Computerized trading, under scrutiny from securities regulators, isn’t spurring broad increases in market volatility even as it sometimes creates “instability” that may lead to crashes, a U.K. government study said.
High-speed automated trading on exchanges and other venues may lead to “feedback loops” that compound problems and efforts to manage risk may worsen those events, the report said. Changes in volume, news reports and delays in market-data distribution can exacerbate the moves, it said. The study predicted that computer-driven trading will increase.
The report, commissioned as part of the U.K. government’s Foresight Project on the future of computer trading in financial markets, aims to assess the “risks and benefits” of automated buying and selling, according to written comments by John Beddington, chief scientific adviser to the U.K. government. He oversaw the project, which includes 16 additional academic papers. Understanding the impact of computerized trading on financial services and the broader economy may help policymakers develop a “resilient regulatory framework,” he said.
“Firms that actively deploy high-frequency trading must not mistake this study for a validation that the practice cannot impact the market negatively,” John Bates, chief technology officer at Bedford, Massachusetts-based Progress Software Corp., said in an e-mailed statement. Bates is a member of the U.S. Commodity Futures Trading Commission’s technology advisory committee. “Trading firms and regulators still largely have inadequate cross-market oversight in place to detect and respond to market abuse and market anomalies before it’s too late.”
May 6 Crash
The role of high-frequency firms in periods of market swings has come under scrutiny since the May 6, 2010, crash that briefly erased $862 billion from U.S. share values. Traders and other professional investors were said to have withdrawn bids as the selloff worsened, according to a Sept. 30 report from the Securities and Exchange Commission and CFTC. Regulators and exchanges later installed curbs to limit market disruptions.
The U.K. report comprises three papers by academics including Dave Cliff at the University of Bristol, Maureen O’Hara of Cornell University in Ithaca, New York, and Terrence Hendershott at the University of California at Berkeley. Academics with expertise in the issues from more than 20 countries helped write or review the project’s papers, Beddington said. The report doesn’t represent the views of the U.K. government or the financial sector, he added.
“Research thus far provides no direct evidence that high-frequency computer-based trading has increased volatility,” the report said. Still, small changes in trading behavior or actions may -- under specific circumstances -- trigger larger events and “lead to undesired interactions and outcomes,” it said.
High-frequency trading is a technique that relies on the rapid and automated placement of orders, many of which are immediately updated or canceled, as part of strategies such as market making and statistical arbitrage and tactics based on momentum. Hedge funds, brokerages and trading firms may use these techniques as part of investment strategies or to execute orders quickly as prices and available bids and offers change.
Algorithms, or strategies that execute bigger orders by breaking them into smaller pieces and sending them to different exchanges, also typically use high-frequency techniques. Mutual, pension and hedge funds employ algorithms built by brokers or vendors to automate their trading instead of manually placing orders in markets or turning to humans to buy or sell blocks.
Majority of Trades
“Computer-based trading has become an increasingly important part of the markets representing the majority of equity shares and futures contracts traded,” James Overdahl, vice president at NERA Economic Consulting in Washington and a former chief economist at the U.S. SEC and CFTC, said in a phone interview. The U.K. government’s working paper and 16 studies reflect the “basic fact that markets have changed dramatically over the past decade in terms of the technology employed for order execution and order entry,” he said.
High-frequency activity accounted for about 38 percent of European equity trading earlier this year, up from 29 percent in 2009, according to Tabb Group LLC, a New York-based financial industry research firm. In the U.S. it provided 53 percent of trading volume this year, down from 61 percent two years ago. In 2006 the strategies represented 5 percent of European trading and 26 percent in the U.S., Tabb said.
Overdahl said the group of papers are likely to become part of the literature about the “effect of computerized trading on market quality.” High-frequency firms have urged policymakers to base regulations on empirical evidence about how automated trading affects markets, he said. Overdahl is an adviser to the FIA-PTG, the Futures Industry Association’s Principal Traders Group, which includes almost 40 U.S. proprietary trading firms.
High-speed trading isn’t likely to merely “fast-forward” markets, according to one paper in the U.K. report. “It seems more likely that despite all its benefits, computer-based trading may lead to a qualitatively different and more obviously nonlinear financial system in which crises and critical events are more likely to occur in the first place, even in the absence of larger or more frequent external fundamental shocks.”
If small changes can produce “widely different” outcomes in the market, then prices and volumes may be “prone to cascades, contagions, instability” and other effects, the paper said. Even temporary results could lead to irreversible events such as bankruptcies if they cause forced sales or trigger penalties in contracts, it said.
Another paper found that the growth of computerized trading has led to lower transaction costs for retail customers and institutions and improved liquidity -- measured by the difference between the highest price at which someone is willing to buy and the lowest sale price, along with other metrics. Faster trading has also led to “greater potential for periodic illiquidity,” the paper said.
Regulators and trading professionals must improve their understanding of how humans and robots interact, especially since “future trading robots will be able to adapt and learn with little human involvement in their design,” according to the third paper. The need to analyze the consequences of these changes becomes more important as computerized trading grows and the number of human traders declines over the next decade.
“The number of human traders involved in the financial markets could fall dramatically over the next 10 years,” the paper said. “The simple fact is that we humans are made from hardware that is just too bandwidth-limited, and too slow, to compete with coming waves of computer technology.”