Maximum drawdown, explained

Maximum drawdown: the short answer

Maximum drawdown (MDD) is the largest peak-to-trough decline an investment or strategy suffers over a period, measured as a percentage of the prior peak. If a portfolio climbs to $15,000, sinks to $9,000, then recovers, its maximum drawdown is 40%. It captures the worst loss you had to endure to earn the return.

How maximum drawdown is calculated

The math is simple, which is part of why it's honest. You walk the equity curve forward, and at every point you ask: how far below the highest value seen so far is the strategy right now? That gap, in percent, is the current drawdown. The maximum drawdown is simply the deepest of all those gaps across the whole period.

Formally:

Work through the classic example. You start at $10,000. The account grows to $15,000 — a new peak. Then it falls to $9,000 before recovering. The drawdown from that peak is ($9,000 − $15,000) ÷ $15,000 = −40%. Even though you never dipped below your original stake, you lived through a 40% collapse from the high-water mark. That number is the maximum drawdown.

Two details matter. The peak must come before the trough — you can't measure a decline against a high that hadn't happened yet. And drawdown is measured against the running peak, not the starting value, so a strategy that doubles and then halves still books a painful drawdown even if you're up on the year.

Why it matters more than average return

Average return tells you what happened on paper. Maximum drawdown tells you whether you'd have still been holding to collect it. Those are very different questions.

Two strategies can post the same annual return while one grinds steadily upward and the other lurches through a 55% crater along the way. On a spreadsheet they tie. In real life, most people sell somewhere near the bottom of that crater — because a deep drawdown isn't an abstraction, it's months of watching an account bleed and wondering how much worse it gets. Return is what a strategy earns; drawdown is what it costs you to stay invested.

• It's the metric you actually experience. Volatility and standard deviation blend upside and downside together. Drawdown isolates the part that hurts — the fall from a high.
• It sets position sizing. A strategy with a −50% worst case demands a smaller allocation than one with a −15% worst case, regardless of headline returns.
• It exposes fragility. A shallow drawdown history across many regimes is far more reassuring than a spectacular return earned in one lucky stretch.

A useful companion rule: recovery is asymmetric. A 20% drawdown needs a 25% gain to get back to even; a 50% drawdown needs a 100% gain. The deeper the hole, the more brutal the math of climbing out — which is exactly why keeping drawdown shallow compounds so powerfully over time.

How to read a drawdown number honestly

A drawdown figure is only as trustworthy as the data behind it, and this is where a lot of published numbers quietly cheat.

• Check for survivorship bias. If a backtest only includes companies that still exist today, it has silently deleted every name that went to zero — the exact events that produce the worst drawdowns. A survivorship-free test that keeps delisted names will show a deeper, more honest low.
• Always demand a benchmark. A −23% drawdown means nothing in isolation. Next to a market that fell −32% over the same window, it's meaningful. Next to a market that fell −10%, it's alarming. Drawdown is a relative statement.
• Ask over what period. A one-year drawdown and a ten-year drawdown aren't comparable. The longer and rougher the window, the more the number has been stress-tested.

The honest way to present drawdown is never as a lonely percentage. It's always paired with the same metric for a passive benchmark over the identical, survivorship-free window — so the reader can weigh the trade-off between how much a strategy earned and how much it made them suffer.

How Coil reads it

Maximum drawdown is the metric Coil treats as sacred, because it's the one that's easiest to hide and most painful to ignore. Coil is long-only and rules-based software you run yourself, and its design goal isn't just to chase return — it's to take less pain getting there. Coil buys weakness in an uptrend rather than chasing breakouts, and it holds cash when the tape says so, because cash is a position. So every performance claim ships next to its drawdown, and the drawdown ships next to the market's. Over 2017 through 2026 H1, the leadership-rotation backbone Coil is built on backtested +638% versus SPY's +282% — survivorship-free, delisted names included, next-open fills, costs modeled — with a worst drawdown of −23% against SPY's −32%, positive in 9 of 10 years. The honest rider matters more than the headline: through the end of 2025 that backbone ran roughly even with SPY at about one-third less drawdown, and the outperformance concentrates in leadership regimes. These are research backtests, not live or client returns, and the engine is newly live. Nothing here is investment advice — it's educational, a way of showing that a drawdown tells you the price, not just the prize. You can see how the top-down read that shapes it works on the how-it-works page.

People also ask

What is a good maximum drawdown?

There's no universal threshold — it depends on the strategy and the market it ran through. The real test is comparison: a good drawdown is meaningfully shallower than a passive benchmark's over the same period, ideally while keeping pace on return. A −20% drawdown is excellent against a −35% market and poor against a −5% one.

What is the difference between drawdown and maximum drawdown?

A drawdown is any decline from a running peak. The maximum drawdown is the single deepest of all those declines over the period you're measuring. There are many drawdowns in any equity curve; there is only one maximum drawdown per window.

Why does maximum drawdown matter more than volatility?

Volatility measures swings in both directions, including upside you'd happily keep. Maximum drawdown isolates only the downside — the peak-to-trough loss you actually have to sit through. It maps directly to the moment investors panic and sell, which is why it tracks real-world behavior better than standard deviation.

How do you recover from a maximum drawdown?

Recovery is asymmetric: gains have to outrun the percentage lost, on a smaller base. A 20% drawdown requires a 25% gain to break even, and a 50% drawdown requires a 100% gain. This math is exactly why limiting drawdown depth matters so much to long-term compounding.

Related terms

Trailing stop · Survivorship bias · Trend following · Risk-on / risk-off · full glossary →