Slashdot Mirror
Algorithmic Investors on Wallstreet
eldavojohn writes "Recently, setting up prediction markets that people play was the big thing to guess the future. But is there a chance that computers will replace investors? From the article: 'Quantitative investment managers use a model to identify sets of characteristics for their investments. Computing power is now relatively cheap. Obviously, computing power can access data almost instantaneously and simultaneously. Asset classes and financial instruments within those asset classes can then be screened and investments are selected. They reflect the manager's views.'"
My work here is dung.
Big deal, computer models have influenced trading for decades. And not only would it be "irresponsible" to fully automate trading (as the article states), it would also be "illegal". Computer-driven market analysis and prediction is a huge industry -- the big firms spend vast amounts of money on it. I'm not seeing what's newsworthy here, for slash or for El Reg.
Slashdot quality declines as the number of hot grits posts decreases. - Provolt's Law, Apr-09-2005
This is nothing new, and it's not even something that's restricted to the world of money managers. It's being used by individual investors now, and has been for years; it's called "technical investing". The definitions of combinations of factors (market cap, financials, etc.) are called 'screens', and are a common source of discussion on forums like those found on The Motley Fool. There's software for sale, priced for individual investors, and there are websites that will even allow you to save your screens to use periodically, looking for new possible stocks to buy into (or to check and be sure that your existing portfolio matches the parameters you want).
For your security, this post has been encrypted with ROT-13, twice.
Wow, great summary... The computers wouldn't be replacing investors, but 'investment advisors'... That's a whole different rung on the ladder. If they replaced the investors, there'd be no money and the stock market would die.
As for replacing the advisors... Even the article tells you that isn't going to happen. "They reflect the manager's views." Oh... So if there's no manager, there's no view... and the computer does nothing. So you can't drop the advisor.
This is simply another tool. It's not going to change much. My father will still complain bitterly when his portfolio loses money, and complain a little less when he's almost back to where he's started... again. And again.
The fact is... If everyone made money, the stock market would be an impossible thing. Some people will lose while some will gain. No magic piece of software is going to change that.
"If you make people think they're thinking, they'll love you; But if you really make them think, they'll hate you." - DM
What they're talking about is arbitrage and trading, not investing. Their trades are designed to be in the short-term. Sometimes, very short-term - within a second.
We all know when it comes to predicting markets they will both be beaten by a gang of monkeys and a dartboard.
12:15 Press return.......
The key comment was:
More importantly, the models provide insight into market inefficiencies to be applied rapidly across asset classes and the vast number of financial instruments within those asset classes. Whole markets can be analysed daily for buy and sell indications at an individual instrument level. This enables portfolios to contain a larger number of instruments and reduce risk through greater diversification of the portfolio.
As inefficiencies are identified (such as when the return / risk ratio is not correct) provides an opportunity to increase returns by taking advantage of them. Of course, as more people use models the inefficiencies will be corrected quicker, leaving less opportunities to exploit. In effect, the market fixes itself. This, of course, is nothing new - markets adjust to new technologies all the time and eventually the opportunities they offered disappear; for example when the telegraph first came out no doubt someone discovered they could buy an item at one place for less then the same item where they were and arbitrage the prices - but as more people started doing that the spread disappeared.
I'm a consultant - I convert gibberish into cash-flow.
...cos I can already see the first search result output in front of me: "double the stockkiller delete select all"
;)
Would be interesting to see how an advisor would interpret -that-
Three rings for the Elven-kings in the sky
Why would it be "irresponsible"?
And no, it's not illegal, but there are some exchange rules on it - depends on the exchange. See here: Program Trading
I gues this is why the SEC is pushing to apply tags to various financial statement data submitted in the annual and quarterly filings. Now the computers can consistently analyze the financial statements and trigger trading, all without human intervention. Great.
Ibid.
