."

This is the most hypocritical statement you have made so far. If you have no interest, why are you expending so much effort in the process?

Because I'm a pedant. We could be arguing over the ratio of sodium to chlorine in table salt for all I care. What I do care about is you playing semantic games wrong. If you can't be a pedantic bitch about your terminology and debate, then don't play semantic games.

You made a claim about what "capitalism" means. I thought the argument stupid. I've been arguing about your argument this entire time.

Further, I repeat: I gave you simple facts that you can verify yourself. If you don't understand the role of facts in a debate, as opposed to rhetoric, then it is not I who is getting it wrong.

The Austrian school has espoused the downfall and collapse of the economic market forever, and surprise, surprise, when it actually does fail, they claim credit for having seen it coming. If you're always predicting doom, then eventually you're going to be right, because at some point, everything comes to doom.

You want to debate facts and not engage in rhetoric? They why would you even get involved with the Austrian school. Their praxeology, around which the entire school was built, is all about rhetoric rather than mathematical, and scientific models...

Re:Traders

[snowgirl]

Further, I repeat: I gave you simple facts that you can verify yourself. If you don't understand the role of facts in a debate, as opposed to rhetoric, then it is not I who is getting it wrong.

Nostradamus predicted the promotion of a cardinal. (Or something to that effect.) See? That means Nostradamus's predictions are CORRECT and they're RIGHT. So we should believe all his predictions!

Or wait. Perhaps he just got lucky... and perhaps Austrian school just got lucky... you know, but always predicting doom...

Re:Traders

[snowgirl]

"Oddly, I don't know how you could then make the claim: "Further, as far as I know nobody (except for you) has tried to make a credible claim that Austrian economics is essentially unfalsifiable." when the very article that you referenced yourself contained details and knowledge about someone other than me (and significantly more credible) making the claim that Austrian economics is unfalsifiable."

The key word there is "credible".

I don't care if God made the claims, if they contradict verifiable facts. That's not credible.

But it's not just the Nobel laureates and I making this assertion either:

Ludwig von Mises wrote of his theory, "its statements and propositions are not derived from experience... They are not subject to verification or falsification on the ground of experience and facts." F.A. Hayek wrote that any theories in the social sciences can "never be verified or falsified by reference to facts."

So, are the very designers and foundation layers for Austrian School now not "credible" either? They seem to agree with me that it's unfalsifiable...

Re:Traders

[PoopCat]

Clearly the problem lies in trusting the bank. Remove that trust, and the customer doesn't accept the non-money the bank doesn't have, meaning you don't get paid, and the bank really has to find 4 real actual dollars from somewhere. So the bank, being unable now to meet its obligations, goes bankrupt (or recently, gets bailed out - but that's another story). Net effect? no more bank, and thus direct trade between you and the customer, able to negotiate prices without any worries about non-existent money.

Tell me again why this is a bad thing?

Re:Traders

[Opportunist]

Because my customer is not going to pay me 100 bucks for the promise that I'll at some point in the future actually deliver a product. Oddly, there's enough trust in banks, though, to hand them 100 bucks for the promise of getting 4 more bucks in a year.

There's also the occasion when you don't yet know what you want to buy. Wouldn't it make sense for you to put the money somewhere where you're going to get more money back in the meantime?

Also, I as a producer need more money than I can put on the table myself. Now, I could go and search for investors. Or I can let someone else do that job. And here's where a bank comes in. Or an investment broker, which is pretty much the same deal.

I don't think we should get rid of banks. That's not going to work out either. But we need to take control of the economy back. Right now, the economy is in the grasp of investment bankers and fonds managers, and even more so in the hands of rating agencies, who can over night cause a company and even countries to go bankrupt. Seriously, if Moody's decided to send the US into bankruptcy, they could. Instantly. At will. Just drop the rating and the US is bankrupt. Even easier with any other country.

And that's way more power than any private organization should have.

Re:Traders

[Jane Q. Public]

"Ah... so now we redefine "credible" to get out of the moral quandary of you being either a deceitful liar, or an ignorant moron."

Look, fool. One last time: "appeal to authority" does not trump verifiable facts.

You tried to pull the "appeal to authority" thing on ME before, yet don't recognize when you are doing the same.

I really shouldn't be answering you at all, because as I stated before, this is obviously a waste of time.

Re:Traders

[Jane Q. Public]

"But it's not just the Nobel laureates and I making this assertion either: Ludwig von Mises..."

Well, first off, it isn't "his" theory, although he has added to it and promoted it.

But secondly: do you even know what falsifiability is?

von Mises aside, Keynesians like to claim that Austrian theory is "unfalsifiable", even when their own theory was actually "falsified"! (The stagflation issue, for just one example.)

Logically, the claim of "unfalsifiable" is false on its face, because any theory (where it it physically possible, unlike String "Theory" until recently, and certain other issues in physics) can be falsified by simply showing that it has no ability to predict.

And I have already given you historical, easily verifiable instances where it did not just predict future events, but explained exactly how ans why they would come about.

So you can talk about unfalsifiability all you want, but it won't wash, because exactly as I have stated several times already, it contradicts known and easily verifiable facts.

If you want to refute that, then you will have to find some way to refute the facts. Offering opinions -- by ANYBODY -- simply doesn't hold water in comparison.

Re:Traders

[Jane Q. Public]

"I've simply objected to your definition being a good definition."

Well, it has only been pretty much exactly the same definition for about 300 years. You go ahead and argue with it, if it makes you feel better.

Re:Traders

[snowgirl]

Ok, you just made my brain pop from your stupid... I'm leaving with my sanity intact...

Re:Traders

[snowgirl]

"I've simply objected to your definition being a good definition."

Well, it has only been pretty much exactly the same definition for about 300 years. You go ahead and argue with it, if it makes you feel better.

Funny that your definition is no longer in use by pretty much anyone, except in strict academic speech registers...

Re:Traders

[snowgirl]

"Ah... so now we redefine "credible" to get out of the moral quandary of you being either a deceitful liar, or an ignorant moron."

Look, fool. One last time: "appeal to authority" does not trump verifiable facts.

You tried to pull the "appeal to authority" thing on ME before, yet don't recognize when you are doing the same.

I really shouldn't be answering you at all, because as I stated before, this is obviously a waste of time.

It is not appeal to authority to pull up text of another person's objections to a theory when you claim that I am the only person making such an objection.

If you say that I am the only person who says that the sun is purple, but I pull out an account from George Hisphaphaphaph and note that he says that the sun is purple, then obviously I am not the only person claiming that the sun is purple, and thus your claim that I am the only person making such a claim is false.

I don't get how you can't understand this. I'm not using these other sources as proof that Austrian Economics is unfalsifiable, I'm using these other sources as proof that I'm not the only one making that claim. Namely, their assertions could be as false as my example above that the sun is purple, because the only reason I'm quoting what they've said is to show that someone else said what I said. In such a case, quoting them shows first-order evidence that my claims are being made by others.

None of this specific line was about if Austrian Economics was falsifiable or not... rather the entire debate shifted to be exclusively about if I were the only person making such a claim. Showing you that other people are making this claim is thus a perfectly acceptable presentation of evidence, and not an appeal to authority.

You are REALLY bad at debate and logic, you know... you're so hyped up about defending your pet theory, that you've completely lost track of what I'm actually arguing about...

Re:Traders

[Jane Q. Public]

No, CREDIBLE objection. You keep forgetting that part.

Re:Traders

[Jane Q. Public]

Why does your brain have to "pop" before you can grasp very simple -- and long acknowledged -- rules of evidence? This is what is a mystery to me.

If something is testable, then it is NOT unfalsifiable. This is simple logic.

If a theory's ability to predict is tested -- and it turns out that it does in fact work -- then it is not unfalsifiable. Why are you having such trouble with this simple concept? You have been giving me peoples' OPINIONS, and I have been giving you verifiable facts. YOU weigh these things, and have decided that the opinions are worth more than demonstrated fact.

And yet * I * am supposed to be the stupid or weird one here? I'm not the one breaking all rules of logic, you are.

Re:Traders

[Jane Q. Public]

Then why does it say that in all the dictionaries and encyclopedias? Far be it from me to "appeal to authority" in the manner you have kept doing, but in a case like this "authority" is the only way to settle the matter. And the authorities disagree with you.

Re:Traders

[Jane Q. Public]

What's funny though is I really did expect you to look it up and call me on that 300 year thing, because it's not true. It was bait.

But 150 years is good enough for me.

Re:Traders

[snowgirl]

Then why does it say that in all the dictionaries and encyclopedias? Far be it from me to "appeal to authority" in the manner you have kept doing, but in a case like this "authority" is the only way to settle the matter. And the authorities disagree with you.

Oxford English Dictionary: "an economic and political system in which a country's trade and industry are controlled by private owners for profit, rather than by the state."

So, a country's trade... like I don't know... the stock market trade.

Re:Traders

[snowgirl]

No, CREDIBLE objection. You keep forgetting that part.

The sound of the goalposts scraping over the ground as you redefine "credible" is already jarring enough. I suspect that no matter what criticism I present to you, you're just going to declare it to be non-credible, and if I ask for "why?" You'll give me an answer, and once I produce criticism conforming to that "why", you will then just call it again non-credible.

Your theory is based on praxeology, and dialectics not rigorously defined quantitative models. You may as well be trying to sell me on homeopathy...

Re:Traders

[snowgirl]

Why does your brain have to "pop" before you can grasp very simple -- and long acknowledged -- rules of evidence? This is what is a mystery to me.

If something is testable, then it is NOT unfalsifiable. This is simple logic.

If a theory's ability to predict is tested -- and it turns out that it does in fact work -- then it is not unfalsifiable. Why are you having such trouble with this simple concept? You have been giving me peoples' OPINIONS, and I have been giving you verifiable facts. YOU weigh these things, and have decided that the opinions are worth more than demonstrated fact.

And yet * I * am supposed to be the stupid or weird one here? I'm not the one breaking all rules of logic, you are.

Nostradamus's prophesies are testable. Creationism is testable. However, both are not FALSIFIABLE, because even if you prove one part of them wrong, apologists can concoct a new set of rules that explain why the test failed.

"Creationism says the Earth is 6,000~10,000 years old!" We show that the world is significantly older, by many orders of magnitude. Response? "It was just designed to look that way!"

Because Austrian school doesn't use quantitative and mathematically models, we cannot actually lay out what the proper claims are and prove the theory itself false beyond all doubt. All of the methodology behind praxeology is to present things with dialectics and words and argument, rather than true rigorous models.

Let me ask you this. You declare that Austrian school is falsifiable. In order to be falsifiable, there must be evidence that would make you reject the theory and seek an alternative. What evidence can you purpose that would completely falsify Austrian school and make you reject it. (Note, the evidence need not exist, it must merely be possible. For instance, evolution could be shown to be false if fossil records showed no species-level variation of fossils, as well as by a randomized distribution of all fossils in the ground, rather than being layered according to the age of the medium. Neither piece of evidence actually exists, but if it were to exist, then it would disprove evolution. More drastically, we could falsify the existence of time in physics if everything stood still. Completely fanciful, but it is a logically possible condition.)

I await your answer to what kind of evidence could possibly disprove Austrian school.

Re:Traders

[Jane Q. Public]

"Because Austrian school doesn't use quantitative and mathematically models..."

See, there you go. Appealing to authority again. How do I know? Because this isn't an accurate statement about Austrian economics.

If you go back and look at that Wikipedia entry again (I'm not going to bother finding other references), you will see that in fact A LOT of the theory and math that is used in modern economics was invented by Austrians. To say that their model is not based on sound theory and math is complete bullshit... again, as the historical FACTS show.

This exaggeration stems from the fact -- I am compelled to repeat myself here -- that Austrian economics rejects certain mathematical "relationships" claimed by other theories such as that of Keynes. And the reason they do so is that they assert that the mathematical models are too simplistic; there are actually far more variables involved. So they simply could not work in the idealistic manner claimed by those economists.

I already gave you an actual example of exactly that: the Phillips curve, and how it was blown out of the water in the 70s.

But to say that Austrian economics is not based on math, or any solid principles that can be nailed down, is simply in contradiction of the facts.

And I also gave you ACTUAL examples of falsifiability. In fact the ultimate example: the ability to predict. You keep failing to grasp that.

If I could SHOW YOU that homeopathy actually worked (I won't try because we both know it doesn't, but hypothetically), then you would be irrational to deny it.

Yet you have been given actual examples of Austrian theory demonstrably working, and yet you continue. I am really amazed at this.

Re:Traders

[Jane Q. Public]

To be even more specific: you were given examples of Austrian theory demonstrably working, even when the pet theories of the people making these accusations against it equally demonstrably did not work.

And yet you deny real world examples and continue with the rhetoric instead. For what reason?

Re:Traders

[Jane Q. Public]

"Trade" with no qualifiers is a very vague term. Are they referring to any trade, in an abstract sense, or trade for tangible goods?

Wikipedia:

"Capitalism is an economic system in which the means of production are privately owned and operated for profit, usually in competitive markets."

Production. "Capital" goods. That's what "capitalism" is all about.

Re:Traders

[snowgirl]

Of course it's vague, that's how people use it, and you know this. Otherwise, you wouldn't have been harping about the "historical meaning of 'capitalism'". Is there anything you can actually debate that doesn't involve mangling semantics?

Re:Traders

[snowgirl]

I'm not talking about proving that homeopathy works. You can't PROVE that a model works, because all models are flawed. What you can do is show how a model is falsifiable, which is by presenting some form of evidence that YOU as a proponent would need in order to stop believing in that theory.

I don't give a shit if you think that I would dismiss it as irrational, I want to know what evidence could possibly exist that would prove to you that Austrian school is false.

Re:Traders

[snowgirl]

Austrian economists reject empirical statistical methods, natural experiments and constructed experiments as tools applicable to economics, saying that while it is appropriate in the natural sciences where factors can be isolated in laboratory conditions, the actions of human beings are too complex for this "numerical" treatment as passive non-adaptive subjects. Instead one should isolate the logical processes of human action. Von Mises called this discipline "praxeology" – a term he adapted from Alfred Espinas (but which had been in use by others).[44]

The Austrian praxeological method is based on the heavy use of logical deduction from what they assert to be undeniable, self-evident axioms or irrefutable facts about human existence. The primary axiom from which Austrian economists deduce further certain conclusions is the action axiom, which holds that humans take conscious action toward chosen goals.[45] Austrian economists focus on goal-directed action and say that it is undeniable because in order to deny action, one would have to employ action in the act of denial.

Methodology is the one area where Austrian economists differ most significantly from other schools of economic thought. Mainstream schools such as the neoclassical economists, the Chicago school of economics, the Keynesians and New Keynesians, adopt empirical, mathematical and statistical methods, and focus on induction to construct and test theories—while Austrian economists reject this approach in favor of deduction and logically deduced inferences. According to Austrian economists, deduction is preferred, since if performed correctly, it leads to certain conclusions and inferences that must be true if the underlying assumptions are accurate. However Austrian economist Robert Murphy has stated that those using Austrian theories can still err in their interpretations of history, even if based on a theory formulated by deduction.[46] Caplan makes a similar point about quantitative significance, explaining that a theory, such as one which logically relates minimum wage and unemployment, tells nothing of the approximate quantity of change in unemployment one can expect upon minimum wage increases.

Either you're not practicing Austrian economics or I am right that they reject experimentation, rigorous models, and quantitative methodologies.

Re:Traders

[snowgirl]

To be even more specific: you were given examples of Austrian theory demonstrably working, even when the pet theories of the people making these accusations against it equally demonstrably did not work.

And yet you deny real world examples and continue with the rhetoric instead. For what reason?

Because I'm not arguing that point, I'm arguing about if it is FALSIFIABLE. That means, fuck all evidence that it's "true", because it only matters how it could possibly be false. And you're dodging this entire question. What would prove Austrian school false?!

Re:Traders

[Jane Q. Public]

But you ARE arguing that point, by definition. If it was shown to work, then it is shown to be not false.

Only things that can be shown to be false or not false are falsifiable. Either one will do.

Re:Traders

[snowgirl]

But you ARE arguing that point, by definition. If it was shown to work, then it is shown to be not false.

Only things that can be shown to be false or not false are falsifiable. Either one will do.

Are you just being willfully ignorant, or are you just completely unaware of what "falsifiable" means? Let me break it down for you. ("false" + "ify") + "able". It means "able to be proven false." This does not mean that showing that it works means it's falsifiable. Asserting that "it has been shown to work, therefore it is valid" is actually falling prey to confirmation bias. You need to be searching for that which would invalidate your theory, and prove it false, otherwise you simply continue to assert constantly, "it's true, it's true, it's true".

I can demonstrate quite easily how creationism can be shown to work. This does not make it falsifiable, because the precise way that it is shown to work is that it can literally explain any and all evidence to support itself. Thus, it is not falsifiable.

Showing that your theory predicts an event correctly is actually in fact worthless... nearly every theory works at first. Caloric theory provides useful and valid predictions that work (under specific conditions). The problem is that there are experiments that can be done to show that they do not work in all conditions, and thus the theory is actually wrong, despite producing some models that work.

What experiment can you think of that you could run that given a certain specific and clear result, would prove that Austrian school is false. The experiment doesn't have to actually produce the given specific result, it just has to be a possible outcome. Note, that such experiments can only produce definitive results in the negative. There is no way to yield a definitive positive result for a scientific theory. Either a theory stands falsified, or it stands yet-to-be-falsified. We have no theory at all that is known and proven to be true, and I certainly hope you are not asserting that Austrian economics has been proven true, because it's the biggest red flag of pseudoscience out there.

Re:Traders

[Anonymous Coward]

If Goldman Sachs has their way, the next "big thing" to over-invest in will be carbon tax credits.

Re:Linux by default?

[NFN_NLN]

Not surprising. Wall street is basically a hot bed of criminals, so it's not surprising to see them using an operating system that is built from stolen Microsoft technology.

That's a twist ending even M. Night Shyamalan would be proud of...

Re:Linux by default?

[Anonymous Coward]

nice!!

Stress

[MichaelSmith]

I used to co-locate in the same building as the local stock exchange. One day, very late, I took the elevator down to the car park, which was where the computers were. There was a guy in the elevator who looked absolutely wrecked. He was sweaty, shaky and not taking things in. He got off at my level and stumbled off towards a porsche which appeared to be similarly young and in equally bad condition.

The thing is, I work in air traffic control, where the stakes are even higher. The difference is that the operational people have an absolutely obsessive approach to managing their workforce. Traffic controllers are just not allowed to get upset or stressed. In many environments they have unlimited sick leave.

The difference, I suppose, is that traders personally stand to take home a lot of money. You could do this in any field: offer ridiculous compensation for ridiculous effort. But if you work it out, I doubt the long term returns justify what this does to people.

Re:Stress

[Anonymous Coward]

Unlimited sick leave? Wow, imagine that :) /europe

Re:Stress

[hedwards]

The difference is that air traffic controllers aren't generally vicious psychopaths willing to let the country go to hell so that they can make a quick buck.

Honestly, anybody that goes to work for the psychos on Wall Street deserves all the abuse they get, and more. It's not like HFT is anything other than a gigantic scam.

Re:Stress

[GameboyRMH]

The difference, I suppose, is that traders personally stand to take home a lot of money. You could do this in any field: offer ridiculous compensation for ridiculous effort. But if you work it out, I doubt the long term returns justify what this does to people.

If I had the opportunity I'd work for a short period for ridiculous pay - even a year or two of half-million dollar salary is equal to decades' worth of a mere mortal's pay - then I'd quit and do something I like (with short hours) for a sane salary. That's understandable. Working your ass off year after year to rack up ungodly amounts of money, not so much.

Re:Stress

[tolkienfan]

Utter nonsense.
HFT is not a scam. It's not the cause of the recession (depression - whatever). It makes a small profit doing what people would otherwise do, only faster.
HFT is also inevitable. As soon as the exchanges were automated and networked some kind of algorithmic trading appeared. It caused a reduction in the bid ask spread and an increase in liquidity. These things are considered by experts to be beneficial.
When I say "small" profit - it's a small amount compared to the size of the actual trades that take place. Often a small fraction of a cent per share, for example. Much lower than your broker charges.

Re:Linux by default?

[earls]

lol, perfect.

I am an HFT programmer

[Anonymous Coward]

I'll be posting anonymously, but I think many here have a very poor understanding of what we do. Most of that is because we do tend to be a very secretive group, but if you were to sit down with some of us, you would see that we really do very normal (and useful!) things in the market.

I work on the algo and core infrastructure. I wrote price feeds that take 1/5th of a microsecond in C++ and (a little slower) in Java. I understand in fine detail how cache and the the PCI-e bus works. I have a very good understanding of algorithms and the constant-time tradeoffs. I know when to make something simple, when to use and avoid threads, and I can debug in minutes and push out a new version in the seconds before market open (not many people can handle that level of stress well). I read the C++ and Hotspot assembly, and know how to program for superscaler architectures specifically. If you really need me to, I can even crank out some VHDL code.

On top of that, I understand market microstructure and derivative pricing. I work 12 hours a day on average and do 100 hour weeks. I am on call during Asian hours and need to come in sometimes on holidays when other markets are going nuts and we need to plan.

I also hope to make $500,000 this year.

You always hear about Google programmers being the best in the industry, but I've been to a couple Google interviews and turned them down both times because the engineering quality just isn't there. I'd put the average HFT programmer up against the best in Google anyday.

Ask away, and I'll answer to the best of my ability.

Re:I am an HFT programmer

[Anonymous Coward]

Clever troll, bragging about how amazing you are and how awful everyone else is, under the guise of being helpful. I'm sure you'll get some bites.

Re:I am an HFT programmer

[MichaelSmith]

I can debug in minutes and push out a new version in the seconds

You must be taking a huge gamble by doing that. I don't see how your new version could be considered safe to use if it is deployed that fast.

Re:I am an HFT programmer

[demonlapin]

I also hope to make $500,000 this year

That's all? Demand more, if your skills are what you say. You should be pulling in $2M/yr minimum.

Re:I am an HFT programmer

[Anonymous Coward]

how do you feel about having all that talent and technical ability, yet produce absolutely nothing of value to society and instead spend your time allowing psychopaths to beat other psychopaths by fractions of a second, all to the detriment of everyone else?

Re:I am an HFT programmer

[Anonymous Coward]

the way you describe it, i bet you have very little of a social life.

I would take googles job over yours any day. even if the engineering quality is lower.

Re:I am an HFT programmer

[Anonymous Coward]

Could you illustrate using a typical use case where running a program in 1/5th of microsecond instead of 1/2 can be beneficial and how? Also what mathematical background is necessary to deal with the algorithms? Any reading you suggest to prepare a general programmer to start HFT programming (Assume basic C/C++ and machine knowledge.) Thanks.

Re:I am an HFT programmer

[PmanAce]

I can debug in minutes and push out a new version in the seconds before market open

No peer reviews and/or staging deployments before going to prod? This sounds fishy...

Re:I am an HFT programmer

[Anonymous Coward]

I work 12 hours a day on average and do 100 hour weeks.

I hope you get time to enjoy your $500, 000 because there are only 168 hours in a week.

Re:I am an HFT programmer

[PmanAce]

Hmm, I went and read some stuff about writing in one of these languages (functional languages like Smalltalk and Haskell). This means that you can prove the correctness of the code, presumably automatically, so presumably very quickly...no need for peer reviews and test deployments. Interesting.

Re:I am an HFT programmer

[NFN_NLN]

...we really do very normal (and useful!) things in the market...

You can't tell me that a company changed direction in 1/5th of a microsecond. You're not allocating capital investment to companies based on actual merit, you're skimming money from other investors.

John Templeton and Warren Buffet didn't get rich through micro-transactions. They allocated capital to companies, allowed them time to prosper and *if* they did, then they were rewarded.

You've convinced me of your technical skills but not that "[you] really do very normal (and useful!) things in the market".

Re:I am an HFT programmer

[Anonymous Coward]

Uh... no it doesn't. I've heard this before from such people, and it's not like the company is going to be out of the market for days(or even a few hours) simply to do some testing. This is why things occasionally *do* fail... but you've got people on staff to fix it and get it running immediately should it happen.
Yes, mistakes happen, but

Re:I am an HFT programmer

[ub3r n3u7r4l1st]

In some jobs you earn twice as much, but end up looking twice as old and feeling twice as old....

Re:I am an HFT programmer

[Anonymous Coward]

Just doing it faster than the other guy.

If somebody find an arbitrage opportunity in an hour or sees a press release and submits a trade an hour later, that used to be considered too quick and the telegraph and phone where seen as unfair.

If an hour is fine, why not a minute, or a second? There is just no real dividing line. If we push the markets together a millisecond faster, there isn't much social worth, but it is that competition for that tiny timeslice that gives what we do benefit. That competition makes sure oil is priced the same (minus friction) all over the world in all different currencies, for example. If you made it impossible to profitably do that, prices would be incoherent and often off the mark (e.g., ETFs rely on that arb opportunity to stay pegged to their true value). I don't see how you can keep that benefit without allowing this speed arms race.

For math:

For 70% of what you do, algebra is about it in equties and futures. Once you venture a little past that into options or scalping strategies for example, calculus comes in a little. It is kind of a little embarassing, but even the pure quants spend most of their time in algrabra and calc.

Re:I am an HFT programmer

[Anonymous Coward]

Another HFT programmer here. I once had to make a run-time modification to an algorithm to keep about $100 million from going at a lower price than what the traders wanted. Sometimes market conditions change so fast that the traders demand the ability to make rapid adjustments to the algorithm. They're willing to take the risk. They can't wait for the safe development cycle.

Re:I am an HFT programmer

[russotto]

I can debug in minutes and push out a new version in the seconds before market open

No peer reviews and/or staging deployments before going to prod? This sounds fishy...

*sigh*. These young ones, with their peer reviews and staging deployments. HFT programmers evidently hark back to a less bureaucratized age, when you could be a hero by hot-patching the running image in the production system. Wouldn't mind doing it myself, as my temperament runs more to cowboy than code review, but 100 hours a week is too much, even for $500K.

Re:I am an HFT programmer

[pauljlucas]

This means that you can prove the correctness of the code, presumably automatically, so presumably very quickly...no need for peer reviews and test deployments.

If that's true, then why wasn't it apparently done in the first place to prevent the bug that he supposedly fixed?

Re:I am an HFT programmer

[Anonymous Coward]

Beware the advice of rich people. They hate competition despite all their public professions of love for the "free market". They like the market to be anything but free. With their advice, there's usually some crucial information missing or it's a dramatic oversimplification. Outright lies that if followed are designed to eliminate the competition. Like that Google psychopath talking about doing dumb things. Yeah, *YOU* go ahead and do dumb things while he sends out lobbyists and has people over a barrel with his cock up as many asses as possible.

Re:I am an HFT programmer

[Anonymous Coward]

$500,000 divide by 52 weeks = $9,615 / week
$9,615 divide by 100 hours = $96.15

Re:I am an HFT programmer

[Jayson]

To some extend yes, but the order management and routing systems have checks and there are circuit breakers to prevent total trainwrecks. It is a cost/benefit thing. How likely am I to lose $10k immediatey verus make $50k on the market close? One very important skill is in being able to estimate how likely you think you are right.

I wouldn't do this for huge programs or where the lost can be gigantic or I couldn't evaluate the risk.

Trying to cut corners, I did lose $1.6 million one day because I had a bug in my code. A lot of people have these stories when working on some high-risk projects (something HFT place usually try to stay away from). I learned from my mistake and more than made up for it a week later.

Re:I am an HFT programmer

[Antique Geekmeister]

I can tell you that the investors who bought the stock _first_ after the company's announcement of better-than-expected annual earnings was published, and started to push up the value of the stock, might indeed make a lot of money. Being the stock trader that handled the trades for them is also a reliable source of revenue. Being able to sell your stock clients a slight advantage in profits, one that you can measure, can easily bring in an unreasonable amount of extra business.

The business is very strange, and I'm afraid vulnerable to changes by the SEC in how such transactions are allowed. It's inherently unstable: the feedback loops are nearly impossible to trace because they're hidden behind the concealed trading algorithms of numerous companies, and it's very easy for a set of very modest delays to feed back and cause a massive positive feedback loop, unless the system is very heavily damped. Right now, it's not very damped, and this sets the possibility of "high frequency trading" become "high frequency oscillation" until it slams against the floor of the stock market assets.

Re:I am an HFT programmer

[bugs2squash]

Where do you get these price feeds from (or route the prices to) ? surely you could save 5e-7s just by using shorter cables or putting the mic and speakers closer to the traders.

Re:I am an HFT programmer

[Anonymous Coward]

Perhaps not. I was interviewed for an HFT position and given a programming challenge and when I was asked "quote is this ready for production" I said no, as a software engineer I would want to test it first", I was, actually, yes truly, yelled at. During the interview!

Bizarre world.

Re:I am an HFT programmer

[Jayson]

FYI, I'm not the original poster, but another HFT developer. That is the only >$1mm loss I've ever had, but have had smaller, and it wasn't exactly for what some consider HFT, but for a trend prediction system. It was for an overseas illiquid market across a couple instruments and we had a bug elsewhere that didn't count our position properly.

Re:I am an HFT programmer

[Anonymous Coward]

$9,615 / (40 hours + (60 hours overtime * 1.5)) = $73.96 hourly wage

Re:I am an HFT programmer

[Anonymous Coward]

Ex-banker here, but not an ex-developer. What he describes is BAU for a quanitative trading desk. There are endless audit kerfuffles around the fact these guys often have simultaneously dev, uat, prod segment access, run modeling software off the dev seg without a corresponding prod seg version, admin access to their prod seg PCs, no four eye deploument rules, blah blah blah. The business management then ignores the audit issue because of the profits involved.

Re:I am an HFT programmer

[MichaelSmith]

It sounds a bit like:

• If I change nothing I am out of business
• If I make a mistake I am out of business
• If I change the code and it works then I am in business

So you change the code and take the risk.

Re:I am an HFT programmer

[Anonymous Coward]

It depends on the situation. Opportunity cost matters. Releasing code with limited (or zero) review or testing is of course risky and might end up losing money, but not releasing isn't risk-free either.

You wouldn't want to release code like this all the time, but it can occasionally be the lesser of two evils.

Re:I am an HFT programmer

[Anonymous Coward]

You have it right. I'm still relative young, but we do tend to be cowboy coders, and this is why HFT places tend to go after very good programmer and problem solvers. We also need to estimate how likely we are to being correct and how likely we are to take a huge hit and determine if it is worth the risk.

Also there are other parts of the system to catch erroneous orders or stop the algo when position limits are reached or it goes aggro spewing out orders.

Re:I am an HFT programmer

[Anonymous Coward]

No peer reviews and/or staging deployments before going to prod? This sounds fishy...

I work on the infrastructure side of high frequency trading. Basically it means squeezing out micro and nanoseconds from everything below the app.

There are *countless* times that we've discovered problems in the morning and barely fixed them by 9:30. There's simply no time for second guessing. Peer review consists of an IM session and a conference bridge and everyone doing a once over. You need to be confident, but not overly so. And most of all, you need to be able to type fast with 10 people over your shoulder.

We'd be out of business if we had to wait for people to approve everything we did.

Re:I am an HFT programmer

[Anonymous Coward]

Got any more work? I'll work for $150k and you can go on vacation :)

Re:I am an HFT programmer

[Anonymous Coward]

Indeed, it must be very stressful working 12 hours a day, 8.3 days a week...

Re:I am an HFT programmer

[pyite]

Where do you get these price feeds from (or route the prices to) ? surely you could save 5e-7s just by using shorter cables or putting the mic and speakers closer to the traders.

Nearly all market data is transmitted from the exchanges via IP multicast. Typically you will have servers in each exchange to trade on that exchange, but you also will have links pulling in market data from every other relevant venue as well.

See, for example, Spread Networks who made a lot of money by digging a really straight trench from Chicago to New Jersey in order to get CME data to the NJ metro area as fast as possible.

Re:I am an HFT programmer

[Mr. Freeman]

"They can't wait for the safe development cycle."

This, from the people in charge of trading amazingly large amounts of money in a market which influences the global economy in a big way. (remember that billion vs. million mix-up awhile back that caused some pretty big problems until it was fixed)

You call it risky, I call it reckless. You try what you're doing in any other field and you'd be fired pretty damn quick.

That said, I'm not so much angry at you as I am at the people who ask you to do this.

Re:I am an HFT programmer

[Antisyzygy]

You may call yourself skilled but I call you lucky. Too many in finance mistake luck for skill. You cannot possibly understand all of the problems and assess all the risks that is the market (equities, futures, forex, whatever) without a lifetime devoted to study of advanced probability and game theory. But then again, Im just a mathematician and we tend to be anal about what it means to understand and be sure of things.

Re:I am an HFT programmer

[Anonymous Coward]

Incorrect in an important way --- while it's certainly possible to write code that's provably correct
--- by writing it in a theorem prover then extracting the computation component to a functional
language --- it's extremely difficult and time-consuming to do so... And the specifications even
  for simplest functions are quite complex in their own right, and may need double-checking, too.

(Also, since when is Smalltalk functional?)

Re:I am an HFT programmer

[Anonymous Coward]

Do these algorithms actually compute marginal use or economic value?

Re:I am an HFT programmer

[sjames]

If you mess up, some pampered asshole takes a temporary financial setback and you're fired.

If the guy who designs nuclear reactor control systems messes up, an entire state becomes uninhabitable for decades and countless people die.

When Apollo 11 was executing the moon landing, a warning light for the computer came on. Some guy had 30 seconds to make a go or abort decision with millions of dollars and two lives on the line and the entire world watching.

Surgeons regularly have people's lives in their hands. They often have only seconds to alter a once well planned surgery to keep the patient from dying.

A janitor has lousy hours, poor pay and little thanks. There's no pressure there, but his day's work makes things just a little bit more pleasant for others so at least he makes a positive contribution.

We would all be just a little better off if high speed trading was made a felony.

Re:I am an HFT programmer

[Anonymous Coward]

Yet another HFT programmer here. Let me address your concern about the risk to the market. HFT systems have tons of controls which govern our ability to trade. We have what we call 'doors', which act like circuit breakers. When a door is closed, it prevents any further trading. Doors can be triggered by any number of conditions (abnormally high P&L, too much exposure, etc.), and there are strict compliance policies governing how and when they can be reopened. We also have throttle-style controls ("speed bumps") which effectively limit how many trades we can make in a given time window. These controls can be configured per security, per name, per exchange, or globally. They are extremely robust, and we maintain them meticulously.

We take these matters very, very seriously. We are at the mercy of the exchanges: if we screw up and don't contain the damage, they will revoke our ability to trade, effectively shutting us down.

Re:I am an HFT programmer

[Bing Tsher E]

Warren Buffet didn't get rich through micro-transactions.

Warren Buffet gets rich by loudly proclaiming that estate taxes should be increased. Then he buys distressed companies out from under the family of the founder when they prove impossible to pass on to the next generation.

And he buys companies distressed in other ways to 'fix' them or to chop them up and sell the parts. Truly an American Hero, worthy of the adulation he gets as the poster boy 'good capitalist' by the left.

Re:I am an HFT programmer

[Anonymous Coward]

Rarely.

There are some classes of algos that rely on you computing some different scenarios ahead of time (and then maybe at runtime, refining them - and sometimes "soft" defintions like "10% of the spread between the current price and the moving average"), but for the most part they are more in the line of arbitrage or technical trading (scalping, looking at trend lines), or exploiting statistical relationships.

Re:I am an HFT programmer

[n8r0n]

I have a question. How can a guy make $500k and obviously think the world of himself while being under the impression that 12 hours x 7 days equals 100 hrs/wk?

Sorry in advance for the sarcasm, but slashdotters do know what you do. If you paid attention, there have been lots of HFT threads since 2008. The question is, "do you understand what you do"? Namely, destabilize markets, manipulate prices for your clients, participate in systemic frontrunning, and divert talent away from solving engineering problems that actually matter to humanity, for the sake of enriching yourself off one of the biggest examples of what game theorists call a "Prisoner's Dilemma" that I can imagine. Your use of the word "useful" to describe your work leads me to think not.

