A backtest hands you a pile of numbers, and it is tempting to grab the friendliest one and call it proof. Two figures deserve to be read together rather than in isolation: the Sharpe ratio and the maximum drawdown. One describes the quality of the returns; the other describes the pain you would have endured to collect them. Read either alone and you can be badly misled.
Sharpe: reward per unit of risk
Sharpe is not a profit figure. It measures how much return a strategy earned for each unit of volatility it took on. A higher Sharpe means the returns were steadier relative to their swings. It is useful precisely because it refuses to reward a strategy for taking wild risk to reach a headline number.
Drawdown: the worst it got
Maximum drawdown is the largest fall from a high-water mark to the trough that followed. It answers a blunt, human question: how deep was the hole you would have had to sit in? A strategy with a strong average can still have a drawdown deep enough that no real person would have held through it.
Reading them as a pair
The two numbers check each other. A high Sharpe with a shallow drawdown is genuinely encouraging. A high Sharpe with a brutal drawdown is a warning that the smooth average hides a cliff. A modest Sharpe with a tiny drawdown might be perfectly tradeable for someone who values calm over flash. The point is to refuse to let one number speak for the whole result.
And only out-of-sample counts
Both metrics are easy to inflate on tuned data. The figures that matter are the ones that survive on data the strategy never saw — the out-of-sample folds of a walk-forward run. A great in-sample Sharpe that collapses out-of-sample is the signature of overfitting, not skill.
Follow the steps above against your own results, then read how to spot overfitting in a strategy and walk-forward testing explained. To score a live track record with the same Sharpe lens, the tools pillar has a calculator. Backtests do not guarantee future results.