September 28, 2012

Few can have missed the fact that Germany is set to go it alone on curbing high-frequency trading, although it looks like the rest of Europe won't be too far behind. The proposed European legislation will insist upon a minimum resting time for HFT orders before they can be amended or cancelled.

This has sparked a fair amount of discussion here at Fidessa Towers on how it will (or probably won't) work. Presumably the stop watch will be in the hands of the venues themselves, but trading firms won't necessarily know when an exchange has hit the start button and so they won't know when they can send an amendment or cancellation. This is because the way trading platforms work means that the messages they send back are often out of step with each other. Acknowledgment that an order has been received often arrives after it has actually entered the system (sometimes it is even after the order has been filled). The boffins call this asynchronous communication and it looks like being a rather nasty fly in the ointment that regulators are trying to rub into the HFT "problem."

One approach is that the venue simply queues all order amends and applies them once the 500 milliseconds is up -- but in which order? And how do they protect themselves from claims that they didn't quite get it right for a particular trader's order?

On top of this the processing overhead of making all these measurements may create anomalies in how the trading platforms behave. And, finally, European venues do not run on one standard universal clock, and yet many trade the same stocks, so problems may arise here for market participants too.

Whatever happens, it looks like this piece of legislation will cause yet another slew of unintended consequences which will doubtless be followed by more regulation. The real problem in all this, though, is that the current cycle we are in simply drives up complexity and costs for the market as a whole and this inefficiency means everyone suffers.

Steve Grob is director of group strategy for Fidessa and this editorial originally appeared on the firm's Fragmentation blog.