I had no idea stock markets operated so quickly now:
High-frequency traders may execute 1,000 trades per second; exchanges can process trades in less than 500 microseconds (or millionths of a second).
In addition to showing off just how blazingly fast financial transactions have become, this also demonstrates just how much more precise and reliable some networking hardware is, when compared to consumer stuff.
For the sake of comparison, I sent four packets from my home computer to the server that runs this site. It took them an average of 92 milliseconds to make the journey: 184 times longer than the rate at which exchanges can apparently process trades. Indeed, the difference between the quickest and the slowest packet to return was itself six times longer than the total processing time.
Clearly, those on dial-up connections need not apply.
The invisible hand is going to get arthritis!
It’s possible that all this makes markets behave more like the idealized forms economists have faith in. By eliminating lags and eliminating price disparities through arbitrage, trading at high speeds might make markets more rational and efficient.
I learn such interesting things, reading your blog.
Many of these firms pay large sums for the right to physically locate their trading systems inside the stock exchange building, precisely to eliminate latency. Speed-of-light time from New York to Hong Kong is 43 milliseconds – which is forever.
http://www.wallstreetandtech.com/resourcecenters/low-latency/showArticle.jhtml?articleID=199702208
Very interesting discussion of the social merits or otherwise here:
http://infoproc.blogspot.com/2009/08/high-speed-trading-8-billion-per-year.html
Thanks for the information.
“It’s possible that all this makes markets behave more like the idealized forms economists have faith in.”
Why? Isn’t equal access to markets an assumption required for idealized markets? Isn’t this the opposite of that?
Basically, by making the markets operate in a smoother way, by eliminating disparities between prices. If someone in one place is selling an asset for a different price from someone somewhere else, rapid trading could bring the two quickly into line. This could conceivably make markets behave more like models that expect immediate transmission of price information.
The issue of unequal access comes up especially in relation to flash orders, as described in the linked article.
Equal access to markets, i.e. no poverty, was something Adam Smith was quite insistent on. He probably passes the “not morally repugnant” test to a higher degree than Jefferson or Marx.
Adam Smith, The Wealth of Nations, Chicago: University of Chicago Press, 1976 (original 1776). An excerpt (Book I, ch. X, p. 111):
The whole of the advantages and disadvantages of the different employments of labour and stock must, in the same neighbourhood, be either perfectly equal or continually tending to equality. If in the same neighbourhood, there was any employment evidently either more or less advantageous than the rest, so many people would crowd into it in the one case, and so many would desert it in the other, that its advantages would soon return to the level of other employments. This at least would be the case in a society where things were left to follow their natural course, where there was perfect liberty, and where every man was perfectly free both to chuse what occupation he thought proper, and to change it as often as he thought proper.
Adam Smith is hardly the cutting edge of economics.
I was referring more to thinks like the Efficient Market Hypothesis.
I don’t think most high speed traders are incorporating meaningful new information into stock prices. Rather, the linked article makes it seem more like they are exploiting statistical tricks.
The ‘$8 billion per year tax’ article that Mark linked seems to support that idea.
Two defences of high speed trading:
a bestiary of algorithmic trading strategies
The three stooges of the high frequency apocalypse
For the record, I don’t think high speed trading does any real harm, it just diverts a lot of talented engineers into a largely pointless arms race. It almost certainly does make markets more efficient in a technical sense, but really not in a way that matters to the real economy.
That’s a good point. Investment banks employed a number of the cleverest people I knew at Oxford at one time or another, and it seems likely that their skills could have been applied to greater human benefit elsewhere.
Saying markets are internally efficient is about the dumbest thing I’ve ever heard. Markets are inefficient because they don’t price externalities. Saying markets are efficient if you ignore externalities is like saying AK-47s are safe as long as you ignore the bullets that hit people.
The particular meaning of efficiency is important here.
Basically, if someone had a way to consistently predict how the market would move, they would be able to make masses of money. The fact that they cannot means that, as far as gaming the system goes, all the relevant information about an asset price is already incorporated.
(It’s a bit more complex. Sometimes, you can identify an asset as being overvalued or undervalued, but you cannot stay solvent long enough for the correction to occur.)
It’s not about crafting good societal outcomes, but rather about the game theory circumstances players find themselves in.
