The Calmar ratio is annualised return divided by maximum drawdown over the same period.
Calmar = annualised return ÷ |maximum drawdown|
A strategy returning 30% a year with a worst peak-to-trough loss of 15% has a Calmar of 2.0.
Why it exists
Sharpe penalises all volatility equally — a sharp rally hurts a Sharpe ratio as much as a sharp selloff. But investors don’t experience upside volatility as risk. What they experience as risk is the drawdown: how far the account fell from its high and how long it stayed there.
Calmar measures return against the thing that actually causes people to abandon strategies.
Reading it
Above 1.0 means the strategy earned more in a typical year than its worst loss. Above 3.0 is strong. Below 0.5 means a single bad stretch wiped out more than two years of typical gains.
The original definition uses a 36-month window. Shorter windows produce flattering figures because the worst drawdown hasn’t happened yet.
Its weakness
It depends on one number — the single worst drawdown — which makes it noisy and sample-dependent. Two strategies with identical return profiles can have very different Calmar ratios because one happened to be running during a crash.
That’s why it’s read alongside Sharpe and Sortino rather than instead of them. Each measures a different thing people mean by risk.