The use of computers models to predict what to buy has been around for some time. The absolute belief in these models caused Long Term Capital Management to go under in 1998 ( see When Genius failed ). I also highly recommend reading Fooled by randomness
back in 1987 when automated selling by computers was blamed for making the collapse worse
The most popular explanation for the 1987 crash was selling by program traders. Program trading is the use of computers to engage in arbitrage and portfolio insurance strategies. Through the 1970s and early 1980s, computers were becoming more important on Wall Street. They allowed instantaneous execution of orders to buy or sell large batches of stocks and futures. After the crash, many blamed program trading strategies for blindly selling stocks as markets fell, exacerbating the decline. Some economists theorized the speculative boom leading up to October was caused by program trading, while others argued that the crash was a return to normalcy. Either way, program trading ended up taking the majority of the blame in the public eye for the 1987 stock market crash.
http://en.wikipedia.org/wiki/Black_Monday_(1987)
what no one has heard of Long Term Capitol Managment? Wasn't that their schtick too? The stock market can be reduced to a formula? Didn't they almost collapse the entire US economy!!! Nope, sorry...i'm not buying it. The market has always been about fundamentals over the long term (10yrs+) and emotions over the short term. This is just a way for money managers to abdicate responsibility for poor performance. "But the software said it should go up..."
ps. here's the Printer Friendly link
The efficient market hypothesis states that the price of a stock reflects all that is known about a stock. It is therefore impossible to outperform the market on a long term basis. http://en.wikipedia.org/wiki/Efficient_market
Warren Buffet has outperformed the market over many years. http://en.wikipedia.org/wiki/Warren_Buffett
Buffet understands the economy and invests accordingly. The computer programs only understand the market. In other words they can't really respond faster than the rest of the market. Buffet can be years ahead of the market.
So when a person is born, using models and best decision practices, the system(s) can predict what the earning potential (investment wise) for the new born is. Compiled with analysis of their genes, the life expetancy and health of the new born can also be determined. Then, the system(s) can decide if the new born will be a drain on society or not. Given enough computing power, data models, and algorithms, mankind's future will soon be predictable.
There are no loopholes. It's either legal or it's not.
I fail to see how a computer can have the same style that our fat bellied barrow-boys-in-suits have. A couple of years ago I was on a trading floor, one of the ForEx traders was in a heated conversation on the stentophone. I only heard his end of the conversation that went something like this:
Give me 2.22 you c*nt
No, I said 2.22 you c*nt
2.22
2.22
Can you f*cking hear me? I said 2.22 you c*nt
What part of 2.22 don't you understand you little pr*ck. 2 f*cking 22.
2.22 you c*nt
2.22 you c*nt
2.22 you c*nt
*pause*
There you go you jumped up little c*nt, didn't take too f*cking long did it? 2.22 done.
"2.22 you 11011100110011ing 1011101101" doesn't quite have the same ring to it.
Let's be honest here, the "human factor" was what made some huge enterprises possible. Because some humans believed in something the "numbers" alone didn't predict. Do you think Google would have found an investor in its early days?
Investment by numbers is by definition a rather conservative way to invest. In other words, put your money where there already is money. Risk investment is usually something done by visionaries, not by bean counters. And yes, 9 out of 10 times the idea bombs. But the one that works pays hundredfold.
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
http://science.slashdot.org/article.pl?sid=06/08/2 1/1646238
Until you can look at the numbers of a company and know everything about that company with certainty (meaning that a human _could_ do it if they had an infinite amount of time), or until we have computers that are great at telling when people (enron) are bluffing, I'll stick with investmant companies that rely predominantly on humans.
Mozy, free online backup service
--
Many times I have played with the idea to try to find patterns in stock history to aid prediction, but where to find the data set to work with? I'm not prepared to pay for it, of course.
Swedish plasma phys. PhD student; MSc EE; knows maths, programming, electronics; finance interest; seeks opportunities
Hedge funds are sometimes seen as the "smart money,'' and their managers hailed as market iconoclasts whose quirky, daring trading styles are central to their success. But some of the smartest traders are often beaten by an unlikely foe: the room full of man-high Hewlett-Packard computers that are the brain of AHL.
http://tinyurl.com/ke2ey
Let's assume that we could do that, setup a bunch of algorithms wich only purpose is to make money. And let's assume that this AI is much better than a human at making market decisions, it always pick the best choices, it always wins.
If that happens, and if all the investitors has access to such software, my bet is that the whole point of investing on market will flop. Since everybody will always "win", nobody will actually make a huge profit, the money will enventually be equally distributed among all investitors.
