Last week, a UK panel came out in favor of a tax on high frequency trading (HFT). As I’ve written previously, I think this is a good idea. Amazingly, some people disagree with me. Here are two objections that have been raised –
HFT provides market liquidity. Um, no. The markets were liquid before HFT became technically possible, and they have remained so. Proving this argument will be difficult, since it amounts to proving a negative. You would have to find a situation in which previous market behavior could have been expected to lead to a stop in trading (or other pricing discontinuity), but the actual HFT activity prevented this from happening. Good luck with that.
HFT is ill-defined, and therefore impossible to regulate. Again, no. Life, its beginnings and ending, is very ill defined, and yet is the subject of much regulation. We regulate a vast number of ill defined things, and by regulation they become better defined.
Spamming the network with millions of “indication of interest” messages because your algorithm is trying to bluff someone else’s algorithm is pretty clearly defined. Executing one million trades in 24 hours in the same security – HFT, baby. You want to avoid paying the tax? Show that you were executing the trade for a legitimate holder of the security, not trading for your own acount.
Yes, it can be done. The schedule of behaviors deemed HFT will have to be flexible, and even defined in a standard, machine readable way, so that when it is updated, algorithms can adapt. Ultimately, the watchdog of the algorithms wll be another algorithm. Left to themselves, algorithms will evolve rapidly through cycles of “arms races” into styles of behavior that benefit no-one. We can foresee that, and we can avoid it, and we should.