Max Drawdown

A drawdown is any decline from a running equity peak. The maximum drawdown (MDD) is the deepest one in the tested period:

drawdown(t) = (peak so far − equity(t)) / peak so far
MDD         = max over t of drawdown(t)

An account that grows to $12,000, sinks to $9,000, then recovers, took a (12,000 − 9,000) / 12,000 = 25% max drawdown, regardless of where it ended up.

Why it may matter more than return

Drawdown is what your psychology and your solvency actually experience. Recovery is asymmetric, too: a 25% drawdown needs +33% to get back, and a 50% drawdown needs +100%. Leverage multiplies drawdowns to the point of liquidation long before it multiplies long-run returns.

Reading it in a backtest

Compare MDD against the strategy's annual return (that ratio is the Calmar ratio), and check duration as well as depth. A shallow drawdown that lasts two years is harder to live with than a sharp two-week one. Remember that the historical MDD is a sample, not a bound; the future's worst drawdown is usually worse. Monte Carlo simulation over reshuffled trade sequences gives a more honest range.

On AlphaProve

Depth is only half the story, so the tearsheet pairs the equity curve with an underwater chart — the running distance below the prior peak — and a drawdown-duration card that measures how long each dip took to recover, not just how far it fell. Drawdown is also something you can act on before the fact: the portfolio kill-switch max_drawdown_pct halts new entries in the simulation once the account falls a set percentage from its peak, alongside a daily-loss limit and gross-exposure caps.