A Bootstrap Certainty-Equivalent Approach to Stable Performance Evaluation Under Non-Normality
Agenda
Presented by guest speaker Boris Klebanov, Ph.D
The traditional Information Ratio (IR) assumes normally distributed active returns, an assumption violated in practice. diBartolomeo (2023) shows that applying volatility equivalence – a method that maps non-normal distributions with the same volatility – produces an IR that is unstable when performance expectations are anchored to a fixed positive alpha.
We propose a simple, model-free alternative: the bootstrap certainty-equivalent (BCE). This method evaluates a manager’s track record directly through the lens of expected utility, using resampling to avoid explicit estimation of higher moments and any artificial normal mapping. The resulting metric is stable, economically meaningful, and naturally yields a manager’s skill probability. A numerical illustration demonstrates the robustness compared to the volatility-equivalent IR.
About Boris Klebanov
Boris Klebanov is an independent consultant with many years of experience dealing with complex investment analytics.
He holds a Ph.D. degree in Mathematics.