It’s not clear to me why high speed trading would automatically increase the magnitude of negative externalities.
The individual in society
SIR – In his defence of economics, Robert Lucas, of the University of Chicago, maintains that: “the term ‘efficient’…means that individuals use information in their own private interest. It has nothing to do with socially desirable pricing; people often confuse the two” (Economics focus, August 8th). These two notions of efficiency may be distinct, but the entire justification for a free-market economy is that it should be easy to confuse the two. When the objectives of societies and markets have so clearly diverged, something has gone badly wrong.
Philip Roe
Ann Arbor, Michigan
The SEC yesterday came one step closer to banning “flash orders” in an effort to restore some fairness to daily stock trading. Flash trades, as the WSJ explains, “give certain large traders sneak peeks at market activity” for up to half a second before they commit to a transaction. According to the WSJ, “SEC Chairman Mary Schapiro said flash orders may result in a ‘two-tiered market’ and noted ‘the interests of long-term investors should be upheld as against those of professional short-term traders when those interests are in conflict.’ ” The Washington Post reports that the proposal to ban flash orders is part of a larger initiative to level the playing field for traders. Up next in the SEC’s sights are high-frequency trading practices and so-called “dark pools, which allow traders to buy and sell big positions without disclosing ahead of time what they’re looking to trade,” the newspaper writes.
Time is money
If traders are located 100 miles away from an exchange, they face a delay of one millisecond whenever they seek to trade a price via their computer screen. Few serious investors can afford to be that late to prices that flash so quickly. The blink of a human eye takes 300 milliseconds; many traders now operate in the smaller realm of microseconds.
Here is the longer story. I liked this bit:
Mr Greifeld said there might have to be measures to ensure speeds within the co-location facilities were the same. “We might have to give everybody the same length cable, believe it or not,” he said.
American stockmarkets
High-speed slide
Nov 12th 2009 | NEW YORK
From The Economist print edition
What is good for cutting-edge traders may be bad for the market as a whole
AMERICA may have been the epicentre of the global financial earthquake but it still boasts the world’s deepest and most liquid capital markets, unparalleled when it comes to nurturing young companies. But for how much longer? This week, as bulls cheered the Dow Jones Industrial Average touching a 13-month high, they were forced to digest a sobering report from Grant Thornton, an accounting firm, on the structure of America’s stockmarkets.
The report’s authors say that America is plagued by a “Great Depression” in the number of listed firms that stretches back over a decade (see chart). For all the talk of a revival in initial public offerings (IPOs), domestic markets are on track to add a mere 50 or so new companies this year, one-seventh of the level needed to offset the average annual loss of listed companies in recent years.
The slide in listings began in the mid-1990s, at around the time that America saw an array of regulatory changes designed to advance high-speed, low-cost trading, such as the introduction of online brokerages and new order-handling rules. An accidental victim of this technological revolution, the report says, was the ecosystem that helped bring small firms to market and then nourished them once there. “It’s a bargain-basement market today,” says David Weild, a co-author of the report. “You get what you pay for, and that’s nothing but trade execution.”
“Say you’re working for the commodities desk of a big investment bank, and a major client — a pension fund, perhaps — calls you up and asks you to buy a billion dollars of oil futures for them. Once you place that huge order, the price of those futures is almost guaranteed to go up. If the guy in charge of asset management a few desks down from you somehow finds out about that, he can make a fortune for the bank by betting ahead of that client of yours. The deal would be instantaneous and undetectable, and it would offer huge profits. Your own client would lose money, of course — he’d end up paying a higher price for the oil futures he ordered, because you would have driven up the price. But that doesn’t keep banks from screwing their own customers in this very way.
The scam is so blatant that Goldman Sachs actually warns its clients that something along these lines might happen to them. In the disclosure section at the back of a research paper the bank issued on January 15th, Goldman advises clients to buy some dubious high-yield bonds while admitting that the bank itself may bet against those same shitty bonds. “Our salespeople, traders and other professionals may provide oral or written market commentary or trading strategies to our clients and our proprietary trading desks that reflect opinions that are contrary to the opinions expressed in this research,” the disclosure reads. “Our asset-management area, our proprietary-trading desks and investing businesses may make investment decisions that are inconsistent with the recommendations or views expressed in this research.”
