← All posts
Metrics6 min read

The Sharpe Ratio Is Necessary. It Is Not Sufficient.

A Sharpe of 1.8 looks great on paper. But if your strategy draws down 40% in a single quarter, your LPs will not care about the annual number.

The Sharpe ratio is the lingua franca of quantitative finance. Every allocator asks for it; every backtest reports it. And it is genuinely useful — but only as one data point in a richer picture.

What Sharpe measures

Sharpe = (R_p - R_f) / σ_p. It rewards strategies that generate excess returns with low volatility. The maths is clean and the intuition is sound: more return per unit of risk is better.

What it misses

The formula assumes returns are normally distributed. Most strategies — especially those involving options, tail-hedging, or crisis-alpha — have return distributions with fat tails and significant skewness. A strategy that makes 0.5% every week for 51 weeks and then loses 28% in week 52 can still report a Sharpe above 1.5.

The metrics that complement Sharpe

  • Sortino ratio — penalises downside volatility only
  • Calmar ratio — annualised return divided by max drawdown
  • Max drawdown duration — how long did it take to recover?
  • Tail ratio — the 95th percentile of gains vs. the 5th percentile of losses
  • Skewness & kurtosis — shape of the distribution matters

The regime question

A strategy with a Sharpe of 1.6 that achieved that number entirely in 2017–2021 (the longest low-volatility bull market in modern history) is a very different animal from one that maintained 1.4 through 2008, 2020, and 2022. Always decompose performance by market regime.