The sad thing about capitalism is that there's always a fixed amount of money on the market, so to win somebody else has to loose. If we make a super AI, that always wins... nobody will, because to win someone has to loose.
In the end this could mean a good thing, and we can turn our attention from pure profit to things that really matters.
---- You know how some doctors have the Messiah complex - they need to save the world? You've got the "Rubik's" complex
I suggest you go out and read "Debt of Honor" by Tom Clancy. As always, he fabricates and exaggerates, but all the same it will give you a very good idea why reliance on a computer trading model is a "Very Bad" thing. I.E. - if you know how a computer program works then you know how you can break it or cause it to react. If you can do that, then suddenly you have the power to control the market. Clancey went a bit further theorizing that a programmer was bribed to place an easter egg in the system, but all the same the ideas are there and are valid.
1. Mathematics is the language of the universe.
2. Everything around us can be represented and understood through numbers.
3. If you graph these numbers, patterns emerge. Therefore: There are patterns everywhere in nature.
So, I guess Max Cohen's theory seems to be panning out in real life. And as cool as that sounds to be able to model the stock market like that, I'll bet that they didn't accidentally stumble onto the structure of life while they were doing it.
Someone's going to write a trade-bot in VB.net and it's 1929 again.
Can I have an algorithmic manager? I wouldn't tell the difference.
When everyone is crunching numbers on their head, the computing might give an advantage. If everyone is computing with compters, the advantage goes to the one with better algorithm. And the better "algorithm" might actually turn out to be thinking and looking at the global picture instead of madly computing.
When huge trades are decided by these algorithms it is almost like a huge herd of milling cattle. When the stampede, just get out of the way. But if someone could trigger a stampede and send it over the cliff, there will be rich pickings later. And in free markets, if there is a way for someone to make money, someone will.
I think algorithmic investing is the new name for the old "programmed trades" and it might actually make the thinking and studying fundamentals investors richer.
The code is
if( InputDataIsGood() && AlgorithmIsGood() && AlgorithmIsBetterThanOthers()){
Output(profit);
}else{
Output(loss);
}
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
We've long used IT to spot and forecasts weather trends. I have long wondered why we couldn't make a system to predict market trends/events given any number of variables.
A friend of mine tried writing his own auto-investment code to see if he could actually make a living at it. His plan was to dedicate a computer to doing nothing but scanning the market and making investments for him. Well, he's not doing that today, and he's probably one of the most intellegent people I know. I also rememeber a concern that, if everybody used automated investing, the market would become highly sensitive to change (or...unstable) as you'd run into situation where most, if not all, of the algorythms out there would react the same way to certain market changes.
Still, it's intriguing isn't it? I mean, one of the things I use computer programming for is to learn how things work. I look at it this way; a computer is rock-stupid, but it does exactly what you tell it to do. So, if I could write a code to do market analysis, I'd be learning the intricacies of how to do it along the way. Sure, most invesment sites have tools for you, but there is value in learning the underlying mechanisms.
Seems the best approach would be to write such a program to simply do the analysis, then you make the final commit to buy or sell. You'll have a good idea how to interpret what you get back because you told it what to do in the first place and you should be able to spot errors/weaknesses in your strategy. It could be downright symbiotic.
A goal is a dream with a deadline
About 20 years ago I had a home 8-bit Z80 computer (Amstrad) running CP/M. I also had a board game called Winalot which involved buying and selling shares in four companies (Oil, Rubber, Gold and Diamonds). Share price was changed after each round according to up and down cards held by the players. What players bought and sold in each round was obviously influenced by what cards they held, and by seeing what others bought and sold.
I programmed the Amstrad to act as another player. It dealt itself "cards", you told it what the humans had bought and sold by typing it in, it displayed what it wanted to buy and sell for itself, and at the end of each round it displayed what "cards" it had held and you typed in what cards the humans had held. It also did the tedious job of banker, displayed the current share prices, and told us who had won at the end of the 20 rounds. That was usually itself!
Other human players (none of whom had ever seen a computer before) did not like it because they thought it was "fiddling" the bank - they could not understand how it resisted the temptation to transfer bank funds into its own account!
