Slashdot Mirror
Have We Hit Peak HFT?
CowboyRobot writes "There was a time when people wanted the fastest networks so that they could trade at lightning speeds. They deployed the smartest formulas at trading venues where no one could know who was asking for that big block of stocks on the other end of the deal. It was a wild time and people made a lot of money along with some very unwise decisions. Wall Street seems to be acting out the lyrics to a Don Henley song. The party's over, the hangover is raging and no one really knows what happened the night before. The number of shares traded via high-frequency trading are down and politicians want to roll out a tax to serve as a speed bump. Iowa Senator Tom Harkin and Oregon Representative Peter DeFazio want a .03 percent tax on nearly every trade in nearly every market in the U.S. Some are wondering if microsecond dealings are poised to fade away. As the founder of HFT firm Tower Research Capital Mark Gorton puts it, 'The easy money's gone. We're doing more things better than ever before and making less money doing it.'"
They would know that people who execute a lot of trades get big discounts. HUGE discounts. If you are doing more than 20,000 or 100,000 trades a month, you start to trade at a fraction of the cost of a normal investor.
http://www.interactivebrokers.com/en/index.php?f=commission&p=futures3
With this fee structure, a HFT can get in and out and make money. A normal trader would lose money on the exact same trade. For the average guy, the trade might be profitable, but the trade commission is greater than the profit. Exchanges give a lot of incentives to traders who bring big volume.
I have a major issue with a system where two different traders make the exact same trade, but one loses money while the other makes a profit.
I'm not suggesting eliminating the lower fees for more active traders. I would like to see the gap between the highest commission and the lowest commission close. This simple change would work wonders to level the playing field. If the highest commission was limited to 130% of the lowest commission, the high frequency traders would lose most of their advantage.
If HFT is the latest shtick in "providing liquidity", then it really is arbitrage. And by its very nature, do enough of that and the opportunities exhaust themselves. In fact, the infrastructure built to support HFT improves access for other traders, too.
I think taxes may not be necessary, and probably would do more damage than good anyway. The big trouble with HFT isn't the trade volume, but the "probing" volume, offers made then withdrawn so quickly nobody can take'em.
I'm with the NANEX bunch on this; simply require that offers be good for a minimum timespan so that completely fake offers are unaffordable, for someone might take the offer. That saves the cost and administrative overhead of another senseless tax that hits far wider than the problem it purports to solve, but then doesn't.
Mostly true, but simplistic.
HFT is a symptom of a deeply broken system. It brings liquidity, which is good, but it has tended to move the market in unreasonable (and sometimes dangerous) ways. So many trades today are HFT versus HFT that it obscures the real markets. We've seen several clear instances where computers are terrible at predicting the real market values when the primary data they're competing against are the actions of other computers. HFT doesn't lead to good price discovery, it obscures it.
When talking about long-term stewardship of institutions, we really need to move away from unreasonable earnings expectations every quarter. When the focus of the market is on ever narrowing periods of time, corporations simply cannot properly invest in the future. Yes, some corporations are lead by unusual people who can leave market panic behind them and truly lead... but most companies are run by very smart idiots.
Ultimately, stock/commodity/bond markets are about profit. There is nothing inherently wrong with that. The question for the rest of us is: what is the social value? Wall Street has some social value, but it also has considerable social expense. Sometimes we forget that. We also tend to forget that "the markets" are still based upon similar principles that brought us the great depression. Yeah, they're more sophisticated now. Yes, there are "regulations" (occasionally enforced). But the market CANNOT bring prosperity by ignoring it, and letting it run on its own.
Stock/commodity/bond markets are not capitalism. They are merely outgrowths of one particular variation on capitalism. They're finicky, hard to properly balance, and break if you try to exert too much control.
I won't join Slashcott. OTOH, If Beta goes live, I just won't be back until it's fixed. Sorry Dice.
The main reason for HFT is to "front run" the market, to game the traditional customers of the price discovery mechanism, and make a risk-less profit. This dis-incentivises the market for everyone else, who see it as corruption and move their money elsewhere.
The big picture though is one of a big liquidity event, in which the velocity of money is rapidly falling as everyone tries to save up enough cash to ride out the oncoming greater depression. The rapid printing of money is showing up in the 17% growth of the M3 Money Supply, but is getting hoarded up by banks and corporations as rapidly as it's getting created. This is the only thing keeping inflation from at bay, for now.
Once the Tsunami of dollars starts to find its way to main street, and chasing goods and services, an inflationary wave will hit us all, and we'll learn to get used to $10/gallon gasoline, and they start to remember it fondly not much later.
Ah, but you didn't consider the effect this will have.
The result will be that share prices become normalized in a certain sense. Not too large to allow trading, not too small to allow reasonable trade taxes.
If Pandora's box is destined to be opened, *I* want to be the one to open it.
The share tax in France and Italy doesn't affect us investors much. But if you're in to make a 30% return, you don't care about the 0.03%+0.03% buy and sell taxes anyway. It's only the traders, and they add nothing to the investment market, they're just skimming off some of the margin between buyer and seller.
IMHO, the biggest problem with the stock markets is delayed quotes. The big guys have the real time prices, the small guys have a fake price, from 20 minutes ago. This is a big problem with European stock exchanges. How can you have a market if you're lying to people about the current prices?!
That's what I'd like to see ended. The two tier price quote system.
