High-frequency trading has been a popular stock market villain for more than a decade. Known as HFT, the controversial but legal practice of seeking tiny profits on rapid buying and selling of stocks was blamed for the 2010 flash crash and became the focus of Michael Lewis’s 2014 nonfiction bestseller Flash Boys.
For all that notoriety, nobody really knew how much HFT was costing investors, or how HFT companies competed with one another to skim profits. Chicago Booth’s Eric Budish, with Matteo Aquilina and Peter O’Neill of the UK Financial Conduct Authority, found a way of studying winners and losers in the hundreds of thousands of daily races by traders to profit from minuscule moves in individual stocks.
They estimate that in typical years in global stock markets, HFT winners may gain $5 billion at the expense of other market participants. They estimate the figure is $7 billion for 2020, a highly volatile trading year. While those are relatively small sums in global markets that have a collective $95 trillion in value, it’s still a lot of money coming out of investors’ pockets, the researchers write.
Further, HFT takes place in other markets as well, including those for futures, options, bonds, currencies, and cryptocurrencies. The researchers’ estimates apply only to the global stock market, implying that the full price tag associated with HFT could be significantly larger.
HFT arms race is a symptom of market instability and not a cause—and that the public debate has overlooked a fundamental flaw in the structure of financial markets: continuous trading.
An Alternative to High-Frequency Trading
On this episode of The Big Question, an expert panel discusses how speed affects financial markets and what, if anything, we should do about it.
Are Financial Markets Too Fast?The problem of understanding HFT has been incomplete records. If a fast-trading institution successfully bids on a stock, there is a trade execution record. But for every winner, there might be many more institutions that put in a bid but weren’t fast enough to make the trade. Aquilina, Budish, and O’Neill found a work-around: they obtained nine weeks of electronic-message traffic between market participants and the London Stock Exchange from 2015, all involving the 350 biggest stocks traded there. The messages included all new orders and order cancellations sent to the exchange and, crucially, whether these requests succeeded or failed. This provided more information than what exchanges usually release to researchers, which is only successful orders and cancellations. “Traditional data are like seeing Usain Bolt run the 100 meter dash and win, without seeing all the other amazingly fast runners who run against him and lose,” says Budish. “Our data are like seeing the full race.”
The average stock in the FTSE 100 Index was involved in 537 trading races a day, with an average race time of around 80 microseconds, the research demonstrates. (A microsecond is one-millionth of a second.) Winners beat losers by only five- to ten-millionths of a second—“less than 1/10,000th of the time it takes to blink your eye,” they write. The three largest HFT players dominated the market, making up most of the winners and losers on each trade, Aquilina, Budish, and O’Neill find.
High-frequency trading races account for about one-fifth of FTSE trading volume, and are so fast that they have to be measured in microseconds, or millionths of a second.
Aquilina et al., 2021
Trading profits were small, but the huge volumes involved led to substantial totals, according to the researchers. They calculate that winning the average race was worth about £2 ($2.75), and that HFT races accounted for more than 20 percent of the volume on the London Stock Exchange.
HFT races accounted for about a third of the bid-ask spread, the researchers find. That spread between the price buyers are offering for a stock and what sellers are asking is a key measure of the cost of transacting in the market, thus races effectively impose a tax on investors, the researchers argue.
If markets were to be designed in a way that eliminated the negative aspects of HFT, investors could save 17 percent of their liquidity costs on each trade, the researchers find. That savings might not amount to much for individual investors, but it would add up to a significant trading cost for pension funds, investments funds, and other large investors, according to the researchers.
“Flawed market design drives a significant fraction of daily trading volume, significantly increases the trading costs of large investors, and generates billions of dollars a year in profits for a small number of HFT firms and other parties in the speed race, who then have significant incentive to preserve the status quo,” they write.
Matteo Aquilina, Eric Budish, and Peter O’Neill, “Quantifying the High Frequency Trading Arms Race,” Quarterly Journal of Economics, forthcoming.
With the transformation of asset markets over the past two decades, ‘flow trading’ could offer a more flexible and fairer way to trade.
A Blueprint for a Better Stock Exchange
Chicago Booth’s Raghuram G. Rajan joins hosts Bethany McLean and Luigi Zingales to explore risks in the financial system and possible solutions.
Capitalisn’t: Why the Banking Crisis Isn’t Over
Some degree of concentration in subprime credit markets could be good for consumers.
When It Comes to Subprime Car Loans, Competition Drives Rates UpYour Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.