Chicago-area research and analysis firm Nanex is the creator and developer of a high-performance ticker plant called NxCore, software that monitors pretty much every single trade made on the open market, from equities to options, in real time, down to the millisecond. NxCore is used by a variety of active traders and researchers as a means to get a complete picture of market information.
But besides developing and marketing NxCore, the firm wears the hat of a watchdog, monitoring trading activity, identifying suspicious activity, and reporting that activity to the media. In June, Nanex spotted something strange, and did its thing. The firm reported to CNBC that it suspected data from the ISM Manufacturing Report on Business was released a fraction of a second — about 15 milliseconds — too early. The data from the report were transmitted from Thomson Reuters to various high-frequency, algorithm-based traders who were able to act on the data before anybody else.
At a glance, 15 milliseconds may not seem like a lot. It is certainly an inconsequential amount of time for most day traders and retail investors. For some context, the average time it takes a human to complete a blink is about 300 milliseconds, or three-tenths of a second.
However, in the world of high-speed trading, 15 milliseconds — even just 1 millisecond or 2 milliseconds — is all it takes to make or lose millions.
Nanex was able to show that there was a spike in trades of the SPY ETF 15 milliseconds before the news was officially supposed to be released from embargo, exactly at 10 a.m. The firm calculated that as many as 30,000 shares traded in just 1 millisecond, and that about $28 million worth of shares were exchanged in the 15 millisecond period in question.
The cause of the “leak” was traced back to an improperly synced clock in the Thomson Reuters server responsible for issuing the information to clients. ISM and Thomson Reuters had established a deal in which the media agency would receive information early, but would only be able to send it out to clients no earlier than ISM did. ISM, for its part, does not distribute its information directly to low-latency traders and typically uses the standard PRNewswire service.
All told, the issue was pretty minor, but like most technical market errors, it raised a battery of difficult questions. Speaking to CNBC, Eric Hunsader, who heads Nanex, said: “How is it we live in a world of microsecond trading where a news agency sells premium advanced news data using a clock with an accuracy of plus or minus 15,000 microseconds? The problem with releasing data early is that it adds uncertainty to the market. Markets abhor uncertainty.”
This was certainly not the first or the last time that there was an issue with the early release of market information. Last week, shortly after the U.S. Federal Reserve surprised markets by reporting that it was not tapering asset purchases, Nanex came forward with evidence it says proves that information — that the Fed would not taper — was released early to some high-frequency traders.
“We’ve learned that the speed of light (information), takes 1 millisecond to travel 186 miles (300km),” explains Nanex in a post on its website. “Therefore, the amount of time it takes to transmit information between two points is limited by distance and how fast computers can encode and decode the information on both sides. Our experience analyzing the impact of hundreds of news events at the millisecond level tells us that it takes at least 5 milliseconds for information to travel between Chicago and New York. Even though Chicago is closer to Washington DC than New York, the path between the two cities is not straight or optimized: so it takes information a bit longer, about 7 milliseconds, to travel between Chicago and Washington. It takes little under 2 milliseconds between Washington and New York.”
Because it takes a certain minimum amount of time for information to travel from Washington to New York and Chicago, even high-frequency trading algorithms have a delayed reaction to the release of news from the Fed. In practice, these algorithms could never actually react to the news exactly at 2 p.m.
The only way they could, Nanex explains, is if the information that the Fed would not taper was already stored on servers located in those cities. Graphs provided by Nanex show a tremendous surge in trading activity exactly at 2 p.m., which they argue is evidence that certain firms had privileged access to information.
There are two ways this could have happened, the firm explains.
First, the data could have been released by a news organization that had access to the data ahead of time. The Fed allows certain reporters access to the information ahead of the embargo time so that they can write a story to be published immediately following the release of the data. “But the Fed news was released from a lock-up room which prevents transmission of any information to the outside world,” Nanex says. “Given that several large news organizations were recently caught doing this we think it’s less likely they would do something so bold, so soon. That leaves us with possibility number 2.”
Possibility No. 2 is that the data were simply leaked ahead of time to some large Wall Street firm. The firm then chose to sit on the data until exactly 2 p.m., at which point trading machines would execute the enormous flood of buy commands that was observed.
“The Fed news was leaked to, or known by, a large Wall Street Firm who made the decision to pre-program their trading machines in both New York and Chicago and wait until precisely 2pm when they would buy everything available,” says Nanex. “It is somewhat fascinating that they tried to be ‘honest’ by waiting until 2pm, but not a thousandth of a second longer.”