Dec. 26 (Bloomberg) — Finally, a bit of evidence, rather than anecdote, about the costs of high-frequency trading.

In a new study, Andrei Kirilenko, the chief economist at the U.S. Commodity Futures Trading Commission, along with researchers at Princeton University and the University of Washington, examined high-frequency trading in a futures contract called the e-mini S&P 500, between August 2010 and August 2012.

The study looked at only the expiring contracts (which trade electronically on the Chicago Mercantile Exchange) that are used to bet on the direction of the Standard & Poor’s 500 Index. The researchers also did something they’d never been able to do before: Use actual trading data from individual firms, though none were identified.

What that data does is help explain the frenzy in today’s markets: The most aggressive firms tend to earn the biggest profits, hence the incentive to trade as quickly and as often as possible. Furthermore, these traders make their money at the expense of everyone else, including less-aggressive high-frequency traders.

The study found that the most hyperactive trading firms earned an average daily profit of $395,875 in the e-mini S&P 500 contract over the two-year period. First and foremost among those on the losing end: small retail investors. The study found that, on average, they lost $3.49 on every contract to aggressive high-frequency traders.

Big Puzzle

A big puzzle, though, is why these small investors lost at all. One high-frequency trading technique is to use computer-powered algorithms to glean the investing patterns of other traders, whether mutual funds or securities firms trading for their own accounts. High-frequency traders can use this information to execute profitable investments. But small investors usually don’t have a discernible strategy. We hate to say it, but more study is needed.

And the big question remains: Does high-frequency trading benefit anyone beyond those who engage in it? Yes, markets create winners and losers, but they also are supposed to allocate capital to aid wealth creation and economic growth, not simply serve as casinos.

It’s also unclear whether high-frequency trading lives up to its promise of increasing market liquidity. By all appearances, the most profitable traders actually reduced market liquidity rather than enhanced it, according to the study.

One thing the study didn’t attempt to do is delve into the equities market, where the biggest computer-driven trading blowups have taken place.

Flash Crash

High-frequency trading was at the heart of the so-called flash crash of May 2010, when the Dow Jones Industrial Average plunged more than 1,000 points and rebounded, all in less than an hour.

Related trading snafus have bedeviled the stock markets since then, including the failure of electronic-trading exchange operator BATS Global Markets Inc. to use its own system to take its shares public, and the near collapse of Knight Capital Group Inc. an electronic trading firm, after faulty computer software made huge, money-losing transactions before anyone noticed.

These scenarios raise a final question: Why isn’t the Securities and Exchange Commission, which sets the ground rules for the stock exchanges, undertaking the same kind of study as the CFTC? High-frequency trading probably plays an even bigger role in the equity markets, where it accounts for more than 50 percent of the trading volume.

So far, the SEC has held hearings and taken halting steps to develop a central repository for high-frequency-trading data on the stock exchanges. But a functioning system is probably years away, and even then it won’t deliver real-time information.

Understanding who wins and who loses in the equity markets might prompt the agency to get serious. The bottom line is it still needs to draft rules to ensure that the fastest traders on earth don’t pile on in the event of another flash crash that might burn us all.

To contact the Bloomberg View editorial board: view@bloomberg.net.