Re:I am an HFT programmer

[Anonymous Coward]

Interesting, because I'm the one in the other anon post discussing possibly working on a masters thesis examining whether HFT algorithms actually compute economic value and the possibilities of the limitation of computing based on price. This was partly inspired by a blog post by Timothy B. Lee regarding the idea that prices may in fact destroy information in markets, which would explain why we see weird successes that die off quickly (think fads, pet rocks, etc).

Re:I am an HFT programmer

[Anonymous Coward]

You always hear about Google programmers being the best in the industry, but I've been to a couple Google interviews and turned them down both times because the engineering quality just isn't there. I'd put the average HFT programmer up against the best in Google anyday.

Arrogant much? I'll take the bait. Aside from pure self-centeredness, what makes you judge HFT traders as so elite compared to Googlers, other than you apparently not liking a Google interviewer?

- Ability to write efficient and correct algorithms under pressure? You mean exactly the thing that programming contests like TopCoder measure? The same TopCoder that is just dominated by Googlers at the upper level?
- Ability to understand complicated math and algorithms? You meant the kind of thing that the thousands of Google engineers with PhDs would be experts in?
- Ability to willingly do 100 hours per week? Fine, you can have that distinction.
- Ability to do the development part of engineering, rather than just the day-by-day hacking? Apparently not, since you never once mention that as a relevant skill.

Yeah, I'm a Google engineer. And I have a strong enough math background that multiple HFT firms tried to recruit me. I chose Google. And your post just makes me very glad that I did. I neither want to be you nor work with you.

Re:I am an HFT programmer

[Anonymous Coward]

Not HFT here, but I'm a finance quant. Testing is a critical part of banks' infrastructure--for the obvious reasons but also regulatory reasons. For example you'll find derivative pricing models subject to high levels of scrutiny. The standards may be lower at hedge funds and other unregulated (in the legal sense) entities.

But you have to be able to fix stuff in the middle of the day. Sometimes things break, or market are going crazy, or maybe the trade you thought you booked wasn't the actual trade you signed. And you need to push out a fix ASAP because you might be losing money, or otherwise causing huge amounts of risk.

Quants are hired to be able to do this kind of coding AND NOT SCREW UP. A lot of the work we do (not all, but a lot) could be done by Joe Programmer if it weren't for the requirement that you can't screw up. Trading is a dangerous occupation and that's why they try to hire the best people.

Re:I am an HFT programmer

[Anonymous Coward]

(Different AC here)

You can't tell me that a company changed direction in 1/5th of a microsecond.

I can. You're not taking into account that every trade in a stock is a piece of information about that stock--it contains information of what other investors make of the stock. Just because there's not an earnings report every day doesn't mean there isn't new information every day.

Anyway, if you know some math, you can actually check for yourself whether stocks exhibit unnatural volatility. If they had unnatural volatility, you would see mean reversion since, loosely speaking, that excess volatility must disappear after a period of time. Incidentally, if you find mean reversion, you can trade it and make money--perhaps start a hedge fund.

You can see for yourself, but I'll tell you the answer is that stocks don't have much mean reversion in them anymore. There definitely was in the 1980's, before computers and mathematicians came on the scene. Now with all the quants trading against the mean reversion, there is much less of it left in the market. So basically all these HFT have actually REMOVED the unnatural volatility caused by trading.

Re:I am an HFT programmer

[bertok]

Just because you're good at what you do doesn't mean that what you do is good.

Microseconds? Why not nanoseconds? How about femtoseconds? Why wouldn't that make just as much sense?

What could possibly change in the underlying value of a corporation made up of flesh and blood humans and capital with decades of depreciation in a fucking microsecond? Here's a hint: nothing. You are not investing, or trading, but simply racing other gamblers. Investing doesn't benefit from microsecond response times, and trading doesn't need it either. People could buy IPO shares just fine over the phone. Nobody ever needed a microsecond response time to buy a thousand bushels of wheat, and never will, because bread is baked daily, not a million times a fucking second.

If politicians had two braincells to rub together, they'd enact a law to prevent trades faster than some tick, say, an hour. Your 'trading' company would go out of business in a week, and nobody would care. Farmers would still sell their wheat, and bakers would still buy it, but without you leeches skimming off the top.

Re:I am an HFT programmer

[jampola]

I concur. Stuff like this makes me angry. It's just stupidly reckless. Reading OP's essay, it sounds quite glamorous in a geeky sense but seriously, I don't think any amount of money would woo me into that kind of recklessness with any kind of responsibility attached to to. In regards to what happened a few years back, if I recall, that was some guy who accidentally entered an extra 2 0's assuming that he needed to include the decimal at the end. Not completely his fault, since I would've thought the system he was using had some kind of system where it spits out the amount in words? Or maybe not since my trading platform allows for one click trades.

Re:I am an HFT programmer

[MAXOMENOS]

I know when to make something simple, when to use and avoid threads, and I can debug in minutes and push out a new version in the seconds before market open (not many people can handle that level of stress well).

This has me wondering how well you QA your software. I'm understandably concerned about this; your code might be handling my investments.

Re:I am an HFT programmer

[Anonymous Coward]

You reek of the arrogance that's caused crashes in the past. "We're the best at everything and we know what we're doing.... oh shit help us Uncle Sam!"

Re:I am an HFT programmer

[Anonymous Coward]

Yet another HFT programmer here. Let me address your concern about the risk to the market. HFT systems have tons of controls which govern our ability to trade. We have what we call 'doors', which act like circuit breakers. When a door is closed, it prevents any further trading. Doors can be triggered by any number of conditions (abnormally high P&L, too much exposure, etc.), and there are strict compliance policies governing how and when they can be reopened. We also have throttle-style controls ("speed bumps") which effectively limit how many trades we can make in a given time window. These controls can be configured per security, per name, per exchange, or globally. They are extremely robust, and we maintain them meticulously.

Would that be before or after May 6, 2010? Seems from this side of the desk that having all the robots simultaneously withdraw their bids didn't work out so well.

We take these matters very, very seriously. We are at the mercy of the exchanges: if we screw up and don't contain the damage, they will revoke our ability to trade, effectively shutting us down.

Hmm. Same question, slightly-different context. Didja take it that seriously before 5/6/10 or was that something that wasn't supposed to be possible?

No hard feelings either way, I'll concede that it was a blip over the long term. I'm just curious.

Re:I am an HFT programmer

[zenyu]

I am a New Yorker so I've considered going the banking route a few times. It sounds like there are interesting problems to be solved and I enjoy constructing efficient code. But what always dissuades me is the talk of 12 hour days. Are there enough jobs north of 200k where you put in normal work hours to make it worthwhile to switch specialties?

I for one wouldn't mind an engineering job at Google. But the pickings in NYC appear to be pretty slim and I'm getting tired of startups & getting bumped "up" to management when what I really enjoy is writing code.

Re:I am an HFT programmer

[Anonymous Coward]

But you have to be able to fix stuff in the middle of the day. Sometimes things break, or market are going crazy, or maybe the trade you thought you booked wasn't the actual trade you signed. And you need to push out a fix ASAP because you might be losing money, or otherwise causing huge amounts of risk.

...Quants are hired to be able to do this kind of coding AND NOT SCREW UP.

Everyone makes mistakes. If you had proper quality assurance you could find and correct those mistakes before they roll to production. Then you wouldn't be in the situation of rushing a fix out in the middle of the day.

Re:I am an HFT programmer

[Anonymous Coward]

I work for Google. I made $300,000 last year (much to my surprise...my salary+bonus+stock compensation tripled in four years), and so far I'm on track to do better this year. My typical week is 50-60 hours, and I come home happy most days. I'm proud of the work I do not just for its technical merit but also because I know (many) millions of people enjoy using my product every day in their personal and professional lives. I have a loving, funny, talented, and gorgeous wife who I spend hours with every day.

Let's assume for a moment you are a much better programmer than I am, your coworkers are much better programmers than mine, and you'll actually get the $500,000 you're hoping for this year. Here's my question: why would you choose your life over mine?

Re:I am an HFT programmer

[sxpert]

could you also explain how this stuff serves the economy ?

Re:I am an HFT programmer

[Anonymous Coward]

Definitely. 12 hours is just because I love what I do. I'd say 10 is probably the industry norm. And the 100 hour weeks are not the usual for me, just when things need to get done. Those are the front office (actually working on the trade desk often and developing the algos and often parts of the infrastructure).

Back office is more your 9-to-5 and deals with all the support systems. The closer you get to the money, the more you make so front office pays more than back office. And you work for bonus, not salary, so you might get a $130k/yr salary and $35k semi-anually in bonus pay.

Re:I am an HFT programmer

[rdnetto]

What can you tell us about the trading algorithms? I understand that it's mostly trade secrets, but if you could even say what broad area it's in, I'd appreciate it.

Re:I am an HFT programmer

[Anonymous Coward]

$500,000 divide by 52 weeks = $9,615 / week

you missed this:

I also hope to make $500,000 this year.

Re:I am an HFT programmer

[Anonymous Coward]

In what way is that a normal or useful thing to do in the market? How does your work benefit society?

Re:I am an HFT programmer

[Anonymous Coward]

If you really need me to, I can even crank out some VHDL code.

Huh? Where does VHDL come into this? Why not the other HDL language? What compiler do you use here?

Re:I am an HFT programmer

[Anonymous Coward]

Ah, finally somebody with knowledge.

Re:I am an HFT programmer

[roman_mir]

What do you mean "under the impression that 12 hours x 7 days equals 100hrs/wk?"

It is much more likely that it's actually 9.5hours x 5.5 days.... equals 100hrs/wk on the invoice.

Re:I am an HFT programmer

[Lanteran]

It doesn't- for perspective, I recently read an article about a new trans-atlantic cable being lain that would decrease latency to something in the low double digits (ms). I thought "this is amazing, think of the use it could have in game servers!". Then, I read that it was a private cable designed to shave a few milliseconds off trading times overseas. Just the shock that reached me when taking that in was... amazing. Such a waste of resources to the detriment of society.

Re:I am an HFT programmer

[Anonymous Coward]

Dang!!! $500K. I'm on the wrong side of the financial sector. I'm a NYSE guy. Programing all kinds of fancy stuff, put more hours than you and only make $130K with bonus.

Re:I am an HFT programmer

[dragonturtle69]

Completely unrelated in experience to HFT, but, many times I have deployed solo.

• There is a major problem. Money/department respect is being lost by the minute.
• Mgmt. requires an immediate solution.
• One person gets to be an instant hero, or instant zero.

Yes, stressful it is, supporting non-technical people.

Re:I am an HFT programmer

[Anonymous Coward]

And all of that ability and knowledge is wasted on gambling instead of actually PRODUCING any added value that could later gets sold on the market. Very disheartening.

Re:I am an HFT programmer

[Anonymous Coward]

It doesn't. It's a cost.

Re:I am an HFT programmer

[aliquis]

The flash crash wasn't a "oh I accidently sold billion of stocks when I wanted to sell millions"-accident.

If they screw up they screw up for themselves. No normal person need to sell at the lower price or whatever. Don't use stop-loss and do your own judgement and you're rather free?

Re:I am an HFT programmer

[Roger Lindsjo]

On top of that, I understand market microstructure and derivative pricing. I work 12 hours a day on average and do 100 hour weeks.

Is that mathematics specific to microstructure and derivative pricing? For the rest of us 100 hour weeks would be 14 hours a day average.

Re:I am an HFT programmer

[Anonymous Coward]

I heard this great explanation about high frequency trading, why it won't cause oscilation.

High frequency traders make profit in a volatile (lots of movement in prices) market.
If high frequency traders can cause volatility you will get a perpetual profit machine.

Trading is not magic, it is founded in math, like motion a perpetual profit machine cannot exist.

Market Making a subset of high frequency trading I like to compare with a second hand clothes store (or second hand music store). A market maker buys any kind of good clothes from anyone, and then sells those clothes on to someone else. He makes money on the difference between the buy price and the sell price. He tries to keep his stock as small as possible because he doesn't want to pay rent for storage space and doesn't want to have to clothes go out of style, and he wants to have a high turnover so he buys and sells for pretty good prices.

Other HFT firms are just slightly more aggressive, if we extend the analogy of the second hand clothes store. This trader will go house to house asking people for how much they want to sell their clothes or buy his clothes, and when the price is right he immediatly trades with them.

Normally buyers and sellers will have it difficult to find each other and the spread between a published buy and sell price is wide and traders are few. Market Makers because they want to trade as often as possible make a good price for both buyers and sellers and can directly trade.

They need speed to lower the risk of trading at bad price, and to compete with other HFT firms. They are actually not in competition with institutional investors or the retail investors at all. As a retail investor you can either trade directly with a market maker at a good price, or if you have more time you can put in a buy or sell order and wait until a HFT firm trades against your set price; in both cases you will trade quicker and on the price you wanted.

Even in an online game like eve online there are market makers. In the big markets like Jita there are people who just put up buy and sell orders for the same product hoping to trade a lot of them. Anyone who wants to buy or sell in volume (or expensive items) goes to these big markets to sell directly because the price is good. Without these market makers the people who buy and sell in volume have to put up buy and sell orders and wait until someone hopefully trades against them. In fact in 0.0 security traders are asked to make markets there because people want a market where you can buy and sell things immediately.

Re:I am an HFT programmer

[aliquis]

Are there any time when you get "a preview" of one order someone will put into the market where you can for instance pick up all that's available for that price and then sell for that order?

A.k.a. stealing?

Re:I am an HFT programmer

[aliquis]

Bug as in program crash or bug as in "now it did what I told it to but I would had preferred if it had done something else"?

Re:I am an HFT programmer

[Alex Belits]

I am most likely better than you at each and every aspect of software (and HDL) development you have mentioned. Except, of course, "debugging in minutes" -- that kind of irresponsibility would get me fired. I also have to work long hours, and have to have clear understanding of complex issues unrelated to software.

Except I do embedded software and FPGA development for professional audio equipment. Each device I worked on, each firmware release, each line of code, does something useful for many, many people. Some of those people don't even know that audio equipment, leave alone software, is involved with what they are hearing. Large fraction of my work ends up being free/open source, too -- platform, drivers, etc.

I also don't have any problems with posting here under my real name. Or with telling you, and people like you, to die in a fire.

Re:I am an HFT programmer

[Anonymous Coward]

"Ask away, and I'll answer to the best of my ability."

I'm a financial programmer myself, having been part of a few hedge funds. But I'm not on the HFT timescale. Here's a question I always wanted to know:

Is there a large variety of strategies? On the longer timescales that I run, it's unlimited how creative you want to be. Well, not completely, but having a look at your timescales, there's hardly any time to do anything? What is the idea? Simple venue arb? ETF basket arb? Getting first on the touch price when it moves? What exactly does your system do? I figure the more time is important, the more common the strategy (otherwise you'd never have competition for a trade).

Also, I always assumed in HFT the programmer WAS the trader. Why would you be taking orders from traders? Surely they can't react much, other than to decide how much risk to take on your system? The run I run my shop, we programmers are looking at the output, positions, PnL all day.

Hope you find time to answer this for me. Posting anon due to moderating.

Re:I am an HFT programmer

[Anonymous Coward]

I make more per hour for much less effort and can take 3 weeks off multiple times a year and actually TRAVEL to Africa, Asia, Europe, or throw the family into a motorhome and DRIVE to the Grand Canyon for a 2 week vacation.

No internet, no smartphone. I'm "off the grid".

Lifestyle matters. A-LOT!

Re:I am an HFT programmer

[Anonymous Coward]

Why does his work have to benefit society?

Does a plumber's work benefit society? No, it benefits the guy who can now flush his toilet.
What about a programmer at any other firm? It benefits the people who buy it.

Re:I am an HFT programmer

[Anonymous Coward]

I'm not HFT but work pretty close to that crowd and I've been doing so for 20 years. Trading is, always has been and always will be a blood sport filled with psychopaths and outsized egos Trading systems are best done by people who: 1 - Have an absolute knowledge of what they are doing (tech and business) 2 - Have the confidence to do it under the gun repeatedly 3 - Understand the cost (personal live,etc) of doing it but are willing to do it anyway 4 - Have the balls the deal with psycopaths that are traders and developers like themselves. It's a kind of artistry and craftsmanship that few people ever get to see and they are worth every penny. As other posters have mentioned - the devs that can't handle HFT go get jobs at Google or whatever the hot tech company of the days is - do 9-5 and become Dilbert managers. Having skills, education and aptitude is just the entrance fee. You have to be good enough to be among the best AND willing to bet your job every trade. And after a year or so - it's not about the money. Like HFT guys, I could take the money I've made in my little niche and go sit on a beach. But that would last about 20 minutes. I'd blow the lot trying to get back on the trading floor. Hey - don't you guys get why Steve Jobs is Steve Jobs?

Re:I am an HFT programmer

[dbIII]

I think the best description of the HFT field for this audience is "man in the middle attack".

After all, usually the idea is to work out what people are going to buy, get there first and onsell it to them.

Re:I am an HFT programmer

[Anonymous Coward]

Just because you're good at what you do doesn't mean that what you do is good.

Microseconds? Why not nanoseconds? How about femtoseconds? Why wouldn't that make just as much sense?

Because the signal at maximum travels 300nm in a femtosecond. This limits the speed at which CPUs can operate reliably.
In a microsecond there is plenty of time to do operations. A 1GHz system can execute 1000 CPU cycles in a microsecond.

It should be obvious that the times spoken of here has nothing to do with why but is only limited to what is possible.

Re:I am an HFT programmer

[JAlexoi]

Wow! That is low.... An SAP consultant will make €800 per strictly 8hour day with half that brain and less stress and travelling expenses(where appropriate). If that person can squeeze out more hours he can handle 2 contracts at the same time for 80 hour week...
That is calculated: €208000 for a 40 hour week and €416000 for an 80 hour week. In addition to that his expenses will be minimal, since he'll be staying in a hotel paid by his client!

Re:I am an HFT programmer

[JAlexoi]

If a janitor falls ill - the office knows about it immediately, if the manager falls ill - the office does not care...

Re:I am an HFT programmer

[Anonymous Coward]

Approximately the same hourly wage as a union crane operator with a GED makes. You deserve more.

Re:I am an HFT programmer

[Anonymous Coward]

what goods or services does your company provide to customers? What kinds of customers do you have? what goods or services does your personal efforts provide to your business and customers?

Re:I am an HFT programmer

[rtaylor]

[quote]If you had proper quality assurance you could find and correct those mistakes before they roll to production.[/quote]

That's impossible. Often it is interaction with other automated tools at different companies that cause the loss of funds.

If they tweaked their algorithm over the weekend specifically to play on a weakness in yours; then there is no QA that can be done on your side to know that.

Losing money is also defined as not making as much as possible. If you were making $1000/minute on Monday doing something and only $100/minute on Friday (many bots only run for the first 5 to 10 minutes of the market); then that is also defined as a loss of $900/minute.

Another part may have adjusted their algorithm to be a more aggressive market maker and is taking all the trades.

Re:I am an HFT programmer

[adamofgreyskull]

For some more perspective, I recently heard about a man buying a Porsche Cayenne. I thought "this is amazing, think of the use it could have as a taxi!". Then, I read that it was a private car designed to shave a few minutes off of his daily commute. Just the shock that reached me when taking that in was...amazing. Such a waste of resources to the detriment of society.

The rights and wrongs of HFT aside....what an assinine and idiotic argument to try and prove that point! Seriously, are you 9 years old? "They add nothing to society! Apart from that time they spent millions and created work for hundreds of people by paying for a new trans-atlantic cable." You were hoping perhaps that they would pay for a new trans-atlantic cable, then let a few million WoWtards and Counterstrikers loose on it free of charge? Out of the goodness of their hearts? (Contact me off-list, I have a few bridges I can sell you.) Or is your objection that they have the money to spend on shaving milliseconds off their trade times?

This profitable (for them) arms race to get the lowest trade time has the capacity to drive technological innovation. That is a gain for society, no matter how incidental. And of course, whether that is enough to offset the negative effects of what they do is another thing but it's a wee bit disingenuous to say that they are mere leeches.

Re:I am an HFT programmer

[Anonymous Coward]

Whose does? Yours? How many peoples' work is of "benefit to society"? Most people aren't working on the cure for cancer/aids/common cold, or striving to invent a cheaper MRI machine or engineering an uncrashable car, they're doing work that benefits a very small subset of society. Just as these traders are doing. What they do benefits the people whose money they are working with. You say they're evil because they're exploiting technology to beat their competitors to try and make as much money as they can. Ok. So a baker who buys a Bake-Matic 9000 so he can bake 10x the number of loaves his competitors can is evil? He hasn't benefitted society either, probably not even his customers, unless he drops his prices. He's exploited his ability to buy a faster machine. He's exploited his newfound advantage over his competitors to edge them out. Ah...no, but he is benefitting society really, because he's a baker, and he doesn't earn as much as you so baking bread is "noble". Not like those flash-harry HFT programmers...the bastards.

Re:I am an HFT programmer

[Anonymous Coward]

There is kind of a core set of HFT algo-types, the meat and potatoes of the industry, and then there is the crazy stuff people try. There is no illegal front-running, we don't get to see orders before anybody else (there is 1 exception to this), unless you could just being closer to the exchange.

The basics you will find at most HFT shops are:

- Basic arbitrage. When people think of the "stock market" they often think of it as a single entity, but it is really a collection of exchanges around the US (and the world beyond domestic stocks). These can be as simple as buying GOOG on the New York Stock Exchange and selling on the NASDAQ. You see a bid for $450.00 on NASDAQ, so you put out a bid for $400.01 on NYSE. Somebody sells to you on NYSE and you immediately sell back to the NASDAQ customer. Pretty simple. However, these do get more complicated as you increase the number of "legs" or types of instruments being traded. For example you an buy gold (GLD) as an ETF on in the stock markets, and sell gold futures on the CME futures markets in Chicago. Or sell it in London. Or you can buy something in Canada prices in Canadian dollars, and immediate sell Canadian dollars in the foreign exchange market and sell the same company on the NYSE (this would be three legs), and they get much more complicated.

- Statistical Arb - These are similar to arb, but are defined by statistical relationships instead of the simple algebraic definitions of basic arb. Two instruments may be highly correlated, so you buy and sell them as in the same way as arb, but you have a model of how this relationship works. Usually these stat arb plays are much more complex than basic arb in terms of number of legs (not always constant) and they are a little more resiliant to speed issues or missing your exit. Stat arb used to run a lot slower than HFT and be different, but the trend has been to making them like more into HFT programs.

- Earnings / News event trading - While arbs just take prices and kind of mirror them onto another exchange, increasing market depth at each individual exchange and decreasing volatility, these are algos that cause pricing changes. There are specialized services that will release earnings data electronically and you can subsribe to them and have your system setup agead of time (or APPL beating earnings by $0.02 / share, then do blah blah), and if you are the first one in, you stand to make a lot of money. The same goes for the employment situation report, GDP numbers, etc. If you've ever seen a chart where it looks like the market went straight down in less than a second from a economic statistic release, this is most liley HFT programs.

- Rebate trades - When you put an order on the market and it doesn't match anybody, then sits on the display screen, this is called adding liquidity. When you send an order and it matches something already on the screen, this is called removing liquidity. Exchanges have an interest in having a lot of quotes on the board, so they are willing to tilt things in your favor if you post an unmatched quote. This makes sense since those trying to buy or sell will go to where they can see volume. To do this, many exchanges will charge you $0.003 to remove liquidity, but pay you $0.002 to add it. This gives them a $0.001 profit, but i a case where the remover incentivizes those that post. There are programs that will try to capture this rebate. You can't make a losing trade and have this make it a winner, but you can scratch (buy and sell at the same price) and make a slim profit. As you can imagine, this greatly increases market depth and reduces volatility. The can also turn over some amazing volume stats when well tuned.

- Iceberg hunters. There aren't a lot of these out there, but they seem to get a lot of attention. When an instituional investor wants to dump a lot of shares, they will give it to this program that tries to trade through out the day in small quantities so as to not be noticed. These HFT strategies try to pick up on that pattern (I find these

Re:I am an HFT programmer

[Anonymous Coward]

I'm a software guy who eventually became an HFT trader. HFT does produce something of value, because there's almost always liquidity in the market because of HFT. I'm not just "skimming profits", I am actually taking a risk so real investors can get in and out. Regular people transfer their risk onto me and I manage it properly so they can sleep at night and stay in cash, or get into positions and manage their long term portfolios.

Keep in mind, HFT profit per trade is less than a penny at the highest frequencies; however, mom and pop can cash out their 401ks cheaply and with minimal spreads. Several years ago they were paying fractions of a dollar. Now, they can pay less than a penny spread and go straight to cash almost instantaneously with almost no overhead except to their broker.

Re:I am an HFT programmer

[pauljlucas]

To me, "correctness of the code" means it's doing what you thought it should be doing. If not, then it's a bug. Bug does not have to equal crash.

Re:I am an HFT programmer

[guanxi]

Your opinion of your self sounds like many people from Wall Street. Why does the culture promote such arrogance? Don't people see the consequences of it?

Re:I am an HFT programmer

[Xyrus]

And with that, you get a giant red "Fuck The World" button placed directly in your hands. One little mistake and you can set off a massive cascade reaction that wipes out billions to trillions of dollars in the span of 15 minutes.

I'm so happy sociopathic assholes are willing to take the risk of wiping out OUR savings and impacting OUR economy in return for making a buck for themselves. But you're just following orders, right?

Re:I am an HFT programmer

[Antique Geekmeister]

I see why you're posting this as an anonymous corward, because it's typical HFT nonsense. You've not actually addressed the underlying issue, such as occurred on http://www.bloomberg.com/news/2010-05-06/electronic-trading-to-blame-for-stock-market-plunge-nyse-s-leibowitz-says.html, where a single significant stock event triggered automatic sales, which devalued a stock, sent other HFT automatic systems into a spiral of selling, with ripple effects spread throughout the HFT systems triggering massive flurries in unrelated companies and markets.

The existence of other types of trading can provide some stabilization against such feedback spirals, but most certainly does not prevent it. Your metaphor is also misleading because it involves second hand clothes, a commodity that most people can live without if the market evaporated tomorrow. A better metaphor would be mortgages, evaluation based partial ownership of actual goods where the trade among them can itself be profitable.

And we all know how safe from ridiculous collapses poorly monitored, high speed and high volume trading was in _that_ market, don't we? The opportunity for "due diligence" simply does not exist in HFT. It's been poorly managed arbitrage, and the feedback cycles are too fast and too dynamic to successfully model. I'm surprised we haven't had repeats of that collapse last year: companies have been _exiting_ HFT because of it.

Re:I am an HFT programmer

[Anonymous Coward]

Tell me where they pay programmers time and a half for overtime?

Re:I am an HFT programmer

[Man Eating Duck]

If politicians had two braincells to rub together, they'd enact a law to prevent trades faster than some tick, say, an hour. Your 'trading' company would go out of business in a week, and nobody would care. Farmers would still sell their wheat, and bakers would still buy it, but without you leeches skimming off the top.

I read an excellent article a while ago that, in summary, described how HFT basically is just ripping of regular companies in regular business (you know, the people who actually create the value in the first place). He made a compelling argument for how imposing even a mandatory 60 second acceptance period for any trade would allow a human to intervene and reject a trade which is obviously a very bad deal for the seller (and thus a very good deal for the scum of the earth that are high frequency traders). He also provided several specific examples of how this could have saved traditional "honest" companies substantial losses which, while they didn't break those companies outright, at least made a serious dent in their bottom line.

According to him HFT gets it profit from exploiting margins in trades which no sane trader would accept if they had the time to respond to it, such as issuing and cancelling offers tens of times in less than a second to nail the price limit which a trader has put on a buy request.

Obviously I don't know much about the subject, and sadly I can't find that specific article now (turns out there is a vast amount of articles written from both perspectives), but if anyone can point me to it I'd be grateful.

Re:I am an HFT programmer

[Anonymous Coward]

No, you're wrong. Every time a trade is made in the stock, that's a new piece of information: other people are revealing what they think of the company. That is just as important as a company's earnings release. Furthermore, new information about the economy as a whole emerges every day, which also affects everybody's stock price.

Re:I am an HFT programmer

[Anonymous Coward]

I sleep good at night. I have a lovely wife. I travel frequently. I do interesting work. I am learning new technology all the time. I get a paid a lot of money to do what I love to do. And I have investments that are set up so that I can retire at 45.

Tell me you have all those things.

Re:I am an HFT programmer

[Kopion]

I'll be honest, this post smells like a whole bunch of BS. I feel like this guy just googled "HFT job requirements" and put them in sentence form in a very trifling manner. The fact that he said "not many people can handle that level of stress well" tells me he just read some article about HFT and posted it. The fact is that millions of people have equally (if not more) stressful jobs than HFT programmers... Doctors, Special Forces, Lawyers, CEOs, Fledging Start-Up founders. Also, if you're some hotshot HFT programmer, why are you spending your Saturday night in NYC posting a very broad overview of a general HFT programmer's responsibilities... pretty sure if this guy was legit, he would be out courting women to the nicest restaurants in NYC Even if this guy really is a HFT programmer, I don't care to ask him anything even though I am fascinated by HFT. I prefer not feed these kind of pompous jackasses exactly what they want.

Re:I am an HFT programmer

[Anonymous Coward]

I love you.

Re:I am an HFT programmer

[hedwards]

There is no misunderstanding. You're working for fraudsters that engage in a high stakes race to see who can fleece the individual investors the quickest. And in the mean time you destabilize the market and suck all sorts of money out as a leech would blood.

Re:I am an HFT programmer

[acak]

For an large enterprise developer who has worked about five years on .NET and Java, could you outline a path towards becoming an HFT programmer?

Thanks.

Re:I am an HFT programmer

[Anonymous Coward]

Smells seriously like jealousy. You want to care of your family? Go and get a job like what he's got.

Re:I am an HFT programmer

[Anonymous Coward]

Thank you. You actually have a respectable job, while the grand parent is just a fucking scumbag. "Just following orders" I'm sure.

Re:I am an HFT programmer

[Anonymous Coward]

The key is to hire developers with the experience and expertise to minimize that risk. Everyone makes mistakes, but these developers command the salary they do because they make mistakes less often.

My only objection is in the great-grandparents complaint about the engineer quality at Google. Not to defend Google or anyone else, but I'm guessing this HFT development doesn't follow the highest quality engineering practices as well, not by changing code minutes before release, anyways. Software engineering is about the whole lifecycle. For a quality engineering environment [even with a few mistakes], look at NASA, it's exactly the opposite of this HFT environment.

That being said, it would be an interesting place to work.

Re:I am an HFT programmer

[Renevith]

Do you ever buy or sell stock? Perhaps indirectly, through a mutual fund or 401k type plan? If so, then you benefit from high liquidity in the market. HFT and other Wall Street shenanigans do skim from the top, but they also provide liquidity. It's almost certain that the liquidity benefits small market players more than the skimming hurts them. In other words, the money they're skimming comes from the banks and brokers rather than you and me.

For example, take the stock of Red Hat (http://finance.yahoo.com/q?s=RHT). Yahoo Finance right now shows that, as of the last time the market was open, I could buy 100 shares for $42.56 (that's the "ask" or best current asking price), or I could sell 300 shares for $42.09 (the best available "bid"). That's a bid-ask spread of about 50 cents. That spread is a hidden cost to either buying or selling stock: If you buy and then sell RHT, you will have paid about 50 cents per share just for the privilege, even if nothing in particular happens to the company. Let's split that 50-50 and say that every stock transaction in RHT (buy or sell) costs you 25 cents per share in implied fees.

Those bids and asks are set by individuals and companies who are competing. They want to get a good deal for either buying or selling the stock, but they also know that if they set asking price too high or their bid too low, they'll never make any trades. The more competition there is, the tighter the bid-ask spread will be. HFT and other algorithmic approaches allow firms to set prices on tons of stocks without requiring human attention for each one, which dramatically increases the competition and thus tightens the bid-ask spread.

In this example, if you outlaw HFT and similar trading strategies, maybe RHT will have a spread of $1 intead of 50 cents. Maybe you'll be happy that HFTers aren't making ther 5 cent skim off the top anymore, but it'll be cold comfort when you're paying 25 cents more on each transaction and it's just going to a different Wall Street firm.

If you think I'm exaggerating the effect of computerized trading of the spread, have a look at slide 8 (page 4) of this study: http://fisher.osu.edu/~diether_1/b822/trading_costs_2up.pdf. Starting in 1960, the average bid-ask spread has ben dropping steadily every decade to a small fraction of what it used to be.

Background: I am an actuary trained in quantitative finance. I've never worked in Wall Street or done any HFT or other algo trading.

Re:I am an HFT programmer

[Anonymous Coward]

http://i.imgur.com/y7Hm9.jpg

"I don't always test my code... But when I do, I do it in production."

Re:I am an HFT programmer

[Anonymous Coward]

So you think that many of the intra-day stock market flash crashes (see Nucor Energy, 09/14/2010) are made from these types of untested modifications?

Re:I am an HFT programmer

[Anonymous Coward]

He and Bill Gates have been very clear on their position regarding dynastic wealth and the "ovarian lottery". It's not just a business scheme for their own benefit.

Also, if the company was distressed before the founder died, the company was not sustainable. If the company became distressed after the founder died, then the family was unable to manage it. In either case, the value of such a company is low.

Re:I am an HFT programmer

[Anonymous Coward]

Can you tell me how you developed all these skills?

Re:I am an HFT programmer

[Anonymous Coward]

its superscalAR. and assembly is not cpp.
i bet google engineers know that. that said im neither. google and traders both are doing morally wrong stuff for money. turned em down.

Re:I am an HFT programmer

[Anonymous Coward]

I know when to make something simple, when to use and avoid threads, and I can debug in minutes and push out a new version in the seconds before market open (not many people can handle that level of stress well).

This has me wondering how well you QA your software. I'm understandably concerned about this; your code might be handling my investments.

Judging by the number of spelling and grammar errors in his post - let's call them syntax errors - I understand your concern.

Re:I am an HFT programmer

[Anonymous Coward]

You realize there are plenty of large websites (Flickr, Etsy) that roll into production multiple times per day, with minimal impact over many years?

It's not just those reckless Wall Street types doing it.

Long QA cycles are a cumbersome reaction to quality problems.

Re:I am an HFT programmer

[Anonymous Coward]

Including 9 holidays, where the markets are closed and 10 days vacation (yes I know, probably doesn't take vacation)
this brings the hourly rate down to $68.56.

Re:I am an HFT programmer

[Anonymous Coward]

(Different AC here)

Sorry, how is that useful? Who gives a poo about audio quality in this age of earbuds?

Or maybe we should learn to not insult people based on gut feelings, and get educated on the actual benefits of people's work instead.

Your post reads like "He makes more money than I do, he must not deserve it". You should be ashamed of posting under your real name.

Re:I am an HFT programmer

[Renevith]

Each device I worked on, each firmware release, each line of code, does something useful for many, many people. Some of those people don't even know that audio equipment, leave alone software, is involved with what they are hearing.

Awfully arrogant of you to claim that some people benefit from your work without realizing it, without acknowledging that the same could be true of his. If he adds liquidity to the market, then you save money every time you buy or sell stock (or your 401k/mutual funds do). It's not necessarily true that he does, but you certainly aren't in a position to know. You could read more here, but I suppose that would just get in the way of your populist rage so maybe you'd rather not.

Oh, and you follow it up with another stunning display when you criticize him for not using his real name and then you wish him death in the very next sentence. Wow.

Re:I am an HFT programmer

[Anonymous Coward]

...fleece the individual investors the quickest.

“You can’t con an honest man” — grifter’s maxim.

Re:I am an HFT programmer

[Anonymous Coward]

Where in the world do programmers get paid overtime?

