Deflated Sharpe Ratio
The Deflated Sharpe Ratio (DSR), introduced by Bailey and López de Prado (2014), adjusts a strategy's observed Sharpe ratio to account for the number of strategy variations tested. If you tried 100 parameter combinations, one of them will look good by chance alone — DSR quantifies exactly how much that inflates your result.
What does DSR tell you?
DSR outputs the probability that the observed Sharpe ratio is above zero after adjusting for selection bias. A DSR below 0.95 suggests the result is statistically indistinguishable from luck at the 5% significance level. A DSR above 0.99 indicates a very strong, likely genuine edge.
How do you compute it?
The inputs are: the strategy's annualised Sharpe ratio, the number of independent strategy trials tested (T), the number of observations in the backtest (N), the skewness and kurtosis of the returns series, and a target Sharpe ratio (typically 0). The formula computes the probability that the maximum of T independent Sharpe ratio estimates exceeds the target when the true expected Sharpe ratio is zero.