Sharpe Ratio

The Sharpe Ratio, introduced by William F. Sharpe in 1966, is the most cited measure of risk-adjusted return: excess return divided by total volatility. It is universally understood but quietly assumes that volatility is a fair proxy for risk — true under Gaussian returns, less true once tails enter the picture.

Formula

Sharpe = ( R_p − R_f ) / σ_p

where R_p is the portfolio annualised return, R_f the risk-free rate, σ_p the annualised volatility. Daily returns are typically annualised by multiplying by 252 (R_p) and √252 (σ_p).

How to read the number

• ·Sharpe < 0 — strategy underperforms cash, not investable.
• ·Sharpe 0-1 — typical for a long-only equity benchmark over a full cycle.
• ·Sharpe 1-2 — well-diversified multi-asset with active risk management.
• ·Sharpe > 2 — either an alpha-generating strategy, a short measurement window, or in-sample data mining. Treat with scepticism.

Where Sharpe breaks

• ·Fat tails: a strategy that sells crash insurance has a beautiful Sharpe until the day the insurance pays out. Use CVaR alongside.
• ·Skewness: Sharpe is symmetric; it cannot distinguish "win small, lose big" from "win big, lose small" at the same volatility.
• ·Time-window sensitivity: the same strategy can show Sharpe 1.8 over 5y and 0.6 over 20y. Always disclose the window.

How MEDGE Capital uses Sharpe

Sharpe is reported in every backtest alongside Sortino, Calmar, and the rolling 63-day Sharpe ribbon. Max Sharpe is one of the 12 optimization objectives, with a constraint to avoid the well-known "Sharpe trap" of pushing weight onto a single high-Sharpe asset. The rolling chart shows how stable the Sharpe is over the window — a flat ribbon means real edge, a spiky one means a few good months.