Re:I am an HFT programmer

[Anonymous Coward]

Warren Buffet didn't get rich through micro-transactions.

Warren Buffet gets rich by loudly proclaiming that estate taxes should be increased. Then he buys distressed companies out from under the family of the founder when they prove impossible to pass on to the next generation.

And he buys companies distressed in other ways to 'fix' them or to chop them up and sell the parts. Truly an American Hero, worthy of the adulation he gets as the poster boy 'good capitalist' by the left.

What a load of bull. Buffet got rich by recognizing the leveraging power of insurance companies, where the buyer pays in advance for a product they may never need. That creates a huge amount of 'float', money that is waiting to be used. He invested that into more diverse ventures, and that is how he got rich.

But that's ok- you just go back to listening to talk radio. Don't bother thinking for yourself.

Re:I am an HFT programmer

[Anonymous Coward]

You need not be angry at them. Usually the engineer is the one that properly understands the risks, so it's his bad decision that is risky. :)

Re:I am an HFT programmer

[Anonymous Coward]

Mr. Genius,

You know everything, but how about spelling the word "superscaler"? What a moron!

Re:I am an HFT programmer

[Anonymous Coward]

You always hear about Google programmers being the best in the industry

We do?

The only place I've heard that is from Google employees, but I've also heard Microsoft employees say that about their company, and Apple employees about theirs, and so on. Now I hear it from HFT programmers about their industry -- no surprise here.

the engineering quality just isn't there

Anybody who's tried to use Google Groups knows that!

Re:I am an HFT programmer

[Anonymous Coward]

I am most likely better than you at each and every aspect of software (and HDL) development you have mentioned. Except, of course, "debugging in minutes" -- that kind of irresponsibility would get me fired.

Sounds like you are much better at being a dick

Re:I am an HFT programmer

[Anonymous Coward]

I also don't have any problems with posting here under my real name. Or with telling you, and people like you, to die in a fire.

I don't think he can hear you over the sound of that $500,000/yr he's making.

Re:I am an HFT programmer

[Anonymous Coward]

dude, i'm sorry to break it to you, but you're underpaid.

Re:I am an HFT programmer

[Anonymous Coward]

Or with telling you, and people like you, to die in a fire.

so much for the moral high ground. do the retards here on slashdot really think you're less disgusting than that HFT fellow?

Re:I am an HFT programmer

[Anonymous Coward]

$96/hour is not bad. It's not great either.

I make around $150 an hour and I work fairly normal hours. I bet the only way a normal human can do 100 hour weeks for any length of time is with a pretty serious substance abuse problem.

Re:I am an HFT programmer

[Anonymous Coward]

what is your educational background and how did you learn what you know? books? websites? forums? any recommendations?

Re:I am an HFT programmer

[Anonymous Coward]

For your hard work and skilled labour, I, as a happy consumer, thank you for providing value. I wish every hard working skilled person had your attitude.

Re:I am an HFT programmer

[mikael]

Carnegie-Mellon University have open white papers on "stochastic calculus". There's enough public available material to keep anyone busy for months. No different from fractals and Brownian motion, except that share prices are one-dimensional. Everything else is basically code optimisation from the device drivers upwards.

Re:I am an HFT programmer

[Anonymous Coward]

I also don't have any problems ... telling you, and people like you, to die in a fire.

Awesome, sounds like you have the personality type for employing HFT programmers.

Re:I am an HFT programmer

[TheRaven64]

You haven't really explained why that's a bad thing. If the spread is more, then that means you can't make money from he noise as easily, you have to actually invest your money based on your expectation of the long-term performance of a company (or invest it in unusually volatile stocks, if you really like gambling). It seems that this would make it easier for companies that had good long-term business models to get investment, which would be good for the economy overall.

Re:I am an HFT programmer

[Anonymous Coward]

There is at least one trading company where the founders made their fortune arbing between exchanges via cell phone when the technology was new and fresh. They took usually what was a 24-hour cycle and turned it into a 1-hour one.

HFT takes what was a 1-hour cycle and turns it into a 1-millisecond one.

As long as products are correlated, there will always be money to be made by taking advantage of the differences when they go out of whack. The question is speed. If you ban HFT, well, people will just go back to using cell phones. Ban cell phones? Back to land lines. Ban land lines? Back to the pit.

You're fighting against fundamental economic forces here. If gold is trading at $1600/oz. on NYMEX and $1600.10 on LME, someone is going to take advantage of that.

Re:I am an HFT programmer

[TheRaven64]

Google does have an office in Manhattan. You should definitely apply there - I don't think it's an interesting company to work for anymore (and their recruiters are second only to Facebook's in how pathetically desperate they seem when they contact me), but the interview process is really fun.

Re:I am an HFT programmer

[aliquis]

Since it's financials it can do what it's supposed to do under normal consequences but something which isn't optimal for the current situation due to some exception.

Though I guess they have plenty of "exceptions" coded in to.

Re:I am an HFT programmer

[TheMadGamr]

The hourly rate does put it in perspective. Unless they happen to get paid with some kind of stock grant or benefit program, most of that gets hammered by the tax man at the %39 bracket (and probably state tax as well). They probably do get a couple weeks vacation, but if they live in NYC you lose a ton to rent and living costs. Which as an employee isn't that bad, but you can do better per hour at a Microsoft, Google or Apple. And you won't burn out after a year.

Re:I am an HFT programmer

[Anonymous Coward]

For the flipside here, I'm a Google SWE.

> I work 12 hours a day on average and do 100 hour weeks. I am on call during Asian hours and need to come in sometimes on holidays when other markets are going nuts and we need to plan.

I have a life. I spend time with my wife and kids. I make almost exactly half of what you do, but I also work a 35 hour week most of the time, so I'm making about 50% more than you, hourly. I'm on call about every 13th week, and will likely never actually have to go into work for it. Quit killing yourself working, or at least get more for it.

> but I've been to a couple Google interviews and turned them down both times because the engineering quality just isn't there.

I was going to come to the defense of Google here - but you know what, I'm going to let you think what you will. (I'd rather have you think we're all tools than have to work with with someone with your grandiose attitude) Also: you misspelled superscalar.

Re:I am an HFT programmer

[Anonymous Coward]

Above a certain level of wages, the best way to take care of your family is to be there. You'll hardly see them if you're working 100 hour weeks.

Re:I am an HFT programmer

[Anonymous Coward]

I'm a pretty fast learner who always had an interest in markets. I kind of accumulated them, sometimes through random events. When I went to uni, I was a computer science major and was under the naieve belief that i didn't need to know the nitty gritty of performance and that optimizing interpreters and compilers would solve it all for me. I specialized in AI.

When I left school, I worked at some small startups and always seemed to be put on projects that required some aspect of performance, weather dealing in large datasets or rewriting some library to be more generic and faster. One of my early projects was a string processing system that processed a lot of strings in concurrently, not my like using threads, but by finding the common pieces of work and doing those once, then piecing together the answers. (Kind of like common subexpression elimination on a grand scale). I did some small Linux hacks after than and between the two, my interest and ability to use memory efficiently really started to take hold.

I'm not sure if it is just the way my mind works, but I seem to have a good intuition of how to make things run fast. I think I take that extra little step of not just trying to understand why my code is trying to do, but what the code compiling or running my code (software and hardware) is going to try to do to it. This isn't some special ability, it is just the direction I took early.

I had written extremely large and fast caching infrastructures, ETL systems. and even my own language interpreter (something I'd really like to go back to actually - maybe someday I will go back to that and write a langauge similar to C# that will make what I do easier, because I think there is a lot of room for languages to grow in this regard).

I got a call one day from a recruiter, went in an interviewed, and was hired the next day. It was one of the most brutal interviews I had ever done.

From there, my desire to do well in work lead me to just keep learning and filling in the gaps in my knowledge. I now know the hardware and networking aspects far better than I could have ever thought I would. I actually have about cache-coherence protocols, bus latencies, specific NICs, PIM, etc.

I hired somebody to teach me VHDL to learn a little more about how that can be used. I've hired a few instructors, and I think they will help me tremendously, even if I'm not using the skills right now. I actually wish I had more geeky friends to join in. Like I had one friend get tutored in VHDL with me for a few weeks, but he had to move, and I thought it would have been cool to bring in a few more and have a mini-class since learning completely new skills after college can sometimes be hard if you don't already have a basic introduction.

It is mostly just an accumlation and my intense desire to keep learning and my extremely competitive nature.

Re:I am an HFT programmer

[Anonymous Coward]

It doesn't bother you that with all your technical ability, you waste it by not BUILDING anything?

You may talk down on Google engineers (and as a startup engineer sometimes we do to...) but at least they're making products. They're building usable software; software that affects the course of the world and many times improves the lives of hundreds of millions of people every day.

Your software generates millions of dollars in profits by exploiting systematic behaviors and flaws. My software generates profit by providing services for people, services those people pay for.

That's the difference.

Re:I am an HFT programmer

[Anonymous Coward]

Plus, this programmer who sounds good would actually be doing something useful.

Re:I am an HFT programmer

[Epic Wizard]

Question from someone about to finish college and wondering where you picked all of that up. I'm getting a degree in Game Design and Dev and I've always been curious about the actual efficiency of the functions I'm using. When should I worry about memory usage, access times, should I store this variable or calculate it the two dozen times it's going to be needed. Stuff like that. I've picked up little bits here and there but it always seems to be something someone got at random from someone old enough to have found that out on their own over many many years. Any good ideas?

Re:I am an HFT programmer

[Anonymous Coward]

I stopped counting the amount of times I saw "Atlas Shrugged" on trader's desks a long, long time ago.

Re:I am an HFT programmer

[Anonymous Coward]

Sucks that 90% of the commenters on this thread would rather yelp politik than ask questions to someone who does technically interesting and not well understood work. So let me give one a shot: what kind of testing processes do you guys have in place and use in normal situations? Are you regularly running market simulations (integration tests) in addition to unit testing your code? Thanks for the openness.

Re:I am an HFT programmer

[Anonymous Coward]

You are a sad, sad person.

Re:I am an HFT programmer

[kwoff]

What does "serve the economy" mean? Sounds like some fascist slogan.

Re:I am an HFT programmer

[Anonymous Coward]

I make $100 / hr doing database consulting, which is a much easier job with pretty much zero stress and 40-50 hr weeks.

Re:I am an HFT programmer

[JamesP]

Ok, here goes nothing:

- Do you guys do any kind of testing prior to deployment? Either running with past data or current data but without sending the orders, or some kind of mock 'signal'?

- Why Java, really? Unless it's something using the JVM like Scala or Clojure

- Any prior 'rule setting' such as, only dealing with a kind of security, only within a certain range of values, etc

- Do you guys use any special libraries (decimal math comes to mind). SSE/SSE2 or other SIMD optimizations?

- Do the programs have internal safety checks (like, see how much this is going to cost, if it's within market rules or is not gonna buy millions of dollars in Enron stocks)

Re:I am an HFT programmer

[s73v3r]

Losing money is also defined as not making as much as possible. If you were making $1000/minute on Monday doing something and only $100/minute on Friday (many bots only run for the first 5 to 10 minutes of the market); then that is also defined as a loss of $900/minute.

I find that statement quite disgusting, and also indicative of why we're in the trouble we are as a country.

Re:I am an HFT programmer

[s73v3r]

HFT does produce something of value, because there's almost always liquidity in the market because of HFT.

That's questionable at best. When you increase the liquidity (and in doing so, volatility) faster than an actual person can react to it, there is no value there, except to other HFT. You are just skimming profits at that point.

And don't even try to hide behind "real investors". You are fucking them over with your shit.

Re:I am an HFT programmer

[s73v3r]

I have a conscience.

Re:I am an HFT programmer

[s73v3r]

There is just no real dividing line.

Sure there is. The point at which a normal person cannot possibly react.

If we push the markets together a millisecond faster, there isn't much social worth, but it is that competition for that tiny timeslice that gives what we do benefit.

Meaning you don't provide anything to society, just to your overlords.

Re:I am an HFT programmer

[Anonymous Coward]

You apparently also don't have any problems being poor. You can use your real name, and be ground underneath my goddamn feet while I abuse my millions of dollars to fuck with you. Remember, good guys finish last. I can pay money to have you roughed up, and you can do nothing but complain self righteously. Money drives the world dickhead. I can have you driven off a cliff.

Re:I am an HFT programmer

[Anonymous Coward]

how do you feel about having all that talent and technical ability, yet produce absolutely nothing of value to society and instead spend your time allowing psychopaths to beat other psychopaths by fractions of a second, all to the detriment of everyone else?

How does it feel to have very little talent or technical ability and waste people's time with your ignorant comment, to the detriment of those more intelligent than you?

Re:I am an HFT programmer

[BitwizeGHC]

You:

I also don't have any problems with posting here under my real name.

HFT programmer:

Fuck you, that's my name. You know why, mister? Because you drove a Hyundai to get here tonight; I drove a $80,000 BMW. That's my name.

These guys don't give a shit about your piddly system of ethics. They're making more money than they know what to do with writing software that the guys who run America rely on every day.

Re:I am an HFT programmer

[Anonymous Coward]

I doubt you're better than him. He's a master at large-scale real-time distributed processing. You're good at programming one piece of hardware where you control everything. He's working on a CPU on a bus that talks to other busses that talks to network cards that talk to other network cards that talk to switches or routers or other busses that talk to CPU's ... and he has to make things happen really fast when he doesn't control most of the architecture.

Re:I am an HFT programmer

[Anonymous Coward]

Everyone makes mistakes. If you had proper quality assurance you could find and correct those mistakes before they roll to production. Then you wouldn't be in the situation of rushing a fix out in the middle of the day.

Another wall street geek here.

The problem is that by then the tip-off that somebody got playing golf with the CEO of [censored] would be common knowledge.

Re:I am an HFT programmer

[BitwizeGHC]

It should be noted that personally I admire guys like you who do the hard work on technically exciting stuff that means something to people. I was just quoting Glengarry Glen Ross to illustrate that when large amounts of money are being moved around, to money people all else is irrelevant.

I'm nowhere near your level, but I work in the robotics field and to see the code I write move and control an autonomous device is cool beyond measure.

I have been called upon to "debug in minutes" though, as when one of our underwater vehicles is running a survey and it bobs to the surface on a mission abort, finding out what went wrong and fixing it if it's trivial is essential if you don't want to burn daylight.

Re:I am an HFT programmer

[IICV]

You realize that, by your own numbers, you're making something like $96/hr? ($500k / (100 hour work weeks * 52 weeks in a year))

That's the equivalent of making $200k per year, and working 40 hours per week - which won't burn you out in a year.

Re:I am an HFT programmer

[Anonymous Coward]

"Cocaine is a hell of a drug!"

Re:I am an HFT programmer

[Anonymous Coward]

Really interesting thread here on all sides.

I've worked in and around finance and trading in a dozen different roles over twenty years, including trading rooms, but not HFT as such. Where I work now, the trading room has its own programming staff, paid modestly better than most equivalent positions outside, and required to follow pretty much no process guidelines. Just as an observation, the skill levels I see there are no better than average, but their ability to respect process was, even before moving there, already much worse than average.

YMMV.

Re:I am an HFT programmer

[Anonymous Coward]

Programmers and other highly paid professionals don't generally earn "time and a half" like that.

Re:I am an HFT programmer

[Anonymous Coward]

Okay...

Here are my questions... First Regarding the role of liquidity...

From Naked Capitalism this morning ( http://www.nakedcapitalism.com/2011/07/time-to-take-stock.html?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+NakedCapitalism+%28naked+capitalism%29 )

" Why do we think liquidity is a good thing?” Answer, because it facilitates trade around the exchange of information. “Information about what?” one might then ask. “The company in which the investment is being made,” is the answer. Does algorithmic trading exchange information about the performance of the company? No, it is only working off information about trading behaviour. Ergo, it may increase “liquidity” but it is not fulfilling the purpose of liquidity.

That kind of shift to traders working mostly off what traders do, rather than assessing the value of what is being traded, has become an absolute plague. It has taken over most Western financial markets. Hedging, for example, used to be all about hedging bets to protect the underling exchanges (usually wheat, or pork bellies or physical things). Now, hedging is all about reading behaviour, which then leads to other hedging strategies that are based on reading the hedging behaviour, and so on.

So the disappearing point is part of the problem. In our “anthrosphere” we are increasingly staring at each other’s navels in the financial markets, trying to make money and sustainable wealth out of ether. That is part of the problem. Regulators forgetting what the PURPOSE of financial markets is..."

Your answer to this?

Next question: ...relative to gross domestic product, the equity market capitalisation of the US, Europe and Asia had not grown since 2000, suggesting that “the contribution of equity markets to economic growth has been static”....

Do you believe your endeavor contributes to real economic growth?

Next question:

"Speed increases the risk of feasts and famines in market liquidity. HFT [high-frequency traders] contribute to the feast through lower bid-ask spreads. But they also contribute to the famine if their liquidity provision is fickle in situations of stress."

Do you believe this?

Next question:

(the conclusion from the same article)

"the misalignment in time between financial market instruments and that which they are supposed to represent is not just extremely dangerous, it is fundamentally inhuman."

Do you agree? Do you feel any responsibility? How do you feel about the people who drove the trains to the concentration camps in Nazi Germany? Did they know? Should they have cared? (The trains in Nazi Germany were very efficient...)

Re:I am an HFT programmer

[Anonymous Coward]

I'm also an HFT developer, and I can only speak to how my own firm operates, but we are very conscious of what sort of risk we are taking on. It is often necessary to make intraday changes to various pieces of software, for example if we find a bug - it's often either fix it or stop trading, so it gets fixed. The reason this seat of the pants development is OK is that any mistakes are backstopped by a risk infrastructure which operates in the opposite manner. It's infrequently changed, rigorously tested, and a required component for any application that is going to trade. If you're operating inside of that sandbox, there is a very limited amount of damage any application can do before it gets shut down.

I think HFT firms as whole are getting much more risk conscious, especially when it comes to operational risks like fat fingers, runaway robots, etc. Our industry is famous (or infamous) for not being very transparent, but risk controls are one area where the public and regulatory bodies *should* be shining the spotlight to prevent major issues, and they have been, especially in the past year since the flash crash.

Re:I am an HFT programmer

[chrissandvick]

Or with telling you, and people like you, to die in a fire.

And people are saying the HFT guys are psychotic. sheesh.

Re:I am an HFT programmer

[Anonymous Coward]

nobody on earth can work 100 hrs a week, 7 days a week every week of the year. a much better number for quick rule of thumb is to divide a yearly salary by 2000 = $250 per hour. and that's not a charge out rate thats delivered to the programmer (b4 tax and including a bonus by sound of it) i challenge any of you SAP programmers to make close to that. also he won't spend time looking for new contracts

Re:I am an HFT programmer

[Anonymous Coward]

Quants are hired to be able to do this kind of coding AND NOT SCREW UP. A lot of the work we do (not all, but a lot) could be done by Joe Programmer if it weren't for the requirement that you can't screw up. Trading is a dangerous occupation and that's why they try to hire the best people.

So does this means you spend a ridiculous amount of time in the testing & reviewing your code?

Or do you rely on the hiring process and the managers involved to find the people who don't put out bad code-- much...

Re:I am an HFT programmer

[Anonymous Coward]

I work on algorithmic trading systems at big investment bank. Milliseconds matter to us buy not microseconds. We write the most efficient java code I've ever seen and it rivals the bank's best C++ systems. I make over $200k/year and work 45 hours a week.

Re:I am an HFT programmer

[Anonymous Coward]

Spot on - this is precisely my problem with all of this. Mind you I think my problem is with the entire system as a whole which allows stuff like this to just happen on a daily basis.

Re:I am an HFT programmer

[JavaRob]

Err... he didn't say he does 100 hour weeks *on average*, just that he does them.

I.e., his average week is 84 hours (assuming he works 7 days a week; otherwise 72 or 60 hours a week), and sometimes he works as much as a 100-hour week (7 14-hour days, or some other split).

Come on now,

Re:I am an HFT programmer

[Anonymous Coward]

Trivial answer: Just tax them. 1% sales tax on all financial transactions, half paid by the seller, half by the buyer.

HFT would stop.

AC.

Re:I am an HFT programmer

[Anonymous Coward]

I am an ex-Google programmer.

I think you (the HFT programmer) have a very poor understanding of what Google programmers do.

I'm not sure I would describe any of what you said up there as "very normal", and you certainly haven't demonstrated that what you do is useful. (You haven't demonstrated that it's not useful.) Now, clearly, *someone* finds it useful, but I think that most people outside your company and its investors are measuring utility as something other than "making money at the expense of others" which is most of what people perceive HFT to do.

One of the things that Google programmers have in much greater numbers than almost any other community of programmers/engineers I know is a desire to create things which will improve the world in some way. If I had any desire to go back to being an engineer, that quality would take me back to Google (or to work with other ex-Googlers I know.) How many HFT programmers are interested in something other than making money?

In my experience (and I live in Manhattan now, and have close friends who have described what it's like to work at a quant fund in detail that's difficult to get unless you're being recruited to work in the field), that's the driving factor. And I guarantee you that many of my colleagues at Google back in 2000 believed that it was worth foregoing some money to work there.

I question how much of an impression you could get of the quality of Google's engineering from the interview process. While I won't claim that it's perfect (because that would be a lie) some of the stories I've heard of engineering practices at quant funds make me laugh. An ability to work 12 hours a day, 100 hours a week, combined with pushing code written in the minutes before market opening does not make you useful.

But I'm glad that you're making a half million a year; hopefully you'll retire from that soon and do something that's actually useful with that money.

Re:I am an HFT programmer

[JavaRob]

He didn't say he works 100-hour weeks on average, just that he "does them" - assume "sometimes" or "occasionally", since he DID say he averages 12-hour days.

He also gets vacation, even though sometimes it is interrupted by emergencies.

Let's assume 2 weeks off even including the 10 or so federal holidays (likely he actually gets more than this!) minus 2 days of emergency work.
He averages 12-hour days, an estimate probably based on a 5-day week (the stock markets are closed on the weekend, and he'd mention it if he had a non-standard work week, right?).

So: 50 weeks * 5 day week + 2 days lost vacation = 252 days * 12-hours = 3024 hours.
500000/3024 = $165.34/hour.

If "average 12-hour days" was within a 6-day work week => $137.97/hour
7-day work week => $118.37

Re:I am an HFT programmer

[JavaRob]

He doesn't say he averages 100-hour work weeks, just that he does them (presumably in case of emergencies).

Re:I am an HFT programmer

[MindStalker]

But but, they add liquidity. //Well they do, but its pretty worthless.

Incase you are curious what I mean by liquidity. In a traditional market if I want to sell my stocks especially if I have a lot of them, it might take hours to find buyers for all my stocks. The speed traders make the market more liquid so I can sell my stocks anytime instantly (though not necessarily as good as a price I would have gotten by waiting). If I don't wish for that liquidity I could sell higher, but might it might not sell or I'd have to wait the more traditional time frame at least.

Re:I am an HFT programmer

[Anonymous Coward]

How much free reign do you have? Are you programming an algorithm specified by someone else or watching the patterns in the signals you see and coming up with new variations on your existing work, or new ideas altogether, and testing them out?

Do you give your algorithm a small amount of money to start with or run it with funny money before letting it go with a larger amount of capital?

I work on control systems. sometimes we make changes to the process while it is running. what changes can be made without testing, simulation, or stopping the process is a judgement call. Programming a machine or HFT is different from programming a website, mobile app, etc because you can't necessarily test, fix, recompile and test again to the same extent.

Re:I am an HFT programmer

[Anonymous Coward]

If you really need me to, I can even crank out some VHDL code.

So where does VHDL come into this? Why not the other HDL language? Whose chip and whose design flow do you use?

Re:I am an HFT programmer

[Anonymous Coward]

I think your definition of value -yours, and the entire HFT business- is based entirely on how much money you earn (not much, if you break it down to hourly rate, btw), and the money you earn to others (does not reflect the _actual_ value of the of product of somebody, since money degraded to math equations in the end) does not add up.
I don't mean to offend you, but do you seriously think this will make anyone happy? Given that the money-happiness relation is actually scientifically proven to be wrong to be determined by such activities and goals, as a (presumably) highly intelligent person: how do you accept the fact that you (also presumably) knowingly do something that's wrong to the mankind?

Again: I'm not judging, just being curious...

Re:I am an HFT programmer

[NoOneInParticular]

Why would information about a company only change on some tick, say, an hour? Wouldn't that be unfair to firms that are not a phone call away? Why not make it 24 hours, or a quarter? Open the trade floor once a year for 15 minutes. You make a mistake, wait a year. Probably often better than waiting an hour.

So, what could possibly change in the underlying value of a corporation in a microsecond? Nothing, you are right. The stock market is however not trading the underlying value (as this value is unknown and unmeasurable, and given that the underlying value also includes expectations about the future, it probably doesn't exist at all), but rather the information about the underlying value. And that information can change by the microsecond. Each trade establishes a new datapoint about the underlying value of a corporation. So one single person, or algorithm, that thinks that this particular stock should trade a quantum higher or lower, makes a statement about the underlying value of a corporation.

Your idea of a trade an hour would make that tiny bit of information aggregate into a big thing. Big things lead to massive swings in the stock market. People get scared and do wild stuff. So, if anything, HFT has removed volatility, not added to it. Yes, everyone also found out that if one person makes a stupid bet, all those suckers that use stop-losses to protect their investment get creamed. It used to be the market maker that would clean out the suckers, now its the algorithm. Nothing has changed. Don't use stop-loss, use options for protection.

In the past, it used to be who had the fastest horse. The New York stock exchange and the Chicago stock exchange could be out of sync for weeks. If you traveled from New York to Chicago with some important news, and beat the other guy, you could make millions. This news, when brought by horse, leads to massive changes and immediate bankruptcies as there is no time to react to such large bits of news. But.... buggywhip makers rejoiced. Then the telegraph came. It become important to have the fastest morse signallers. Then the phone came, and then it became quiet for a while. The limits of the technology were reached and the playing field was leveled. Trades could now be done by professionals and regular people. Then the computer came, and the network. We are now again in a speed race that will last until some physical limit will be reached. Lots of energy goes in there, but not more than the gains. When this all settles down, there will be a steady source of arbitraders that will remove inefficiencies in nano-seconds. Brilliant engineers will leave the field, and another level of artificial volatility is removed from the markets, leading to less artefacts created by the discrete nature of trading. Time is made continuous. And that makes this whole trading thing more stable.

Is it worth billions? Well, yes. The cumulated effects of the inefficiencies in the market apparently add up to such sums. If you leave these in the market, they will be translated to unnatural volatility.

Re:I am an HFT programmer

[Anonymous Coward]

It will gap up or down each hour. Just look at some Asian exchanges with lunch break.

Re:I am an HFT programmer

[Anonymous Coward]

What's changing rapidly is not the underlying value of a corporation made up of flesh and blood humans, but rather the knowledge about the world as it affects that corporation.

Re:I am an HFT programmer

[waveman]

1. Can you offer any insights into what happens with these meltdowns? If that actually HFT related and what goes wrong? Is it amateurs in action of is there a more fundamental issue with market structure? Eg http://www.zerohedge.com/article/how-hft-quote-stuffing-caused-market-crash-may-6-and-threatens-destroy-entire-market-any-mom

2. Can you offer any suggestions as to how a retail investor might deal with the presence of HFTs and not get taken to the cleaners? Eg in the light of things like this "Evaluation of the 'Adaptive-Aggressive' Trading-Agent Strategy Against Human Traders in CDA: AA Wins" at http://lscits.cs.bris.ac.uk/docs/AAMAS_CAMERA_READY.pdf

Re:I am an HFT programmer

[Anonymous Coward]

Im not trying to turn this into a pissing contest. But it struck me that if I worked the same hours as you, Id make more money and I have a pretty standard consulting job. I can totally understand spending a couple of years working as much as you do, but in the end there are better things to do out there. Save enough to buy a house, get a normal job and you should be set for life.

My .02

Best,
K

Re:I am an HFT programmer

[Anonymous Coward]

righteous +1

Re:I am an HFT programmer

[Anonymous Coward]

You call it risky, I call it reckless. You try what you're doing in any other field and you'd be fired pretty damn quick.
That said, I'm not so much angry at you as I am at the people who ask you to do this.

That's kind of funny coming from you, Mr. Freeman

Re:I am an HFT programmer

[Anonymous Coward]

Can you add some more specifics?

Re:I am an HFT programmer

[Anonymous Coward]

I wrote another comment on some of the basic types of strategies here:

http://developers.slashdot.org/comments.pl?sid=2357190&cid=36938820

It is definitely more constraining, but that is opening up a little. For example stat arb used to be slower timescales, but the more competition the more the need for speed since you are competitive for even smaller slices. Now stat arb is merging into the basic HFT real. Options are another area where that is happening. I think this will be the continual path (within limits). It is kind of like that saying that it is always AI until it is well understood, then it is just software engineering (how game tree search and chess used to be more AI-like and is not an undergraduate project or how statistical NLP used to be cutting edge AI, but is not also pretty basic in some ways).

So yeah, you got the right idea for all the arbs. We'll arb anything. I've even heard of a friend arbing gold and crude distaillates by going through a statistical tie between gold and crude (brent or west texas, depending on price) and then using the cracks to get to distillates.

And yes, the more used the legs of the trade are, the faster and simplier the strategy needs to be (or find other entries and exist). For example, you may have an amazing idea to trade crude and see an opportunity, but since crude is so heavily traded and moves to quickly, your amazing idea might not work in practice because you miss half your hedge orders from other people doing another trade involving crude. This also depends on the time of day (you might be able to run that strategy in the off hours).

Also, I always assumed in HFT the programmer WAS the trader. Why would you be taking orders from traders? Surely they can't react much, other than to decide how much risk to take on your system? The run I run my shop, we programmers are looking at the output, positions, PnL all day.

I made some references to this in a post, but yes, in the HFT world, the line between trader and algo developer is very blurry (eg, I have my Series exams). We work very closely and both of your asses are on the line for PNL. And their job is mostly to contain risk and manage position (and do a lot of accounting work that I have done, but really hate). I constantly have the trading screens up all time and watch what is going on and yell back and forth to my traders when we see things that look a little funky. We both share a lot of the same responsibilities for getting the strategies to work.

I didn't go into the role of a HFT traders, but it is different than other trading roles. It has a huge data analysis component to it. We both pour over tick by tick logs (and have tools to help with this) trying to make sure everything happened the way it should have and see if we could have done something better. He'll sometimes sit behind me and we'll "pairs program" (sometimes a few traders, since a couple might be using the same base strategy with tweaks). He can write scripts and stuff and has a good knowledge of computers, but he couldn't write the code I do. He can certainly read it and pick out bugs or follow along to make sure we are writing what we want.

I hope I answered everything. Commenting is becoming hard because Slashdot keeps throttling me.

Re:I am an HFT programmer

[Dixie_Flatline]

I wish I could score you up higher. Thanks for saying what everyone else is thinking.

Re:I am an HFT programmer

[hexagonc]

What could possibly change in the underlying value of a corporation made up of flesh and blood humans and capital with decades of depreciation in a fucking microsecond? Here's a hint: nothing.

Not exactly true. If companies were islands isolated from each other and other external factors then you might have a point. However, as we all know that is not true. The value of a company can indeed change within microseconds and less because some external factor outside of that company's control might affect its value. For example, suppose some disruptive event occurs in the world, such as a coup in Iran or a UFO landing on Washington. The value of a lot of companies will change instantly (well, really at the speed at which information can flow, which is on the order of nanoseconds). I imagine our HFT friends here would want to be amongst the first to act upon this information, either minimizing loss or maximizing gain from such an event. This doesn't strike me as voodoo or particularly nefarious/underhanded/unethical, especially since there would probably be obvious losers depending on the type of event.

Re:I am an HFT programmer

[Delwin]

Wait - what programmer actually gets a differential for overtime?

Re:I am an HFT programmer

[Anonymous Coward]

Damn good post.

Re:I am an HFT programmer

[Anonymous Coward]

Some questions / your views on:

1) In 'the race to zero', surely your huge skills are destined to break the market? (- by design)
2) Dark Pools - meaning the people running your algos are fleeing from the consequences, whilst raking in vast profits of other people's money. Define why Dark Pools exist, and if your algos are allowed on them
3) Nanex, and other companies, have been studying the effects of your algo's [flash crashes] - are you really so naive / full of hubris that you can't foresee the info-war of competing algos, and what they're likely doing to the market?
4) Hyper-parasitism between algos - worried about it? (you should be)

Re:I am an HFT programmer

[Anonymous Coward]

You give IT professionals a bad name

Re:I am an HFT programmer

[Anonymous Coward]

I'd like your comment on the relative importance of stat arb to latency arb in your line of work. An argument can be made that stat arb adds liquidity and provides information to the market through changed bids faster more rapidly than would be the case otherwise. Latency arb seems to be about stealing pennies from pensioners. The impression I had gotten is that HFT started as stat arb until someone figured out how to make money through latency.

Re:I am an HFT programmer

[Lanteran]

I don't have a problem with private high speed cables- I have a problem with the ridiculous waste of resources to gain literally a 5ms advantage over the other psychopaths. True though, work is created. But that's about the only benefit to anyone other than HFT-ers. Of course, if the process is ever outlawed or something, we'll get a flood of programmers who can write a port of asteroids from Z80 assembly without minimum requirements jumping to an i7.

Re:I am an HFT programmer

[zopf]

We're in huge trouble as a country because a completely different sector of the financial industry was irresponsible in the way they lent money. Making high-frequency trading the scapegoat is not going to solve our problems.

As for high-frequency trading itself, I don't see the cause for disgust. Traders of all kinds (that is, people who exchange things with willing participants) try to optimize their revenue, as do most for-profit businesses. If your argument is that for-profit activities in general are the reason our country is in trouble, then I suppose I won't convince you of anything. If your argument is instead that doing business faster than others is immoral, I'd like to see a little more of the rationale.

Re:I am an HFT programmer

[POMOSexual]

Thank you for initiating this fascinating topic. I apologize for the aggressive nature of some of the posts below and I hope people realize the opportunity you are offering them for learning. I am a small investor only, but have been intrigued by the occurence of "flash crashes" probably caused by these HFT programs. They seem to be getting very sophisticated and adaptive, I have a lot of questions I hope you can answer. 1) Do you know if your programs (or others) have caused these flash crashes? Is it a desired or undesired effect, if so? 2) Do your programs look/hunt for stop loss triggers? 3) Do your programs resist normal market conditions (price fix)? How do they do this, if so? 4) When graphed, the bid/ask prices migrate sometimes in "blocky" and sometimes "curvilinear" patterns (within milliseconds). Is this a programmer style or a function of market conditions, do you know? 5) Do your programs often compete directly with other algos? How do they find each other in these very short time frames, if so? 7) Can you explain the general mathematical principles of the algo(s) to a lay person?

Re:I am an HFT programmer

[Anonymous Coward]

...

5) Do your programs often compete directly with other algos?

[6)] How do they find each other in these very short time frames, if so?

7) Can you explain the general mathematical principles of the algo(s) to a lay person?

I suggest http://christianmarks.wordpress.com/2010/05/25/mathematical-logic-finds-unexpected-application-on-wall-street/ for more info from highly knowledgeable practitioners.

Re:I am an HFT programmer

[Anonymous Coward]

. I read the C++ and Hotspot assembly, and know how to program for superscaler architectures specifically. If you really need me to, I can even crank out some VHDL code.

You can't spell superscalar. And most RTL code these days is in Verilog unless you work in certain parts of the aerospace/defense.

Re:I am an HFT programmer

[Anonymous Coward]

I can debug in minutes and push out a new version in the seconds

You must be taking a huge gamble by doing that. I don't see how your new version could be considered safe to use if it is deployed that fast.

Before I place judgement what kind of testing is being done? Or is it simply compile and push? If it's the latter then I can't see how he can claim high standards of engineering quality. If he has an automated test harness and audit-able flow controls to production, then perhaps you need to ask yourself what's wrong in your organisation that you can't push to production more quickly?

