Delayed Reaction

Somewhat related to the last post is the findings of a study by Alasdair Brown titled Evidence of in-play insider trading on a UK betting exchange. It's not a new study, having been published in December 2010, but it's conclusions are perhaps more relevant now than previously due to the scale of the insider (court-sider) trading.

Having confirmed the proposal that:

informed bettors derive their advantage from observing the action before the rest of the public.

Television pictures typically transmit with a few seconds delay and therefore bettors with a faster transmission, or present at the game, are able to trade on an informational advantage. To capitalise on this fleeting advantage, bettors would need to feed this information into a computer and initiate bets via an algorithm. In these circumstances, a high frequency trading strategy (trading after each point, for example) would generate significant returns.

the author comments. 

Another issue is that Betfair does attempt to nullify the private information that accrues to those with a viewing advantage. For in-play markets there is a 1-5 second window (after a bet is matched) during which a liquidity provider can cancel the offer. The length of the window depends on the company’s estimation of the delay in television transmission. The aim of this window is precisely to deter trading on private information during a match. Our results, however, suggest that this window is either insufficient to remove the viewing advantage, or at least is perceived by those providing liquidity to be insufficient.

The author erred in saying the delay is after a bet is matched - however, it is unlikely that Betfair are unaware of this study, which again raises the question of why they are not doing more about the issue. In a Forum thread bemoaning the lack of liquidity in baseball markets, someone commented:

this happened to volleyball, where we had liquidity and, now vball has gone. This is happening to basketball too, and this is the worst liquidity on basketball I have ever seen! 7 years ago was more backers than now!.

Although my personal study is not academic or peer reviewed, my trip to San Francisco did offer the opportunity to compare TV and local radio with price movements on Betfair. With the play-offs currently in full swing, the markets are remarkable liquid considering the disadvantage that most players are at, and it is clear that court-siders are active in this sport. As I suggested in my previous post, this state of affairs is unlikely to continue. As popular a sport as NBA basketball is, constantly losing money is likely to dampen enthusiasm and participation at some point. The next review by the Gambling Commission into in-play betting promises to be interesting.

The study concludes:

To capitalise on this fleeting advantage, bettors would need to feed this information into a computer and initiate bets via an algorithm. In these circumstances, a high frequency trading strategy (trading after each point, for example) would generate significant returns.

Apparently I wan't the first to relate CST to HFT.