Still a few weeks to go until the start of the Premier League for 08/09 but I thought I would lay out the expectations for the performance of the Funds for next season. 

After improving the algorithm, the greatest decision i’ve had to make is what level of risk to take.  On the one side, increasing stakes lead to higher returns but at the same time adds to the volatility of the Fund.  The biggest change in the current model over the model used for last season is a vastly reduced stake size.  The question is, how much volatility am I now comfortable with?

In the table below are some figures to illustrate the effect of changing the maximum stake size on risk and return over the 8 historical seasons 2000/1-2007/8.  The % max. stake is the maximum bet possible as a percentage of the Fund balance.  The Fund balance begins at £100.

The figures show the performance of the model using backtested historical data.  The performance of the Premier League Fund next season could therefore be higher or lower.  I do expect the Trading Fund to outperform these figures as it can allow for major injuries and other non-quantifiable data.

*Note season returns do not follow the normal distribution due to % stake betting.  As for a normal distibution, where 95% of expected returns lie within -/+2 standard deviations of the mean, this is not the case for the model’s returns.

**Max. potential loss from any one bet as % of Fund.  This is below the max. stake size due to value betting; i.e. each bet is reduced based on the amount of value the price is calculated as offering.

***A measure of how long the Fund ’stays under water’.  A figure of 40 would suggest the Fund is expected to reach a new high every month (10 matches/week).

****The r-squared measures the tightness of fit between the Fund balance and an exponential line of best fit over time.  A figure of 1 is a perfect fit, 0 no fit at all. As all gains are reinvested and the model uses % stakes we would expect the returns to be exponential.  The r-squared is a measure of how close the returns are to this exponential line of best fit.  You can calculate this in Excel one of two ways.  Either create a graph of the Fund over time and add an exponential trendline, Excel then allows you to show the r-squared figure on the graph.  Alternatively, you can calculate the natural log of the Fund balance over time and calculate the r-squared of this figure against the sequential match number.  The divergence of the model from 1 is mainly due to the last 2 seasons where the Fund really takes off.  The exponential power in the best fit line underestimates the returns! See graph below.

Observations

The 5% stake model is low risk but hardly gets the pulse racing.  It takes 8 seasons to build the Fund tenfold but you can expect a new high to be reached every 3 months.  The 20% stake builds the Fund 48x over the 8 seasons but a new high can only be expected around every 5 months.  The new highs however can be significant with a 141% gain coming in just 1 month.

A point worth noting here is that only the last 2 seasons use actual betting exchange prices, these are taken from Betfair.  Previous to this, I have created prices based on traditional bookmaker prices adjusted to remove the profit margin. Whilst these prices are believed to be accurate of expected exchange prices they are not as good as using real exchange prices.  Gladly, the % returns for the model in the last 2 seasons are similar or higher than in the earlier seasons.

The model uses % stake betting; i.e. stakes are based on a % of the Fund balance, so the bankruptcy risk is minimal.  Coupled with the proven performance of the model in backtesting, monte carlo testing and last season’s performance I am opting for the 20% max. stakes for both the Premier League Fund and Trading Fund.

Whether the season will be profitable or not for the Premier League Fund I don’t know.  But I will be disappointed if the Trading Fund doesn’t make a decent return.