Why High-Frequency Trading Is So Hard to Regulate

Photo
Stock prices for eBay on Nasdaq in 2012. The S.E.C., in a case against a high-frequency trading firm, cited its trading in eBay.Credit Mark Lennihan/Associated Press
White Collar Watch
View all posts

Related Links

Stock manipulation is one of the great bugaboos among investors because it generates fear that the market is as rigged as a three-card monte game on a Bronx street corner. Since the publication of Michael Lewis’s book “Flash Boys” earlier this year, high-frequency traders have come under increased scrutiny for their use of sophisticated computer programs that can whipsaw prices by flooding the market with orders in milliseconds.

The challenge in pursuing charges against these firms is that they are taking advantage of changes in the technology underpinning the markets to profit from quick trades, which is not illegal. But regulators can find it difficult to draw the line between acceptable trading strategies and manipulation because of the complexity of the strategies.

Last week, the Securities and Exchange Commission announced a settlement in what it called the “first high-frequency trading manipulation case” against Athena Capital Research, a New York firm. The administrative order settling the case stated that Athena used a sophisticated algorithm nicknamed Gravy to “carry out a familiar, manipulative scheme” to enter orders in the final two seconds of trading to push the closing price of stocks up or down so that other trades by the firm came out profitably.

High-frequency trading firms like Athena are not in the market as long-term investors, and have no interest in the fundamentals of the companies whose shares they trade. The firm designed the Gravy program so that it could profit from rapid trading without holding shares once the market closed. Thus, the S.E.C. said, the program’s primary goal was to shift the price of the stock a few pennies to eke out a small gain — at the expense of other investors.

Proving market manipulation requires showing that the person acted with the intent to artificially affect prices, not just that the trading sought to generate a profit. The S.E.C. was aided immensely by e-mails from managers at Athena discussing how the goal of the Gravy program was to change stock prices.

One e-mail pointed out how Athena’s stock trades one day were insufficient to affect share prices, so “make sure we always do our gravy with enough size.” Another message forwarding an alert from a regulator about suspicious orders at the close of trading said, “Let’s make sure we don’t kill the golden goose.”

The S.E.C. noted that even though Athena was a relatively small firm, its trading in the last final seconds before the market closed accounted for 56 percent of the shares it accumulated, and 73 percent of the entire Nasdaq volume for stocks traded in that tiny window of time. Athena maintained a spreadsheet showing the price movements caused by an early version of the Gravy program. A report that one day’s trades resulted in a profit of $5,300, led a manager to e-mail about “going to Vegas tonight.”

What is more difficult to figure out from the case is where Athena crossed the line into illegal manipulation, despite the statements in the e-mails. The administrative order goes into great detail about how the firm entered orders to buy or sell shares 10 minutes before the end of the trading day at 4 p.m., and then flooded the market with orders on the opposite side of that trade in the last two seconds of trading. The S.E.C. used trading in eBay as one example, showing six orders to buy shares in the last few milliseconds to push the price up about 3 cents.

The S.E.C. did not identify whether specific trades were improper, or if the entire Gravy program constituted market manipulation. The administrative order recounted internal e-mails describing the firm’s strategy as “dominating the auction” and “owning the game.” But Athena also provided a service through its trading by filling orders that might not otherwise have been executed at the close of trading, so at least some of the trades aided investors.

The settlement did not involve an admission or denial of liability, which is the standard way the S.E.C. resolves cases even after it adopted a new policy in 2012 to require an acknowledgement of guilt in some cases. In a statement issued after the settlement, the firm did not sound contrite. It noted that while it did not deny the charges, “Athena believes that its trading activity helped satisfy market demand for liquidity during a period of unprecedented demand for such liquidity.”

The S.E.C. case rests on the proposition that the Gravy program was intended to have an artificial impact on stock prices, and that the trading rose to the level of manipulation. But affecting stock prices is true of many trading strategies, not just those of high-frequency traders.

William A. Ackman’s presentation in December 2012, for example, accused Herbalife of being a Ponzi scheme to help the $1 billion short position against its stock taken by his hedge fund, Pershing Square Capital Management. But that is not market manipulation, even though the presentation was intended to drive down the stock price. Similarly, when an analyst puts a target price on a stock, there is an expectation that it will have some effect on the shares to help the brokerage firm’s clients profit.

Athena agreed to pay a $1 million civil penalty for its violation, although the S.E.C. did not explain how it arrived at that figure or Athena’s profits from the Gravy program. The maximum penalty for a fraud violation at that time was $725,000 if it involved substantial losses to investors.

The administrative order also states that the violations took place from April to December 2009, so the settlement came five years after the manipulative trading. This highlights the challenge the S.E.C. faces in dealing with sophisticated trading strategies that may result in artificially affecting stock prices that come to light only months or even years later.

In a statement accompanying the settlement, Andrew J. Ceresney, the director of the S.E.C.’s enforcement division, said, “Traders today can certainly use complex algorithms and take advantage of cutting-edge technology, but what happened here was fraud.” Finding the line between acceptable trading strategies and fraud may be difficult without helpful e-mail evidence and a program nickname that fairly screams market manipulation.

The case is a warning that programs aimed at nudging prices a little bit for a quick profit can easily be viewed as fraudulent. So the message for high-frequency traders is to tread carefully in designing algorithms to avoid the appearance that the sole goal is to affect stock prices. And don’t send e-mails gloating about a “golden goose.”

Correction: October 20, 2014
An earlier version of this article misspelled the surname of the director of the Securities and Exchange Commission's enforcement division. It is Andrew J. Ceresney, not Ceresnay.