The Sharpe Ratio, introduced by William Sharpe (1966), measures return per unit of total risk:
Sharpe = (R − R_f) / σ
where R is the annualized portfolio return, R_f is the risk-free rate, and σ is the annualized standard deviation of returns. Both the numerator and denominator are annualized — for daily returns, multiply mean daily excess return by 252 and standard deviation by √252.
The Sharpe Ratio is the de facto standard performance metric in quantitative strategy evaluation, though it has important limitations:
- Normality assumption — it assumes returns are normally distributed. For non-normal returns (fat tails, negative skew), the Sharpe overstates risk-adjusted performance, which is why the Deflated Sharpe Ratio and Sortino Ratio exist.
- Symmetric penalization — it penalizes upside variance equally with downside variance.
- Data-mining sensitivity — it is easily inflated when many strategies are compared without correcting for multiple testing.