The past year in U.S. stock markets earned a place in the volatility Hall of Fame, coming in just behind the full-blown financial crisis years of 2008 and 2009. For this, credit (or blame) goes to Europe’s debt mess, U.S. political paralysis and upheaval in the Middle East.
But something else was at play: the increasing dominance of so-called high-frequency trading driven by computers and programs that thrive on and exacerbate wild market swings.
U.S. securities regulators have been examining the trading records and computer codes of brokers, hedge funds and banks. This welcome inquiry delves into a world where buy-and-sell orders are executed at speeds normally associated with astrophysics, and traders profit from opportunities that may last a few millionths of a second.
At this point, it’s beyond doubt that high-frequency trading contributes to volatility, fueling perceptions among retail investors that insiders have the game rigged. One need only look at the so-called flash crash of May 6, 2010, when the Dow Jones Industrial Average plunged almost 1,000 points in an hour and bounced back just as quickly. High-frequency traders may not have triggered the crash, but researchers have found that they kicked it into overdrive. Investors pulled almost $20 billion out of U.S. stock mutual funds that month, the most since the dark days of October 2008, according to the Investment Company Institute.
A loss of confidence on that scale might, over time, diminish liquidity, increase trading costs and erode investor returns. Ideally, the Securities and Exchange Commission will accelerate a full-scale monitoring program that gleans all trading transactions. Such “audit trails” might help the agency spot wrongdoing. But the securities industry, which will bear much of the potential $4 billion cost to hire analysts and buy computers that can keep up with the traders, has balked at having such information provided on a real-time basis.
The SEC and the Financial Industry Regulatory Authority have an added burden of drawing up rules in the absence of clear guidance from the Dodd-Frank law, which gave short shrift to high-frequency trading. But it’s crucial that the market overseers take steps to limit the possible harm that high-frequency trading can cause, without compromising its benefits, which are lower trading costs and increased liquidity.
Here’s one suggestion: High-frequency traders can bombard markets with buy-and-sell orders and then instantaneously issue cancellations. This might be a legitimate tactic for price discovery in much the same way that an auctioneer seeks to find out how much money potential buyers are willing to offer. Yet this also might be used to skew a market, perhaps even leading to mini-flash crashes in specific securities. Regulators, therefore, might impose some discipline by requiring traders to stick with bids or offers outside a certain price range for a specific period of time, say 1,000th of a second or longer, before canceling.
Such a requirement would be consistent with something else the SEC might consider. High-frequency traders now do much of the work once performed by humans on the stock-exchange floors. Regulators could ask that they fulfill an obligation to make markets, helping maintain the functioning of the stock exchanges by offering to buy and sell securities when others refuse.
Overloading market trading systems with excessive buy-and-sell orders offers another potentially unfair advantage to high-frequency traders. Some traders may be able to exploit the minuscule lag between the time a transaction is recorded on a local exchange’s electronic ticker and when it shows up on a feed that consolidates information from all the exchanges. By stuffing the system -- or another trading firm’s computer systems -- with quotes, this lag widens, yielding trading opportunities. Regulators might be able to combat this by imposing a fee on traders who introduce excessive messages into the data stream.
Then there’s the issue of co-location. Exchanges rent space to trading firms in their computer warehouses, letting high-frequency traders place their own computers just a few feet from the ones used by the exchanges. The minimal distance between the computers can slice a few microseconds off a transaction, which can mean the difference between a profit and a loss. It’s hard to see what the answer is -- proximity in markets has always offered an advantage. We just hope rulemakers consider ways to ensure that no one has an unfair edge simply because they can pay the rent or find the space at all of the exchanges’ computer centers.
As a general principle, we favor free markets. We also believe markets function best when rules promote transparency, and enforcement ensures integrity.
To contact the senior editor responsible for Bloomberg View’s editorials: David Shipley at firstname.lastname@example.org.