A sprinter who jumps the gun gets disqualified from the race. But selling gun-jumping technology to Wall Street is big business for suppliers of economic data.
Wall Street banks are paying Thomson Reuters to receive important data earlier and faster than the rest of the market. The early-released numbers include the results of the University of Michigan Consumer Confidence Index, according to my CNBC colleague Eamon Javers.
(Read More: Thomson Reuters Gives Elite Traders Early Advantage)
Meanwhile, Thomson Reuters offers reports from the Institute for Supply Management (ISM) on extremely fast connections known as "ultralow latency releases."
Thomson Reuters has contracts with the University of Michigan that allow it to release the consumer sentiment data early to elite customers. It's long been known that although the general public doesn't get to see the information until 10 a.m. on release day, the data is distributed five minutes earlier on a conference call for Thomson Reuters' paying clients, who are given certain headline numbers.
Less well-known is that the highest-level subscribers get the information two seconds earlier, at 9:54:58 a.m.
Information is one of the Street's most valuable currencies—all the more valuable when it is not in wide circulation. Wall Street banks equipped with ultrafast computers can exploit even small timing advantages to trade before the market moves on economic data.
Unlike the Michigan consumer confidence data, the ISM reports are not released to elite customers earlier than the general 10 a.m. distribution, but that doesn't mean banks aren't buying faster access.
According to the marketing material of Thomson Reuters, subscribers to its "News Direct Feed" can receive ultralow latency delivery of "key market-moving indicators, such as the Institute for Supply Management Reports On Business, Thomson Reuters/University of Michigan Surveys of Consumers and the Purchasing Managers' Index (PMI) data produced by Markit."
What's the advantage of ultralow latency? Thomson Reuters says it will enable customers to "be the first to react."
Think about it this way: If Thomson Reuters, which is the exclusive distributor for the ISM numbers, dropped the manufacturing report in the mailbox for some customers at 10 a.m. and emailed it to others at the same moment, it could claim it released the information to everyone at the same time. But some would get it moments later, while others would wait for days. Obviously, getting it by email would convey a big advantage.
The times involved in ultralow latency releases are much smaller—fractions of a second. But with high-speed trading, passing and momentary advantages are very valuable.
How valuable? Well, if firms agree to shell out for early access or delivery, it's clear that Wall Street, at least, believes that getting information early or faster is very valuable.
And Thomson Reuters isn't the only one providing faster access to those willing to pay hefty fees. Rival Bloomberg and the exchanges themselves sell low-latency access.
A variety of regulatory rules prevent some kinds of market moving information from being selectively disclosed. Public companies are required under Reg FD to avoid releasing material information except through approved channels intended to ensure equal access. Brokerages are required to prevent research departments from providing information about upcoming reports until they have been released to clients.
The differential release or delivery of economic reports seems to violate at least the spirit of rules aimed at creating a level playing field for traders and investors, even though it appears to be perfectly legal under current regulations.
There is a family resemblance between this practice and the "trading huddles" in which Goldman Sachs' research analysts met to develop tips that were passed on to its traders and most-favored clients.
Goldman paid tens of millions of dollars in fines to the Security and Exchange Commission and to Massachusetts regulators over the huddles, and had to change the way it operates. And in an unusual move, the firm actually admitted to certain facts in the regulatory complaints rather than employ the "neither admit nor deny" language usually found in Wall Street settlements with regulators.
Similarly, regulators have tried to crack down on the data race that drove the "expert network" business. Hedge funds paid large fees to expert networks in an effort to get better information about companies and markets—until the SEC and the Justice Department cracked down on the business.
The main difference is that when it came to the huddles and expert networks, regulators had clearer rules on which they could pin their allegations. The SEC said the huddles were inadequately supervised and created the risk that analysts would tip customers off to things like ratings changes before reports were published. Some of the expert networks became transmission mechanisms for nonpublic material information that led to insider trading, law enforcement experts claimed.
The differential release and delivery of the Michigan and ISM data, on the other hand, does not appear to violate any existing rule.
There's an arms race aspect to the early and fast provision of data. The availability of ultralow latency delivery and early release means that market actors with deep pockets feel compelled to buy access to keep up with competitors, much like some professional athletes feel they have to take performance-enhancing drugs to stay even with their rivals. This can result in an inefficient and unhealthy use of resources. Perhaps everyone would be better off if the accelerated information were simply banned.
(Read More: Early Data Underscore Moral, Technology Dilemma)
Trading on privately produced information is of course allowed. Hedge funds employ researchers to produce data about the markets. Wall Street banks can exclusively release reports to their brokerage clients. This is how it should be. The production of information about the economy, markets and companies is expensive and socially beneficial, helping investors price assets better. Because of that, we tolerate some unleveling of the playing field.
But big picture numbers produced by Michigan and ISM are different. They are privately produced but quasi-official. Michigan is a state university, for one thing. What's more, both sets of data are designed to eventually be released to the public and relied upon by many to better understand economic conditions. Arguably, this creates a responsibility to the public about how the numbers are handled and disseminated.
What if ISM and Michigan were required to put in place restrictions that prevented the data from being released early or disseminated to some players faster than to others?
Richard Curtin, director of the Thomson Reuters/University of Michigan Surveys of Consumers at the Survey Research Center of the University of Michigan, said that if it were unable to sell early access, there would be no money to fund the study.
"This research project is privately financed," he told Javers. "Without the support from Thomson Reuters, no data would be collected, the project would no longer exist, and the public benefit would disappear."
My guess is that this is wrong—at least in so far as it applies to the need to provide data a few seconds early. The consumer sentiment survey has been around a long time—long before the era when a two-second advantage would have conveyed the kind of benefit it can now. If Michigan could fund the survey then, it should be able to fund it now without this advantage.
We should undertake a cost-benefit analysis. Does social value of the consumer sentiment survey outweigh the costs of providing uneven access to the data? In the age of high-speed trading, it may be that this particular data set is valuable mostly because of the advantage it gives to deep-pocketed traders. In which case, maybe it should be allowed to pass from the Earth.
It's probably time to rethink the rules around this kind of quasi-public economic information.
—By CNBC's John Carney. Follow him on Twitter @Carney