Re:I am an HFT programmer

[Anonymous Coward]

"I read the C++"

oh really?

troll!

Re:I am an HFT programmer

[Anonymous Coward]

Parent commenter again. You're damn right it is reckless. I spend half my time screaming at the people who approve these things but to no avail. Ultimately, the right course of action is to exit the industry and do something that actually adds to the world's wealth than drain it into a small group's pockets.

Re:I am an HFT programmer

[Anonymous Coward]

First, your welcome. It has become harder to reply to people since Slashdot has started to throttle me to about a post every four hours. I just wish more would have asked questions like this, since my industry does tend to be very secretive and you aren't going to get many opportunities to peer inside.

The flash crash last Many from the CFTC report was caused by algorithmic, but not HFT. Somebody at Wadell entered the wrong quantity and there were no curbs to slow down its trading so the program thought it had to dump a ton of E-mini contracts on a short time. Many HFT shops shut down because they feared trades would be broken if they took the other side (they were right), but the ones that stayed in, from what I hear, did very well.

1- Yes, I have had bugs that caused some market disruptions. I have know people that have had some terrible things happen where huge numbers of traders had to be repriced or rolled back that spread across multiple instruments. It happens if you've been in the industry long enough. It is very costly and not a desired effect at all. It can be hugely costly and take a ton of time to trade out of these positions when they get handed to you because of a problem in the algo. You also tend to get fired or put on a very short leash for a while after that.

2- I have never written any algo that tries to set off stop orders, and I don't know of anybody that has done so either. Everybody has heard of the theory, but it just seems very impractical. I've heard some rumors or people doing it at even numbers like maybe 500 for GOOG or 15000 for the Dow, but these are unconfirmed, but I have no doubt it has been tried. You would need to set off a flood of orders, the other side of the book would need to follow up or down, and you would need to make enough to cover the very large spread left, the price to set off the avalanche, and take into account only selling part of your holdings at the new price since others will be in there too. (The way the story usually works involved the options market.)

3- As a general principle, you never try to go against the market. There have been a lot of firms that have tried to do that and it really doesn't work too well. You take what the market gives you and admit defeat. The market for things like crude or S&P index are so deep and interconnected that to push crude around, you would also need to push crude next month (and the months after that) around too, and nat gas, and distillates, and the crude ETFs because there are HFT shops that link those together and will clean up if you ever tried to do that. It is a common conspiracy theory, but if it were that easy, everybody would do it.

4- I'm not sure what you are asking. Bid-ask spreads are a function of volatility, the more volatility the wider they are since they are the price of taking on risk. The market maker you are transacting with (and HFT firm in this example) collects the spread as a sort of charge or the risk they need to take on. If volatility goes up, the spread goes up. I'm not sure what you mean by blocky though, sorry.

5&6- Yes, and I can usually tell when more competition enters the market. They may not be running the same algo, but they might be doing one that involves one of the instruments in my algo making me miss on orders I send. You can see see it in the stats pretty clearly. We don't hunt them out, and we don't care actually, as long as we the price we want.

7- There aren't any general principles. It is mostly algebra at runtime. To development the model and tune it calculus and and stat is in there too. That is for equities. For derivatives, you will get the calc in the runtime too (kinda, since often you use an approximation that is easier to calculate).

Good luck out there.

Re:I am an HFT programmer

[Bob Cat - NYMPHS]

The ones who get paid by the hour, like I did. Are all programmers on salary these days? Suckers.

Re:I am an HFT programmer

[Rebelgecko]

More importantly...
100 hours a week divide by 12 hours a day...he works 8 and a third days a week.
Mr. Anonymous sounds legit— definitely no blatant exaggerations and/or fabricated bullshit here.

Re:I am an HFT programmer

[MasterHundinco]

But in the end i have to ask does what you do really matter? You're making a better arm lever for a slot machine system that greed driven people game and manipulate. You might not care what you build, but you are part of a system that promotes a status quo of apathy. The work you do it just another vector that put us on a path to the world wide economic shit storm we have now. What is 500,000 dollars really worth when you could have taken those skills and gotten us to the stars? You could have wrote a program that solved the folding errors of prions. You could have wrote a program that could calculate similar ways of producing the same drugs cheaper. I say you are a wasted talent and it saddens me that the majority of graduated with math degrees and comp sci degrees from top universities are joining you in further bringing a faster "real-time" end to any shining bright academics solving issues for the good of humanity. All because they rather be paid 500,000 for their soul.

Re:I am an HFT programmer

[Xphile101361]

These guys don't give a shit about your piddly system of ethics. They're making more money than they know what to do with writing software that the guys who run America rely on every day.

No wonder we are in such a sad state of affairs

Re:I am an HFT programmer

[Anonymous Coward]

How did you get started on this track? Did you have to know everything going into the interview or did you learn a lot of programming on the job? I only ask because I'm a sophomore undergrad right now and am interested in quant trading. Did you learn all these things on your own (google, tutorials, experience) or were you trained? Thanks.

Re:I am an HFT programmer

[Anonymous Coward]

Speaking as a developer working for a stock exchange; we've all heard more than enough stories about the cowboy programming mentality amoung HFT's. They like to think that they're worth the big bucks and some of them are; but most are just doing what the traders did in the Great Recession - taking massive risks with other people's money and hoping that it works out.

Humans screw up, end of story; no matter how much you get paid or how well trained you are - you always screw up. You can increase your chance of screwing up by encouraging sleep debt and increasing stress. Anyone who claims you can work massive hours and reliably not screw up is a deluded moron with zero understanding of human biology or psychology.

You're playing with fire. Still. After screwing up the world economy once; you still haven't learned any lessons and people are still letting you drive. Incredible.

Re:I am an HFT programmer

[Anonymous Coward]

"I wrote price feeds that take 1/5th of a microsecond in C++ and (a little slower) in Java."

Define "price feeds" please: which protocol, which kind of data, which kind of hardware?

I also wrote price feeds and there is no way you can fully decode a native exchange price message and exploit it in a trading application under a microsecond in java with standard hardware.

Re:I am an HFT programmer

[mayanksinghal]

LOL! You just made it sound so depressing!

Re:I am an HFT programmer

[Anonymous Coward]

Particularly with the hours worked and regularity with which on-call is exercised. I know sr. / multi-country execs that make barely more than you do. Maybe you're new at this?

Re:I am an HFT programmer

[RatOot]

Assuming you worked 7 days a week, then a 100 hour week equals over 14 hours a day (average). So you either don't work 12 hour days OR you don't work 100 hour weeks OR your week consists of 8.33 days... If your maths was as good as your hyperbole perhaps you would earn even more.

Re:I am an HFT programmer

[Edmund Blackadder]

You do realize that for your dumb conspiracy theory to work you have to show that Warren Buffet has actually caused any increases of estate taxes. He hasn't. Furthermore, the owners of Buffet companies usually run their companies for a long time after Buffet buys them, so it is doubtful they were distressed because they could not pass them on to their children.

Warren Buffet is a hero to many people because he actually did something extremely unselfish, for the benefit of the world and Americans specifically. Naturally this is something you cannot appreciate.

Re:I am an HFT programmer

[ren-n-stimpy]

you sound like an idiot, full of piss and vinegar. How in the hell is VHDL relevant given you don't build hardware? superscalar is ancient, basic technology, if you care about hand-optimized assembly on current architectures, you should be addressing vector, multicore, and GPUs. you mention threads in combination with usec timing requires, which seems bizarre, given the thread scheduler has timeslice quanta orders of magnitude beyond that. for real-time trading, i'd expect event-based programming (full scheduling control) not threads, like even simple high-performance HTTP servers. you're proud of working 100 hour weeks? how about work smarter dude, that's simply not necessary at sustained rates, for anything, period. lack of sleep + poorly-tested changes = bugs. I could go on. I am a programmer who grew to hire and lead programmers for a living (which is actually way harder than simply writing the software yourself.. it becomes all about architecture), and, I assure you, you sound nothing like a top-of-breed programmer. go get a PhD and learn what you don't know. you quants in NY live in a bubble of big money and shitty software. take away the money, and you got shit.

fail.

(*) I'm sure you can clarify many of these things to us, I'm sure you do indeed write working software -- but, that you can so poorly explain it speaks volumes.

Re:I am an HFT programmer

[Anonymous Coward]

yeah, and most websites are in php, unless it's in asp, jsp, ruby, or any other language.

Re:I am an HFT programmer

[Anonymous Coward]

Some basic math tells me that 100 hour weeks are not doable with 12 hours a day:

7 * 12 = 84

Re:I am an HFT programmer

[Anonymous Coward]

Tips for a beginner?

Re:I am an HFT programmer

[Kakihara]

I wonder if these desperate-sounding recruiters would chase you so hard if they knew how much of your working day you spend on slashdot. Which is not to say that I'm not grateful for it - I am.

Re:I am an HFT programmer

[TheRaven64]

Heh, maybe not as many. Mind you, I live somewhere where the cost of living is pretty low - my regular expenses are lower than what I'd make in a minimum-wage job here - so I can afford to turn down work. And, as a freelancer, I can work erratic hours, so my working day may not be when you think it is. My clients are often on different continents to me, so I often end up working in the evenings, rather than during the day, and I may be working very long hours for a week and then not at all for most of the next month. On the plus side, I do get to work on a lot of interesting projects. This year I've taught a university module, written two books, implemented the low-level parts of three different programming languages, written code for the FreeBSD core system, and taught a few tango classes.

Re:I am an HFT programmer

[Anonymous Coward]

Note to self: Never offer to answer questions on Slashdot. It just brings the morons out.

High frequency trading smooths volatility. It moves money backwards and forwards between the companies that specialise in it. It doesn't affect your trading account. Unless you're trying to trade over anything under a few years. In that case you're just giving me money. Thanks.

Re:I am an HFT programmer

[Alex Belits]

The device I am currently working on, has 6 CPUs (4 of them are microcontrollers though) and 2 network interfaces. And plenty of other components.

Re:I am an HFT programmer

[Alex Belits]

If he adds liquidity to the market, then you save money every time you buy or sell stock (or your 401k/mutual funds do).

Thanks to people like him, I don't have stocks or 401(k) anymore.

Oh, and you follow it up with another stunning display when you criticize him for not using his real name and then you wish him death in the very next sentence. Wow.

I do not expect people to wish me death for what I do. He does, or at least should.

Re:I am an HFT programmer

[Alex Belits]

What about it? He is a loathsome being that causes harm to other people, there is nothing wrong with wishing him to die in a fire.

Re:I am an HFT programmer

[Anonymous Coward]

It requires no intelligence to skim which is still infected with gambling. If on the other hand you can consistently PREDICT you can make REAL money. But that requires INTELLIGENCE. People with intelligence don't work for a living.

Re:I am an HFT programmer

[Anonymous Coward]

>modification to an algorithm to keep about $100 million from going at a lower price than what the traders wanted

questions...
<sarcasm>
- did you have to use one of those relatively difficult 'if' test logic thingys?
- so they didnt want to sell at $98million but for $99million?
</sarcasm>

Re:I am an HFT programmer

[Anonymous Coward]

he didn't say he ALWAYS works 100 hr weeks...

Re:I am an HFT programmer

[unabashischizm]

I am most likely better than you at each and every aspect of software (and HDL) development you have mentioned. Except, of course, "debugging in minutes" -- that kind of irresponsibility would get me fired. I also have to work long hours, and have to have clear understanding of complex issues unrelated to software.

Except I do embedded software and FPGA development for professional audio equipment. Each device I worked on, each firmware release, each line of code, does something useful for many, many people. Some of those people don't even know that audio equipment, leave alone software, is involved with what they are hearing. Large fraction of my work ends up being free/open source, too -- platform, drivers, etc.

I also don't have any problems with posting here under my real name. Or with telling you, and people like you, to die in a fire.

not only there is no god , but try getting the plumber at home on weekends........

Re:I am an HFT programmer

[Anonymous Coward]

I would like to know if software/information security is an important factor in HFT shops. If an attacker could tamper with your algos or even slow down (through DDoS or otherwise) feeds, etc... what effect would that have on what you do? Most large banks' security controls are mandated by things like GLBA, SOX, etc. They are also motivated by loss from criminal elements because they are an obvious target. What about you guys? You aren't regulated, so what does infosec look like in your industry?

Re:I am an HFT programmer

[John Courtland]

That's actually front-running. It's illegal but it happens.

Re:I am an HFT programmer

[debruce]

Well, how much do brilliant people who spend their lives trying to get you to click on ads add to society? Not to mention something awful like Zynga.

Still those people enable things like Google that do add a lot.

The issue with HFT is, in the old days the marketmaker got first look, and in exchange was limited in what they could do in terms of front-running and market manipulation, and was supposed to maintain an orderly market. The HFTs took over the liquidity-providing functions but it seems pretty wild west, and if the market really goes nuts they step away.

Still, the only reason there is big money available to build things like Google is because there is a liquid stock market for them, and one that learns and evolves, which HFT is part of.

Re:I am an HFT programmer

[Anonymous Coward]

Maybe your system does, nd maybe it does it extremely well. But can you say that about every HFT system out there? And are you 100% certain how your system will react to a sudden market crash or other anamolies that, if not accounted for, could trigger huge problems in the market? How about the others?

This isn't really about your programming skill, it's about human nature. The markets are complex systems. HFT's are complex systems. There are a million points of failure and to think you have them all covered and nobody will ever make a mistake or "program with malice", for lack of a better term, is just naive.

As for the useful things you guys do, please, tell me what they are. You slowly suck money out of the system for profit. That's it. Provide liquidity? Puh-lease. There is plenty of liquidity in the market and somebody having to wait a few extra minutes or OMFG hours!!! is trivial.

Re:I am an HFT programmer

[Anonymous Coward]

Exactly. If the toilets are dirty, everybody is mad. If the boss is gone and another stupid meeing gets cancelled, everybody is happy.

Vive le janitor!

Re:I am an HFT programmer

[Anonymous Coward]

Wow...a whopping $96.15 an hour! On the contracting market, a reasonably skilled web developer or project manager can earn $100/hr., work 40-50 hrs. a week, and have weekends off. Which would beg the question, "Exactly how smart are these HFT programmers?". Looks like the only real smart people in this equation are the ones that hire them to work non-stop for ridiculous wages.

Re:I am an HFT programmer

[ansible]

Well, let's look at the fundamental motiviation of HFT: making money.

OK, nothing wrong with that. However, we, as a society, have reason to question how everyone makes their money. Lots of people don't want a minority of people to make money clubbing baby seals. The value generated by that practice (nice fur) is not high enough overall benefit to society to justify it.

What is the net result of HFT? The people doing the HFT make money, sure, but what value are they generating? Nothing, as far as I can tell.

HFT constitutes a tax on everyone else in the market who is going to hold onto a position for more than a couple seconds. Anyone who is a long-term investor. Every sale you make, will generate less profit. Everything you buy will be more expensive. HFT is just siphoning off money from everyone else.

The markets work because they allow more efficient allocation of capital to useful functions, and they move money away from non-useful (or at least not money-making) functions in society. This is good. HFT is sand in the gears. And don't give me arguments about how they provide market liquidity, the markets were doing just fine before this practice started.

Re:I am an HFT programmer

[Anonymous Coward]

Wow. You're really dumb. Well, no, not dumb - arrogant - arrogant in assuming your ideas work. Even if you've worked in the industry, you would not be qualified to make blanket statements like ' prevent trades faster than some tick, say, an hour'. You have no idea what the implications of such a rule would be.

To the asinine individual who wrote the phrase: 'produce absolutely nothing of value to society': stop making these stupid broad statements. For what it's worth, I would rank the 'productivity to society' of any HFT programming team above the team that wrote the Angry Birds game any time. In any case, why do you care about 'value to society'? We work to provide value to ourselves. Anyone who truly wishes to work to provide value to society should give all earnings in excess of the simplest living expenses to charity or research. How many of you are willing to do that?

Re:I am an HFT programmer

[Anonymous Coward]

Do you demand to be paid per transaction? I see no reason why you should not be skimming a significant amount of revenue from the skimmers.

Re:I am an HFT programmer

[Anonymous Coward]

Interesting. I have the same kind of skillset as you (very good low-level hardware knowledge, C/asm and VHDL), although no stocks/investment know-how. I make roughly the same amount as you, but it sounds like my stress level is at least two orders of magnitude lower, and I am a contract design engineer, which I always considered pretty high-stress.

Re:I am an HFT programmer

[s73v3r]

We're in huge trouble as a country because a completely different sector of the financial industry was irresponsible in the way they lent money. Making high-frequency trading the scapegoat is not going to solve our problems.

I am not trying to pin it entirely on HFT, as that feeling I mentioned is pervasive throughout the entire economy. Sorry if that wasn't clear.

Re:I am an HFT programmer

[tolkienfan]

Not true. The issue with CDSs and mortgage bonds were that they were incorrectly valued and a balloon resulted. The crash that resulted had nothing to do with HFT, and was inevitable. People had been selling stuff at way over value. Those came down to life and blood people using flawed models of risk.

Re:I am an HFT programmer

[Anonymous Coward]

The spread would actually be much, much higher. HFT accounts for roughly 70% of volume on our exchanges. If you got rid of it, exchanges would have to charge such that they could survive on 30% of the current volume, so that's triple-fees at minimum, assuming their costs scale linearly (which they don't - they grow logarithmically.)

So if you're on like Scottrade and you normally pay $7 to trade 100 shares, now you have to pay $21 (though I'd guess closer to $50 personally) to trade those same 100 shares; the stock would have to move up or down a dollar before you broke even just on the transaction. In that case, good luck on a $20 stock.

Re:I am an HFT programmer

[Anonymous Coward]

that's pretty harsh - the "die in a fire" part.

The FPGA's you use are made by publicly-traded companies (XLNX and whatever the ticker is for altera ). If there were no stock market, no chipmaker would have the capital to build a billion-dollar plus foundry. If you've ever used a Tektronix scope, an amphenol or molex, connector, or copper wire, it was an equity market or commodities market that made it possible.

HF traders can actually make the world a better place, not because they do meticulous research about the stocks and other instruments they trade, but because they make prices cheaper for everyone. They lower the "spread", the prices of what you can buy or sell something for.

If you buy a car, drive it off the lot, and sell it to another lot, the spread is huge, probably thousands of dollars. But what if you could buy the car and sell it back for pennies under what you just paid for it ? That's what HFT's allow you to do. In the old days, market makers would jack up the spread and slow down trading. HFT's are making them obsolete.

I will agree the OP comes across pretty cocky, but having worked in that industry, I can tell you most people in it are cocky. In fact he's so cocky I could almost guess he's in a hedge fund in chicago about two blocks from the willis tower. Dont' want to say the name of it, lest i'm wrong.

Re:I am an HFT programmer

[Alex Belits]

The FPGA's you use are made by publicly-traded companies (XLNX and whatever the ticker is for altera ). If there were no stock market, no chipmaker would have the capital to build a billion-dollar plus foundry. If you've ever used a Tektronix scope, an amphenol or molex, connector, or copper wire, it was an equity market or commodities market that made it possible.

s/stock market/slavery/g , and it would sound like a perfectly valid argument when slavery was the dominant way of running the economy.

Re:I am an HFT programmer

[Anonymous Coward]

"Except I do embedded software and FPGA development" - which was pioneered by companies that are owned by shareholders and TRADED in the free market.

Re:I am an HFT programmer

[Anonymous Coward]

Now, now, don't make the lad think about what he's doing...

Re:I am an HFT programmer

[Alex Belits]

Which used science and mathematics developed in the Middle East while some European assholes were trying to "liberate" a fictional relic from there. Let's all pay tribute to Arabs for Al Jabr (Algebra)! In its turn, it used foundation of the same developed by ancient Greeks while some other assholes were busy annexing them to their empire. This means, we should immediately restore the practice of slavery. I am sure, Egyptians had some say in this, too -- so hail Pharaoh!

In reality, I am a part of unbroken chain of scientific and engineering development that goes for thousands of years into the past and likely will continue indefinitely into the future. It survived all known and forgotten forms of social, political and economic systems. Sure as Hell, it will survive a giant Ponzi scheme that is current version of Capitalism as implemented in US.

Re:I am an HFT programmer

[Anonymous Coward]

The decrease in the spread due to cheaper infrastructure would also happen if trades were forced to happen on ticks as someone else in this thread suggested. You only need the algorithm part, not the high frequency bit. Then you wouldn't have programmers constantly under pressure trying to hotfix their code whilst it's running. This would mean overall less risk and fewer occasions where bugs cause massive artificial sways in the market.

Re:I am an HFT programmer

[zopf]

Your argument that HFT constitutes a tax without adding value is a little shaky.

Every sale you make, will generate less profit. Everything you buy will be more expensive.

Short rebuttal:

Every trade you make will likely execute faster, because an HFT was willing to take the other side. You may miss out on price improvement, but you will be able to dispose of your position risk faster, because the HFT is accepting that risk in the hopes of turning a profit.

Elaboration:

Any serious investor who does not want to immediately lose his or her shirt should be using limit orders. For example, I might want to sell 100 shares of AAPL at $398.00. I don't care who takes the other side; I just make an open contract to sell AAPL at that given price, and some anonymous market participant accepts that contract. In this case, the value that an HFT offers me is the liquidity (yes, I said it), whereby my order to buy those shares might be crossed nearly immediately, even though it was a resting offer (i.e. it did not cross any existing buy order in the marketplace). As an investor, I am able to sell quickly at the price I want.

The HFT will then try to sell those shares they've bought from me at a higher price. So - what has the HFT done to deserve said profit? They've taken on risk. I was able to sell my shares at $398.00 like I wanted, and now the HFT is stuck managing the position that I no longer wanted to hold. They're now long 100 shares of AAPL at $398.00, and while they're betting that they can sell them to somebody else at $398.01 or $398.05, it's also quite possible that AAPL will decline to $397.50 over the next few minutes, in which case the HFT will likely have booked the loss at $397.9x and moved on to their next trade.

In this case, the HFT is acting as a market-maker, which is a function that traders at financial firms have provided since markets have existed. What's the difference between today's HFT and yesterday's phone brokers? They do it faster; and they don't typically get the huge commission breaks or shoulder the liquidity obligations of traditional market-makers.

Another typical role an HFT plays is that of an arbitrageur. Arbitrage is buying and selling a set of instruments that results in an immediate profit with little risk; the only major risk involved is execution risk - that someone would beat the arbitrageur to the prices he saw in the market. Again, this might seem a useless activity, sucking up pennies by getting information faster. But consider a simple example: two futures exchanges, one in the US (the CME) and one in Europe (the ICE), both list heating oil. I, a hotel owner, want to hedge my building's exposure to heating oil price changes, so I submit a limit order to buy 10 heating oil futures (delivering contracts at the CME at $3.0000/gallon. I can only afford to pay for access to the CME, but an arbitrageur has access to both CME and ICE. The arbitrageur sees my order to buy 10 contracts at $3.0000/gallon on CME, but also sees an order to sell 30 contracts at $2.9999/gallon on ICE. He decides to sell me 10 contracts at $3.0000 and buy 10 of the 30 available contracts on ICE at $2.9999, locking in a whopping $42 profit.

So what value did the HFT provide to the market? He created an effective liquidity channel between the two disparate exchanges. And what specific value did he provide to me, the business owner? He gave me access to liquidity on an exchange that I could not afford to access. In a sense, I paid him $42 to do this. If I could afford to do it more cheaply myself, I would have. In this sense, he actually provides this liquidity channel for cheaper than I could have achieved otherwise. And what risk did he take? He took the risk that one of his orders did not execute against the market price he saw. Imagine that he is US-based as well, and sold me those ten contracts at $3.0000, but someone in Europe happened to buy those 30 c

Re:I am an HFT programmer

[Anonymous Coward]

If it spreads more, it means that people who trade more often and don't need liquidity (professional traders) can make more more _easily_ from the noise, at the expense of the people who need to sell more quickly (people who need the money).

If you are paying $25 for something that nobody is willing to pay more than $24, or getting $24 for something that people are asking $25, somebody is getting $1 at your expense.

Re:I am an HFT programmer

[Anonymous Coward]

Not to mention something awful like Zynga.

Totally OT, I realize, but why is that so bad? Gives people a way to destress and relax and do something they enjoy. Stop being so uptight. If you don't like it or are too superior for it, then ignore it, dickhead.

Re:I am an HFT programmer

[Anonymous Coward]

w-wow you're a pompous dick, Alex Belits

Re:I am an HFT programmer

[Anonymous Coward]

These guys don't give a shit about your piddly system of ethics. They're making more money than they know what to do with writing software that the guys who RUIN America rely on every day.

There - fixed it for ya.

Re:I am an HFT programmer

[Anonymous Coward]

It seems very challenging. How to start? How to become an HFT programmer and find positions in the market? Especially for a guy outside of the USA or Europe.

Re:I am an HFT programmer

[Anonymous Coward]

What about taxes?

Re:I am an HFT programmer

[Anonymous Coward]

Seriously? "I'm better than you at everything, go die in a fire"?

I'm glad you're working in a field that I'll probably never land in. Thanks for reminding me how non-shitty my coworkers are.

Re:I am an HFT programmer

[smellotron]

Are there any time when you get "a preview" of one order someone will put into the market where you can for instance pick up all that's available for that price and then sell for that order?

If any trading firm receives customer order flow (e.g. Goldman Sachs, Interactive Brokers, and countless others), the previews are technically available in the form of direct customer orders. However, trading on that knowledge is front-running, and it is illegal. These types of companies are expected to separate their proprietary trading desks from their broker desks to avoid this sort of behavior, but I wouldn't bet my life on it.

In a proprietary trading firm there are no customer orders, and all trading decisions are based on public knowledge. The doom and gloom you typically hear about "HFT front-running" typically refers to HFT algorithms reacting with lower latency than other algorithms or human traders. AFAIK, there's no ability to actually front-run in that situation.

Re:I am an HFT programmer

[smellotron]

For an large enterprise developer who has worked about five years on .NET and Java, could you outline a path towards becoming an HFT programmer?

Start by looking for an opening at an exchange or an ECN. Check out the job requirements of any major exchange located in your area.

Re:I am an HFT programmer

[smellotron]

Can you offer any suggestions as to how a retail investor might deal with the presence of HFTs and not get taken to the cleaners?

As a retail investor, you should largely ignore HFTs. Trade infrequently, and you won't be subject to short-term volatility. As somebody else mentioned, use options to hedge instead of stop-loss orders.

Re:I am an HFT programmer

[smellotron]

The problem is that by then the tip-off that somebody got playing golf with the CEO of [censored] would be common knowledge.

First of all, that sounds like insider trading, and that is illegal. Like, Martha-Stewart illegal. Second of all, if you need to "push out a fix" to your platform react to a generic "up" or "down" signal from an insider, you're doing it wrong. This is an operational control change, not a software logic change.

Re:I am an HFT programmer

[Alex Belits]

Seriously? "I'm better than you at everything, go die in a fire"?

I'm glad you're working in a field that I'll probably never land in. Thanks for reminding me how non-shitty my coworkers are.

My co-workers are decent, talented and responsible people, who, like myself, do all kinds of work that provides something useful to other people. I respect that, and I would never wish them to die in a fire.

On the other hand, scumbags draining the wealth of everyone on Earth by exploiting loopholes in financial systems that would've been closed if those scumbags weren't in control of the rules governing such systems... They and their minions have no good reason to be alive.

Re:I am an HFT programmer

[Anonymous Coward]

If you are talking about the place that starts with a G, then I actually interviewed there, but took another offer. They are a great firm, and I know people who work there though. I must say too, they treat you wish some mad respect from the first class air travel and dope room at the Trump for the interview days plus let me stay the weekend there.

Re:I am an HFT programmer

[Anonymous Coward]

I'll be posting anonymously, but I think many here have a very poor understanding of what we do. Most of that is because we do tend to be a very secretive group, but if you were to sit down with some of us, you would see that we really do very normal (and useful!) things in the market.

I work on the algo and core infrastructure. I wrote price feeds that take 1/5th of a microsecond in C++ and (a little slower) in Java. I understand in fine detail how cache and the the PCI-e bus works. I have a very good understanding of algorithms and the constant-time tradeoffs. I know when to make something simple, when to use and avoid threads, and I can debug in minutes and push out a new version in the seconds before market open (not many people can handle that level of stress well). I read the C++ and Hotspot assembly, and know how to program for superscaler architectures specifically. If you really need me to, I can even crank out some VHDL code.

On top of that, I understand market microstructure and derivative pricing. I work 12 hours a day on average and do 100 hour weeks. I am on call during Asian hours and need to come in sometimes on holidays when other markets are going nuts and we need to plan.

I also hope to make $500,000 this year.

You always hear about Google programmers being the best in the industry, but I've been to a couple Google interviews and turned them down both times because the engineering quality just isn't there. I'd put the average HFT programmer up against the best in Google anyday.

Ask away, and I'll answer to the best of my ability.

Fortunately for us, people developing in the nuclear power plants, medical or spatial industry don't work in the same way... even though they earn less income...

Re:I am an HFT programmer

[Anonymous Coward]

such bull***t

if you know so much, work so hard, for so long, for so little relative money, you are the biggest tool ever.

Re:I am an HFT programmer

[Anonymous Coward]

I work 12 hours a day on average and do 100 hour weeks.

What kind of average do you use, typically?

Also, the major point of our discussion is that while you may be able to do all that you say, it's unethical, and you ought not do it.

Re:I am an HFT programmer

[alexo]

But what if you could buy the car and sell it back for pennies under what you just paid for it

Where's the added value?

Re:I am an HFT programmer

[alexo]

w-wow you're a pompous dick, Alex Belits

He's also right.

Re:I am an HFT programmer

[alexo]

I want to buy some stock and willing to pay, say up to $1/share.
Somebody is willing to sell me that stock for, say as low as $0.99/share.
Hey, I'd gladly pay $9.90 as it's less than my limit, but as soon as the order is sent, some HFT algorithm gets in before me, snatches the stock at $0.99 and resells it back to me at $1.00

Re:I am an HFT programmer

[Anonymous Coward]

Industrial programmer who formerly wrote life-critical code here.

Maybe you didn't notice how frequently your doors have failed lately, or how extensive the losses have been.

Re:I am an HFT programmer

[libertytoast]

I also work in the finance industry analyzing and testing performance software performance. I would love to hear more about how you do 1/5th of a microsecond price feeds if you're game. This is clearly not possible without some custom hardware (you mention VHDL). Could you please tell us more about this hardware and what kind of logic resides in it vs regular software running on the hosts cpu(s)? BTW, this is a terrific presentation on HPC Finance using Java: http://www.infoq.com/presentations/LMAX

Re:I am an HFT programmer

[Anonymous Coward]

Thank you.

Re:I am an HFT programmer

[aliquis]

Thanks. So not common and hopefully not happening then :)

Translation

[Chas]

They're paid not to steal (or at least not get caught).

yeeeeeah

[Anonymous Coward]

I'm past the whole "a-holes can yell at me if they pay me enough" phase.

Re:yeeeeeah

[TheRaven64]

I don't mind it. On the other hand, most of the people I work for are on different continents, so if they're rude I can easily ignore them until they apologise.

Re:Linux by default?

[bmo]

>so it's not surprising to see them using an operating system that is built from stolen Microsoft technology.

Funny how the NYSE and the LSE both tried Microsoft solutions and moved off of it.

Microsoft is an example of how not to do something. Darl McBride, Microsoft's attack chihuahua, called this "negative know-how" and thought that since Linux devs learned how not to do something, somehow SCO was owed money.

Softies: confirmed for total retards.

--
BMO

from an HFT developer's view

[Anonymous Coward]

I'm a developer at an HFT firm and can say that it is some of the most interesting work I've done in my career. You need to wear many hats for the job.

On a daily basis I multitask between making low-level kernel modifications to reduce system latencies, to refactoring our high level marketdata->prediction->execution engine. It's a never ending balance/conflict between making things as fast as possible while trying to maintain good design principles and not take shortcuts in the sole interest of efficiency.

To those that say HFT is an anathema to the economy, I think there is some merit, but there are two sides to every story. I've definitely met some other HFT shops that focus on some really shady and morally questionably tactics (essentially DDOSing markets to gain an advantage). However, the amazing advancement in network technology and all these crazy RDMA and 10G user-space network stacks is literally due to the HFT community, and I'm already seeing these new technologies start to penetrate other markets. This is capitalism at work.

Also, believe it or not, HFT has reduced spreads due to competition, which means you pay less when you want to buy or sell a stock RightNow instead of using a limit order (and while spread reduction is good for investors, it sucks for us because it is not as profitable as it used to be a few years ago).

The "Billions" in profits going to HFT firms right now previously went to the big investment banks and the old classic broker taking orders over the phone and charging a ridiculous market spread.

Re:from an HFT developer's view

[earls]

Excellent counter-point. Thanks for the info.

Re:from an HFT developer's view

[Anonymous Coward]

How did you get into HFT?

Re:from an HFT developer's view

[pyite]

However, the amazing advancement in network technology and all these crazy RDMA and 10G user-space network stacks is literally due to the HFT community

This is completely true. Financial applications are driving HPC components in a way that educational and government spending haven't in years and years of use.

Re:from an HFT developer's view

[jovius]

It seems that we need to couple Large Hadron Collider with stock market to advance science. It's somewhat tragic that money is the driving force of development.

The human reward system is amazingly rigid.

Re:from an HFT developer's view

[Anonymous Coward]

I think there is some merit

Like what? I call BS.

Re:from an HFT developer's view

[LanMan04]

To those that say HFT is an anathema to the economy, I think there is some merit, but there are two sides to every story.

Yes, but those two sides do NOT necessarily have equal merit. See the evolution "debate".

word on the street...

[kervin]

As some of the interviewers in the article hint at, HFT is a small niche that's getting way more than its share of attention.

I'm not saying HFT isn't dangerous, or something to keep an eye on. But if you're looking for a job in Manhattan financial sector, you'll more likely be working in web or other "high level" stacks.

Java is *huge*, .Net is also very popular. Web is, of course, very important with just about every stack, from Ruby to ASP.Net represented. I see a surprising about of MatLab for analysis ( I guess I just wasn't expecting to see any with QuantLib and similar floating around ).

Re:word on the street...

[Anonymous Coward]

Dont forget Cobol, Never forget the Cobol!

Re:Linux by default?

[bmo]

>Linux stole from minix and unix

Boy oh boy, this story is bringing out the softies tonight.

Do you get paid to do this clown act or do you just do it to attract the children?

--
BMO

A couple more things to point out

[Anonymous Coward]

1 - A lot of stuff you hear about HFT (using supercomputers, using functional languages, using clusters, using nosql databases, etc...) are either back office, B-rate HFT firms, or BS. In my experience at the higher end places I've worked, Java and C/C++ are the standard for any trading system or order routing system. To push something out in single-digit microseconds, you really need the control of those languages (yeah, Java doesn't quite have the same level, but you can get it close enough by knowing enough about HotSpot). (back office are those that do the behind the support stuff like keeping track of risk, order flow, etc.. while front office does the actual algos and trading systems)

2 - We tend to rewrite everything at the upper levels. A lot of firms will use things like 29 West or other messaging, order routing, parsing libraries, but when you get to head of the pack, you tend to do it all yourself. This goes doubly true for the language APIs. I can't remember the last time I used STL, Boost, or the JDK util classes.

3 - We don't front run, we try to not predict over any appreciable amount of time. We have very high Sharp ratios (risk adjusted return) because we don't do those things. HFT is very low risk and we like to keep it that way. There are some HFT firms that can a full year without a losing day.

4 - In HFT the line between a front-office developer and a trader is very blurry. I'm often responsible for developing my own trading algos and running them all myself. Various shops have different break downs, and some don't even employ traders beyond the front office programming staff.

Re:A couple more things to point out

[Anonymous Coward]

What's the advantage of doing all this for someone else instead of just doing it yourself (and thereby increasing your own profit)?