Obviously I had written a simple algorithm for its decisions on what to buy and sell - a very simple one. It was therefore no more "clever" than I was. Yet this thing would beat me (and most others) in the game 3 times out of every 4! Simply because I was using a bit of intuition, even unconciously, whereas it was sticking strictly to its algorithm.
As a party piece it went down like a lead balloon!
Thanks to the inputs of human labor and intelligence bringing increases in productivity into the economy, the stock market is not a zero sum game. Even accounting for inflation, there has been real growth in the overall value of the market, just as there has been real growth in the overall value of the world economy. So it IS possible for everyone to make (a little) money.
On the other hand, if everyone is making money in the same proportion to their level of investment, your subjective perception of your wealth won't be as high as your actual gain in purchasing power. That's because relative to everyone else, you haven't moved. The fact that you now have an HDTV and a new car seems less like a gain in wealth because everyone else in your neighborhood has one too.
And once you factor in commissions, account fees and so forth, you can see that the small investor is still more likely to lose money than larger investors, even if they are able to implement the same investment strategy.
We are the 198 proof..
The problem with developing a perfect algorithm to predict the market is in recognizing that your algorithm is also a participant in the market. Any actions it takes will also affect what it's predicting. It doesn't introduce much error on small scales, but on any significant investment levels, the distortions on a single stock will be pretty chaotic. It's probably impossible to both perfectly observe the market and to participate in it.
What a strange bird is the pelican, his beak can hold more than his belly can.
I recommend every aspiring trader have a read of A Mathematician Plays the Stock Market by John Allen Paulos.
The book is pretty much as it sounds. While the author doesn't *actually* invest in stocks, he *is* a mathematician and he plays through (mostly with logic) ways to get ahead in the stock market game. As you would probably guess, it's not easy.
A great read. Sadly, my dreams of a quick fortune by computing stocks were quickly squashed by his well presented arguments.
Cool, but useless.
If fundamentals were *all* that investing was about, Oh! how simple. Just read Bernstein on asset allocation. And wait a decade or two.
Every algorithm -- read: trading strategy -- is inefficient in it's own way. Markets are random walks that remove inefficiency, at least over the long term. One can't eliminate extraordinary popular delusions and the madness of crowds. Let's program around them!
In the short run it's all about crowd psychology. In the markets, money is made in the short term by *taking* it from another investor.
This will go nowhere except for book and secret method sales.
It's called risks management (or whatever the traduction is...), and you use statistical stuff to do it... then go to a spreadsheet, put some infor and make some functions and voala, you get a result saying you should invest or no.
All new about this new, is that you don't have to tipe all the info on the spreadsheet.
Yahoo's financial site has data. (The only free place I've found too.) finance.yahoo.com, look up a quote, click the graph, on the bottom of the page: "historical prices." Set ranges and frequencies for data, then at the bottom again, "download as spreadsheet," CSV.
The snow doesn't give a soft white damn whom it touches. -- ee cummings
Just find the 216 digit number.
Just plain bogus. Enron would have looked great on paper and to any computer algorithm. A keen investor however would have interviewed employees and performed other tasks a computer can't easily do to find that, hey, something is really F*#CKED UP at this company. Same for Krispy Kreme a couple of years back, and a thousand other examples. I can see (and use) computer programs as screening tools, but I still perform personal research to decide if a "screened" company is worth investing in.
First, there are patterns. You are right that those patterns have a limited capacity to absorb trading and that anyone who finds a pattern would do well to keep it a secret. The EMH (Efficient Markets Hypothesis) is best lampooned by the following old joke. Two economists are walking down the sidewalk when one of them spots a $100 bill in the grass. The first economist starts to pick up the $100 when the second economist tells him, "Don't bother, if it were a real $100 bill, someone would have already taken it." Moreover, EMH makes predictions about the statistical distribution of price movements and the volume of trading that are empirically false.
Second, the price dynamics are not entirely caused by exogenous factors. Investors, speculators, the media, and government officials do watch the prices. People make buy and sell judgments without any fundamental basis such a stock being "expensive" just because the stock is $300/share (never mind understanding the relationships between price per share and capitalization). Humans also have instinctual beliefs in patterns such as trends or momentum that are self-fulfilling. If enough traders believe in trends or momentum, they will trade in a way that creates trends.
