Thomson Reuters will stop giving high-speed traders a two-second head start

When two seconds makes all the difference.
When two seconds makes all the difference.
Image: AP Photo/Charlie Riedel
We may earn a commission from links on this page.

Thomson Reuters will stop releasing economic data early to high-speed clients—at least for now—according to a source cited by the New York Times, suspending a practice we first reported on last month.

Until now, the data distributor has been releasing the University of Michigan Consumer Sentiment Index to traders with special, high-speed and inordinately expensive connections at 9:54:58 a.m. Eastern Time—two seconds before investors that subscribe to its basic services. Media begin reporting the data after they’re released to ordinary subscribers at 9:55:55, though technically, the index isn’t published for the general, non-subscriber public until 10:00:00.

The problem is not that companies shouldn’t be able to sell data at a price. After all, the price is what gives trade unions, universities, polling organizations and other private institutions the impetus to collect better data in the first place. The issue is that the data are only available to one specific kind of investor: according to the contract, only to subscribers on an “ultra-low latency distribution platform for purposes of algorithmic trading” [emphasis ours].

New York attorney general Eric Schneiderman and his team have been investigating the early data releases because they believe they might violate the Martin Act, a New York law that allows the state to prosecute firms that attempt to “defraud” the general public of investors. But that’s a pretty broad definition, according to the New York Court of Appeals (pdf):

In a broad sense, the term [fraudulent practice] includes all deceitful practices contrary to the plain rules of common honesty. The purpose of the law is to prevent all kinds of fraud in connection  with the sale of securities and commodities and to defeat all unsubstantial and visionary schemes in relation thereto whereby the public is fraudulently exploited. The words ‘fraud’ and ‘fraudulent practice,’ in this connection should, therefore, be given a wide meaning so as to include all acts, although not originating in any actual evil design or contrivance to perpetrate fraud or injury upon others, which do by their tendency to deceive or mislead the purchasing public come within the purpose of the law.

Whether prosecutors will be able to accuse Thomson Reuters of this kind of “fraudulent practice” remains to be seen. Certainly, the data company didn’t advertise to regular investors that they weren’t getting the data as fast as others, but the early release of that data wasn’t precisely a secret.

Markets aren’t fair; wealthier investors have always had access to better information, better research reports, and more data. Whether or not regulators can—or should—draw a line somewhere is still an open question.