Calmar ratio = annualized return / absolute value of max drawdown. A strategy that returns 30% annually with a 15% max drawdown has a Calmar of 2.0.
Better than Sharpe for trend-followers because it ignores upside volatility entirely. Sharpe punishes big winning months because they raise the standard deviation of returns; Calmar doesn't care how rough your equity curve looks on the way up, only how far it falls from peak to trough.
Common benchmarks (descriptive, not prescriptive): Calmar above 1.0 is solid, above 3.0 is exceptional, above 5.0 is usually small-sample or overfit. CTAs and managed-futures funds frequently market themselves on Calmar precisely because their style of return (occasional big winners punctuated by long flat periods) reads poorly under Sharpe but cleanly under Calmar.
The catch: Calmar is only as honest as the window you compute it on. Most strategies haven't lived through a real max drawdown yet — their reported Calmar assumes the worst hasn't happened. A 36-month track record with a 10% reported max drawdown is suspect; the real number is usually 2-3× worse over a full cycle.