Is it because of the infrastructure they provide? I assume you have a better connection than a retail person could get by co-locating?

Re:A couple more things to point out

[earls]

Instant access to hundreds of millions of a dollars that makes these super short trades possible.

Re:A couple more things to point out

[Anonymous Coward]

The infrastructure, and other talent, but most importantly the seed capital.

There are a lot of small HFT places though. When outsiders think HFT they often think big banks, but in reality the top HFT firms tend to be small couple hundred people places. I even worked at a 12 person company, and I was late to the party on that. If you want to specialize in a particular market or type of trade, you can be successful when small. If is hard to make a general purpose trading infrastructure that goes to these extreme limits. But if you want to optimize for futures trading on CME or trading in Japan, you can have your little slice of pie.

Re:A couple more things to point out

[aliquis]

Another advantage (maybe not for HFT but rather trading for someone else) is probably also that _if_ you screw up it wasn't your money. As long as it works you get money. If it doesn't it's not really your problem.

Re:A couple more things to point out

[s73v3r]

I'm often responsible for developing my own trading algos and running them all myself.

If that's true, then why aren't you taking the bulk of the earnings?

Selling the soul?

[Anonymous Coward]

Basically they are getting paid to be someone's bitch. As in selling the parts of their souls, that is, their professional integrity. The truth is that is the case in most of 'programming jobs' where one is acting on someone's whims and wishes, however inarticulate they are. Inept managers, bosses, indebted traders, all with no clear thoughts and power to concentrate envision, plan and/or solve engineering problems, but with enormous greed, driving us all "forward"...

Not all HFT programmers are alike

[Tony Isaac]

A while back, I hired a former HFT programmer. He was bright, but not six-figure talent.

If you're the best and brightest in any field, you can make a very nice salary. HFT requires some special skills, but in my experience hiring programmers, intelligence, hard work, and ability to learn quickly are more important than specific resume points.

Re:Not all HFT programmers are alike

[Anonymous Coward]

Man, that's what HFT is all about -- it's a resume bullet that encapsulates just what you're looking for. HFT has a new problem every day and you need to deal with hardware, OS kernels, network issues, prediction models, and resilient software that handles an amazing amount of data where downtime costs real money. Plus most of us work crazy hours because we love it.

One word of advice: Don't hire a shy quant who "worked in finance but was recently let go". There are these people that literally have 3 or 4 degrees in economics, finance, and math yet are seriously incapable of being able to handle the requirement to handle whatever issue comes up or whatever path leads you to improving the system.

How long to programmers stay in that type of job??

[Cutting_Crew]

I get e-mails ALL the time from these people up in NYC. If i so much as wink at my resume online differently I can always be certain that I will be receiving many opening jobs in the trading and banking industry. In fact, throughout my career, this has always been the case. I live in Florida. Why in the world would they be trying to recruit me from florida and why have I always gotten e-mails from them. Is it because they have a lot of people leaving? Do people get frequent burnout? Are they just spamming?

The last "recruiter" to e-mail me wanted to know if I was looking for a 'great' opportunity in the banking industry writing C++ applications with a big whopping salary of $100,000 with a 12 month contract. Since i was tired of these people wasting their time trying to contact me I e-mailed them back and asked why they thought I would just get up and move from Florida to New York for a 1 year contract and i tried to get confirmation of a starting salary of $100,000 a year. He e-mailed back stating that $100,000 was indeed the starting salary with great 'benefits'. So i e-mailed back and informed him that i was already making $75,000 there and no state taxes so $100K really wasn't that great of a deal, especially considering that I would have to make 4 times my current salary that I am making now just to come close to breaking even in NYC. Incidentally I never got another e-mail back from him.

When you have a family, there is no way I am giving up a house to move to NYC to live in a one bedroom closet for the same price as my 4 bedroom house. - with less income coming in.

Re:How long to programmers stay in that type of jo

[malakai]

You do it for the bonus, not the salary. But either way, you don't do it for a 1 year contract.

Insider Trading at its Technological Best

[Anonymous Coward]

The field is based on various "low latency" multicast services, and being able to make your trades microseconds before other companies before your competitors. It's expensive to set up, but done properly it gives you a lead in manipulating huge sums of money at slightly greater profits to your own company. The result is effectively shoving the money into the pockets of the few companies that can afford that "insider" advantage of the low latency.

Unfortunately, it's all about to go away. The use of FPGA's directly where the fiber optic leaves the stock market, or even inside it, to pre-process and merely send along the "buy" or "sell" signals of relevant stocks is about to cut the the high frequency trading market to ribbons, and expose the house of cards of extremely expensive hardware investments for such trading as a completely wasted capital investment.

Re:Linux by default?

[WorBlux]

POSIX: Portable Operating System Interface for UNIX.

Linux tries to keep pretty close to to it. 'nuff said.

It's also no secret linus learned OS design a large part from hands on operation with minix, and that the minix file system was the first file system linux used.

Re:How long to programmers stay in that type of jo

[Cutting_Crew]

yeah but the cost of housing costs about 3 or 4 times as much for what i am living in now. how much does a 3,000 sq ft home cost in the NYC suburbs anyway? I am also aware of the fact that if i want to maintain that same sort of lifestyle then i am not going to be living in city which is not going to work with kids - but maybe if people do end up working 100 hour weeks, then it is probably more suited to people that have no family at all to take care of. So what are we talking about here? A 1 hour commute one way? No thanks. I also doubt that the bonus would still not make up for what I would get to break even - just a hunch.

What about the sysadmins?!?

[ewwhite]

Unfortunately, us systems/network engineers in the HFT industry don't get the same level of respect (or pay) as the developers... It's better than other industries, but yeah...

Re:What about the sysadmins?!?

[Anonymous Coward]

Unfortunately, us systems/network engineers in the HFT industry don't get the same level of respect (or pay) as the developers.

You're at the wrong firm, my friend.

Re:How long to programmers stay in that type of jo

[Anonymous Coward]

When you have a family, there is no way I am giving up a house to move to NYC to live in a one bedroom closet for the same price as my 4 bedroom house. - with less income coming in.

Not to mention NYC is a crowded, crime-ridden, often filthy, concrete jungle of a depressing nerve-wracking shithole place to live that's full of rude assholes and people who think even the most casual eye contact is an advanced form of aggression. Those are the decent neighborhoods and it gets much worse. And if you ever have to drive a car, I sure hope you love traffic! Oh and if any sort of natural disaster or anything of that sort were to happen, a big city with at most a 3-day food supply is the very last place you would ever want to be.

Maybe you love money more than I do. Maybe you're a masochist. Or maybe you don't know what you were considering. But there's just no way in hell I'd consider moving to NYC. Not even just to screw around with an ambitious recruiter. The place is a monument to everything that is soul-less and dehumanized about society.

Ada!

[Javaman59]

When I worked on the Canadian Automated Air Traffic Control System (CAATS) Ada skills were at a premium, and I was recruited from Australia on $80,000, in 1998, so that's probably $150K plus in 2011. The project had nearly two hundred Ada engineers recruited from around the world, on similar, or better, salaries. Plus, it was a great project, great team, moderate pressure, and in one of the worlds most beautiful cities!

I shifted to C++ and then .Net, but I think the Ada market has kept high salaries, even while it has been shrinking, especially for those with a security clearance in the US.

Re:Ada!

[n8r0n]

I worked on defense systems in Ada for a number of years. While I love working in C# and Java, and sometimes Objective-C, I have to say that Ada had some really great features (some of which clearly influenced later languages).

It might seem like Ada would be a great language for financial services, as it was designed with security and reliability in mind. But, alas, the financial sector is really less interested in those features, and vastly more interested in skimming their clients' money while telling them how lucky they are to have the liquidity. I don't remember Ada including a library for that (system.scheming.ponzi, maybe?).

Re:Ada!

[Anonymous Coward]

Canadian IT salaries haven't grown for a decade, so it is likely that you would only be paid $80k today if the same job existed. Nav Canada salaries are terrible.

you know who else posted bullshit analogies

[decora]

on internet forums?

that's right. Adolph Hitler.

He wrote hundreds of pages 'exposing' the 'truth' behind 'power'.

It was called 'the protocols of the elders of zion'. of course, the whole thing was bullshit. made up by some anonymous author, possibly the Russian Tsar's secret police, to support yet another pogrom.

But Hitler took this and ran with it. Over and over this document, and many others, including his book Mein Kampf, were given out by the tens of thousands. People were happy to 'forward' these 'revelations' to others, often tweaking details here and there, or changing the attribution of the author(s).

And what happened in the end? Trillions of children were killed. I'm a student of history too. I have over 5 billion books published. On the internet. You can look it up.

Re:you know who else posted bullshit analogies

[Opportunist]

That book was passed around, but hardly anyone ever read it. Imagine what would have happened if people did and noticed what a loony that idiot was before he got enough power.

Seriously. If you want to know just how much "truth" one of those people, who feel the need to explain easily how complicated things work, really has to tell, read one of their books. You'll regret the expense, but then again, who pays for ebooks?

Re:you know who else posted bullshit analogies

[Anonymous Coward]

That book was passed around, but hardly anyone ever read it. Imagine what would have happened if people did and noticed what a loony that idiot was before he got enough power.

His support would've disappeared so fast, you could actually hear it evaporate. I think the sound it made would've been something like this: "WHOOSH"

Re:Linux by default?

[bmo]

> Using a specification
> "stealing"

No. Just no. You may as well claim the same thing about Windows, because Windows idiots hop up and down about how POSIX compliant Windows is.

>It's also no secret linus learned OS design a large part from hands on operation with minix

You have no idea what Tanenbaum had to say about this, do you?

http://en.wikipedia.org/wiki/Tanenbaum%E2%80%93Torvalds_debate

>Implying that a file system specification has any bearing on how a kernel is designed.

Don't know if you're trolling or just stupid. I suspect both.

--
BMO

Re:How long to programmers stay in that type of jo

[Anonymous Coward]

As to how long programmers stay in these kinds of jobs, it depends. I'd say it's a split: many programmers stay in finance for a long time, some leave after less than a year, and there's not a lot in between. It tends to be a "love it or hate it" proposition: those that love it stay, and those that hate it figure it out pretty quickly and leave. Some programmers won't have the chops for HFT but will want to stay in the financial sector, in which case there are plenty of other jobs to choose from.

enough lies please

[decora]

we all understand what 'arbitrage' is. when the synthetic CDO market calls their deals 'arbitrage' we all know its fucking bullshit.

when the sales guys in the brochures talked about the 'AAA' ratings on these pieces of 'arbitrage', it was all bullshit.

when Lloyd Blankfein calls it 'hedging, not betting', its fucking bullshit. ]

there is absolutely nothing, whatsoever, 'valuable' behind a credit default swap. it is a bet. that is a fact, and its not rocket science, and its not a conspiracy theory, and its not "the ignorant and alarmist" decrying some nefarious boogey man. its the basic fucking fact of what fucking happened.

I beg of you. stop lying. nobody believes you anymore. this is like the scene in Shattered Glass when Peter Saarsgard has to finally explain to Hayden Christiansen that the whole charade has ended.

the financial industry has no clothes. we all know it. there is no point in pretending.

Re:enough lies please

[pyite]

there is absolutely nothing, whatsoever, 'valuable' behind a credit default swap. it is a bet. that is a fact, and its not rocket science, and its not a conspiracy theory, and its not "the ignorant and alarmist" decrying some nefarious boogey man.

You're completely wrong. If you buy a bond from company X, it certainly makes sense to have insurance that if company X goes out of business, you still get your money. And hence, the credit default swap was born. The fact that someone may use the instrument to speculate is a separate issue.

People speculate on everything. It's what you do when you stock up on cans of soup when it's on sale. You speculate that the price is going to go up next week.

Re:enough lies please

[UnknowingFool]

Yes it makes sense to buy insure on something that you own like if you buy a mortgage; however, it's a bet if you speculate for or against something that someone else owns.

Re:enough lies please

[obarthelemy]

there's a strong moral risk: if you can both buy a company's debt, and insurance in case they don't repay it, why not keep doing that endlessly ? you make money both ways, nobody is really interested in the underlying debt quality, and bankers make fat commissions on both sides of every deal.

oh, wait...

Re:enough lies please

[peragrin]

sure but the moment you take out insurance on your insurance to cover your insurance on your default swap, with each insurance only covering the 50% of the riskiest credit the last bit of insurance is covering 75% of bad insurance and will collapse, which then forces all the other insurances to collapse.

tada that is exactly what happened. those big banks jumped into an insurance pyramid scheme of their own making and where to stupid to see the forest for all the green they were growing.

They didn't break any laws because no one really knew what they were doing including themselves.

Re:enough lies please

[JamesP]

Yes, but you wouldn't be able to buy it without a speculator. That's what they do, they sell these instruments.

it's a bet if you speculate for or against something that someone else owns.

If this was something specific, yes, but it's usually a combination of bets.

The amount of financial instruments, and derivatives of derivatives has gone beyond the level of ridiculous, no doubt about that.

But we always think the other guy is the speculator and we are the investors.

Re:enough lies please

[JamesP]

there's a strong moral risk: if you can both buy a company's debt, and insurance in case they don't repay it, why not keep doing that endlessly ? you make money both ways, nobody is really interested in the underlying debt quality, and bankers make fat commissions on both sides of every deal.

oh, wait...

If you do that, you end where you started.

Mathematically, the insurance price is derived from risk.

The money is on selling insurance, of course. Still, most of the time, nothing happens. And of course there are a lot of investors willing to buy insurance for crap and compromising themselves even more.

Re:enough lies please

[Hognoxious]

if you can both buy a company's debt, and insurance in case they don't repay it, why not keep doing that endlessly ? you make money both ways

Bullshit, and shame on the asstards who modded you up.

Replace "company's debt" with "house" and "they don't repay it" with "it burns down".

You only gain the money from the payout if the first event (which is a loss) happens. And you're down by the amount of the premium anyway.

Re:enough lies please

[UnknowingFool]

There were two major problems with calling credit default swaps as "insurance". First of all by regulation, insurance policies have to be backed up by real assets. If it was real insurance, the company which insured has to have a portion of the future settlement available for payment. As such, the insurance company has an interest to make sure that the policy is a good policy. CDS had no financial backing other than the name and reputation of the company issuing them. Second, insurance is normally issued by companies as a contract directly between two parties. ABC Insurance sells to XYZ mortgage company insurance covering their risks. In the case of credit default swaps, Acme Trading Co or anyone can bet for or against XYZ by buying from ABC.

Re:enough lies please

[JamesP]

Yes, CDS (or any kind of swap or derivative) are not insurance in the 'strict sense' of the word, like car insurance.

But everything that's bought/acquired/engaged to minimize a risk can be called insurance.

Of course, if it works as intended is a different problem altogether.

Re:enough lies please

[UnknowingFool]

But that's the thing. If CDS is not insurance, then what is "insurance" is lost if people use that term to describe things that it is not. It was called insurance as a ploy to allow buyers to think that what they were purchasing was safer than it was. CDS are side bets; they are hedges. As such they are very risky.

Re:enough lies please

[JamesP]

They are using the word 'insurance' because it's the word most likely understandable by the public.

I'm not sure CDSs are sold 'directly to consumers' to qualify for a certain regulation, I'm guessing they're sold as part of funds, etc

CDS are side bets; they are hedges. As such they are very risky

Well, they're more risky for who's selling them, as opposed to who is buying them.

Re:enough lies please

[EdgeyEdgey]

CDS had no financial backing other than the name and reputation of the company issuing them.

No, a CDS contract usually requires collateral to be deposited equal to the change in the value of the contract. This way the reputation of the issuer is irrelevant. This was the downfall of AIG as they didn't account for all the collateral they had to post when lehman & other banks went south. P.s. In the case of CDO's the problem was working out the current value of the contract, nothing to do with collateral.

Re:enough lies please

[UnknowingFool]

As a derivative, CDS are really under no regulations at this time. They were sold under hedge funds but those in themselves are derivatives. They are supposed to be regulated under Commodities Futures Trading Commission. In the 1996s the head of CTFC, Brookely Born started learning about OTC (over the counter derivatives) which were essentially unregulated. Under the mandate given by Congress, her agency was supposed to regulate them. She got enormous push back from practically all the major players including Fed Chairman Alan Greenspan against any sort of regulation. What got her attention was not so much the market wished to be unregulated was that they fought furiously against any disclosure about big the market was.

Greenspan and the banks, pressured Congress to essentially gut her agency of any regulatory power. Still she warned Congress back then that it had power to take down the financial markets as the banks were not required to disclose how much money they owed or to whom they owed money. In 2008, she was proven right when one after one investment banks like Lehman Brothers and Bear Stearns started to collapse due to over-leveraging.

Because they are unregulated those that invested in them essentially had no idea how sound the purchase was. At the time of their collapse, Bear Stearns were leveraged $13.40 trillion dollars in derivatives but only $395 billion in assets. It was only after their fire sale to JP Morgan that saved them. Since disclosure was not required, how many funds and companies had money in Bear Stearns and Lehman Brothers that they thought were "insured" against risk.

Re:enough lies please

[khallow]

there's a strong moral risk: if you can both buy a company's debt, and insurance in case they don't repay it, why not keep doing that endlessly ?

So what? The moral hazard in insurance is that someone is encouraged to make riskier decisions. So it's a problem - for the insurer not for the insuree. If that hazard turns out to be a bad problem, then the solution is to either raise the price on the insurance or not offer it at all.

Everyone remains interested in the risk of the debt because it affects the cost of the insurance and the chances that the insurer will default or other harmful events happen.

Re:enough lies please

[jschlesinger]

The problem with CDSs is that the are effectively insurance contracts, but without the limitation that you have to have an interest. This rule was added to insurance back in the 18th century. Without the rule, you can take insurance out on your neighbour's house and then arrange for it to be burned down.

That is exactly what some of the more ruthless financial services companies were doing. It is why Goldman Sachs paid out $550m in a fine and probably breathed a sigh of relief at how small it was.

If the CTFC had been allowed to regulate CDSs as insurance contracts, we would most likely not have had the credit crunch.

Re:enough lies please

[obarthelemy]

who is "everyone" ? everyone is only interested in how much money they're making for themselves, and their bonuses are pretty much instantaneous, whereas the risk can manifest years later.

the individuals who brokered those deals made millions, even though they cost billions to taxpayers in the end. their bosses didn't come off too badly either. their companies/shareholders, a bit worse sometimes, especially for those who were not propped up.

Re:enough lies please

[UnknowingFool]

Reputation matters if the there is not backing. For insurance if you buy it today, the policy is collectable as soon as the insurance company gets the payment. For the investment banks like Bear Stearns they had derivatives that were liable up to 13.4 trillion but only had assets of 395 billion or about 36:1 ratio. AIG is mostly an insurance company; their securities division over-leveraged their assets. Reputation matters as these derivatives were purchased from companies they thought were trustworthy. Yes they should have done more research but in the unregulated derivatives market, disclosure is not mandatory.

funny you mention potatos - see NYMEX

[decora]

there is a great book that just came out, The Asylum, by Leah McGrath Goodman , which explains how the potato market became a cluster fuck of manipulation and greedy assholes absolutely stealing from the ordinary person.

it also explains why certain industries were banned from trading this shit. why? because you cant operate a society where the price of basic commodities fluctuates by several hundred percent a year just so that a handful of a few dozen traders can make massive amounts of money through manipulating the market.

Re:funny you mention potatos - see NYMEX

[TheRaven64]

Hopefully the book also mentioned the fact that Goldman Sachs got an under-the-counter exemption from the normal rules that prevent out-of-control speculation on commodities...

how much cocaine do traders use?

[decora]

and how many prostitutes do they kill, on average, per year?

Re:how much cocaine do traders use?

[JAlexoi]

I even think that narcotics trade is a "bit" more profitable with a little less stress. Though a bit more legal issues.

Re:how much cocaine do traders use?

[sgt scrub]

I agree they contribute most of the same things to society but I don't know about fewer legal issues.

stock+trader+arrested == About 184,000 results

Re:how much cocaine do traders use?

[gninja]

programmer+arrested: 3 million results

Who likes to be screamed at

[Col. Klink (retired)]

I interviewed once for a trading firm. They were honest at the interview and told me that someone would probably scream at me for something that wasn't my fault. I declined.

I had another friend who did take a job up there (different firm and not a programmer). She was ready to quit and was complaining about how mean they were. I tried to console her and told her she had no way of knowing what it would be like. Then she told me that they actually made her cry during the interview but she took the job anyways.

Re:Who likes to be screamed at

[Anonymous Coward]

People scream at other people for no good reason all the time. Thickening your skin up a little can only be good for you.

Re:Who likes to be screamed at

[Anonymous Coward]

So it sounds like they aren't paying enough, if people are quitting from the working conditions. The money is supposed to make up for that.

Re:Who likes to be screamed at

[jcr]

I've heard stories like this, but I never encountered anyone yelling at me in all the time I spent in the financial industry. This was over several years in NYC and Chicago, at Salomon, UBS/Warburg, JP Morgan, and Phibro energy. The only time anyone ever did yell at me at work, I just turned my back on him, went to my desk, and calmly packed up my stuff while he and his boss were frantically apologizing and begging me to stay.

-jcr

Re:Who likes to be screamed at

[Anonymous Coward]

If you get screamed at during work by your boss, calmly tell him that they are disrupting the work environment and threaten to call the police/security if they can't compose them self in a professional manner.

Re:Who likes to be screamed at

[spire3661]

People dont scream at me, I simply do not allow it. If a professional is screaming at me, he better be prepared to back it up with fists. Im a man, and you will treat me with respect, period. No amount of paycheck will change that.

Re:Who likes to be screamed at

[Anonymous Coward]

What good is an apostrophe if you can't put one where it belongs!

Re:Who likes to be screamed at

[mikael]

That would be true for a lot of industries that are IT related and have "production" pipelines (game/film/TV development, manufacturing). Even help-desk support is going to have that. A help-desk floor would be quiet as a desert for hours. Then someone's network card would fry and jam the whole segment (known as "jabbering"). All hell would hit the fan as every user would call us up in every way possible and ask if we knew the network was down.

Defensive "programming" requires that you have sanity checks on every input that comes your way, log everything that you've done and send the data onwards with timestamps and ID's.

Re:Who likes to be screamed at

[Christian Marks]

Never encountered anyone yelling? Do you work in credit risk? That's more sedate than trading. I sure as hell have had people yell at me at every Wall Street trading desk I've worked at. Here is a small sampling of incidents:

1. 1) at an options trading desk by a future Forbes billionaire and his blonde assistant;
2. 2) at a trading desk at Merrill Lynch when debugging the code of an options trader who was on his honeymoon cruise on the day his options were expiring;
3. 3) at CitiBank by a random trader who angrily demanded that I fix something to do with Word Perfect--I had nothing to do with it.

This doesn't include violent outbursts by traders on the floor. I left Wall Street. I developed tinnitus and could no longer hear the yelling.

Re:Who likes to be screamed at

[jcr]

>Never encountered anyone yelling? Do you work in credit risk?

No, The risk analysis I worked on was more along the lines of balancing the firm's global exposure like being long on a given currency in Zurich, but having that hedged by offsetting positions in Chicago. I also worked on derivatives pricing, soft-dollar accounting, and order clearing.

> a random trader who angrily demanded that I fix something to do with Word Perfect-

Heh. I've have told him to go pound sand up his ass.

-jcr

I am a Google engineer

[Anonymous Coward]

[Posting anonymously since we are talking money, etc.]

I am a senior engineer @ Google. I work regular 40 hour weeks, am rarely on-call and work on projects that really interest me. I work with a great team where I have a lot of fun with my co-workers, enjoy fantastic free meals, great health insurance and numerous other benefits (such the latest and greatest geek toys, and working from home any time I want). And yes, I am sure to make $500K this year.

Maybe you should consider accepting the Google offer?

Re:I am a Google engineer

[Anonymous Coward]

Continuing with this whole anonymous posting trend, I'll do the same thing. I'm currently employed by a Wall Street firm and although I'm not making 500K, my salary would be considered quite high by some standards. I am familiar what kind of skills and work ethics are required in finance IT, and I would second OP's opinion. One thing that I might add, is that HFT programmers are not the only ones earning the big bucks, though in other fields such in-depth knowledge of low-level tech might not be as critical as the understanding of regulations, business procedures, and financial models; and/or the complexity of piles and piles of different not-very-compatible IT systems. While I am naturally very happy with the Wall Street salaries, the CS guy in me would want to work in a company similar to Google. The few things that do stop me from even looking in that field are: 1) accumulated and otherwise useless knowledge, and 2) how much such company would be willing to pay. So, the question to you, Anonymous Google Employee, how common are such salaries at Google?

Re:I am a Google engineer

[Anonymous Coward]

I dont work at google, but I find this site useful sometimes:
http://www.glassdoor.com/GD/Salary/Google-Salaries-E9079.htm?sort.sortType=BP&sort.ascending=false&filter.jobTitleFTS=software

Re:I am a Google engineer

[Anonymous Coward]

$500K+ is not that uncommon (I believe) among engineers at my level (and above of course) @ Google, but of course I do not know the total compensation of other engineers at my level. Bear in mind though that this includes bonus and stock options / grants. Also, there are not that many engineers at this level of seniority at Google.

At the cost of going off in a tangent, I should also mention that sometimes, the perks are not that obvious. I personally have a strong preference for working for a company with strong ethics, and irrespective of what you may have heard from others, as a long time Google employee, I can tell you that Google is the only company I have worked for in 15+ years of my professional career where I see every major decision being considered carefully to ensure that we are doing the Right Thing (TM). The amount of transparency and freedom of expression, especially in engineering, is incredibly high. I value this immensely; the day this stops being the case, me (and many other senior engineers @ Google) will start looking elsewhere.

Re:I am a Google engineer

[The MESMERIC]

Instead of lazying about - how about fixing Google Local?

Huh?

Scrapping bad data and posting it as official.

Making it IMPOSSIBLE for owners to edit/change.

How many other businesses similarly affected? Including Hospitals, Shelters and Emergencies services? Not impressed.

Shame on you Lot !!

Re:I am a Google engineer

[Anonymous Coward]

Ethics @ Google?
They are one of the biggest UK/EU tax dodgers. They don't give back to society. Ever hear about the Double Dutch Sandwich?
I am an employee for one of the biggest American ex-HF/Bank in London, but from a moral perspective I feel Google is as transparent as the banks.
Use your head.
Google are common thieves and are like all the fat cats we love to hate, but really if we look inside things we all know how they can afford to give you free swimming pool or pay you 500K USD per year...
Just don't go around like a peacock as the typical American who likes to show his salary, thinking you really are paid with don't be evil money.
When you know the true reality about all this huge Ponzi scheme which is the global economy, and know Google is part of it, then probably I guess you'd feel a bit ashamed.
What a sad sad world we live in.

Re:I am a Google engineer

[Anonymous Coward]

I'm glad you've got a well-paying job that you enjoy. It's certainly a rarity these days.

But the HFT guy might well make his money in cash (depends on whether he's bank or prop shop), and he might very well have the right to participate in the strategy. His potential upside from this is astronomical. Not sure how that compares to IT jobs.

Re:I am a Google engineer

[Anonymous Coward]

I'm also a Google senior engineer. And if you're making $500K/yr as a senior engineer, Google is screwing me big time. Glassdoor suggests otherwise, though.

Re:I am a Google engineer

[Anonymous Coward]

You work for an advertising company. How can you consider the mental pollution google shovels around the "Right Thing"?

Re:I am a Google engineer

[Anonymous Coward]

Ethics @ Google?

Yes

They are one of the biggest UK/EU tax dodgers. They don't give back to society. Ever hear about the Double Dutch Sandwich?

You mean unlike other ethical companies like Microsoft?

I am an employee for one of the biggest American ex-HF/Bank in London, but from a moral perspective I feel Google is as transparent as the banks.

Then clearly you have never worked @ Google.

Use your head.

Google are common thieves and are like all the fat cats we love to hate, but really if we look inside things we all know how they can afford to give you free swimming pool or pay you 500K USD per year...

I dunno, maybe because we put out high quality products that people actually like using?

Just don't go around like a peacock as the typical American who likes to show his salary, thinking you really are paid with don't be evil money.

Of course, the reason I posted anonymously is because I want to show off my salary. BTW, I am NOT American.

When you know the true reality about all this huge Ponzi scheme which is the global economy, and know Google is part of it, then probably I guess you'd feel a bit ashamed.

This is rich coming from a guy who admits he works for an American financial institution. Clearly, as everyone knows, Google is responsible for the state of the global (and US) economy, and not the American financial institutions .

What a sad sad world we live in.

I agree, which is precisely why I am thankful that I can work for a company like Google. Maybe you should consider not working for an American bank in London?

Re:I am a Google engineer

[Anonymous Coward]

Because we build awesome products? And no, I do not consider making money off advertising evil if that means we can give great products for free to billions of people around the world.

Re:I am a Google engineer

[Anonymous Coward]

I used the word "senior" in a generic sense - I am not actually a Senior SWE if that's what you mean.

Re:I am a Google engineer

[russotto]

They are one of the biggest UK/EU tax dodgers. They don't give back to society. Ever hear about the Double Dutch Sandwich?

The Double Dutch sandwich is kind of funny. It's sort of like a law which says if you go to the tax office the normal way, you pay 20%. But if you take the long route walking backwards while balancing a beach ball on your nose and playing a kazoo, you only pay 3%. So any company with sufficient resources hires a few circus performers to go to the tax office the long way, walking backwards while balancing beach balls on their noses and paying kazoos, pays a flat rate for that, and saves 17%. It's not unethical; the rules are silly, that's all.

Re:I am a Google engineer

[JamesP]

Yes. I can see the Monty Python sketch already in my head

The problem with regulation (and increasing it) is that :
- bigger companies have more resources to fulfill them
- closing a loophole always hurts someone small/well-intentioned
- small business have to deal with a regulation bag of hurt

Solution? I really don't know. But for GEs, Starbucks, etc regulation is only another excuse to increase prices, and quickly brushed off. While small business owners cant afford dressing as a clown bad playing with a seal before paying the taxes.

Re:I am a Google engineer

[Anonymous Coward]

Yes, I get paid all in cash. I make $250k in base salary and this year am expecting about 100% bonus (semi-annual payout). This is my first full year with the firm. The highest bonus i ever made in terms of percent of salary was about 250%, but my base was lower since I was relatively new to the field at the time (3rd year). I'm probably above-average in my pay for my group, but not even close to the top. Being lead and getting a Director level title would get me up close to the $1,000,000 mark. Behind head of the group and you are talking some serious cash.

Re:I am a Google engineer

[JavaRob]

Right; the bottom line is that there are known loopholes that all of the major international corporations use to avoid taxes... and the governments could certainly close the loopholes, if the corporations didn't have such massive influence over the entire political process.

So every once in a while there's a big "exposé" of one loophole or other, and various politicians start bills which are all destined to die or be completely neutered -- as hoped even by the politicians flogging them, because of course if you're the one who successfully pushes through the law that closes a serious loophole, you're screwed.

Re:I am a Google engineer

[JavaRob]

WRT Google -- the moral question of whether to use the loopholes is not completely clear-cut. The loopholes are perfectly legal, and all of their competitors use them. I suspect they could be sued by stockholders if they decided to optionally pay far more taxes than they were legally required to pay.

There's also the question of what the taxes will actually be used for -- this isn't so bad with EU taxes, but the biggest chunk of any US taxes you pay go straight to the military.

Morally, the best solution might be for Google to publicly post the amount of taxes they are not paying in different locations thanks to legal loopholes (thus putting pressure on governments to actually close them), NOT lobby for keeping the loopholes open, and to use at least some portion of the "loophole money" to do some direct good in the regions affected.

Re:I am a Google engineer

[horza]

Wow that was one of the most bitter unsubstantiated sad rants I've seen in a long time. And it comes from some banker waiting for his anticipated bonus on top, for +1 Hypocrite. Unlike yourself, Google actually adds value to society in many ways. What a nutter.

Phillip.

Re:I am a Google engineer

[Anonymous Coward]

So you claimed to have a very specific job title at Google and are now falling back on claiming that you just used the title 'senior engineer' generically?

I think you are making stuff up. :)

HTF programmers, this is YOUR fault

[Anonymous Coward]

You people enable this. There is zero difference between the arms dealer who arms the cartels who then shoot up an innocent town, and you, the programmer, who craft weapons of light and electricity and plastic and silicon and then give them to these billionaire psychopaths. You are not only part of the problem, you are the problem's enablers.

Stop. Now. Stop thinking of the measly 500K you make doing the grunt work that allows these parasites to vampire-suck hundreds of billions of dollars from "we the people." They are laughing at you. They routinely steal more than that in an afternoon using your work. And you could bring this whole thing to a screeching halt if you just. Stop. Giving. Them. Weapons.

Why don't you?

Re:HTF programmers, this is YOUR fault

[Anonymous Coward]

um, I essentially risked my entire savings to cofound an algo fund. It's been well worth the risk. I don't think I've ever grown so much intellectually due to the intricacies involved in combining machine learning and predictive systems with just as important OS tuning of low-latency systems , and emotionally from dealing the the stresses and problems that arise out of nowhere when trying to start a business.

I don't consider myself a a billionaire psychopath parasite at all. Am I really that much of a dick?

Honesty, where is this moral judgement and hatred coming from? What jobs do you list on your "morally acceptable list" and how exactly did you come up with it? Are you just be jealous that you don't have the breadth of social and technical skills for what it takes to do work in the small financial subsector (by employees, not trades) of HFT.

If you've got an issue with buyers or sellers in the market and think it is unfair, then please, by all means, DON'T INVEST. There are any number of much less risky means to make a capital investment.

How sad

[Anonymous Coward]

Such talented people doing such silly and meaningless work

Re:How sad

[Anonymous Coward]

This is a common refrain in this thread. I moved from marketing into finance. Who is worse? All the programmers at Google working on convincing consumers to buy more shit? Or the HFT programmer? What about burger flippers contributing to the obesity epidemic? You want to rail on them too?

Good Choice! Stay out of NYC or get out if you can

[JakFrost]

Lived in NYC, worked there in Fortune 10 (yes ten) companies as a Sys Admin, got out when I got a chance and a nice offer far-far away in a land of sanity, relaxation, no state taxes, affordable housing, and am never coming back to NYC. The quality of life anywhere else is so much better in all the ways than the daily grind in NYC no matter how much they pay you. The disgusting NYC subway system shows the true underbelly of the city.

NYC financial recruiters (aka head hunters) and companies always get their talent (aka willing slaves) on short term 6-12 month contracts and never or hardly ever as employees unless they are doing a direct transfer from another financial company and want to retain their corporate title and position Associate, AVP, VP, Director, Managing Director, etc. Their favorite ploy is to dangle conversion to full-time employment and eligibility for bonuses as a way to whip the consultants and contractors to work long overtime hours on straight-time (aka no overtime pay).

After years of being away from NYC these recruiters still contact me with contract offers for NYC and I just keep telling them to take a flying fuck off a bridge and remove me from their contacts database.

Been there, done that, never again, it isn't worth any kind of money to live in the technology ghetto that is NYC.

I work for a financial firm, but not HFT

[Anonymous Coward]

I work in a financial firm, but not in the HFT department, so while my salary isn't at some of the levels mentioned, I'm doing pretty well. I don't code really fancy math and I'm not determining prices or anything, but what think is the critical skill in my group is being able to think on your feet when people are calling you about something gone wrong with or how to pull off for the first time TODAY a deal valued at north of $50M, and being able to pull it off on a regular basis and with confidence. There is constantly lots of pressure and dealing with people who are under even more pressure. You have to know the systems involved so well that you can make changes to code which moves millions of dollars around, test them, then release them right away and still sleep at night. I've dealt with lots of other programming shops who seem terrified to actually release code, and that just won't fly at a financial trading firm. You get used to a lot of responsibility and performing under pressure without a net. It takes a lot of confidence and experience. You also have to have the business sense that when a situation comes up that the code doesn't handle and the business types tell you not to work on it because they've decided not to ever do that again, you make sure it's the next thing you get done because it'll get mentioned again someday, but on that day it'll be a high energy financial type calls you and tells you the go/no go decision has to be made within the next 30 minutes for that day.

