High-frequency trading has a way of staying in the news. These days, there's lots of talk about regulation on both sides of the Atlantic. The EU is expected to take up the issue and propose a series of reforms. Mifid II just might end up proposing some sort of direct regulation. Transaction taxes have also been talked up in Europe as well as the U.S. IOSCO, a group of international securities regulators, has come out with a report that notes the salutary effects of high-frequency trading while also warning regulators of the need to monitor the activity, suggesting that some might lack the resources for this. The group has published guidelines to this end.

So, where to begin?

To better monitor, much less actually regulate, high-frequency trading, we need a working definition of the practice, which has been all too elusive for some regulators. Scott O'Malia, a CFTC commissioner, tells the Financial Times:

"I don't believe the CFTC has ever attempted to define HFT, so this is the first attempt out of our building. I'll admit that I don't have all the answers, but I hope this kicks off a debate. [The] first step is to define the activity. Next is to make relevant entities reportable and study their trade behaviour before making a determination about regulation."  

He has come up with a seven-point test, as noted by the FT:

• The use of "extraordinarily high-speed order submission/cancellation/modification systems with speeds in excess of five milliseconds or generally very close to minimal latency of a trade."
• The use of algorithms for automated decision-making "without human direction for each individual trade or order."
• The use of co-location services, direct market access, or individual data feeds offered by exchanges and others.
• "Very short" round-trip time-frames.
• High portfolio turnover.
• The use of cancelled orders, of within milliseconds.
• Ending the trading day close to flat. 

These attributes certainly describe high-frequency trading. In the end, we'll see more and more firms that meet several but not all of these attributes, raising questions about what to do about firms that meet say three out of the four criteria. At some point, you could see traditional buy-side firms qualifying on some of the items. The distinction between market-maker and non-market-maker is also pretty critical from a regulatory point of view. But in terms of a definition, this is an interesting first start.

In the end, regulators would be wise to avoid trying to precisely define a group for purposes of subjecting them to specific rules. It runs the risk of being overly formulaic and leaves open the possibility for all sorts of exemptions and loopholes. It may be better to rethink market structure from the ground up, instead of focusing in on particular trading styles. -Jim