The profitable patterns do exist (and so do a large number of profitless pseudopatterns). People with a very sound understanding of both market psychology and statistics can and do succeed.
Two wrongs don't make a right, but three lefts do.
Vectorvest has been doing this for years. -ted
shouldn't that link read "that computers will replace CEOs"? :)
... how probable it is to rain.
ona more serious note: i believe it is possible to
have toaster like device at home doing "intraminute"
trading with ~garunteed gains
acctually having a few terra flops, enough storage and
a good compression algorithem should enable you to
"database" track all money going in and out of the market.
if you can track that, it would be like if you were able to track
total water evaporation
of course our planet doesn't create water, but some countries
create (read print) money (out of thin air), so tracking that might
not acctually work.
You forgot that governments print money. They make it out of thin air and devalue currencies. The US government is doing exactly that right now. There are inputs to the economy, the sun, oil, human inventiveness which mean it isn't a zero sum game, it's only zero sum on the very shortest timescales.
Sorry, you're young... You haven't yet figured out that markets, competition are all based on the desire to reproduce. It isn't ever going away while there are human beings.
Deleted
I have a friend who worked in the hedge fund game for a number of years. He's a brilliant mathematician, and worked on the models they used to inform their trading. The group he worked with was quite successful, and make a heck of a lot of money.
One of his most interesting comments: "The model can inform your decisions, but you have to know when to NOT trust the model." Another of his comments on a completely different talk: "Mathematical models are never perfect, but they can be useful."
The trading system can be modeled, but you can never capture all the true complexity of the real world. If you leave the model to do it's thing, if I know how it's going to act, I can game the system. If the world changes in a way that the model builders did not predict, then the system will also act inappropriately.
I can't imagine ever getting rid of all the traders out there, though I imagine expert systems will become more 'expert' as time goes on.
Headlines read: COMPUTERS PREDICT ROULETTE ROLLS WITH MATH!!!!! Then: ALGORITHM PREDICTS ANYTHING THAT EVER HAPPENS IN THE FUTURE !!!
Anything can, could, and will happen.
Use mmm, a bayesian classifier to rate news announcements for positivity/negativity towards a particular stock. Rather than trying to develop an AI, let evolution do it for you. Download the news, stock info, classify the news pages, hand all the info on the market to a couple of hundred thousand simple genetic algorithms, kill off the bottom 10% or so and mate & mutate the rest. Over the generations some sophisticated algorithms will develop which can use factors like news information and the likely response of the market to that information.
Huge amounts of CPU power and time required, but you end up with algorithms wandering the probability landscape of the market looking for money. Not necessarily intelligent, AI but pretty sophisticated and highly optimised nevertheless which "understand" the market as well as any human possibly can.
Deleted
I'm sorry, but there is just no way possible for you to make the statements you've made and call them "authoritative."
You're assuming that since you haven't found a pattern, and since nobody has suggested a successful pattern, that one does not exist.
Honestly, if you found a reproducible, reliable pattern in stock market data, would you give it away? How long would you take advantage of the situation before releasing your research in the name of science? Even if you're not a greedy person, who would give away the goose that lays the golden eggs? Perhaps more to the point, as you said yourself, who would *slaughter* the goose and then eat it for dinner? It's almost "Heisenbergian" that the usage of such research on a mass scale would change the dynamics of the market and render your pattern useless.
Your assertion carries about as much weight as saying "Aliens do not exist." Your evidence is that people have searched and nothing has been found.
I'm not suggesting that there are people who have identified or developed such an algorithm but I am suggesting that it is not beyond the realm of possibility. This is classic game theory. There are mountains of research available on how this might be possible, all that's left for a financial-programmer-genius is to apply the theory and fill in the gaps. A tall order, perhaps, but success would mean untold treasure. Just as people tromp all corners of the globe looking for buried treasures and lost cities, I'm sure many have devoted their talents and fortunes to this.
A college friend has been doing this for the last few years with great success. Go figure...
That said, it is pretty easy to statistically spot deviations caused by "irrational exuberance". Even so, a computer betting against the "new economy" rally of the late 1990's would have suffered some deep drawdown before it payed off.