You are also dealing with data files received from firms all over the world, and you'd be amazed at how primitive some of those firms are and their files just make random changes from time to time and you've got to fix for those changes NOW and get things running again, not an hour from now.

So it's about the ability to work with confidence, under stress, and be able to maintain intimate knowledge of broad systems which must deal with all sorts of real world exceptional cases.

I know this isn't going to sound all that special to people not in that situation, but it really does raise the stakes a lot.

We need Another Word For This

[virb67]

The last I read, 70% of all volume in the NYSE is HFT. We need another word for this because it is not capitalism. It is taking advantage of quarks in the system. Everyone knows these quarks exists and that they're distorting the system, yet these quarks are deliberately left as is.

Re:We need Another Word For This

[Lanteran]

If I weren't so tired I'd make a lame physics joke about quarks. But I am, so you can go about your business. Move along.

Re:We need Another Word For This

[Anonymous Coward]

You keep using that word. I do not think it means what you think it means.

Re:We need Another Word For This

[Anonymous Coward]

Inconceivable!

Re:We need Another Word For This

[Anonymous Coward]

Your words betray your ignorance of market making (HFT) and the services it provides. In a capitalist system, goods and services are privately owned and traded for profit in competitive markets. Markets are competitive when the forces of supply and demand are well balanced.

Market makers bring competition to the financial markets, to the considerable benefit of producers and consumers (individual accounts looking to sell or buy stocks, options, etc.). For any given security on any given exchange, there is a best bid offer and a best ask price. These represent the highest price someone is currently willing to pay for a security and the lowest price at which someone is currently willing to sell a security. In a competitive market, the difference between these prices (the spread) will be low, e.g., a fraction of a percent of the total value. In less competitive markets, the spreads widen, perhaps as high as 5% of the total value. Market makers (high frequency traders) are able to keep the spreads low because their speed allows them to manage their risk very effectively. They can acquire a security at a low spread because when the opportunity to sell that security arises, they have the speed necessary secure that sale quickly--before the market price has a chance to drop in response to the shift in supply in demand. And when they sell, they must sell at a low spread, or else they will lose out to another market maker who is willing to sell at a lower spread. This is true, uninhibited competition.

Moreover, it is not uncommon for a market maker to make no money at all off a trade. How is this possible? Well, for every transaction, there are actually three parties involved: the buyer, the seller, and the exchange. The same security might be traded on multiple exchanges, and some of those exchanges will execute more trades than others. On exchanges where the number of buyers and sellers is well balanced, the spreads are optimal. On exchanges where there are more buyers than sellers, the sellers have the advantage because the buyers must compete amongst themselves for the buy opportunity. On exchanges where there are more sellers than buyers, the buyers have an advantage because there are more would-be sellers than there are opportunities to sell. Market makers bring balance to the exchanges by providing liquidity where it is otherwise lacking; they make it possible to buy or sell at a more competitive price. When the sellers outnumber the buyers on a given exchange, market makers will join the pool of buyers and then turn around and sell on exchanges where the buyers outnumber the sellers. This helps the other buyers on the first exchange to buy at a better price while helping the sellers on the second exchange sell at a better price. They often do this without actually profiting from their positions; they can do this because the exchanges are willing to compensate them for providing liquidity. Nobody wants to trade on an exchange with poor liquidity, so the exchanges are willing to pay market makers to pick up the slack, lest their customers decide to go trade on other exchanges instead.

I defy you to explain how this is not capitalism at work.

As to your comment that 70% of all volume on the NYSE is HFT, perhaps now you have a better understanding of what that means, i.e., it means (assuming the figure is accurate) that for 70% of transactions, one party is a market maker. This, in and of itself, means nothing except that a market maker was able to engage an investor in a transaction which that investor wanted to execute. Market makers are classified differently than the exchanges' customers (e.g. individual investors, "old school" traders, etc.), so their ability to harm individual traders (or any non-market makers) is quite limited. Customers will be given priority status over market makers in any transaction. Market makers compete amongst themselves for the opportunity to do business with those customers, and in return those customers are able to trade in a more competitive market.

Personal programmers for those who need performnc?

[Anonymous Coward]

Are the HFT systems just REALLY high performance computer systems for people that REALLY want top rate performance, and have VERY LARGE amounts of money at stake? If FAR smaller amounts of money were at stake, could trading be done on old XP computers, programmed by secretaries that learned Visual Basic on the side?

I have marveled how fast 2011 technology can be.

If the resources for a HFT system were put into making an original IBM PC Word Processor run as fast as possible, how much faster could it be? A thousand times faster? a million times? A billion? I imagine a human that thinks a million times faster than a normal human, typing away.

Re:Linux by default?

[ralphdaugherty]

POSIX: Portable Operating System Interface for UNIX.

Linux tries to keep pretty close to to it. 'nuff said.

What's your point? OS/400 - i5/OS - i operating system has been POSIX compliant for many years.

And another HFT employee

[Anonymous Coward]

I've worked in HFT and I can say, most just feel privileged with their income. Most that are trading without these tools know better, but do it anyhow. Investing your pension on the stock market is either very smart, or very stupid. It has nothing to do with how solid a company is, but everything with how HFT reacts and what people's sentiments are. If you know how it works, you can make money without the speed, but slower and with more risk. That's what a lot of HFT workers do on the side.

Re:And another HFT employee

[Akima]

What about you? How do you feel about doing this? You can't really know what other HFT programmers feel.

Re:And another HFT employee

[aliquis]

Investing your pension on the stock market is either very smart, or very stupid. It has nothing to do with how solid a company is, but everything with how HFT reacts and what people's sentiments are.

Say what?

If you do your pension saving for two weeks or what?

Can't see how a comment this stupid can be coming from someone who work with HFT.

HFT doesn't matter at all for someone doing pension savings.

You save for 40 years and HFT matters how during those 40 years?

Not many pension savers are day trading their pension ..

Re:And another HFT employee

[Rich0]

Uh, the pension saver doesn't choose how it is invested at all - this might be a language issue.

In the US the word pension refers to a fund set up by an employer to pay the employee retirement benefits at some point in the future. It isn't an employee-owned account, and the employee has very little control over how it is invested.

And in the US, employers invest their pensions in hedge funds all the time. And where do you think those funds make their money? Hint - it isn't from building factories and selling widgets.

Re:And another HFT employee

[aliquis]

So is it traded? Or you mean to keep some cash within the fund for payments?

Here in Sweden we got one part where you choose your funds, that wouldn't lead to multiple buys and sells unless you want to?

There is one other part to though, I don't know how that one work.

If it suck I guess the system should be scrapped and people get to choose how they want to invest their pension money themselves.

I do understand that hedge funds must have someone else to play with (or well, must and must, but to be of any use) and that there will probably be some of them who are successful and some who aren't.

One would still be free to choose whatever one wanted to invest in a hedge fund or not.

You mean the HFT algorithms ride the selling and buying of the hedge funds? Or you mean they gain from their 20% or so of the gains of the hedge funds?

For pension savings buy and hold with re-invested dividends probably work quite well? At least if you got decades and move some over to interests the last 5-10 years before pension.

Re:And another HFT employee

[Rich0]

So, in the US the way a pension basically works is:

1. Employer promises to pay employee some amount of money when they retire.
2. Employer hopefully makes that payment.

Now, that was how it worked until a while back when companies started defaulting on those promises and employees had no recourse (spend 30 years working for a company that goes bankrupt to find out a big part of your total compensation is gone). So, now companies have to actually set aside some amount of money in some kind of fund to try to cover some of these obligations. The problem is that the regulations are pretty weak.

Now, when creating a pension fund a company can either set up a fund that makes a low, safe return, or a high, risky return. Well, the company isn't the one losing their shirt if the investment goes south, but if the return is low the company will be putting a lot more of its own money into the fund. So, companies tend to make risky investments with pension funds.

Now, private retirement accounts do exist in the US. They take the form of IRAs, 401ks, and 403bs, which are all tax-deferred accounts. In these cases the employee actually owns the money and decides (sometimes within limits) how it gets invested. There are usually tax penalties for taking money out early, but the money belongs to the employee and they can cash the whole thing out any time they want to. These tend to be invested in more sane investments, since the employee actually has a stake in the outcome and usually it is illegal for individual of ordinary means to invest in the kinds of financial games that make people millionaires or broke overnight..

Wall Street

[br00tus]

I worked for a Fortune 100 financial company in Manhattan. Some of the nastiest people I ever worked with were there. I also never worked anywhere where the company made more clear to me that I was a disposable cog. Not that that isn't the case elsewhere, but management usually tries to at least conceal it. I worked 60 hours a week, including weekends, and I was on the low end compared to others who were working to move up in the hierarchy.

There was a limited and fixed bonus pool, so if you got less money, others got more. I'm sure this was a plan by management - the firm was so wealthy, they felt it better their workers be divided against one another than working together. They did this in a variety of ways - staff versus contractors, contractor firm versus rival contractor firm and so forth. This encouraged people trying to rip one another apart during weekly meetings, code reviews and the like.

The office politics can be strange too. I used to be on conference calls with a programmer with a huge ego, and who people deferred to because the program suite he wrote was important for the firm. The program was shoddy though, it had massive memory leaks and the like. But he and the team under him were able to throw it together quickly and the business people were happy with him so he was a golden boy. It wouldn't have been so bad if he wasn't always denigrating everyone else's work with a holier-than-thou attitude. I wasn't in a position to say anything though. Lots of stuff like this happens. A lot.

It was not all bad. It was a large environment. Stuff I would have done maybe once a year at a smaller company I was doing every day there. I was surrounded by dozens of people who were sharp and knew what they were doing. There was a feeling of camaraderie among some of us. It is hard to explain the change in quality due to the sheer size of the company, with its ability to spend massively if needed.

One thing I will say - unless you are there to work 24/7/365 and try to make it to be one of these "highest paid programmers", there is no reason to be there. Some of my co-workers joined straight of college, which seemed dumb to me. Anyone who takes a job for less than six figures is a fool - if you don't have the skills to make at least $100k, there is no reason to work there. It is not the place to come in at a low level and hang around.

To repeat a point - anyone who takes a job on Wall Street right out of college is a fool. They recruit heavily on campuses for the same reason Microsoft and Intel and Electronic Arts do: so they can take advantage of suckers with no work experience who don't know any better. Thankfully I didn't make this mistake. I was able to put everything I saw there into perspective. Going from college straight to Wall Street as a programmer is a dumb move for most people.

Re:Wall Street

[Anonymous Coward]

Yea, but what do you think about someone going to work on Wall Street right after college?

Re:Wall Street

[Seth024]

So what do you recommend people finishing college look for in a job? What did you do?

Re:Wall Street

[Anonymous Coward]

Wall Street Project Manager here. Parent is dead on, especially about the campus recruiting.

I inherited my application and on the team is this one really, really bright young programmer who they got out of Stony Brook.

Dude has been here for 5 years and is making $76K. Lives in a 2 room apt with his gal in Bridgewater, NJ these days and commutes into the city every day. He's a brilliant young programmer, but too dumb to leave. I've tried to get him raised.. fighting for the guy. My senior mgt just sees him as a tool, and wont give me any coin for the kid.

It sucks. Really, really sucks.

Re:Wall Street

[unsolicited]

I've seen only 2 types of attitudes in Wall street
1. make money for the firm
2. save money for the firm

Re:Wall Street

[Anonymous Coward]

Ah, but look at the other side of it...

I used to fit your profile perfectly... I was a "really bright young programmer who they got out of Stony Brook." (Yep, SUNY SB class of '03) I was hired by a hedge fund within one of the big wall street firms for a measly $45k... Five years later, I was making $90k... Now eight years out of college, I'm writing portfolio management software for a hedge fund, and making just shy of $200k... I only work 40-45 hours a week with a 20-minute commute to midtown, and though it can get stressful at times, it's not overwhelmingly so.

So taking a wall street job can go either way... If you have the skills AND find the right place to showcase them (Hint: Work on the trading desk, not in the IT cube farm), you can do pretty well, while still writing actual code (One of the few places you can make a decent paycheck without going into management. If you take a wrong turn, then yeah, you end up like the guy in your example, and eventually give up and leave the industry.

The only real downside is the industry itself... Not much job satisfaction, since it's all just about money... Money money money... No direct contribution to society.

Re:How long to programmers stay in that type of jo

[Anonymous Coward]

The housing costs are the big hurdle to get over. As much as everyone bitches about the overall cost of living in NYC, it's nowhere near as bad as I expected--it's mostly just real estate that's out of control.

I live in an apartment in Manhattan, but most of my colleagues with families live in upscale suburbs of New Jersey. IIRC, the guys with a wife and two or three kids ended up paying somewhere between $750k and $950k for their homes. Their commutes are probably around 90 minutes total (each way). Since I'm young and single, I would rather pay $3500/month for a fantastic view and a 12 minute walk to the office. The savings from not owning a car are also pretty substantial.

Quality of life

[Anonymous Coward]

A high salary is one thing, but that doesn't factor in quality of life. Odds are it's quite low.

Re:I am a truck driver

[Anonymous Coward]

I left IT to drive a big truck. I work only 168 days in a year. I make $170,000

Waste of talent

[Anonymous Coward]

Sad to see such talent being wasted in an area that brings society at large no benefit.

It's gambling - not capitalism.

Everyone is on the hook the minute the house of cards fail, but only a small few ever reap the short term profits.

Re:How long to programmers stay in that type of jo

[jcr]

Any recruiter pitching a $100K job in the banking industry is not worth talking to. In my first wall street gig back in 1990, I was making $100/hour writing options pricing apps on NeXTSTEP, with barely six months of hands-on experience on the platform, and no finance or accounting background at all.

-jcr

Yet Another HFT Article

[hsk17]

Disclaimer: I work at an HFT firm.

The implied accusations are flying out of the page like daggers. I wish you, Slashdot readers, could see the world through my eyes. As techno-savvy as you are, you somehow love to hate on HFT without having any idea what it is. Don't get me wrong -- I really don't care if you hate it. What bothers me is that haters have NO IDEA what HFT is doing and basically project their hatred for finance onto it.

I have to say, this article is pretty level-headed. I was expecting more baseless accusations. Of course, the article throws around the typical "HFT was blamed for the huge drop in the stock market in May 2010..." If you cared to look at the linked WSJ article, you would have read that Waddell's desk had sold 75,000 E-mini contracts at the start of the flash crash. If you cared to look at the CFTC report that officially investigated the flash crash on May 6th, you would have read that CFTC blames the flash crash on some trader who executed a large sell order worth $4.1 billion dollars -- why, isn't that just about 75,000 E-minis?

You would have also read that HFT firms actually mitigated Waddell's mistake. They were there to absorb the thousands of E-minis and so dramatically lessened the initial impact. It's really quite admirable the amount of precision coders needed to invoke in order to create a system that executes so quickly and at scale during such a turbulent period. I was hoping that the discourse here would be more along those lines.

Unfortunately, the amount of volume that Waddell executed was too much risk for the traders to bear, so they started getting out of their positions. In fact, no one could handle a trade of such size. It was as if someone predicted the collapse of the US economy and bet $4.1 billion on it. The ensuing chaos was purely the after-effects of the initial destruction caused by Waddell.

New technologies can be used for bad. I bet there's plenty of bad traders manipulating the markets and using speed as an unfair advantage. We need to police HFT, for sure. But I'm also sure that people are using guns to kill other people out of malice, using cars to traffick illegal drugs, and using airplanes to destroy buildings. HFT is a style of trading. It's a technique, not a strategy. The sooner we realize this, the more progress we will make as a society in implementing policies and regulations.

You guys all hate on HFT, but you are really the ones benefitting from this technology. In market making there's a spread -- the difference between the lowest ask price and the highest bid price. Take a look at the most liquid stocks. They are probably trading at 1 cent wide spreads. Compare that to years ago when spreads used to be dollars. Go to a bank and look at the currency exchange rates. I just did a look into Bank of America's spreads. 100 euros gets you 135.35 dollars. Based on recent trading prices on the public exchanges, 100 euros should be able to fetch closer to 143.62 dollars. BoA is charging a spread that is more that 5% of the value of the product! I don't blame them -- the landscape in the currency markets discourages technological innovation and competition.

This phenomenon isn't true just for currencies. It's true for most products that are not regulated or traded publicly. You, the average investor, are being ripped off dearly investing into these opaque markets. The size of spreads is truly a symbol of capitalism. If there's competition, the spreads are tight. If there's monopoly, the spreads are wide.

You complain about HFT being super fast and "shaving off transactions" as if we somehow have access to your accounts and embezzle money a la Office Space. That's like complaining about WalMart having such efficient systems and internal logistics that your cereal is getting too cheap. Yes, I do believe that some traders use speed unfairly, and yes, WalMart probably did shady things we don't know about, but my point is that not every trader is bad. Technology

Re:Yet Another HFT Article

[sxpert]

and what's the point of all this, making the rest of the people super poor and unable to feed their kids... thanks but no thanks, go to hell you and your ilk of bastards. you should all receive capital punishment for what you're doing

Re:Yet Another HFT Article

[Anonymous Coward]

Disclaimer: I work at an HFT firm.

Tell me, did you read anything past that sentence? Judging by your response, no.

Re:Yet Another HFT Article

[hsk17]

Thanks for reading the post, instead of just the first line. The food people are buying to feed their kids is probably cheaper because of HFTs.

Re:Yet Another HFT Article

[Anonymous Coward]

Trading is a zero-sum game.

No, it is not.

Re:Yet Another HFT Article

[benhattman]

Sorry, but I don't buy it. I posted something similar above, but please explain how pushing transaction speed from 50 microseconds to 44 microseconds actually benefits any normal person. I honestly don't believe it does. But, the HFT who has that faster algorithm becomes tremendously wealthy while the 10% slower algorithm is put out of business.

You can state that you are providing liquidity to the world, which you are, but compared to the liquidity we had perhaps 10 years ago, I really don't see how you're work makes anyone's life better except perhaps your own. And it all brings with it the risk that my own stock holdings may be decimated in literally the blink of an eye. The fact is, your line of work is looked down on because you are essentially gaming a system. Comparing it to cars used in traffic is not accurate. HFT is more similar to the people who used their credit cards to buy $1 coins for the frequent flier miles, and then deposited them in the bank.

http://www.creditcards.com/credit-card-news/mint-closes-loophole-ends-credit-card-coin-sales-frequent-flyer-flier-miles-1263.php

Re:Yet Another HFT Article

[Anonymous Coward]

Keep rationalizing this to yourself, so you can conscientiously sleep at night.

Re:Yet Another HFT Article

[Xyrus]

I don't think people here have a problem with HFT technology or understanding it. It's how it is being used (and abused) that make people angry.

The anger directed at HFT is really misplaced. It should be on the banks and firms who use it, shady practices, and dark markets to bend the world over so they can make more money. I think people are also angry with the fact that these companies continue to rake in the cash and profits, while simultaneously bitching about taxes. A recent report cited companies making considerable profit, and yet unemployment rates are not improving all that much. Isn't it their mouthpieces in congress who claim that lower taxes will spawn job creation?

The anger comes from getting economically ass-raped seven different ways till Sunday, and then having to pay the sociopathic assholes for the honor of being ass-raped. That's why HFT, CDOs, big banks, and financial institutions have been the target of all this venom and hatred. It's not the technology or the ideas or the products. It is what was done with them that pissed people off.

That being said, you don't work for some benevolent altruistic company. You don't work for a company who is concerned about the greater good, or furthering the progress of humanity. You aren't performing some great majestic service for mankind. You work for an HFT company, who's only concern is to make assloads of money. You're upper management could give a fuck less about "balancing the markets" and even less about wiping out grandma's savings and throwing her out in the street. As long as it pulls in profits, they'd be happy kicking babies off the top floor of their 30 story trade tower.

So don't come on here and pretend that HFT is some mechanism intended for the benefit of all. HFT is what it is, and that's a money making tool. The fact that it has some fringe benefits for the rest of us serfs is purely a unintended side effect.

Re:Yet Another HFT Article

[hedwards]

You're full of it. It's well established that HFT is just a continuation of practices going back at least a century to the days when brokers would have access to the next days prices. The delay has shrunk significantly to a fraction of a second, but some of the smaller exchanges still offer that peak at the future prices.

Otherwise HFT would be completely unacceptable in terms of risk. The shorter the holding time the greater the risk. The longer the holding time the lower the risk typically. It's a bit more complicated than that, but pretending like HFT is anything other than a scam really misses the point. Wall Street wouldn't be able to afford to do it if it was on the level.

Re:Yet Another HFT Article

[hsk17]

I don't know why you're so opposed to the idea that there are things machines can do better than humans. Perhaps you can enlighten me on where you are coming from? Your arguments seem to belie a deep-seated dislike for everything HFT.

Anyway, being faster is tangentially helpful to consumers, in the sense that we're competing against each other for market flow. Pepsi and Coca-Cola probably worry about producing 12328 bottles per assembly line per day. Who cares if they produce 12328 or 12329? As a consumer, you just want your bottles. In market making, being faster allows us to manage risk better. We stand publicly so that investors can trade against us. Sometimes, someone will put in a massive order (probably some pension fund) and we will all get huge positions that we need to liquidate quickly. The fund gets a cheap price for all its shares, and we have the difficult task of liquidating it as quickly as possible. That's where our speed comes in. It may not benefit consumers directly, but because we can manage risk better, the spreads are thinner, thicker, and pension funds that manage your 401k can get more volume done at once.

I do agree that the arms race is unfortunate. Many in the HFT community wish the arms race to stop. If you think about it, the only people who benefit from this arms race is the arms producers: the companies that dig holes into the Earth to create a network line faster than any existing line. In order to simply stay in business, HFT companies have to fork up some dough for those lines. Trust me, we hate what this arms race has begun.

Also, please try to separate "market failures" and "HFT evils". The flash crash was caused by faulty technology and an irresponsible trader, NOT HFTs. The exchanges were quoting some products at 1 cent... Waddell bet $4.1 billion dollars that the US economy was going to tank... All of these events added to the events. With technology comes trouble. When cars were first introduced, everyone got worried about safety. Do people still die from cars? Yes, it is very unfortunate. But progress is not going to wait.

If you got this far, I commend you for having the patience to read an opposing viewpoint.

Re:Yet Another HFT Article

[hsk17]

You have it backwards... shorter hold times = smaller risk. Longer hold times = larger risk. Do a thought experiment with a random walk. For any given period of time, there is a measure of expected move (0) and variance. If you hold positions longer, you're more likely to suffer more swings.

Your confusion arises from the fact that there are two problematic things here: market failures and HFT evils. Yes, HFT evils do exist. Flash orders are evil. They allow HFT firms to take a peek at orders before they hit the public exchanges. They should be eliminated. Step ups in options are not fair for similar reasons. They should be eliminated.

But something like the flash crash on May 6 was largely due to market failures. Someone bet that the US economy was going to collapse -- to the tune of $4.1 billion dollars. Some exchanges, for some reason, started quoting products at 1 cent. These failures contributed to the massive market swing.

Market making HFT is a continuation of... designated market makers back in the day. People would manually be assigned the task to provide liquidity for certain companies. Again, not all HFT is good. I'm sure there's HFT firms preying on unfair data and naive customers. But let's not throw the baby out with the bath water.

E-mail is used for the so-called "Nigerian" scams. Guns are used to murder people. Technological progress doesn't wait. The most annoying thing about this whole debate is how people are unaware of the bigger issues at hand. Monopolies in the currency exchanges forcing spreads that are 5% the price of the product. Consumers unknowingly buying illiquid, unregulated derivatives that banks charge exorbitant fees for, only to have those derivatives blow up and destroy the entire market.

Re:Yet Another HFT Article

[hsk17]

You bring up the valid points I was trying to get at. People should recognize HFT for what it is -- just a technique. A way to do something. Not a strategy. People used to deliver mail with cars. Now people use planes as well. The fundamental value-add is the same; the way in which it's being done is different.

You have strong feelings against the financial industry in general. I understand. Enron screwed over thousands of people with their shady tactics and outright manipulation of their finances. Banks used their position and superior knowledge of the markets to sell unsafe products recklessly and brought down the housing market. Unlawful HFT firms started using flash orders to their advantage, peeking at client flow before the information hit the public exchanges.

At the end of the day, every for-profit business is trying to make money. So don't pretend that we are alone in that respect. The only difference is that as consumers, you don't really understand the value of a few pennies being saved over billions of transactions. We may be saving the US economy, in aggregate, millions and millions. For any one person individually, it will be insignificant.

Re:Yet Another HFT Article

[tunapez]

What say you to Mandelbrot's unheeded warnings? I've read some Black/Scholes and frankly I don't buy the 'benefit' outside the ones who are 'printing money' every day. How many disparities in markets can be exploited before the available resources become priced artificially high in all markets? This wealth does not materialize from thin air much like when recessions/depressions hit the money doesn't just disappear.

Have you looked at market prices lately? How did the Dow double in value in less than two years of a sour(to say the least) economy? Why are all commodity markets multiple factors higher than they were 2 years ago, 4 years ago, 8 years ago? Market fluctuations and environmental variables for the past 10 years have been extreme, yet prices for goods consistently 'beat the market'. Go ahead and deny absolutely, be a sophist, I expect nothing less.

Re:Yet Another HFT Article

[hsk17]

I have no idea what kind of conspiracy theories you're talking about. Are you complaining about the rampant speculation and panic in many markets? I don't see how that's related to HFT. I'm not defending speculation, nor am I defending market manipulation.

Most firms do not have enough capital to unilaterally push the market in one way or another. If the market's behaving erratically, it's usually because of the majority of players thinking in a certain way. If the value of the dollar falls, everything will rise in price, especially commodities.

Re:Yet Another HFT Article

[hsk17]

Care to elaborate?

For every trade, there is a buyer and seller. Buyer makes X, the seller loses X (barring fees). The sum of values exchanged is zero.

The reason why the stock market exists is that it's not zero-sum utility. Like insurance, some people are willing to lose money in order to hedge out a risk or move risk to different products. Nonetheless, trading is absolutely zero-sum.

Re:Yet Another HFT Article

[hsk17]

Thanks for the ad hominem. If you care to elaborate on any of the arguments posed, I will be happy to hear it.

Re:Yet Another HFT Article

[soks]

Hsk17 is in fact quite correct and you sir cannot be. There are only two ways to "peak at the future prices." Either your computer is faster than everyone elses, in which case the price reaches your logic/screen faster (if you call this cheating... umm... I don't know what to say to you) OR what some exchanges offer is called "flashing."

First thing about flashing, not all exchanges offer it. In fact America's largest futures/options exchange (think crude oil, corn, carbon credits) does NOT offer any form of flashing. As for the exchanges that do offer flashing, it is simply an added service that any of their clients can sign up for. It is a special high speed connection that, if your firm has the technology to handle it, allows you to see orders as closely as possible to when the exchange receives them. This is not cheating, anyone can do it, you just need to have the funds to support the required technologies. No doubt if you know what you are doing you can get those technologies relatively cheap (likely a special NIC w/ whatever hardware can support it (linux!)).

As for risk being related to holding time, that is just wrong. Risk is generally considered in terms of how far can a price go in a given amount of time. An option contract that is farther out (read: longer potential holding time) will cost more due to the increased chance of volatility (read: price change, up or down, good or bad) through that longer period. Thus more time = more risk, as represented by the higher price of a longer contract.

As for what wallstreet can and can't afford, sure, there are imbalances in our economic system that allows certain industries to work with more money than they should. However imbalance usually comes from the use of force against one group to make them do something they wouldn't normally do, thus giving someone more favor (money) and perhaps creating an entire industry that simply shouldn't exist. I urge you to consider my proposition; Any time you see trading in a brand new high volume industry you should not say "trading is bad, look at this false industry" but you should instead say "what is the industry that utilizes this trading from the buy side." That is, who uses this type of trading for their actual business, as a client, not a trader.

See when an exchange offers traders to trade something like corn futures contracts, that means there are people out there (real, hard working people) who benefit from entering said contracts to mitigate price risk. The ability to mitigate price risk allows seasonally impacted industries to average out their costs and provide more stable prices to their consumers. Once this need for a contractual service exists it is the exchange who goes out to market makers (a type of trader) and says "hey, we have people trading, we need you to guarantee them realistic and competitive prices to trade at." They go out to brokerages (not really considered 'traders' but they do trade) and say "hey, we have products and if you bring us clients you can charge them fees for the complex task of processing, placing, and managing orders on our exchange." Market makers are speculators and there are many other types of speculators of which some may or may not be HFTs. If HFTs were harming the market for other brokers or speculators then the exchange would simply slow down their transaction rate (which would be business suicide) and even this would not stop the HFT market. That is, you will still want to have the fastest reaction time to market events possible and you can likely game transaction rate limiting systems as well.

Either way, it's just competition, and if it interfered the exchanges would not allow it.

Re:Yet Another HFT Article

[Anonymous Coward]

At the end of the day, every for-profit business is trying to make money

At the end of the day, the soldier was just following orders. We've already said that's not a valid excuse for what society considers abhorrent behavior, but the "making money" justification still enjoys some protection.

Really, it's not making money that's a problem. Most of us do it in one way or another. It's the rapacious attempt to maximize profits at any expense, preferably externalized to other outside the firm, that people dislike. Add on top of it the current financial crisis and the unwillingness of those who do make extreme profits to help support the system that allows them to make those profits in relative safety, and perhaps you can just about start to understand where the frustration comes in.

But, you admitted that you work inside (and therefore profit from) the system, so you have a vested interest in not seeing it from other points of view.

Re:Yet Another HFT Article

[soks]

If an algorithm is running 10% slower then it must also be some HFT, albeit a slower one. If they cannot compete anymore that just means an HFT went out of business. Since you don't like them why do you take issue with this?

Any trader that trades in terms of minutes, hours, or longer is not going to be affected by someone who trades in terms of microseconds.

Re:Yet Another HFT Article

[hsk17]

You're trying to paint every person working in finance with the "evil brush". There's more to it than that. You're frustrated with the collapsing housing market. Frustrated with corrupt corporations. Not every financial firm is Enron. Keep that in mind.

Re:Yet Another HFT Article

[Anonymous Coward]

"and that's a money making tool."

And so should be subject to taxation.

AC

Re:Yet Another HFT Article

[Anonymous Coward]

what good or service does your company provide to your customers? who are your customers? what does the fruit of your labors provide to the american economy?

Re:Yet Another HFT Article

[fliptout]

Thanks for your insightful, against the grain posts. I'm not a quant, but I have an rudimentary financial engineering background. I think if more people here were familiar with basic financial concepts, this would have been a more fruitful discussion.

Re:Yet Another HFT Article

[hsk17]

The fast market makers buy and sell lemonades at $1.00 and $1.01 respectively what banks are buying and selling for $0.10 and $5.00. Our customers are everyone that trades with us. We tighten spreads and save institutions millions in execution costs.

Re:Yet Another HFT Article

[Jarik_Tentsu]

God, I don't think I've ever been so damned annoyed at Slashdot as I was reading these comments.

Everyone who is going on about how HFT traders are evil, and not benefiting anyone, or not benefiting society - I really think you should shut the fuck up.

How many of you are working for the pure purposes of 'bettering humanity' and 'feeding the poor'. I imagine very few. You're working to make money.

HFTers doesn't hurt the market as much as people seem to be claiming. I imagine it barely affects any of *you* either. If you're investing long term, then the stock will move based on fundamentals. HFT only really affect stuff in short-medium term.

So one HFT company wants to make money, they develop an algorithm to be better than the other HFT company. They're competing with *each other*.

So all your retards who are saying "HFTers should be shot" - well, I wonder what your programming skills are being used for? Maybe a volunteer project for charity? If so, good on you. But I'd be willing to bet most of you hypocrites are programming for your companies to make them more money.

Re:Yet Another HFT Article

[Slackenerny]

I wish I had mod points for this poster. HFT is just a catch-all term used in the media for lots of different things. It's clear that a lot of people here on Slashdot (which is normally fairly insightful) have no idea what they are talking about in regards to this type of trading, and are just blaming 'HFT traders' for their economic woes. Firstly don't confuse algo and HFT trading. HFT is a type of algorithmic trading, but there are lots of others. There are also lots of algo systems that are not prop systems (trading on the brokers own account), but agency algo systems where they trade for a client instead (like trading VWAP over the day). Secondly, HFT is actually a mis-leading term. What most people mean when they talk about HFT is an algo that recognises and responds to market events in a very quick fashion. This may include executing trades or revising limit orders. Actual HFT (where trades occur multiple times per second) is a small subset again of this. Academic are leaning towards thinking that algo trading increases liquidity (as monitoring costs are decreased) which in turn decreases spreads (as discussed above) which makes trading for everyone cheaper. So basically they think its overall impact is beneficial to the markets. They have found that liqiudity does withdraw in times of market tumoil, but then again who wouldn't (machine or otherwise). But don't let actual research get in way of a good bashing. HFT is bad.....hmmmmmm....okay. However, I do have strong opinions on the risk-reward payoff for traders themselves. There is little downside personally if they lose a lot of money (they lose their job and move to another firm), but there is a lot of upside personally if they make a lot of money. So they take a lot of risk unless monitored correctly. Because they potential rewards outweigh the potential risk. Basically buying a call option. And evaluating these risks is what these guys do every day. No wonder it happens. We need strong monitoring and greater downside risk to offset this. Not going to happen though.

Re:Yet Another HFT Article

[Anonymous Coward]

I have to say, this article is pretty level-headed. I was expecting more baseless accusations. Of course, the article throws around the typical "HFT was blamed for the huge drop in the stock market in May 2010..." If you cared to look at the linked WSJ article, you would have read that Waddell's desk had sold 75,000 E-mini contracts at the start of the flash crash. If you cared to look at the CFTC report that officially investigated the flash crash on May 6th, you would have read that CFTC blames the flash crash on some trader who executed a large sell order worth $4.1 billion dollars -- why, isn't that just about 75,000 E-minis?

If you haven't already, I would encourage you to check out the documentary "Money and Speed: Inside the Black Box" by VPRO. In it, Eric Scott Hunsader of Nanex (a company that specializes in HFT data feeds) shows rather convincingly, that although Waddell was the original seller of the E-minis, Waddell was not responsible for dragging the market down. He goes on, again very convincingly, to show that it was in fact other unsophisticated HFT players that had loaded up on E-minis, and then sold them off in massive bursts that dislocated the market. According to the following marketwatch article: http://www.marketwatch.com/story/sec-cftc-blame-algorithm-for-flash-crash-2010-10-01-1246290 , the SEC-CFTC report reaches the same conlcusion:

High frequency traders and market makers were the likely buyers of the orders. According to the report, these traders accumulated significantly large positions, but between 2:41 and 2:44 they ‘aggressively’ sold these positions, taking liquidity away from the markets.

They also directly quote the SEC-CFTC report that states

"[High Frequency Traders] in the equity markets, who normally both provide and take liquidity as part of their strategies, traded proportionally more as volume increased, and overall were net sellers in the rapidly declining broad market along with most other participants"

So to me it seems rather clear that the consensus is that it was in fact HFT, not a douchebag trader, that was the main culprit causing the flash crash.

Re:Yet Another HFT Article

[Anonymous Coward]

Cool story, bro.

Re:Yet Another HFT Article

[hsk17]

You are correct. If Waddell was the proverbial straw that broke the camel's back, the trading that occurred afterwards was the pack of hyenas feeding on the camel. The flash crash would not have happened without Waddell, but without the rabid selling afterwards, the flash crash would probably have been more of a flash blip. So you do have a point.

However, my point still stands. HFT is just a style of trading, not a strategy. You didn't need to be HFT to short sell E-minis that day. You didn't need to be HFT to be Waddell. You didn't need to be HFT to manipulate price action and take advantage of the panic. Any click trader who recognized the plunge would have sufficed. Those who sold the market did so because they were opportunistic traders, not because they were HFT.

