← All 25 termsQuant methodology
PBO (Probability of Backtest Overfitting)
QuantThe probability that the best backtest winner, selected by in-sample performance, underperforms the median out-of-sample. Estimated by CSCV.
Full definition
PBO is the probability that a backtest search overfits. i.e. that the 'best' in-sample strategy underperforms the median OOS strategy, implying the selection was driven by chance rather than signal. PBO is estimated by Combinatorially Symmetric Cross-Validation (CSCV). A PBO below 25% is conventionally acceptable; above 50% means the search has produced more losers than winners regardless of which one looked best in-sample. On this site, PBO is reported as a per-project transparency metric on /proof.
Related terms
Each link opens the long-form definition of a closely related term.
See this term in action.
The /methodology page maps every term to the gate or artifact that implements it on this site.