The stock markets make us little investors deliberately ignorant of the true price, they then sell that ignorance as 'real-time-quotes- to the big guys. They should be forced to charge one fee to everyone for the same quotes, and stop deliberately deceiving investors into the market.
Indeed HFT trades with HFT, there is no money in that. Or at least not much, as they all chase the same fractions of a cent.
The flash crashes as I understand are partly caused by all HFT systems using essentially the same algorithm, and as a result movements amplify really quickly. If there would be several radically different decision making algorithms in the market this shouldn't be much of a problem, as wrong decisions by one are taken advantage of by another, so the other can make a profit (directly punishing the bad decision of the one algorithm), and such market movements are smoothed out.
HFT is a symptom of a deeply broken system. It brings liquidity, which is good, but it has tended to move the market in unreasonable (and sometimes dangerous) ways.
That is a bit simplistic as well. First, stepping in front of a trade because you can (1) see it before the other guy (due to having lower latencies in talking to the exchange or having priority 'fast' data feeds) and (2) quote using fraction of a penny increments as opposed to pennies for the 'regular' guys is not adding liquidity. In fact, from the way flash crashes happen you can see exactly how HFT does not add liquidity with a stock/future going bidless and crashing. Properly supplying liquidity would not allow for something like that to happen.
When talking about long-term stewardship of institutions, we really need to move away from unreasonable earnings expectations every quarter.
Who is we? The short-termism here is the market speaking. The idea of markets setting the 'right' price is highly dependent on your definition of 'right'. As the saying goes, markets can remain irrational longer than one can remain solvent (and bet on rationality). Nevermind the fact that people in the market have different types of interests which are not always aligned to long-term value. See corporate raiders, LBO, and so on - case in point, Icahn.
When the focus of the market is on ever narrowing periods of time, corporations simply cannot properly invest in the future.
That's not entirely the problem. When the upper echelons of the corporations receive stock options with short vesting periods (no, 5 years is not long term compared to typical investment horizons, and 5 years are not really that frequent) the motivation to invest in the long-term future is basically absent. Also, there are industries where that is simply not possible, as the prevailing conditions fluctuate too much.
Ultimately, stock/commodity/bond markets are about profit. There is nothing inherently wrong with that. The question for the rest of us is: what is the social value?
Indeed, profit is the key word. Social value is incidental, if at all.
Wanna bet? :D Without speculative currency trading, the apparent value of a currency becomes disconnected from the actual value. Thus you see in countries with currency controls (e.g. Venezuela) the 'black market' value of a US dollar is several times that of the 'official' price. Speculative currency trading is what allows the liquids in different buckets to flow to their equilibrium levels. And HFT trading just makes that happen faster. I responded in more length about HFT to an earlier comment, which IMHO makes the case for HFT (now that the tech bubble aspect is over) as a useful, even essential part of everyday market equilibration.
IOW without speculative currency trading, when you go to the country next door you will have no idea what the price of goods is. (And as the Euro debacle demonstrates, sovereign countries need to have their own currency in order to have a way to balance their economies. Devaluing a currency is akin to a 'store-wide sale' for a country. To a good first approximation, a currency _is_ sovereignty.)
It's easier to be a result of the past, but more fun to be a cause of the future! http://www.spacefinancegroup.com/
That's not true, unless by smaller "Mom and Pop" investors the author really means semi-professional, day-trading arbitrageurs. I am a small time investor trying to make retirement savings grow by hunting for safe stocks with high dividend yields. What matters much, much more to me is that I can't get a decent interest rate on CDs or money market because of pro-Wall Street and pro-big bank FED policies.
Why does HFT matter to me? It really doesn't. Sure, in theory some HFT algo is going to snap up bargains ahead of me, but they're also shorting the bogeys so I'm always going to get the fair price. Stop the FUD.
NOT DUE TO ANY ACTUAL CHANGE IN STOCK VALUE BUT DUE TO PING TIME.
Wrong. this is no different than arbitrage has been for centuries, only faster. Baron von Rothschild used this to make money on the defeat of Napoleon at Waterloo, since his messengers were hours faster than the official royal messengers.
The 'vicious cycles' were due to the early primitive nature of HFT algorithms, and the fact that there weren't enough of them (so often the HFT was on one side of a spread, and not on the other => big money transfer). TFA shows that the bubble has burst, and HFT is now going to evolve and perhaps already has evolved to a useful, stable part of the overall market. Now when one HFT senses an opportunity and makes a trade that influences the market, another HFT is likely to spot the ripple and make a countering trade. So HFT now just damps the ripples.
I could say that there is an effect analogous to the 'vacuum energy' of quantum mechanics. Let's define 1c in US trading as the 'quantum of money' - the minimum price difference. Then there is always going to be a random fluctuation of +/- 1c at very close to the maximum trading rate for HFT. and there will be random fluctuations at +/- 2c at some fraction of that maximum rate - probably 1/2 the rate. It will follow an inverse power law distribution, so there will always be a 1/N probability of a fluctuation of +/- N cents in the minimum trading period - call it 1 usec. So the distribution will be on the order of T/N for T= a given period of microseconds. This means that there is a tiny, but real, probability _from pure chance_ that a price difference will be large. But it will also be countered, and damped, within a small amount of time - you saw that several times, as sudden market errors were compensated for within a few minutes, and the market basically continued on as normal.
Your bit about buy cancellation has nothing really to do with HFT.
It's easier to be a result of the past, but more fun to be a cause of the future! http://www.spacefinancegroup.com/