The purpose of my post was that with new technology comes responsibility, and the quicker we realize that HFT is just another piece of technology, the faster we'll get to regulating it properly.

Re:Yet Another HFT Article

[pseudorand]

All of you HFT programmers almost had me with the "we minimize the spread" arguments. But a $0.01 spread x 1000 transactions per second is still $10/sec while a $10 spread on a transaction that happens once a day is a hell of a lot less. I read somewhere that before the great recession, there was a year when the financial sector was 40% of GDP. One way or another, you're skimming a huge and undeserved (as the great recession proves beyond a shadow of a doubt) amount of money off the top.

The keys is to the lies you tell yourself is this:
> Trading is a zero-sum game.

Zero-sum games don't, by definition, contribute to the greater good, and no civilized person would ever play one for any serious stakes. If you do, you've either got a gambling problem or you've figured out your opponent's tell and you've unscrupulous enough to use that knowledge to take advantage of him.

Trading should NOT be a zero sum game. It should involve 4 parties: sellers, buyers, workers and consumers. Sellers are people who need cash to buy something else, like a house or retirement. Buyers are people who have more money than then need to spend right now. Workers are people who produce goods and services for wages. Consumers are the people who benefit from what's produced. That's not a zero sum game. In that game, everyone is a winner. It's honest, hard-working people who produce, consume, save and invest to benefit both themselves and their neighbors.

As stated earlier (see the post about wheat and bread), no one in the non-zero-sum economy needs anything traded at the ms or ns level. Once an hour or even once a day is more than frequent enough. And if that means liquidity goes down, that's fine. The only one who needs that much liquidity are the gamblers. If I'm buying stock, it's because I think it will pay a good dividend and it's value will increase over time. It will be years, not nanoseconds, before I sell it. And if you trick me into buying a bad stock that I know nothing about because, hey, it's only $0.01/share, you can always just sell it again, then shame on me for gambling and shame on you for deception.

And if less frequent transactions means the spread goes up, that's fine too. Because instead of paying someone thousands and thousands of pennies on thousands and thousands of transactions, I'm paying an honest stock broker a few bucks on a few transactions. And I'm getting a service in exchange for that price. I expect my broker is watching out for me. I expect he's a smart guy who knows useful things about the markets that I don't know. I expect he's researched the stocks he's recommending and I expect he'll keep track of them and call me with a sell recommendation if the company's management takes actions he find questionable.

But those type of brokers (if they even exist anymore) now have to contend not just with the fundamentals of an investment, but with the large ups and downs cause by the gambling that HFT allows. That can't give sound advice because they can't possibly predict the randomness you introduce into prices.

Don't get me wrong. As a geek, I'm in awe of your abilities. I have no doubt you're smarter than me. I have no doubt that you can program faster and better than me. And I envy those abilities envy. But don't tell me (or yourself) that anything you do is in any way good for the markets or for society. It's not. You're a bookmaker plain and simple. You're like the clerics who argue over how many angles can dance on a pinhead, siting in an ivory tower, gorging yourself on the best food and wine paid for by tribute extracted from the unwashed masses by lies, manipulation and even brutality. You consume a lot. You contribute nothing. And eventually the rest of us will have to get out the torches and pitchforks and come after you.

Re:Yet Another HFT Article

[hsk17]

All of you HFT programmers almost had me with the "we minimize the spread" arguments. But a $0.01 spread x 1000 transactions per second is still $10/sec while a $10 spread on a transaction that happens once a day is a hell of a lot less.

I think you're confused by what "spread" means. It doesn't mean that for every transaction, we're getting a cut of investor money. Think of it this way.

Say MSFT bid is $99.50 and ask is $100.50. That means you can sell for $99.50 and buy for $100.50. What market makers do is tighten that spread so it becomes $99.75 and $100.25, for example. Now in the first situation, if you wanted to buy 100 shares you would have paid $100.50 per share + broker fees. Let's say broker fees are zero. You just paid $10050. In the second case, you would have paid $10025, saving you 25 dollars. Note that broker fees are the same in both cases.

In reality, the MSFT spread is much tighter thanks to fast market makers, so you would probably pay $100.02 per share, so you would actually save closer to $50 on that transaction. Intense competition among different market makers makes it so that every stock has as tight a spread as possible. Think of any business that's competing with price -- offer products at too high a price and your competition will sell for less, adding value to investors and taking profits for themselves.

I read somewhere that before the great recession, there was a year when the financial sector was 40% of GDP. One way or another, you're skimming a huge and undeserved (as the great recession proves beyond a shadow of a doubt) amount of money off the top.

The keys is to the lies you tell yourself is this:

That finance in general has such a large percentage of GDP is unfortunate. But you're getting off-topic. If you're attacking the entire financial industry, I have no responses for you. Banks, hedge funds, etc. etc. I don't associate myself with all of them. My main point is that high-frequency trading is a technology, and it can be used for good and bad. Example of a good is tighter spreads, which I explained above.

> Trading is a zero-sum game.

Zero-sum games don't, by definition, contribute to the greater good, and no civilized person would ever play one for any serious stakes. If you do, you've either got a gambling problem or you've figured out your opponent's tell and you've unscrupulous enough to use that knowledge to take advantage of him.

Trading should NOT be a zero sum game. It should involve 4 parties: sellers, buyers, workers and consumers. Sellers are people who need cash to buy something else, like a house or retirement. Buyers are people who have more money than then need to spend right now. Workers are people who produce goods and services for wages. Consumers are the people who benefit from what's produced. That's not a zero sum game. In that game, everyone is a winner. It's honest, hard-working people who produce, consume, save and invest to benefit both themselves and their neighbors.

As stated earlier (see the post about wheat and bread), no one in the non-zero-sum economy needs anything traded at the ms or ns level. Once an hour or even once a day is more than frequent enough. And if that means liquidity goes down, that's fine. The only one who needs that much liquidity are the gamblers.

I should have been clearer. By zero-sum, I didn't mean that someone always gets screwed. I meant that someone always loses money, someone always wins money. But they can both win at life. You are very familiar with this. You buy food. You lose money. But you gain calories. The store earns money. Both players win, but the sum of money exchanged is zero sum. That is what I meant. If you actually read beyond my comment to the lines below, you would have understood it as such.

The concept is not so esoteric as you make it out to be. In every trade, someone loses money and someone makes money. However, th

Re:Yet Another HFT Article

[pseudorand]

> I think you're confused by what "spread" means. It doesn't mean that for every transaction, we're getting a cut of investor money. Think of it this way.

> Say MSFT bid is $99.50 and ask is $100.50. That means you can sell for $99.50 and buy for $100.50. What market makers do is tighten that spread so it
> becomes $99.75 and $100.25, for example. Now in the first situation, if you wanted to buy 100 shares you would have paid $100.50 per share + broker fees. Let's
> say broker fees are zero. You just paid $10050. In the second case, you would have paid $10025, saving you 25 dollars. Note that broker fees are the same in both
> cases.

I may be a bit confused, and I'd be grateful if you explain it to me, but that was a piss-poor explanation. The example above assumes that HFT is responsible for reducing the ask. But is that really the case? I'll buy that frequent trading reduces the spread, but on an undervalued stock, it seems that the bid would rise quickly with little effect on the ask. The opposite would be true for an overvalued stock (the ask falls quickly while the bid changes little). So your example where there's this nice meeting in the middle probably doesn't happen very much, does it?

Now I'm sure your next argument will be that this is simply Adam Smith's magical invisible hand working hard to find the "true" price of a stock almost instantly. It's magic because there's no effort on behalf of buyers and sellers needed. In fact, there's no effort on behalf of buyers and sellers even possible because the system moves so quickly that when an stock price is changing there's no time to examine the fundamentals and see if the change is warranted. You have to either hope you were right when you took your position (in which case the price will return) or trade without any true knowledge (which is just gambling).

But isn't the end result that stocks are traded at fair value? If I own MSFT and I think it's still a good value at $100.25 but I need to raise cash, I can get $99.75, most of what I think it's worth. A larger spread allow me to get in and out of MFST at minimal cost (i.e. liquidity). Isn't that a good thing? Yes -- if you're a gambler. If you're decisions to buy and sell are primarily based on how prices are changing, you need liquidity. If you need to get in and out of a position quickly to buy some other stock that you think will move more in price and you can therefore make more money on, you need liquidity. If you make big, risky bet on one stock, then knowing your other stocks can be dumped at pretty close to what you paid for them if you need to cover you losses is quite an advantage.

But if you do fundamental research and buy stocks to share in the profits of good companies that make a profit by making useful goods and services, then liquidity isn't of much use.Volatility, however makes your job much harder, because that good company may loose you money if you happen to buy or sell at the wrong time. And while HFT provides a lot of liquidity (which is only really useful to gamblers) it also causes a lot of volatility.

HFT simply keeps tight spreads on stocks, but because prices can change so quickly, big market players can manipulate the markets with big orders and make money on the volatility while fundamental researchers have no chance of keeping up. Prices change not just because of the realities of the economy but because of trading itself.

So I DO deny that market makers add value by offering liquidity and availability. There's nothing magic about Adam Smith's invisible hand. It should be hard working people who decide where money is best invested through research and deep understanding of markets. But at the moment Adam Smith seems to be using his other hand, the one holding a some playing cards. He's betting big, but he may only have a low pair.

Now we can go back and forth on the benefits and drawbacks of volatility and liquidity, but the the fundamental facts remain:
- The financial sector is both a huge portion of G

Re:Yet Another HFT Article

[Anonymous Coward]

In general, I like that we're arguing about more technical details. As I said, HFT is simply a technology. People may or may not use it in evil ways. I'm also glad that we're not seeing this at the extremes. It's disingenuous for me to claim that markets are completely fair and it's disingenuous for you to claim that market makers add no value to society. It's like gun control. Detractors will say it makes it too easy to commit crimes. Advocates will claim the opposite, saying that a thief who knows that someone may own a gun will be much more hesitant to commit a crime. Who knows, maybe guns should actually be banned and so should HFT. But you have to understand the issues fully before arguing either way.

I may be a bit confused, and I'd be grateful if you explain it to me, but that was a piss-poor explanation. The example above assumes that HFT is responsible for reducing the ask. But is that really the case? I'll buy that frequent trading reduces the spread, but on an undervalued stock, it seems that the bid would rise quickly with little effect on the ask. The opposite would be true for an overvalued stock (the ask falls quickly while the bid changes little). So your example where there's this nice meeting in the middle probably doesn't happen very much, does it?

You're being unfair in this comparison. Let's say the stock is undervalued. Then yes, the inside bid would probably at $100.25 and the inside ask would stay at $100.50. But why do you assume you want to buy? If you wanted to sell, you just got a great price improvement! From $99.50 to $100.25! One of the aspects of an electronic market is that all participants are anonymous, so only through trades and quotes can you make yourself known. Moreover, if you're trading small sizes, you effectively don't exist! So it's unfair for you to assume that somehow, based on your preferences, the market will always screw you. The fact of the matter is, even if the market wanted to screw you, it has no idea what your positions are unless you're a big player.

Now in that case, why shouldn't the market move when you put on a big size? Why should market participants subsidize your big order? When traders get caught in a position because an institution somehow gets earnings data illegally before the public announcement and pushes the market up, they lose a lot of money. Now you may cheer and root for this situation, but then you're being disingenuous. You want fairness only if it applies to you. Maybe traders should subsidize long-term investors. I'm perfectly fine debating the pros and cons of such a system. I have nothing personal against taking money from corporations and giving back to society by law. But to claim that the market should be fair and yet be favorable for long-term investors is disingenuous. Either it's fair for everyone or favors long-term investors and is unfair. Choose your stance.

Now I'm sure your next argument will be that this is simply Adam Smith's magical invisible hand working hard to find the "true" price of a stock almost instantly. It's magic because there's no effort on behalf of buyers and sellers needed. In fact, there's no effort on behalf of buyers and sellers even possible because the system moves so quickly that when an stock price is changing there's no time to examine the fundamentals and see if the change is warranted. You have to either hope you were right when you took your position (in which case the price will return) or trade without any true knowledge (which is just gambling).

You talk as if traders can instantly move prices up and down at the scale that would affect long-term investors. In general, such is not the case. In macroeconomics there is something called a monopoly, in which case, indeed, one company has huge market power to screw everyone over. If there's competition and distributed market share, companies can still adjust prices but it's much harder to do so without losing customers. Market makers are extremely competitive an

Re:Yet Another HFT Article

[pseudorand]

I'm not saying having market makers doesn't add value. I'm saying that market makers using HFT specifically doesn't add any more value than market makers using slower trading and, in fact, causes more problems, lessening or even negating the value of having the market maker. An no, I don't have any empirical evidence of that, nor that HFT causes more volatility. But it seems quite logical, so, having no evidence one way or the other, I'm inclined to believe that HFT is a detriment unless someone presents convincing evidence to the contrary. Why the default position should be "it's okay unless we prove otherwise" is beyond me. It's just like use of antibiotics in feed animals and GMOs. Maybe they're perfectly safe. But I'd prefer to error on the side of caution.

And no fair trying to end my anonymous internet rant my agreeing with me! Don't you know that on the internet you're supposed to disagree just for the sake of being disagreeable to keep things interesting.

Re:

[taiwanjohn]

I agree completely. I don't think it's feasible to outlaw HFT, but I'd settle a STET tax to put a damper on it.

But what useful thing do you do in the market?

[Anonymous Coward]

we really do very normal (and useful!) things in the market.

You elaboratly outlined your qualifications but failed to address the quoted thing above.

I understand what you do is "only cooking with water, as anyone else". But what good and useful thing are there microtransactions to anyone (but those who get rich)?

A question

[Anonymous Coward]

How do you get you head so far up your arse and still appear to be able to breath?

Re:

[Jane Q. Public]

What a great idea! I would have to think about it for a bit, but my first impression is close to love at first sight.

Amen!

[Weezul]

You missed one criticism though, the claim that HFT is a "waste of talent". I'll cover that one for ya :

Is there really such a limited supply of talent in the world? Talent is created by intellect, motivation, and experience. We may ignore experience as it's self creating. If banks hire fresh math & physics PhDs with significant programming experience for $100k plus maybe a $50k bonus, that'll just creates more motivation to study math & physics.

We're therefore left with the claim that intellect is limited supply. Is this true? Yes and no. Intellect isn't in nearly such short supply as people imagine. It's also heavily impacted by people's motivation during their youth.

Conversely, there simply isn't enough money in research for all these people. You've seen these online discussions saying : For God's sakes, please don't do a PhD in the humanities!! If advertising and finance weren't slurping up all the excess scientists, then we'd be seeing the same discouragement levied at PhDs in the sciences. And that'd be extremely bad for our society!

Ideally, we want the quantitative fields and algorithms to take over more of the traditionally "powerful" professions, encouraging young people to think more quantitatively and algorithmically, and pushing our society forward.

As an aside, do you realize that anti-semitism was heavily influenced by Jewish bankers, like the Rothschilds, refusing to finance wars when they had interests in both wars? Ain't necessarily all love & roses for those that help society progress.

Re:Amen!

[sxpert]

"like the Rothschilds, refusing to finance wars when they had interests in both wars?" you, sir are mistaken, and should read http://en.wikipedia.org/wiki/Rothschild_family in more details !

Re:Amen!

[AchilleTalon]

Antisemitism was rather than heavily influenced by a fraudulent document entitled: The Protocols of the Elders of Zion.

The document is still distributed today as a genuine document.

Re:Amen!

[Rich0]

Or, we could just collect trades and have them execute every hour on the hour, with the exact time of execution having a standard deviation of a few minutes and being chosen at random and without advance knowledge of anybody. That makes the value of HFT zero. To address your concern about not having enough people pursue physics you then add a one-cent-per-share tax on every trade and give it to the NSF. Now we not only can afford to pay physicists, but we can actually employ them to study, heaven forbid, physics. The only people who lose out in this arrangement are a few insanely rich trading houses.

Employing people to do useless work is not the solution to unemployment. There are plenty of productive things we can have people do instead...

Re:Amen!

[Weezul]

Yeah, good luck with that, either hourly execution or funding research through taxes on trading. You're rich trading houses won't even lose out, they'll simply switch gears back to hiring Yale MBA assholes with powerful families. Instead, you'll wipe out the middle tear all-quant-algo trading houses who cannot retool. All those houses business will flow back to Goldman-Sacks and Morgan-Stanley, who'll widen the spreads and take an even bigger cut.

You know, I'd love seeing all good scientists get good paying research jobs, but 'power' never allocates resources like that. Do you think the soviet union would've fallen if they had? We're infinitely better off just giving the quantitative fields the economic power directly. And instituting an enormous inheritance tax that bleeds dry the really wealthy families like the Waltons.

Primary skills

[Anonymous Coward]

i) Greed
ii) Sociopathy

Re:I am a truck driver

[outsider007]

Are you heading west? Can I get a ride to Fresno?

Re:I am a truck driver

[Anonymous Coward]

mother of god. do you listen to books on tape?

Re:How long to programmers stay in that type of jo

[loufoque]

For HFT, I get proposals of 200k+ salaries all the time, and not fixed contracts.
Your recruiter wasn't very good.

HFT borderline illegal

[loufoque]

HFT is somewhat illegal, since it could be considered insider trading, since you have time to process information before it becomes available to other people thanks to a faster network infrastructure.

Re:HFT borderline illegal

[PacoSuarez]

HFT is somewhat illegal, since it could be considered insider trading, since you have time to process information before it becomes available to other people thanks to a faster network infrastructure.

You probably would also complain if you were playing soccer against people that run faster than you. It's not fair!

There is no insider trading If everyone in the market has access to the same information. You'll try to process it as quickly as possible and do something as smart as possible with it. If you are not fast enough or smart enough, you don't need to be in this business. Trading is not an amateur activity.

Re:HFT borderline illegal

[hsk17]

You're confusing flash orders, which is indeed illegal -- they allow firms to get a peek at customer orders before the orders hit public exchanges.

But being fast as a competitive advantage, given that everyone receives the same information -- that's something entirely different. By the same argument you're making, Google should be fined because they have infrastructure that's vastly superior to the average person. If the average person wanted to compete with Google, he or she would need to build too many things.

I'm being a bit facetious, obviously, but you have to distinguish these concepts.

Re:HFT borderline illegal

[Anonymous Coward]

When you have a lower latency link between you and the exchange centre than others, the information does physically come to you before it reached the others, which is cheating.

Re:HFT borderline illegal

[hsk17]

You're not barred from purchasing machines that are close to the exchanges. Contact your local exchange liaison and he'd be happy to set it up for you.

Is it an "unfair practice" if it's openly available to anyone who's willing to pay? Is it cheating if the CVS store closer to you gets more business than the Walgreens farther from you?

There are many actual instances where trading firms DO cheat. Educate yourself and you'll have much better arguments. Flash orders, for example. But co-location is not.

Re:HFT borderline illegal

[Rich0]

Soccer is entertainment. The stock market is not. And we're talking about an industry that consumes a huge chunk of our economy on what amounts to gaming a complex financial system.

Oh, and trading as it is practiced today affects the lives of everybody - not just some elite who can afford to get ahead in the game.

Re:HFT borderline illegal

[smellotron]

HFT is somewhat illegal, since it could be considered insider trading, since you have time to process information before it becomes available to other people thanks to a faster network infrastructure.

That is an incorrect definition of insider trading. Don't take my word for it, read the SEC definition of insider trading.

I feel sad

[stevegee58]

I'm married, in my 50's with college age kids. I have knowledge and experience in Linux, C++, signal processing and trading. I have a nerdy interest in Cuda and GPU algorithms.
I'm sad because I could be involved in this exciting field but I'm simply too old to up and relocate to the places where this exciting and interesting is being done. With kids in college, and friends and relatives where I live I'm simply not as mobile as I once was.

Ah, to be 25 and single in this environment knowing what I know!

Re:I feel sad

[tulcod]

Oh please enlighten us. I have knowledge and experience in Linux, C++, signal processing and i know a teensy tiny bit of knowledge of stock markets. I have a nerdy interest in CUDA and GPU algorithms but haven't actually used them yet. I'm single (ugh, aren't we all?), 20 years old and am obtaining a very solid background in physics, electrical engineering, computer science and mathematics. What should I do.

Re:I feel sad

[Anonymous Coward]

Enlighten you? How about you strengthen the areas of nerdy interest and trading and put it on your resume? As long as you're conversant with these things you should be able to have a credible interview (provided you have the software chops of course).

When you're young, it's the best time to take risks like this. Listen to the old guy and picture yourself in 30 years. Where do you want to be?

Re:I feel sad

[mikael]

Try the freelancer websites (freelancer.com) - they outsource work for people to do remotely.

Re:I feel sad

[adamofgreyskull]

Seriously? No, seriously, are you kidding? How can you make any money on those sites? I considered it once, as a web developer, when I was really desperate for work, fresh out of University, but it just wasn't worth it. You're bidding against cowboys and school-kids to the point where you'd be better off working at McDonalds. And I don't just mean pay, I mean the hassle of ill-defined requirements and dealing with people who post shit like "Need a groupon clonne, easy to do for write person, budget $200. No time-wasters plz." Having quickly browsed some of the other categories, they seem to be plagued with exactly the same kind of problems.

Re:I feel sad

[mikael]

Sometimes, there is the occasional bargain if you have the source-code for that project at hand. But for something like an online poker game, if you had what they were looking for, you would make more money have your own site.

You are right, no way is it going to be a long term career. As far as the UK goes, you'd need a minimum of 25K to just survive outside London, and god knows how much to live inside London. Not knowing whether or not you are going to win a contract to feed yourself for the next month isn't practical.

False metrics

[dbIII]

All that and yet your job (and mine in a different field) is ultimately less important than that of a first year nurse, a private soldier or the guy in the control room that operates the line in a large factory or steelworks. Vast amounts of money is streaming through but a trading halt is not going to kill anyone or hold up the production of a physical commodity anywhere. It's not even a 24 hour operation. The "mission critical" bullshit attitude burning people out is from false importance being granted due to using the wrong metrics plus poor management that would rather burn out their staff and wear the inevitable mistakes instead of taking a professional attitude. It costs millions to shut a steelworks down for a few weeks for maintainance but it gets done to avoid expensive failures. A stock exchange which isn't even a 24 hour operation is not magically more sacred, despite egos and a bizzare type of Feudalism insisting it is.

Re:False metrics

[Anonymous Coward]

This has got to be one of the dumbest, socialist egalitarian bullshit I have read on Slashdot so far:

All that and yet your job [] is ultimately less important than that of a first year nurse, a private soldier or the guy in the control room that operates the line in a large factory or steelworks

One would think the 2008 credit crunch or looming EU bankruptcy might have taught people something, but apparently NOO! Tens of millions of peoples' savings WIPED OUT, houses REPOSSESSED, because of screw-up of a relatively small number of people - doesn't that sounds like what they are doing is important—or at least dangerous? Oh and don't start with the whole "but they don't create value/add anything that nurse/worker do". Bullshit! If stuff's broke without them, they do add value!

Comparatively, I don't give a rat's ass about your lovely nurse, soldier, steelworks operator. They can cure/kill a person, two, five, maybe 100—who cares. Wall street can ruin nations for generations!

Re:False metrics

[Anonymous Coward]

I really don't get why some people are trying to mock HFT programmers by telling them that they produce nothing of value to society while someone like a soldier does. Seriously, if you take this perspective, any job can be meaningless. What is valuable to society? How is a soldier valuable? War is bad. How is a baker valuable? Does one more cupcake really improve society? Please guys, you have to be more sophisticated than this.

Discrete time

[Anonymous Coward]

Why not divide time into discrete blocks, e.g. one each minute or so? All bids placed in the current time block are then traded simultaneously at the end of the current block. If traders can only look at data form previous time blocks, I think all this BS could be avoided.

Re:Discrete time

[hsk17]

Discrete auctions could work. I can't think of any a priori reason why periodic auctions wouldn't work. They might even make the market place more efficient.

However, if you're willing to listen, I'm inclined to disagree.

First, investors get utility out of a continuous market. It's like having a convenience store open continuously from 8:30AM to 4:00PM, not in periodic intervals.

Second, that system still favors super-speed traders. Think about it. People will submit their quotes, but the market won't match them until after a period of time. That means the last entrant has the most advantage. As a result, you'll still want the fastest programs, best connections, the best programmers, etc. Even if you make everything hidden until the match (as in a dark pool) you can still process worldly information and wait till the last microsecond.

Third, that system will widen the spread, the difference between the lowest ask price and the highest bid price. Think about it. Let me stretch out the time so that it makes more sense in an everyday context. Imagine you're doing a lemonade stand business. You buy from suppliers, sell to consumers. If you were forced to only buy and sell once a month, instead of every day, what kind of limitations would you face? You would probably be extremely wary of overstocking, because the oranges might go bad (do they? let's just say they do for now). You could probably get away with charging more per drink because people know it happens so rarely. As a result, the costs of having to deal with the increased risk of holding the unsold oranges for longer are passed to the consumer.

There are dark pools, which you can read about. In such exchanges, all quotes are private from all participants. Only the machines see the quotes and generate matches. Obviously, you can probably back out information by analyzing your own trades, but because a lot of information is hidden, many institutions prefer to do volume on dark pools. However, as much as dark pools solve issues, they are somewhat bad for the market as well, because a lot of trading is done secretly. Average investors like yourself -- or anyone, for that matter -- has no visibility into the quotes -- the price discovery -- that are being generated.

There are different flavors of exchanges, all trying to solve the problem of risk transfer. You're not alone in advocating auctions. But it might not have the intended effect.

Re:Discrete time

[Tacvek]

Sure, but only up to a point. If for example, you limit trades to once per minute, but break ties by who submitted a buy order/sell order first, then there would still be emphasis on getting your offers in as quickly as possible, (i.e. within the first few microseconds). Some, but not all of of the existing HFT practices would change. the HFT would become every-block traders, and would still still try to be first to the transaction.

Breaking ties by last to submit wins would produce something similar. Randomly breaking ties would be best, but people who accuse the system of being based, even if it were not.

Re:Discrete time

[horza]

It was my first thought also. Then we will probably see sniping (ala eBay) and the HFT technology will be used to be first or last into the discrete block. However with the volume reduced so will the impact of HFT on volatility.

Phillip.

This is not exactly right...

[HerculesMO]

If you're a HFT programmer, you usually don't have traders in your ear, you have a bunch of quants in your ear. They usually don't yell (much).

The programmers that get yelled at by traders do more line of business applications and datapulls (kind of a business analyst role), because traders rely on them to make trading decisions and want their information "right now, you fucking asshole". These programmers usually stick to .NET or Java, though with the invention of Silverlight I'm seeing a lot more of that in the arena lately.

Full disclosure, I have worked in finance for 10+ years and currently work at a hedge fund as a systems architect/engineer.

I know how to correct this perversion

[Anonymous Coward]

Just reinstate the capital gains tax. The government is short of money and if they charged a 5% percent tax on investments held for less than a minute (or day or week or whatever) this type of abuse would vanish in a microsecond.

And if everyone realized the greed that led to the notion of 'supply side economics' was toxic, and that the Republican solution of giving milionaires and bilionaires even more money helped the economy was actually toxic things would improve rapidly.

The reason the economy hasn't improved is "NOBODY HAS ANY MONEY!!!!" (pardon me for shouting) In the last decade all the money has gone into the hands of the m[b]illionnaires who aren't spending it because economic conditions aren't right. Massive numbers of people turned their houses into credit cards to go into debt to keep the economy roaring dyrubg the early 00s but the credit card has been torn up.

Re:I know how to correct this perversion

[Opportunist]

Dead on. That's exactly what happened and exactly what caused the current recession. It's not a lack of goods or supply. It's a lack of money and demand.

It had been cushioned for a while since people were living on their substance for the last decade (or rather, for the last 2-3 decades). They didn't regenerate as much as they spent, but they tried to keep their living standard. So they spent more than they earned, and that's what kept the system afloat for a while. But we now hit the moment when that overspending doesn't work anymore because people simply cannot continue that way. They don't have anything left. Worse, they're in deep debt by now, and even reversing it would not result in an immediate recovery because they'd first of all have to recover the debt before starting to spend again.

Money has accumulated in the hands of a few, who (as you mentioned) do not spend it now because there is little, if any, incentive to do so. Why invest when there's nobody to buy the services you would offer? No, I didn't say "goods and services", I said services. Because that's what drives our economy. Services. Our economy (with "our" being "every country's in what's considered the "western world", don't feel left out, I mean you too!) depends heavily on services. Mostly because they're just absolutely awesome cash cows. It's selling the pure labour, without having to expend resources. Unlike agriculture or the manufacturing of goods, services do not require you to invest any (or at least many) resources. You don't sell what's in or on your ground, you sell your workforce. The beauty about it is that you can sell as much as you can put people behind it, and if there's no shortage in today's economy, it's workforce. At some point you cannot buy more land to grow something, and you might not be able to acquire more raw materials to produce something, but there's always unemployed people who you can pay a few pennies and have them create a service for you to sell.

Unfortunately, services are also a very, very delicate commodity. First of all, they're near impossible to export. Unless you have tourists. You cannot send a haircut abroad, unless the person from abroad wanting it comes to you. And second, and more critically, they're the first thing people start to cut when money gets tight. A new haircut would be nice, but I can do without. Fixing the plumbing would be nice, but that dripping faucet is not really on top of my list. First I need food. Then I need shelter. Then I need a few other things before I start to care about services.

This is why our economy is on its belly. We need money in our people. But don't expect things to pick up any time soon, even if you somehow manage to give them more money than they spend on "existing". They'd spend a good deal of the near future recovering from debt.

And that's exactly why this ain't happening.

Discrete time

[hakonm]

Why not divide time into discrete blocks, e.g. one each minute or so? All bids placed in the current time block are then traded simultaneously at the end of the current block. If traders can only look at data form previous time blocks, I think all this BS could be avoided.

Whats wrong with arbitrage?

[Anonymous Coward]

Whats wrong with arbitrage? You are just matching buyers and sellers in different markets together. If you knew one guy willing to sell item X at 5.30, and another guy willing to buy item X at 5.34, you could HELP both of them out. Buy it from the first guy, and sell it to the second guy. The first guy is happy because he wanted to sell at 5.30 and you made that happen for him (he thanks you). The second guy is happy because you sold him item X at 5.34 (something that he wanted and was willing to pay that money for). In the process you have helped two people do what they want and make 0.04.

Nothing is "evil" about arbitrage. You did real work matching people together. You are a matching expert. And you have risk in this process too. You could buy it from the first guy, but at the same time the second guy found another seller, so now you're 5.30 poorer and stuck with some item you don't want. That is why microseconds matter.

Re:Whats wrong with arbitrage?

[PPH]

A couple of things:

One or both of these people are effectively cutting in to the head of the line. What if I'm willing to pay 5.34 for those shares as well? But you are set up for HFT and I'm not. I'm getting screwed.

The most equitable trading system would match the 5.30 sell offer with the 5.34 buy offer and make the deal for both parties at 5.32.

Re:Whats wrong with arbitrage?

[Anonymous Coward]

You missed the point of the example. They are in different markets so they can't me matched. It is the "you" in this example who is in both markets which is why you're able to make both trades. If the buyer and seller found themselves in the same market, this wouldn't be an issue as they would automatically be matched. But "you" are doing the work to match them yourself. If I want to buy an item for 5.34, and you sell it to me for 5.34 I am "better off." If I was worse off, I wouldn't have tried to buy the item in the first place. I don't care if when you acquired the item you got it for 5.34, 5.30, or one penny, or one million dollars. All I care is that I have the item for the price I'm willing to pay (5.34). You did a good job of getting me the item. I don't care how you got it to me. If I cared about that, I would go look for it myself on other markets, but that would take my time away from me which, to me, is more valuable than the 4 cents I could have saved if I looked harder.

Re:How long to programmers stay in that type of jo

[Anonymous Coward]

If you could have a job like that, why would you choose to have kids? :)

I left the HFT desk...

[Anonymous Coward]

I used to work as a programmer on a core algo engine that was used for a variety of purposes, including:

Market Making
HFT
Index Arbitrage
and a few very specific prop strategies.

Technology was a mixture of C, C++, Java and a heavily modified version of Lua.

Skills for the job are:

Programming, with a capability of eeking out performance where required, but not churning out stuff so baroque and arcane that it is unmaintainable.
Business knowledge - you have to know the products, payoffs, pricing, risk, hedging, the works. If things go bad, you need to know why and how to fix.
Cool head - problems only get worse if you panic or crumble under pressure.

Downsides:

Long hours
Stress
Watching life go by, and realising you produce little of value
One day realising your hourly rate is lower than someone on a 40-hour week in a parallel field, so you leave.

I would actually recommend interested candidates with programming and maths smarts, and no family commitments to try it for a while. I just wouldn't recommend getting stuck in it for too long, or expect to view it as a 'career'.

Give the monkey a banana

[Anonymous Coward]

Here is how the market works

1) Teach a bunch of monkeys that every time they see blue sky they should hit the 'buy' button. Give them a banana if they do.
2) Teach the same bunch of monkeys that every time they see clouds they should hit the sell button. No need to whack them; the absence of a banana will be punishment enough.
3) Put the monkeys in a room, or many rooms spread out around Wall Street and leave then in front of their computers.
4) When the blue sky appears they will all hit the 'buy' button and the price will go up.
5) When the clouds appear they will all hist the 'sell' button and the price will go down.

It's as simple as that. It has only a very loose connection to do with the intrinsic worth of the commodity being sold and everything to do with what the monkey's (trader's) have been told are the signals to buy and sell.

Take one of the monkeys aside and teach him a different set of rules, say, hit buy when you see a spider and sell when you see a bird fly past, and that monkey will be a failure - not because his rules are any more stupid that the rules the other monkeys are following but because he isn't following the rules which the other traders are guided by.

Take all the monkeys (traders) aside and teach them all the new rules and sure enough, you will find they work to make the price go up and down.

Now, the people putting the computer algorithms in place are just more sophisticated monkeys. They think they are reacting to a real world and somehow enriching our lives but in practice they are just following the herd, enjoying their bananas and sharing out the loot between the owners of the commodities and companies in an arbitrary way.

The Penn effect; Community Reinvestment Act

[tepples]

Further, outsourcing is not "fair trade", because it artificially bypasses exchange rates.

Then perhaps the problem is that the exchange rates are lower in the first place. The "Penn effect" is an observation that less industrialized countries have cheaper currencies than purchasing power parity alone would predict. This is traditionally explained by the Balassa-Samuelson model that developed countries are more efficient at producing tradable goods and services. Some goods and services are "local": some goods spoil quickly in intercontinental transit, and personal services such as health and beauty care require face-to-face interaction. Others are far more tradable: durable goods have a shelf life far longer than the weeks it takes for a container ship to cross an ocean, and many of us are aware of offshore outsourced services. The value of a market's currency is related to the market's efficiency at producing tradable goods and services because local sectors have to increase their wages to keep workers from defecting to a tradable sector. See more detailed discussion of this effect. So once a country industrializes, its currency will become more valuable, and wages will rise.

Plain and simple: we need to bring the manufacturing jobs back home.

And according to the Balassa-Samuelson model, the way to do that is to make manufacturing here far more efficient than manufacturing there. This is already the case for automobiles, which I admit are less than perfectly tradable due to the cost of intercontinental shipping (case in point: the major car makers headquartered in Japan have plants in the United States). But you don't see cars of international brands being made in, say, Vietnam because the infrastructure isn't there yet.

You saw billions being spent to acquire companies that had never earned a dime of profit, and did not even have a good plan for doing so.

I'd argue that the eventual plans for the dot-coms were 1. to build up a user base whose demographic was attractive to advertisers, and 2. to offer services at a fee.

Later, banks and finance companies were making bad loans on real estate for a number of reasons

I've been told that one of the justifications for the rise in subprime mortgage lending involved mandates from the U.S. government to make home ownership easier for lower-income people, especially those in groups that were traditionally targets of racist discrimination. See Community Reinvestment Act.

