Algorithmic Trading Close to Insider Trading But Legal

Financial firms pay to obtain market data milliseconds early
June 14, 2013 Updated: June 14, 2013

When it comes to trading in the stock market, some are more equal than others. Financial firms pay high fees to obtain market moving data milliseconds early and their trading programs benefit from it.

Trading on material, non-public information is the common definition of insider trading. The old-school version is mostly used by corporate insiders who know something about the company that nobody else does. They use the information to trade, knowing full well that once the public finds out, it will send the shares either soaring or plummeting.

A good example would be a scientist at a biotech firm that buys shares of his company because he knows ahead of time that a revolutionary drug the company has been developing will be approved by the relevant authorities.

The behavior outlined is illegal and will be prosecuted by the Security and Exchange Commission. There is however, a new form of insider trading that most people don’t know about and that is completely legal.

Trading firms have been using computer algorithms to trade for a long time now. Sophisticated programs screen stocks for specific company data or chart signals and automatically buy or sell. Other algorithms screen the news tapes for specific buzz words and hit the market with orders much faster than any human could possibly do.

Especially the lastest version of a high frequency trading (HFT) algorithm comes close to insider trading. Economic data such as consumer confidence or forward looking industrial gauges move whole markets up or down.

This why the companies that produce the statistics pay special attention to releasing the data to all market participants at the same time, so nobody gets an unfair advantage. Or so the theory goes.

In reality, some HFT firms have been paying high fees to obtain this data early, sometimes just a split second, which is enough to give them a head start and trade on the news before everybody else does.

Thomson Reuters, one of the leading providers of news and data to financial market participants recently had to admit it “inadvertently” released manufacturing data from the Institute of Supply Management (ISM) early.

Reuters blamed the mishap on a clock-synchronization issue. “We have identified that there was a minor clock synchronization issue Monday causing this data to be released 15 milliseconds early,” it said in a statement to CNBC.

A company called Nanex, which tracks HFT activity, spotted the incident and marked the burst in trading activity seven milliseconds after the early release. The company also spotted early trading in several other instances.

A more public arrangement is the deal between the University of Michigan, which reports a measure of consumer confidence and Reuters, which gets the data five minutes ahead of the scheduled release at 10 a.m. Eastern Time. This data has also been released early to Reuters clients ahead of the general public, as data by Nanex shows.

According to the Wall Street Journal, clients pay a fee of $2,000 per month to have that privilege. So why is this form of trading on material, non-public information still legal?

First, there is no specific regulation concerning high frequency trading, although some people think it is a violation of the more general SEC rule of Fair Disclosure.

Also, regulators simply haven’t made up their mind. Bart Chilton, the head of the Commodities Futures Trading Commission put it this way in a talk with CNBC “It’s all about money. Right? So, if you have the money to afford these extra services—if you can have five reporters, if you can have a supercomputer and do high speed trading, yeah, you can be in the markets. Is that really where we want to go …? That you have to have the money, that you have to have the best, fastest computers?”

He also promised to present some concepts on how to deal with the technology in the near future. Until then, this legal form of insider trading will continue.