the audit

NYT Flashes Confusion in HFT Reporting

How much high-frequency trading involves "peeking"?
August 14, 2009

For a brief moment last week, “flash orders” ranked #2 on Google Trends, just below “cash for clunkers car list” and a few spots above “Ryan O’Neal hit on Tatum at Farrah funeral.” Why would an esoteric piece of financial jargon suddenly light up the pop-culture scoreboard? The reason is simple. The Securities and Exchange Commission chairman had said the SEC staff would seek to eliminate the “inequity” associated with flash orders—an order type used on several U.S. exchanges. The media interpreted this as an eventual ban on flash orders. Two days later, last Thursday, Nasdaq and BATS Exchange announced they would withdraw their versions of flash orders by Sept. 1.

What led up to the SEC’s statement, however, is not so simple. It is, in fact, a cautionary tale for both financial reporters and their information-hungry audience of how ominous-sounding but vague information, combined with the public’s already deep suspicion of Wall Street, can lead to unnecessary confusion.

It started when a handful of financial websites fanned the flames of a small but simmering controversy in the trading industry, mixing up two different issues along the way. This brushfire blew up on July 24, when a front-page New York Times article, “Stock Traders Find Speed Pays, in Milliseconds,” in my view, amplified the confusion and—such is the paper’s power—wound up distorting the debate. That same evening, Senator Charles Schumer (D.-N.Y.) upped the ante further, flogging the SEC for allowing regulatory laxity to compromise the fair and orderly functioning of the equities market.

At the heart of this conflagration was something called flash orders. With this flash functionality, if a market can’t fill an order because another market is quoting the industry’s best price, the first market can flash the order for 0.03 seconds to recipients of its proprietary data feed. If a market participant then matches the best price, the market can fill the order and avoid having to send that order to another market and lose that business.

There are several trading and regulatory issues around the use of flash orders, including the main one: that some players get to see some order information that others cannot see.

But the Times story focused on a different and much larger phenomenon, something called high-frequency trading, which now represents at least two-thirds of the equities market volume. The trouble is, the Times implies, although it never explicitly states, that a big part of HFT involves strategies based on flash orders—which the Times refers to as “peeking.” The reader throughout is left to assume that flash-based strategies are integral to HFT—and that’s a problem. In fact, we don’t know how much HFT volume involves strategies built around reading flash orders.

Sign up for CJR's daily email

In an email, the reporter on the story, Charles Duhigg, told us that his reporting found that in fact flash is a big part of HFT. But that assertion never got into the story. That’s too bad, because many experts think that flash-based trading strategies are a drop in the HFT bucket and might have been happy to challenge the claim. Readers, meanwhile, are left with a vague but unsupported sense that a large portion of HFT—indeed, of the market—involves “peeking” at investors’ orders via flash.

This all got started on June 1, when Nasdaq and a smaller rival, BATS Exchange, rolled out their own versions of a flash order that another competitor, Direct Edge, had been offering since 2006. Trade publications jumped on the issue back in May, and I weighed in later with a cover story in Traders Magazine that appeared the second week of July.

Then on Tuesday, July 21, a popular, often indignant financial blog called Zero Hedge ran a series of excerpts from my story, describing it as a “curious” article that “definitively peels off the cover of what truly happens at the pantheon of stock exchanges.” Zero Hedge immediately linked flash orders to high-frequency trading, which, along with Goldman Sachs, has long been a lightning rod for the site’s wrath.

That got the ball rolling. Other blogs, including The Big Picture (by Barry Ritholz, whose book, Bailout Nation, has just been published), The Business Insider, Seeking Alpha, and Daily Kos gave wing to the story.

Three days later, the Times shot off its dispatch. The problem started at the very top of the story:

It is the hot new thing on Wall Street, a way for a handful of traders to master the stock market, peek at investors’ orders and, critics say, even subtly manipulate share prices.

It is called high-frequency trading—and it is suddenly one of the most talked-about and mysterious forces in the markets.

But it’s flash orders that potentially raise the specter of “peeking,” not HFT generally. The first two sentences joined HFT and flash, and the article never explained their differences later.

HFT has been around for many years, and it has its critics. Just like it sounds, it involves using high-powered computers crunching lots of data to execute a large number of trades extremely rapidly. HFT involves dozens of different, unrelated strategies, some of which incorporate flash data. More typically, the strategies are based on analyses of published order-flow information displayed on market centers, along with vast streams of pricing data. What these strategies have in common is not peeking at investors’ orders to try to profit from their contents but an intense focus on speed.

To be fair, the Times doesn’t say that all high-frequency trading firms use flash orders to take advantage of investors’ information. It just suggests it, both at the top of the article in the reference to high-frequency firms “peeking” and in the graphic that accompanied the story. The graphic was about how HFT traders on Nasdaq “often” “peek at orders” for 30 milliseconds to take advantage of those orders. How big a part this is of HFT volume, the reader is never told.

