Inside Asian Gaming

March 2014 | INSIDE ASIAN GAMING 33 Insights Towards the end of 2009 a conversation with my father changed things irrevocably and convinced me of the merits of a Sports Hedge Fund. I took out a 15-leg sporting multi—I had about $3,000 to win about $40,000. It was the last leg, I’d got the first 14, and the Wallabies (Australian rugby team) were playing Scotland. I was on theWallabies, and they lost 9-8 after they missed many kicks for goal and played very poorly. It was the first time that Scotland had beaten Australia in almost 30 years! I said to my dad that I’d lost my $3,000, and he said no, you’ve lost $40,000—all you had to do was lay off and you would’ve had it. The hedge fund was born, and in 2010 I started looking for outside investors who believed in the concept. I applied my skills and knowledge from business school like risk management and portfolio management to sporting markets. The most important thing needed now in terms of attracting investors is education. The biggest change in the last ten years is in-play betting. Sports betting used to have a binary outcome, now you can take a position every minute in football, every ball in cricket. You can invest during a horse race. Odds fluctuate during a match, so you can trade the curve of a team like trading the curve of gold or a stock. Multis allow you to hedge a lot more. What happened on that cold night in Scotland in November 2009 will never happen again. How long did it take to develop the initial database? The first one we built was for horse racing. We started with 20 variables and kept monitoring and adding. You can write programs to scrape data from the Internet, which we outsource to computer science people—the difficulty is deciding which variables matter. For example, in a Test match in cricket, there are factors to be considered like batting first versus second, winning the toss and putting the other side in to bat, etc. All of that needs to be weighted. Data is just a commodity; it’s how you use it that’s important. Every innings, every ball goes into our database and is updated live— player stats as well, like a batsman faced X amount of balls from a right-handed bowler and X from a left-handed bowler. All that is automated, it’s just data extraction. One good thing about sport is there’s so much data. The most time-consuming thing is getting it into a format you can understand and then determining which pieces are worthwhile. The initial database for any sport takes a couple of months. Thereafter it’s a “living thing”, being forever updated. Developing a predictive algorithm takes longer. Firstly it gets back-tested against previous results, testing hypotheses and variables. Once that’s completed it’s then tested “live” with actual stakes—albeit small. At some point thereafter—12 months or so—it’s then considered ready for primary trading. The algorithms are forever monitored for efficacy and changed when needed. It’s an ongoing process but the insights are what help to identify trading opportunities. A good example was in the recent first cricket Test between Australia and South Africa at Centurion. (Australian captain) Michael Clarke declared the second innings closed and left South Africa 480+ runs to win (instead of batting on and leaving a bigger target). South Africa went immediately from $44 to $7. We knew Clarke would declare—he’s an attacking captain. We could back South Africa at $44 and lay off at $7. This clearly showed some of the inefficiencies in the sports markets. The market rated South Africa at $7—a one in seven chance of winning, or alternately a 14% chance. Our model rated South Africa’s probability of winning at less than 0.5%, or in terms of odds $200+. No team had ever scored that many to win, the wicket was very difficult for batting and Mitchell Johnson was bowling well. The latter two points were judgment or intuitive assessments that the model doesn’t assess. With such a significant edge on the market, we opened a maximum position (3% FUM total exposure) by betting against a South African victory. South Africa were very quickly in trouble, never “The biggest change in the last ten years is in-play betting. Sports betting used to have a binary outcome, now you can take a position every minute in football, every ball in cricket. You can invest during a horse race. Odds fluctuate during a match, so you can trade the curve of a team like trading the curve of gold or a stock.”