Banks like Goldman admit this stuff openly, despite the fact that there are securities laws that require banks to engage in “fair dealing with customers” and prohibit analysts from issuing opinions that are at odds with what they really think. And yet here they are, saying flat-out that they may be issuing an opinion at odds with what they really think.
To help them screw their own clients, the major investment banks employ high-speed computer programs that can glimpse orders from investors before the deals are processed and then make trades on behalf of the banks at speeds of fractions of a second. None of them will admit it, but everybody knows what this computerized trading — known as “flash trading” — really is. “Flash trading is nothing more than computerized front-running,” says the prominent hedge-fund manager. The SEC voted to ban flash trading in September, but five months later it has yet to issue a regulation to put a stop to the practice.
Over the summer, Goldman suffered an embarrassment on that score when one of its employees, a Russian named Sergey Aleynikov, allegedly stole the bank’s computerized trading code. In a court proceeding after Aleynikov’s arrest, Assistant U.S. Attorney Joseph Facciponti reported that “the bank has raised the possibility that there is a danger that somebody who knew how to use this program could use it to manipulate markets in unfair ways.””
“Satisfying the technical needs of these speed merchants is hard. High-frequency traders execute thousands of trades a second. Exchanges are required to process trades in microseconds (millionths of a second). They also need to increase the speed with which traders receive market data to feed into their algorithms. These upgrades are expensive. Singapore Exchange’s plan to build one of the fastest platforms in the world, capable of executing trades in just 90 microseconds, will cost an estimated S$250m ($185m). NASDAQ’s system is currently the fastest with trade times of 177 microseconds (see chart).
For the traders themselves, expansion abroad makes sense. HFT uses automated strategies to capitalise on inefficient pricing of financial instruments at blinding speed. As markets in America and Europe have become more competitive—HFT now makes up over 60% of equity trades in America and nearly 50% of British transactions—bid-ask spreads have narrowed and arbitrage opportunities exist for ever-briefer periods. In newer markets traders can use simpler algorithms for higher yields.”
“The high-speed trading exchanges that conduct the business of buying and selling stocks and mutual funds are so fast that hackers can introduce delays of a few microseconds completely unnoticed by today’s network monitoring technology — and manipulate prices in the process to reap millions of dollars to the detriment of everyone else, InfoWorld’s Bill Snyder reports. This kind of activity creates new reason to distrust Wall Street and shows how the computer networks we all rely on for conducting business and moving information are ripe for undetectable hacking.”
But as Andrew Haldane of the Bank of England pointed out in a speech last month, HFT firms may be benefiting at the expense of other investors. This is an “adverse selection” problem. Instead of being better informed about a company, HFT outfits are simply seeing and acting on market prices sooner than competitors. “To be uninformed is to be slow,” he said. “These uninformed traders face a fundamental uncertainty. They may not be able to observe the market price at which their trades will be executed.”
Yes, more of the financial industry is being automated, which is the evolutionary nature of every business (“March of the machines”, October 5th). The good news is that humans are not going away. There is no substitute for human interaction, judgment and experience. I doubt there are any trading operations with zero human oversight; someone has to press the stop button to turn a malfunctioning system off. After all, a fool with a tool is still a fool.
As for the stability of the financial system, particularly equity markets, the answer has been staring us in the face for some time. Does share trading have to be continuous? Is faster always better? A good market needs to be efficient and fair, not necessarily faster. Introducing a periodic single-price auction (spa) trading mechanism in place of the continuous system we have today would reduce the risks of future market meltdowns. It would also obviate the insane speed-trading arms race.
Almost as ridiculous as microsecond trade executions is the speed-bump mechanism used to slow things down. A spa market mechanism would not only be more elegant, it would also provide more liquidity, as it would concentrate trading interest both in time and place.
This would help deal with the question of market fragmentation. Do we really need so many markets to trade? With trade executions essentially a commodity (for equities at least), one could argue that the execution of business in a spa-oriented market should be a regulated utility, whether member-owned or not.
Joseph Rosen
Editor, “The Handbook of Electronic Trading”
New York
https://www.economist.com/letters/2019/10/26/letters-to-the-editor