And, Mechanical Investing, as it's called, is a pretty popular hobby among some fools.
the PHI number is related to stock market prediction in the long run. There are calculations that show the approximation of 1,618 between different volumes of stocks. The only think that is specific about the stock market is that you can't invest based on this if you're into the short term investment. If you're willing to buy a stock for the next one or two years, then maybe these algorithms will be able to help you a lot when analyzing patterns. It all depends on the way you invest and how you do it. These algorithms, in this moment, are useful to help investors analyze data, but not for doing operations automatically. Automatically in this case would be automagically.
-ld
apparently the mods haven't seen the movie.
So yes people are talking about using markets to predict terrorist attacks, or major weather events, etc. (Markets are reasonably good attractors of information.)
And since the dawn of computers people have been searching for the magical program that could predict stock growth with enough accuracy that you give it a little money and retired to a tropical island.
While a computer program might be able to match or beat the average investor in the stock market, that is only because publicly traded companies have to report audited data. And even so I don't expect computers to beat someone like Warren Buffet in the near future.
Computers will not beable to predict terrorist reports nearly as well. If terrorists produced annual reports and quarterly statements listing their plans, then we wouldn't even need predictive markets.
Think Deeply.
... it will buy it.
> Dependency on programming and computer nerds has not replaced the dependency on star investment managers.
I love the way this a$$holes look down on computer programmers. Why not say "dependency on star programmers has not replace the dependency on brokerage nerds?" This is why no one in their right mind is entering software now. Programmers get no respect. But hey, I'm not bitter.
I forgive you for not giving credit since you don't know any better, but your friend should. "All models are wrong, some are useful" is a quote from George Box, who founded the Statistics department at UW-Madison of which I'm a graduate.
http://www.americanbulls.com/
It is based on reading candlesticks (look that up) and ideal info for the swing trader. As with any of these things, you have to know when it is out of whack, but candlesticks have been used for centuries.
You're buying _A_ market, the one encapsulated by the index you buy into. The market that you buy into may, itself, underperform, especially over a given time period.
No guaranteed win, just a conservative and usually-effective risk/reward tradeoff.
Whence? Hence. Whither? Thither.
There's a huge amount of automated trading going on, and there has been for years. The effects on the market are well known. For one thing, arbitrage margins are now tiny. If there are some things known to be related, like the stocks underlying a mutual fund vs the fund price, or crude oil vs. gasoline, or futures prices vs. current prices, program traders are tracking that relationship constantly and will trade if a spread develops. As a result, there are no big spreads in those areas. Once enough traders are doing it, it stops working.
Elaborate attempts are made to find more complex relationships like that to exploit, which is what derivatives are really all about. That, too, has been done to death.
As a result, efforts are made to exploit very tiny arbitrage margins by making huge opposing trades. This works until there's some common failure mode for both side of the trade. (See "Long Term Capital Management" and "portfolio insurance").
Then there's automated velocity trading; if it's going up, buy it quickly. Now that produces market instability. This was popular during the dot-com era. It's less popular now.
None of this is new; it dates back to at least the 1970s.
for a time.
7 29520-1412955?v=glance&n=283155
Check out: "Eudaemonic Pie" http://www.amazon.com/gp/product/0595142362/102-3
and
http://www.thomasbass.com/work2.htm
I have written more than one automated algorithmic trading system for more than one international investment bank. The only human interaction is the trader specifying what position he would like to take, and which algorithm he feels is going to get him that position by the end of his investment period.
After that, the trading system is analysing market data, applying some relatively simple rules and issuing market orders directly on the exchange. No human clicking a 'yes'/'no' button at all.
Illegal - hardly !
Common - extremely
In the short-term, the driving factor is human psychology. Things that are mathematically "supposed" to happen don't necessarily. The smartest mathematicians, economists, and programmers have tried this, and the resulting disaster nearly toppled the global financial system:
Benjamin Graham, the father of value investing and Warren Buffet's inspiration, even talked about this. You simply cannot take the previous behavior of a chaotic system (and in the short term, the stock market certainly qualifies as such) and extrapolate the future. Any predictions you might make successfully will still only result in very small gains, so you'd have to bet enormous sums of money to make any sort of return. Then when you get the unexpected 10, 15, 40% drop in the market (October, 1987 anybody? Nobody saw that coming) you're screwed.
