In our opinion, the Kelly Criterion should be viewed as a thought exercise—a tool to help investors think strategically about where their best opportunities lie and how to size positions relative to the risks and rewards they perceive. The assumption of independent bets, infinite opportunities, and accurate probability estimates can lead to missteps if applied rigidly in the complex, interconnected world of investing. We have just done an article on it if you’re interested.
Thank you! Agree that the efficacy is limited by your accuracy in determining expected returns. I do not assume independent bets and the process I laid out accounts for covariances between assets. Covariances are based on historical data and there is no reason to assume the assets will remain correlated in a similar way in the future. Therefore my process involves recalculating Kelly allocations periodically with evolving data. There are other alternate ways of allocating capital - equal weight, risk parity, etc. None of these are perfect. Kelly criterion has worked well for me
In our opinion, the Kelly Criterion should be viewed as a thought exercise—a tool to help investors think strategically about where their best opportunities lie and how to size positions relative to the risks and rewards they perceive. The assumption of independent bets, infinite opportunities, and accurate probability estimates can lead to missteps if applied rigidly in the complex, interconnected world of investing. We have just done an article on it if you’re interested.
https://open.substack.com/pub/schwarcapital/p/where-gambling-meets-investing-the?r=2m1atw&utm_medium=ios
Thank you! Agree that the efficacy is limited by your accuracy in determining expected returns. I do not assume independent bets and the process I laid out accounts for covariances between assets. Covariances are based on historical data and there is no reason to assume the assets will remain correlated in a similar way in the future. Therefore my process involves recalculating Kelly allocations periodically with evolving data. There are other alternate ways of allocating capital - equal weight, risk parity, etc. None of these are perfect. Kelly criterion has worked well for me