The article’s sole anecdote about high-frequency trading also involves flash orders, further creating the impression that a large portion of high-frequency trading in the market involves information gleaned from flash orders. Quoting an anonymous source, the Times said that traders one day last month anticipated a run-up in Broadcom stock because of good news from rival Intel. Slower traders, the paper said, didn’t want to tip their hand, and, as they often do, broke up their orders into small batches.

The slower traders began issuing buy orders. But rather than being shown to all potential sellers at the same time, some of those orders were most likely routed to a collection of high-frequency traders for just 30 milliseconds—0.03 seconds—in what are known as flash orders. While markets are supposed to ensure transparency by showing orders to everyone simultaneously, a loophole in regulations allows marketplaces like Nasdaq to show traders some orders ahead of everyone else in exchange for a fee.

In less than half a second, high-frequency traders gained a valuable insight: the hunger for Broadcom was growing. Their computers began buying up Broadcom shares and then reselling them to the slower investors at higher prices. The overall price of Broadcom began to rise.

The Broadcom example was also used to suggest how profitable it might be to get advance order information and take advantage of market moves. After noting that slower-moving investors paid $7,800 more for Broadcom than they would have if they had been able to trade as fast as HFT firms (how the anonymous source was able to calculate this is unclear), the article said:

Multiply such trades across thousands of stocks a day, and the profits are substantial. High-frequency traders generated about $21 billion in profits last year, the Tabb Group, a research firm, estimates.

The article doesn’t say that all high-frequency trading profits come from advance information gained through flash orders, but it certainly leaves that impression. After all, the use of flash orders to trade ahead of investors is the only strategy discussed in the story at any length.

There’s another point where readers may have been led astray. In mid-July, Goldman had announced record second-quarter net revenues of $13.76 billion, creating a storm of controversy. The Times article said:

Nearly everyone on Wall Street is wondering how hedge funds and large banks like Goldman Sachs are making so much money so soon after the financial system nearly collapsed. High-frequency trading is one answer.

It turns out that high-frequency trading in equities accounted for less than 1 percent of Goldman’s revenues in the first half of this year, according to Ed Canaday, a Goldman spokesman. Surely Goldman made a lot of money from high-frequency trading in other asset classes, but the article discussed only equities, so readers might have reasonably concluded that a large portion of Goldman’s revenues came from high-frequency trading in the stock market. (Editor’s note: Goldman is a funder of the Audit. Read all about it here, here, and here.)

After the Times article appeared, misimpressions in the media snowballed. Newspapers, magazines, and blogs echoed the Times’ broad statements about flash orders and high-frequency trading. Some congratulated the paper for exposing yet another scam that showed ordinary consumers getting burned by market professionals. Among those applauding the Times, alas, was CJR.

Meanwhile, back in Washington, D.C., another legislator had joined the party. On July 28, Senator Ted Kaufman (D.-D.E.) repeated Schumer’s call for a ban on flash orders from the Senate floor:

Commentators have begun to explain how flash orders work to—quite literally—“pick the pockets” of the average investor. In essence, these traders get a very quick look at all pending orders in advance and, through technology, can trade ahead of those orders.

Kaufman then had the Times article about high-frequency trading read into the Congressional record.

In a note to The Audit, Duhigg, the author, says that his reporting backed up the idea that flash was a big part of HFT:

My reporting indicated that many high frequency traders were seeing—and using—flash order data, so it’s not unfair for readers to get that impression. I would often like to include more details, caveats and explanations in my articles. However, given scarce space resources, I felt that other details—including explaining how high frequency trading works, various perspectives on the practice and providing an example for readers—were topics that deserved the most attention. And, it seems that my focus on flash orders was shared by others with more time and insight into this world: the SEC decided soon after the story was published to push for restrictions on flash orders and, more broadly, review problems posed by high frequency trading. That said, I’m always trying to learn more about everything I cover, and so thanks for your thoughts, and I hope anyone else won’t hesitate to contact me with their perspectives as well.

We appreciate the courteous and thoughtful reply. We also understand the space problem.

But in my view, the article ended up conflating flash orders and high-frequency trading in ways that would confuse a general reader. Instead of noting that some HFT firms incorporate information from flash orders into their data and strategies, the article joined the two at the hip.

Flash orders clearly deserve attention. High-frequency trading, and the effect this trading activity may have on markets, deserves closer attention. Other market-related issues tossed into this melee also deserve attention. But when a conversation starts out as a shouting match, there’s nothing to do but pick sides.

It remains to be seen how the flash issue plays out among ordinary readers, rather than the trading community, blogs, and those keeping tabs on the financial industry. Bloomberg, Reuters, the Associated Press, Dow Jones, the Financial Times, the Wall Street Journal and other news outlets, including the trade press, have been explaining these topics, and have been more careful than the Times to differentiate between them.

But first impressions are lasting impressions. It is important to write about complex topics that affect the lives and pocketbooks of ordinary readers, and to do so in accessible ways. But details matter in trying to understand complicated issues, particularly when writing for an audience that already regards the financial industry—not to mention the press—with deep suspicion.

Nina Mehta is a senior editor at Traders Magazine. She can be reached at ninamehta@aol.com, and on Twitter: ninamehta212.