If you want to make money in the stock market, it's really not that hard. Pick some good companies that you have reason to believe will grow their earnings over the next 5-10+ years, hopefully paying good dividends along the way, and use dollar-cost averaging to minimize losses and maximize gains. But the act of picking the companies in the first place is not a mathematical problem, but one at the intersection of business, economics, psychology, sociology, and politics.
I'd love to see how the approach calculates:
Apple without Steve Jobs
Microsoft without Steve Ballmer
I'm not saying anything about these particular stocks, but you get the idea. The post of research director in Specialty Chemical (e.g. Biotech, Pharmceutical) companies is even more critical.
Also, what happens when eeveryone uses the same approach? Any competitive advantage your investing strategy goes away.
One of his most interesting comments: "The model can inform your decisions, but you have to know when to NOT trust the model."
:-)
That's a pretty trivial observation. If you understand the model, you know its limits. Once you go beyond this limits the model's outputs are just garbage because it's beyond its capabilities.
To give a simple example, at the time of the Iraqi war (that is, when the US troops were moving to Baghdad) most short-term stock market models were useless. The US market was reacting to daily casualty numbers and not much else. You don't have to be a genius to stop trusting your models at times like this.
Another of his comments on a completely different talk: "Mathematical models are never perfect, but they can be useful."
That's a derivative of a much better quote by George Box: "All models are wrong; some are useful".
Kaa
Kaa's Law: In any sufficiently large group of people most are idiots.
or, at least, that most of the investors participating in the stock market are intelligent. That's just not true and it's been becoming less true every year as more and more average folks pump their retirement monies into one of the few sinks big enough to absorb it all.
every time a new version of the interweb comes out, just buy stocks from all those web X.0 companies, sell them a few days later. If the pipes aren't glogged, you can't go wrong with that.
This is totally off-topic, but I thought I'd try for the /. opinion anyway...
Basically, I'm now working for (read: being micro-managed by) a large, successful manufacturing company. I live in my hometown, am close to my family, have about as good of job security as is realistically possible, am reasonably compensated, have a comfortable 12-minute commute. I bought a house two years ago.
A close and trusted friend has been in this "algorithmic trading" business for the last few years. He's basically kicking out on his own, with a small number of other invdividuals experienced in the same. And I'm invited.
Taking this job would mean a huge increase in pay (almost 2x what I make now, 3x if the company is profitable and pays the expected bonus). Plus, being part of a tiny company, I would have an immense of amount of ownership in my work, and actually have a sense of accomplishment at the end of the day. But I'd have to move about 3 hours away from my hometown, family and serious girlfriend.
What do Slashdotters think?
It was actually Black Monday.
My mother always told me not to stare at the sun. So one time when I was six, I did.
I mod down pyramid schemes in sigs.
Another friend was a quant for a while around 1990. He and some coworkers found a few sets of patterns in the market that they were able to arbitrage - it makes you a pile of money for a short time, until the market adjusts to it (that's *how* efficient markets work - people find inefficiencies and exploit them, and the first people to find them can make money if they get back out again before everybody else stomps them.) Having faster computers means you can find and grab smaller inefficiencies and make smaller chunks of money off them sooner, leaving fewer big inefficiencies around.
Of course, that doesn't let you predict whether or when Bush or bin Laden is going to pop up and say "booga-booga!" and jack the oil price or increase US government borrowing even more radically than predicted; you have to be an insider to get that, though it does help to keep a range of models around to predict the effects.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
The novel also shows the end point of algorithmic trading is the end of mankind.
Program trading is HUGE already. The bulk of that trading is probably ETF's and other mutual funds that are automatically rebalancing to match changes in the indexes and what not. Another chunk are programs catching split second arbitrage opportunities. So it's not quite the same type of investment that is implied by the summary, but it's scary nevertheless. Who knows how all of the programs will interact with each other over the open market during a really weird chain of events? It could be disasterous... for people on the loosing end of that day, anyway.
My quick google search for a reference came up with this direct from the NYSE: 70% of trades are program trades in June of 2004. It's an old article, IIRC, I've seen a more recent number that was closer to 90% of all trades being program trades.