September 4, 2019
Technology has changed the world of sport. Whereas the tools of visible analysis, feel and conversation were available in the past, now there are technologies designed to measure all aspects of performance. Data showing how far a footballer has run is fundamental analysis nowadays, while the impact of a tackle or the physical stress imposed on the body can all be scientifically assessed. Coaches and trainers have much more information to work from and no professional outfit ignores this opportunity to seek out the ‘marginal gains.’
However, when it comes to horseracing, data analysis is minimal. We know how fast a horse runs, but we cannot adapt that to determine how fast a horse might run. And obviously, we can’t ask it!
When it comes to buying an unraced foal or yearling, it is even harder to judge. Does the horse look right, is its constitution sound, does it move well, etc.? Breeding is carefully considered, but this is a hit-and-miss concept – a good sire and a good dam don’t always produce a good foal.
I felt there was more that could be done to turn this from an artform into a science, and realised that some of the methods we use in investment analysis at Vertem could be the answer.
The use of algorithms (using a complex set of rules to be followed in problem-solving) as an aid to creating research and analytical platforms is a major part of Vertem’s research programme.
Gary Stockdale and I were fully aware of the added analytical power we could create by adopting modern algorithmic research techniques to Vertem’s unique in-house research process when we established our stockbroking company in 2010.
Over the years, these processes and formulas have helped improve results for our clients and have proven the strength of Vertem’s research programme.
I subsequently started to create models to aid with the selection of prospective racehorse purchases. To say that my methods were greeted with wild enthusiasm among the conservative racing industry would not be accurate; however, as with Vertem, the results have been proven over time.
I adapted the unique analytical style at Vertem to create a unique horse buying model. Horses are acquired not only with prospects on the racecourse in mind but also from a breeding objective, and this is where the algorithmic-based models came into their own.
The focus is to win races but also to ensure that, even if a purchase disappoints on the course, it should still have some value from a breeding angle and can be sold readily. This buy-and-sell culture is not dissimilar to investing in shares. It is often said to ‘not fall in love with a share’, and the hard truth is you must do the same approach for racehorses – although that can be harder said than done.
As with Vertem’s research processes, my horse buying models are continually evolving as more evidence is created as to what works best.
My target is to deliver top-industry returns for clients at Vertem and a strong stable of racehorses to enjoy in the future.
As the late comedian Bob Monkhouse once said: “They laughed when I said I was going to be a comedian…they’re not laughing now.”
Someday, using algorithms to select racehorse purchases might well be taken seriously.
– Advertising feature –