Re:The Penn effect; Community Reinvestment Act

[Hognoxious]

I've been told that one of the justifications for the rise in subprime mortgage lending involved mandates from the U.S. government to make home ownership easier for lower-income people, especially those in groups that were traditionally targets of racist discrimination. See Community Reinvestment Act.

Told by Fox News, I take it.

If you'd bother to read the article you linked to, that myth is well debunked. Stick a fork in it already.

Now if you can explain how the act caused such a fuck up in commercial property, when it only applied to residential, then we'd all be happy to hear it.

Re:The Penn effect; Community Reinvestment Act

[Bob the Super Hamste]

I've been told that one of the justifications for the rise in subprime mortgage lending involved mandates from the U.S. government to make home ownership easier for lower-income people, especially those in groups that were traditionally targets of racist discrimination. See Community Reinvestment Act.

That is a strawman argument. I would think that the more likely cause was that banks could dump loans onto to Freddie Mac and Fanny Mae and not assume any of the risk, but that was only for traditional loans. Now since most people who could get a traditional home loan already had them and there were very few new people who could get them we saw banks come up with more creative solutions since they undervalued risk since they thought housing prices could only go up. I qualified for a $750,000 loan (would have taken about 80% of my income to make monthly payments) 8 years ago and was being told that I buy as much house as I can. I didn't and I bought a $220,000 house with 20% down on a 30 year fixed @ 4.75% (so I am not underwater on my loan but pretty damn close as a few houses in my neighborhood were foreclosed on) which was well within what I could afford at the time. I decided that I would be responsible and did the right thing so my sympathy for those who who took out exotic loans or giant ones that they could barely afford is non existent. I do feel for those who were responsible got a reasonable loan but then lost their job but there are far fewer of those people than you think.

What we really need is for the market to clear, let those home go into foreclosure, I don't really feel sorry for you real estate investors and your loss. Yes this will depress housing prices but you shouldn't be viewing your house as an investment, it is an asset and even if its taxable value drops to $0 (which it won't unless it is really blighted) it still has intrinsic value as you live there. The problem we have now is everyone is trying to keep the market from finding the bottom, but it needs to so it can clear the current inventory. Have banks done some dodgy things, yes especially foreclosing on home that were current on payments, and those the companies should be strung out.

Liquidity

[tepples]

Liquidity of an asset is a service that said psychopaths provide.

Re:Liquidity

[benhattman]

Liquidity is a commodity of diminishing returns. If I put in a sell bid on stocks and you say it will take one month, the team than can sell in a day instead is indeed providing a valuable service. And going from taking a day to trade down to an hour or say five minutes is quite valuable too (though not nearly as much so). But the moment your liquidity is faster than my ability to be informed about it, additional liquidity has ZERO value. It takes me several seconds to click sell on a website, and watch as the site refreshes to inform me the transaction has occurred.

In the 1990s, these people were creating value. Today, they are exclusively leeching money from the rest of us.

Re:Liquidity

[Anonymous Coward]

Why do we think liquidity is a good thing?” Answer, because it facilitates trade around the exchange of information. “Information about what?” one might then ask. “The company in which the investment is being made,” is the answer. Does algorithmic trading exchange information about the performance of the company? No, it is only working off information about trading behaviour. Ergo, it may increase “liquidity” but it is not fulfilling the purpose of liquidity.

That kind of shift to traders working mostly off what traders do, rather than assessing the value of what is being traded, has become an absolute plague. It has taken over most Western financial markets. Hedging, for example, used to be all about hedging bets to protect the underling exchanges (usually wheat, or pork bellies or physical things). Now, hedging is all about reading behaviour, which then leads to other hedging strategies that are based on reading the hedging behaviour, and so on.

So the disappearing point is part of the problem. In our “anthrosphere” we are increasingly staring at each other’s navels in the financial markets, trying to make money and sustainable wealth out of ether.
    -- Yves Smith:

Re:Liquidity

[Anonymous Coward]

At some point things shift from providing liquidity to rent seeking. Economically they'll spend up to the point of hitting their minimum margin to keep winning that race. Which means the only people who benefit from this light speed liquidity are the transaction fee collectors and the infrastructure providers.

Particularly when companies are getting privileged access that helps them be the first past the post, I don't think you can fall back on the righteousness of saying "but we provide liquidity".

Re:Liquidity

[Anonymous Coward]

Volume != liquidity

Re:Liquidity

[mandelbr0t]

+1 nailed the buzzword A lot of the HFT proponents in this discussion are using liquidity as a reason that HFT should not be banned. But encouraging people not to hold money in stocks for any length of time seems to lead only to market volatility. The stock market was created to allow people to make a long-term investment in a company. Expecting to unload a long-term investment at full value right this moment is ridiculous. What else (besides other currency) can you buy of value that you could later sell at full market value in an instant?

Re:Liquidity

[Anonymous Coward]

The fact that an HFT shop can turn trades around so quickly makes me wonder how much of their claim to providing liquidity is just trying to take credit for liquidity that the market already had without them.

Re:Liquidity

[Anonymous Coward]

Sorry, liquidity is not an something of value to society -- and I'm speaking in the most general, colloquial sense of the phrase "value to society". It provides valuation, but don't confuse that with value. Liquidity is a feature, a level of market efficiency -- and completely unnecessary in an ideal world where humans weren't concerned with screwing each other over. If everyone was trying to build something, and that production was for the betterment of society, and we were all working together to build and better society, there wouldn't be as much concern over providing a demand market for an asset or security. Many of those assets and securities wouldn't exist as the need to hedge, insure, penalize, compete and dominate wouldn't be as strong. Of course life isn't like that. We all should be looking out for ourselves, and so we've created markets and professions around accumulating wealth and holding on to it. That's why liquidity is necessary, and that's also why the people who work to provide not only liquidity but market efficiency (for example, taking advantage of price discrepancies on different exchanges) have been legitimized.

Are WS programmers helping people and civilization advance peacefully and rationally? Or are WS traders helping to advance a part of society, albeit very vital to our needs as a society, that has nonetheless shown itself in the past to actually deplete value, ruin people and wealth, and contribute to the very significant commercial and political mountains that stand in the way of smaller, less powerful groups of people producing and realistically competing in today's economy?

WS programmers do not provide a legitimate value to society.

Re:Liquidity

[Anonymous Coward]

If computers are trading, not humans, then the additional liquidity has value. And computers do a lot of trading nowadays.

Re:Liquidity

[Anonymous Coward]

I dont see how not having these people would destroy our market's liquidity. All they do is artificially inflate the price of assets by constantly buying and selling. Oh well there was a run on this stock so the price go's up, but wait it had nothing to do with that companies business practices or having a innovative idea IT WAS ONLY BECAUSE YOU FUCKS MADE A PROGRAM TO BUY STOCKS AND THEN SELL THEM A FRACTION OF A SECOND LATER for an extra cent. Im pretty sure if they didnt exist people would still invest and our market would still be liquid.

Re:Liquidity

[toddestan]

How do they do that? The high-frequency traders only execute a move when they have a seller and a buyer lined up and they can take a slice by acting as a middleman. They add zero liquidity to the market.

Re:Liquidity

[vertov]

No, you're not really getting these programmers working on brokerage accounts. These are prop traders and programmers, and the volume of them in the market means there's probably always a price and market.

Re:Liquidity

[Anonymous Coward]

Liquidity only matters when stocks are hard to trade. When there's a crisis all your "liquidity" goes away and we get LTCM again.

Re:Liquidity

[Anonymous Coward]

The reality is you won't get your money any faster with HFT that without, because settlement is a couple of days after the trade.

I think there is a reasonable argument to be made that the HFT leach less of our money that the old lower frequency kind.

The thing of value would be to design a market where long term providers of capital to companies could interact, and where no value leaked to non-investors.

Re:Liquidity

[tolkienfan]

Not true. Just because you don't understand the timescale doesn't mean it has no value.
The bid ask spreads and the speed of price discovery are better than ever before (almost entirely due to HFT). When you buy or sell stocks you do so at much closer to the actual value of the stock than ever in the past. That may not mean much to you, who would be trading maybe $1000s and paying a broker several percent, but to an institution that's moving $100,000,000s around it's an enormous value. When that institution is your mutual fund it means real dollars to you...

Re:Liquidity

[tolkienfan]

HFT doesn't cause volatility. Indeed HFT algorithms will pull their orders out of the market place when volatility goes too high. The reduction in liquidity actually causes problems.
Yes - HFT companies stopping trading causes lower liquidity, higher volatility, higher cost of trading and higher risk to remaining participants.

Re:Liquidity

[tolkienfan]

Absolutely incorrect. The exchange does what you describe. The HFT market maker will have orders (buys AND sells) on the exchanges book waiting to trade. When you send a market order to your broker, that order will probably execute on an exchange against an HFT market maker's passive order. The market maker is taking a high risk having the order resting on the book, because the value can change and those orders get swept - resulting in an instant loss for the market maker. It's only viable if they make enough money from the bid-ask spread to cover such losses.

Re:Liquidity

[Anonymous Coward]

Exactly how did we get into this mess with the capital markets? A situation where the global stock of derivatives is over $US600 trillion, which is about twice the capital stock of the world. A situation where high frequency trading is over two thirds of the transactions on the NYSE and about the same in the stock markets of the UK and Europe. Likewise they are over half the action in foreign exchange markets and they are rapidly becoming dominant in the futures market. Andrew Haldane from the Bank of England is arguing against allowing high frequency trading — algorithms chasing algorithms chasing algorithms — from being allowed to proliferate pointing at volatility as the problem:

        Speed increases the risk of feasts and famines in market liquidity. HFT [high-frequency traders] contribute to the feast through lower bid-ask spreads. But they also contribute to the famine if their liquidity provision is fickle in situations of stress.

Haldane noted that relative to gross domestic product, the equity market capitalisation of the US, Europe and Asia had not grown since 2000, suggesting that “the contribution of equity markets to economic growth has been static”.

Little wonder, when you consider that companies are putting themselves in the hands of algorithms.

I think this conclusion, which many come to, is to some extent a blind alley. Because the volume of transactions is higher — when it is claimed that it adds liquidity this is a circular argument, like saying “the more we trade the more we trade” — it should mean that “volatility” at least on basic measures, is lower. The trades are made around normative models so they will tend to re-inforce those norms. Right up until the moment when the market does not behave with regard to norms, and then they suddenly start doing weird things, such as the so called “Flash Crash”. So it is probably correct to say that there is less volatility. It is also beside the point.

A better question is: Why do we think liquidity is a good thing?” Answer, because it facilitates trade around the exchange of information. “Information about what?” one might then ask. “The company in which the investment is being made,” is the answer. Does algorithmic trading exchange information about the performance of the company? No, it is only working off information about trading behaviour. Ergo, it may increase “liquidity” but it is not fulfilling the purpose of liquidity.

That kind of shift to traders working mostly off what traders do, rather than assessing the value of what is being traded, has become an absolute plague. It has taken over most Western financial markets. Hedging, for example, used to be all about hedging bets to protect the underling exchanges (usually wheat, or pork bellies or physical things). Now, hedging is all about reading behaviour, which then leads to other hedging strategies that are based on reading the hedging behaviour, and so on.

http://www.nakedcapitalism.com/2011/07/time-to-take-stock.html

Re:Liquidity

[Anonymous Coward]

Millisecond liquidity has value for the psycopaths that trade every second. Second liquidity has value for the psycopaths that trade every minute. Minute liquidity has value for the psycopaths that trade every 5 minutes [...] one-day liquidity has value to us mortals that trade rarely and would trade in one month without it.

So, I would say it is a whole chain of psycopaths providing value to each other, and some of them at the end of the chain use the rest of the chain to provide value to society.

Re:Liquidity

[Anonymous Coward]

What if you have a representation of yourself that for these purposes take these choices for you? An algorithm that takes approximately the same choices you would have made can make use of the additional liquidity.

Subtle Microsoft PR piece

[Anonymous Coward]

I think people who read the article are meant to work their way through it to get to some of the quotes near the end like this one: "Moreover," he continued, "I was very productive using VB6, which enabled me to get a lot more done than if I had attempted to build everything using C++."

They are then meant to think, "Gosh, that Microsoft stuff really can't be so bad after all. I think I might just junk my free software stuff and get myself a licence. It's got to be worth it!"

How does the path to reformation look?

[amiak]

I will agree with the obvious sentiment here: HFT is rubbish! But to be brief, what needs to take place for this unsustainable pattern to end? I surely don't know, I steer clear of financial voodoo.

Re:How does the path to reformation look?

[PPH]

The SEC needs to step in and mandate the release of instantaneous pricing information. In much the same manner as they required companies to make releases of material information available on an equitable basis some years ago.

If I get my stock quotes with a 20 minute delay, everybody should get theirs with the same delay.

Re:Astroturfing

[Anonymous Coward]

Why is this the SECOND post where a person claims or insinuates that they are an air-traffic controller, and compare the stress level between that and algo-traders???

Somebody is astroturfing Slashdot. Nothing new to see here. These aren't the droids you are looking for. Move along, move along.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:I am a truck driver

[Anonymous Coward]

So is my younger brother. After deducting for speed traps, taxes, maintenance, and cargo fees he still lives in my aunts cousin/husband's mother's old trailer.

Just kidding...

About the taxes.

Where does all the HFT hype come from?

[Lazy Jones]

It seems to me that the HFT industry and the wall street crooks who employ it are desperately trying to portray HFT as the sophisticated work of some extremely talented programmers who are just smart enough to help brokers earn more money, while in reality the whole system seems to be half inside trading (having access to information other traders don't know yet and methods other traders cannot employ) and half DoS attacks on regular traders (the same order will not work for them because HFT is preventing it). Therefore, it should (IMO) be illegal and perhaps such spin posts exist mostly to counter sensible attempts to ban it.

Disclaimer: I do not own or trade shares of public companies.

Re:Where does all the HFT hype come from?

[GameboyRMH]

Who said it couldn't be technically fascinating and morally repugnant at the same time?

Re:How long to programmers stay in that type of jo

[Anonymous Coward]

recruiters are about the same level as double glazing or used car salesmen in terms of skillsets and attitude - they are just interested in getting a recuitment fee and it is just a numbers game. Don't even bother responding they will have sent the same exciting offer to 500 people.

Im Still A Bit Confused

[Anonymous Coward]

You deal with high level complex algorithms and engineer some of the most high end elite code in the world that basically manipulate the trading of the world economy yet, what worries me is you average 12 hour work days and 100 hour work weeks.

12 * 7 = 84

Where do you find those extra 16 hours a week?

Re:How long to programmers stay in that type of jo

[Anonymous Coward]

About 18 years ago I got an offer from one of these places. $125k/year. And that was EIGHTEEN years ago, when $50k was a big salary.

I turned it down when I heard about the "high-pressure, high-availability" requirement. I also wasn't going to live in a closet and get yelled at by assholes all day.

boring, but blame society

[Lazy Jones]

I doubt the long term returns justify what this does to people.

What does society measure one's success by? Yearly salary (regardless of lifespan), or perhaps whether the subject will die happily at the age of 95 while having 30 grand-grandchildren? In fact, we glorify those who die young - after becoming famous for some work that took a huge toll on their lives - much more than those who live long in spite of it ...

postscript on ode to precious snowflake

[epine]

I hate it when I get my possessives wrong. It would help in the heat of composition if SAUDIS didn't rhyme with TARDIS in the tips of my fingers.

I also missed one stitch: I should have connected the leverage of the baby boom bulge with leverage in capital markets.

Normally I write for myself. For once it would please me if anyone read past the deceptively slow ignition.

Re:postscript on ode to precious snowflake

[Unoriginal_Nickname]

The vote is tied up by the boomers? Hahaha. I hate to destroy your fantasy, but the boomers aren't the ones propping up the Conservative party. Do you think it's a coincidence that Alberta, the Conservative stronghold, has the lowest median age of any of the major provinces? Do you think it's a coincidence that Alberta's many young, successful professionals, businessmen and tradespeople are voting for Conservative candidates?

It doesn't even make any sense if you think about it. An aging and retiring population has a lot to gain by supporting the NDP. On the other hand, the NDP has absolutely nothing to offer someone who has a significant source of income.

You're just miserable that you didn't get your way, so you're trying to blame it on those stupid boomers and their backwards thinking, right?

By the way, jobs still fall of trees for young people. It was YOUR choice to get a BA in Fuck All instead of researching whether or not employers will care. People with talent, qualifications and social skills are just as busy today as they were in 2007.

Re:Good Choice! Stay out of NYC or get out if you

[fafaforza]

So where are you living now, also in Florida? I ask because I wouldn't mind getting out of NYC myself.

I used to be you

[Anonymous Coward]

But then I quit and I couldn't have been happier. For those of you envying AC's life please don't. The job is miserable knowing the work you do fuels the greediest of the greedy. The same traders you support on the floor are those reckless scums who choose to make short term gains and year-end bonuses at the detriment of the well being of people, the market, and even the firm that hired them. It's overly greedy borderline evil (seriously). There wasn't a day going by when you'd hear about a trader getting jailed for channeling "losses" to his own fictitious counter party account. I started my own thing, which by the way has nothing to do with speculative or transactions trading.

goodman studied NYMEX, others

[decora]

have studied goldman sachs, there are actually several books out now that specifically focus on GS, and several other books that mention its place in the crisis.

The typical HFT job interview

[Christian Marks]

Occasionally I receive calls for HFT jobs from recruiters
who represent elite hedge funds. Here is the typical exchange.

RECRUITER
You’ll be working with astronomically smart people. One
group uses crystalline cohomology to obtain the best
polynomial time approximation algorithms for
intractable problems in HFT. The engineer who did this
was an embryo prodigy who taught himself calculus
within ten to the negative sixty-seven seconds of conception.

ME
Is that the work you have in mind for me?

RECRUITER
No. You’ll be cleaning the group’s digital bed pans.

ME
Perhaps you should recruit a Nobel Laureate.
Thanks for calling.

See? Screaming makes things difficult.

[Anonymous Coward]

The screaming makes things more difficult, why don't they see that and just shut the hell up? Why are they compelled to scream? Less Screaming = More Money, so you should want to scream less. I guess that's why they can't do the programming, they have no concept of logic.

Re:Good Choice! Stay out of NYC or get out if you

[Cutting_Crew]

Yeah i am in the Tampa area. If you can find some software development work, the pay is great. I am making 75K doing some C++, Objective-C, some OpenGL. Even if i were to get offered a job for $200K i STILL don't think that i would break even because for a 4 bedroom house with a family you just cant be living in some 1 bedroom closet close to downtown. That involves riding a train for an hour 1 way, that is assuming that I would even want to bother working on wall street. Also consider the benefits of living in the middle of florida. Disney stuff(not that i am a fan of mickey but still), sea world, busch gardens, multiple water parks, close to a great number of amenities such as the arts and concerts. Because of the 9 month summers that means you can practically go camping or swimming at a good quantity of beaches that are accessible from anywhere. The water stays nice, cool and comfortable starting in March and all the way into October. Theres also great food and great shopping. So whether you are single or have a family its a great place to live. I did some calculations and i would have to make around $185K a year to come close to breaking even, that is the same kind of house. Thats not counting the fact that NYC has state taxes and not counting the fact of having to travel to wherever it is that i were to work would be a 2 hour commute each day. Thats 40 hours a month, a whole work week just commuting. Now take into account you have no where near the family environment as far as fun stuff to do. No beaches(ok well maybe you can go swimming 2 or 3 months out of the year) is a big deal. just food for thought. take it for what it might be worth.

No, it's not about slavery.

[198348726583297634]

It's about the world we want to live in, which stands in ever-starker contrast to the world we do live in.

Do you want to live in a world where the best and brightest throw their efforts away with such mundane, trivial shit?

Again, not where-are-we-now, but where-do-we-want-to-be.

Do you want to live in a world where, thanks to some tricks of law and circumstance, a handful of people have such domination over billions?

Again, not where-are-we-now, but where-do-we-want-to-be.

You don't have to be a slaver (wtf??) to say, the world we live in is wrong, because a lot of people are suffering for what seem like really lousy reasons.

You do, however, have to be a decent human being.

Re:No, it's not about slavery.

[russotto]

It's about the world we want to live in, which stands in ever-starker contrast to the world we do live in.

Do you want to live in a world where the best and brightest throw their efforts away with such mundane, trivial shit?

I don't want to live in a world where it's my business where other people choose to place their efforts (nor theirs mine, obviously). And that's why I think that particular argument against HFT -- that it drains talent from other "more important" fields -- is fatally flawed.

Knuckleheads.

[Anonymous Coward]

Dude. Don't listen to these knuckleheads.

The folks on this board don't have any idea what Wall Street does or what HFT is. Allocation of capital is the LAST thing these folks think about on a daily basis.

Head over to quant.ly if you care to.

narcotics trade

[Anonymous Coward]

I make the majority of my income buying and selling illegal drugs. (Mostly moving marijuana from California to the Midwest/South.) Posting anon for obvious reasons. I make close to 300k a year, and only work about 10-20 hours a week. You can imagine that the ratio of blood pressure to income is one of the best in this occupation.

Re:narcotics trade

[russotto]

I make the majority of my income buying and selling illegal drugs. (Mostly moving marijuana from California to the Midwest/South.) Posting anon for obvious reasons. I make close to 300k a year, and only work about 10-20 hours a week. You can imagine that the ratio of blood pressure to income is one of the best in this occupation.

Unfortunately, marijuana is not commoditized (though certainly it could be) and there's no futures market for it. Thus the speed of weed arbitrage is limited by how fast you can move the actual product, and the spread has to cover the transportation costs. There's no need for HFT systems.

Re:narcotics trade

[Anonymous Coward]

Oddly enough, the price doesn't fluctuate enough for there to be much in arbitrage, and the transportation costs are minimal ($15 for $5000 worth of product)

adtmag sources...

[PGGreens]

Anyone think it was a little strange that half of the quotes in the ADTMag article were just pulled from Slashdot comments? I mean, there are some great comments on here and all, but that seems like some lazy reporting to me.

Bravo Re:Traders

[Fubari]

Bravo, well said!
So... are you running for office? :-)

fascinating Re:Yet Another HFT Article

[Fubari]

Thank you for what you & the other HFT people posted.
Your perspectives are very interesting; especially about the workplace dynamics, the value flow from big banks etc.
This is one of the more interesting threads I've read on Slashdot for a while.

p.s. about the "leech" complainers, *shrug* if I thought they had a plan for building a better world, I'd ask to hear about it.

This is not an ideal world

[tepples]

and completely unnecessary in an ideal world where humans weren't concerned with screwing each other over.

If we eliminated everything "completely unnecessary in an ideal world where humans weren't concerned with screwing each other over", we'd be living in working communism. That won't happen until the tribulation, comrade.

Re:This is not an ideal world

[vertov]

Haha, very good. Hopefully we'll never be plowing our common fields together, but I also hope the hypothetical of an "idealized world" still gets my point across about prop trading programmers. It's not that they're not crucial participants in financial markets, but it's a stretch to think the liquidity provided has a brilliant part in advancing our society. Liquidity is around today primarily for speculation and valuation; it doesn't have the same meaning as it does when you're talking about the foundations of an ideal capitalistic economy.

so what big deal?

[Anonymous Coward]

All those electrons move just like any other program. The difference is, you're getting over paid to do it.

Liquidity

[Pseudonymus Bosch]

Isn't liquidity good?

another HFTer

[Anonymous Coward]

I too am an HFT programmer, but I work on the other portion. I work on statistical models and algorithms to make predictions. No traders involved at any point. The reason for anonymity is not because we are ashamed, it is because our employers would *not* approve of this. Any leak of IP, whether it be mentioning hardware involved, software techniques, model or algorithm names, is *not* appreciated, and can be a terminable offense. I work for a small proprietary shop ( 10) that trades a few percent of the daily volume around the world. The developers, especially the infrastructure folks *must* be good, and are well compensated. I see liquidity as no less noble than selling advertising. Further, if something is predictable, we will find it. Call it time-arbitrage.

I hope to make some single digit multiple of his 500k (1-5MM).

That's not a disclaimer

[Anonymous Coward]

Disclaimer: I work at an HFT firm.

That's a disclosure. Funny how complex stuff comes out of your mouth yet you don't understand the difference in meaning of two words.

Re:That's not a disclaimer

[hsk17]

You're right. I should have used the word "disclosure".

Nope

[abigor]

You misread his comment. He said he normally works twelve hour days (presumably five days a week, so 60 hours a week), but sometimes works up to 100 hours a week.

Yeah, neither do I.

[198348726583297634]

I don't want to live in a world where other people's actions shape the decisions I make throughout the day, either.

Curiously enough, that's exactly why I'd love to see this sort of frivolous, deeply harmful shit stopped. These people are reckless, and the actions they take day after day make my life harder in a million tiny, indirect ways, along with your and virtually everyone else's life too.

(Note: you may not agree or see it that way, but the world doesn't need your consent or acknowledgement to be profoundly broken and in desperate need of structural repair. So if you're getting ready to type some sort of "no I'm cool everything's cool" reply, just spare us.)

It's just another one of those world-we-want vs world-we-have situations. I want a world where a bunch of shits in a room thousands of miles from me aren't chipping away the economy a billion tiny transactions a day.

In fact, I'd go so far as to say, that's the price you pay for living in society. You give up some freedom (can't wreck shit for millions), and in exchange you get police, fire, sanitation, the whole ball of wax.

Do I think I should be able to tell them not to do that, so long as we're all sharing this same society?

You'll get one guess as to the right answer here.

You've completely misunderstood

[dbIII]

Wall Street is a useful level of abstraction above the businesses that provide products and services. Losing Wall Street would be like losing nearly every physical and electronic copy of the phone books covering a large percentage of all US businesses - chaos and annoyance but not total disaster because the physical and majority of financial structures exist elsewhere.
Now are you starting to understand?

but they don't create value/add anything that nurse/worker do

If that was not a deliberate attempt to turn me into a strawman then you need to work on your reading comprehension skills. I was writing about overly inflated levels of importance and never wrote anything about not adding value/anything and especially did not write the text you put in quotes.

Social conditions

[unsolicited]

And I propose taxes be linked to 'social' conditions of those countries http://www.rediff.com/business/slide-show/slide-show-1-tech-apple-workers-forced-to-sign-no-suicide-pledge/20110504.htm and http://www.rediff.com/news/2007/may/03touch.htm

 

Still unfair

[migloo]

Suppose I am the greatest programmer in the world applying the fastest algorithm from my home.
I would still lag about 1 sec behind the competition with my "fast" internet connection.

Since we are speaking microseconds here, is not the main advantage being the closest or even within the final exchange computer?

The competition is obviously biased in favor, not of the best programmers, but of those privileged enough to be located nearest to the apex of the whole system.

These particular banks or trading offices can very safely and easily skim the cream and slowly leech the rest of the world, they do not even need the best programmers.
And presumably this whole looting scheme would collapse if some minimal Tobin tax (say 0.001%) was applied to all ttransactions.

Re:Still unfair

[Xentor]

Since we are speaking microseconds here, is not the main advantage being the closest or even within the final exchange computer?
The competition is obviously biased in favor, not of the best programmers, but of those privileged enough to be located nearest to the apex of the whole system.

Yep, the serious HFT groups run their boxes as close as possible to the exchange... Either they rent space on the floor (Or in the same building), or they're right next-door.

no one calls them "algos"

[ren-n-stimpy]

sorry, this just bugs me. i've worked with coders from asia, europe, and all over the states. from open source zealot hackers in the bay, to buttoned-up microsofties in Seattle, IBM, health care software in Atlanta, Film/TV/games in Northern Virginia, telecom software integration in Madrid, Spain. Worked with remote people in Montreal, and remote people in various places in India. While I haven't haven't worked in NYC, where you are, I know people at Google Manhattan, and a couple quants there, one of whom is an MIT+Berkeley PhD old friend, fellow grad student, and fellow coworker.

and, dude, sorry, but...

NO ONE CALLS THEM "ALGOS"

It's like going to silicon valley and calling SF "San Fran". ugh. just uncomfortable for everyone in earshot.

Use of FPGAs

[Anonymous Coward]

How often are FPGAs used in the industry? Is use increasing it decreasing?

I'm interested in a career in hardware HFT

How often are you home?

[maillemaker]

That's pretty amazing. You are making not quite double my salary in Engineering.

Is it worth it?

google says it all

[MadMaverick9]

NYSE Euronext

"Stealing democracy, one billion dollars at a time."

just hover with your mouse over the text "NYSE Euronext" on the map.

Re:Linux by default?

[WorBlux]

Windows has some add-ons that make it more posix-compliant, but they are implemented in user-space. It's an afterthought at best. On the other hand Linux has a Posix interface in the kernel, and only breaks away from the specification when there is a really good reason to.

The Tanerbaum Torvalds debate is about the design decision of the respective kernels. This does not imply that Torvalds did not consult Tanerbaums Operating System Books or code when he first began his project. The use of the same filesystem clearly shows the influence. Aristotle and Socrates did not agree on everything, yet Socrates was a clear influence upon Aristotle.

But I think all your objection is just because I didn't make my first post all that clear. If you want to say linux "stole" technology (factually copied or adapted), then you should start by looking a Unix and Minix systems. I'm not saying that any copyright or patent was violated, or that linux has added no novel contributions of it's own. On the contrary it's the fastest evolving software project on the planet.

Re:Linux by default?

[bmo]

My, we're disingenuous here.

You're attempting to sound reasonable. But in fact, you're using the same stupid argument that SCO, Microsoft, Dan Wallace, Darl McBride, Robert Enderle, ADTI, etc, all used against Linux to paint it as intellectual theft of Unix.

You are the one that said "stole" first, bucko.

You are intellectually dishonest.

--
BMO

Re:Astroturfing

[Anonymous Coward]

There can only be one? You are silly.

Don't forget who are the people you are part of

[Anonymous Coward]

Hi,

In spite of all these software development skills that you think make you so much better than any other, on your own judgement, comparing yourself to millions of individuals that you don't even know. (I mean google does not say they are the best, you said it mate).
So stop trying to re-assure yourself you're the best, because you'll never be.

So, I'd like to remind you that you work for and you are part of these organisations that play with people's money, savings and pensions and that the likes of you have lost losts, in fact milions and hundreds of millions of the liquidities, svings, pensions of poeple all over the world and in the USA in the first place.

If you had a bit of understanding, if you were a bit smarter, just enough to take your head out of your small universe, you 4 walls, you would figure out that regardless of your technical genius (which is still to verify, I mean I never have to go back from holidays to debug myself and firefight my bugs), then you would realise that your colleagues are not generating much liquidites for putting together projects and companies over the world that would server many real life, useful everyday, purposes.
Instead you help them putting the financial world and thererfore the industry and the rest of the people generating value thanks to their labour, and produced goods, all of us so to speak, in the situation we are in nowadays.

So please, before saying you are the best, look around and ask yourself if you are not contributing to ruining the economy and bothering a few of us with your egocentric, and your own purse considerations.

Re:Don't forget who are the people you are part of

[Anonymous Coward]

By the Way, this is the first time I post on slashdot, don't have an account and don't like to be called an anonymous coward.

I am Romain, I work for a company called Cadence, that provide companies like Intel or AMD with their processors design tools.
You use my work to executes your so fast and so good code. So fast to put to serve wrong or no purpose.

When I think that my work can be used for healthcare applications or traffic management, or audio applications, I mean I could not think of traders in the first place, people that still the 1st categories of people's work... Ridiculous ;-)

Re:Linux by default?

[WorBlux]

Abusus non tollit usum. Some people have used a similar argument to try to do bad thing in the past, however that simply isn't relavent to the accuracy. (Software Patents shouldn't exist anyways (When's the last time someone from Microsoft said that?)) Should I believe Linux sprang fully formed, a system in itself from the loins of Linus? Now for some words from the big cheese himself.

I'd like any feedback on things people like/dislike in minix, as my OS resembles it somewhat (same physical layout of the file-system (due to practical reasons) among other things).

PS. Yes - it's free of any minix code, and it has a multi-threaded fs.

So it's own code, but not exactly a clean room implementation, nor 100% original. The linux kernel has taken BSD code and incorporated it in the kernel. How exactly does this work unless they two kernels work in a similar fashion? "Stealing" in the context of intellectual ideas is always metaphorical and may or may not have anything do with copyrights or patents.(which again, I am no great fan of) e.g IThree years I built some raised gardens with two by twelves. In the next two years each of my neighbors neighbors stole the idea.

Lastly I'll leave you with a bit of advice, though I know you will likely ignore it. Adjectives describing personal qualities never settle an argument, and usually derail it. It's better just to try to find the common ground first. Here's the question. Does linux owe it's current form and existence more to Minix and Unix than any other operating systems? If not show me the evidence, if so then just a matter of settling to which degree.

Re:Linux by default?

[bmo]

>filesystem hierarchy

Filesystem hierarchy is trivial. It is neither patentable nor copyrightable. It is not intellectual property owned by anyone. Copying a filesystem hierarchy is like copying a list of phone numbers from a phone book. Mere listings of facts are not and never were copyrightable. If you are going to say "stealing" with regards to intellectual property, you have to have intellectual property there in the first place.

This is also why following the POSIX standard is not copyright infringement. The POSIX standard is a list of facts that an OS needs to match in order to be called POSIX compliant. How one implements those facts is up to the OS author. There is simply no theft being done.

>cleanroom

Cleanroom implementation is required for originality? Since when? Since when is something ever created in a vacuum? Since when is a "cleanroom implementation" not influenced by outside forces or the thing you're trying to emulate (Phoenix BIOS implementation of the IBM BIOS as a classic example).

Your requirements for something to be called "original" is ridiculous and an impossible bar to pass.

>"Stealing" in the context of intellectual ideas is always metaphorical

No, no it's not. It's not metaphorical. This is the same argument that SCO tried to use against Linux. Either there is identical code or patent infringing algorithms or there aren't. After all these years from 2003, there has yet to be a single line of code shown to be "stolen" from Unix.

Stop redefining words with your own private definitions. This verbiage gaming that you are doing is done by crooked lawyers and politicians everywhere.

>Should I believe Linux sprang fully formed, a system in itself from the loins of Linus?

Argument from incredulity. Also ignoring the history of Linux. It was very rough and very small in the beginning. It was nowhere "fully formed." This "fully formed" bullshit is part of the "Big Lie" that ADTI was using.to discredit Linux as if to say "there's no way Linux could be this functional, there /must/ be stolen code in it."

>Does linux owe it's current form and existence more to Minix and Unix than any other operating systems?

One thing you should get through your brain is that Minix is a microkernel and Linux is monolithic. Your failure to understand this FUNDAMENTAL STRUCTURAL difference in design leads you to think that these are similar. They are not. Your continued arguing in spite of this fact tells me you are basing your entire argument upon a false assumption. For an argument to be valid, the whole chain must be valid. Your argument fails on its most base of assumptions.

There is not a single line of code that is from Minix or Unix in Linux. As shown in court.

>If not show me the evidence, if so then just a matter of settling to which degree.

>Asking me to prove a negative.

My we're full of logical fallacies today.

SCO tried for years to show that Linux had SysV code (you know, actual AT&T Unix, not FreeBSD) in it. There is no "there" there. Your argument has been used by many other people with very large financial interest and consequences if proven true. Yet for years, those with a financial stake were never able to prove a single thing in court. SCO wanted $5 billion from IBM for "stolen code" and were unable to show a single line. They even tried to reopen the USL can of worms and failed.

If you are going to continue to argue SCO's bullshit case all over again, then I suggest you ask Darl McBride how well that worked out for him.

And calling someone disingenuous when someone is indeed disingenuous is not ad-hominem. This is especially true when the alternatives are worse.

--
BMO

Thieves

[Anonymous Coward]

So, it's you that allows the HFT guys to place bids and offers than yank them 200ms later?
Placing said bid or offers without intent on executing it is illegal.