What Is Maximum Drawdown?
Maximum drawdown is the largest percentage fall from a peak in account value to the lowest point before a new peak. It is the single most useful risk number because recovery is asymmetric: a 50% drawdown requires a 100% gain to get back to level.
What is the recovery arithmetic?
To recover a drawdown of d, the required gain is 1 ÷ (1 − d) − 1:
| Drawdown | Gain needed to recover |
|---|---|
| 10% | 11.1% |
| 20% | 25.0% |
| 30% | 42.9% |
| 50% | 100.0% |
| 70% | 233.3% |
| 90% | 900.0% |
This asymmetry is why risk control dominates return-seeking. Avoiding a 50% drawdown is worth more than any strategy improvement that does not.
Why does duration matter as much as depth?
Because people quit on time, not on percentage. A 25% drawdown that recovers in six weeks is survivable. The same 25% lasting fourteen months causes the trader to change the system, cut size at the bottom, or stop — all of which convert a temporary drawdown into a permanent one.
So ask for two numbers: the depth, and the number of months from peak to new peak. A record that reports the first and not the second is reporting the easier half.
What does max drawdown miss?
It is a single observation. The worst drawdown in a sample is, by definition, the tail of what has happened so far — not a bound on what can. The realistic planning assumption is that the future worst is larger than the past worst.
It is sample-length dependent. A longer record has more opportunity to contain a deep drawdown, so a strategy with a 12% max drawdown over one year and one with 22% over eight years are not directly comparable.
Marking matters. Drawdown computed on closed trades only misses the open positions, which is exactly where an unbooked loss sits.
What should you ask for?
Worst peak-to-trough percentage, the dates it spanned, whether it is computed on marked equity or closed trades, and how many months it took to make a new high. A trader who answers all four immediately with specific uncomfortable numbers is showing you the strongest positive signal available.
How does this connect to a track record?
Volunteering the worst period is not a courtesy, it is what a performance standard requires. The CFA Institute's GIPS standards rest on fair representation and full disclosure across every portfolio in a composite[1] — a presentation that omits the bad stretch fails the standard by construction, whoever is presenting it.
The worst period is also the first thing selection removes. A performance study limited to survivors produces spurious evidence of skill precisely because the failures dropped out.[2] And where a ratio like Sharpe is an estimate carrying its own standard error,[3] a drawdown is a realised fact — it happened, at a date, to a real balance.
A metric is a summary of a record, so it inherits every weakness of that record. Computed from trades selected after the fact, it is a number about the selection. kappi commits each trade before it resolves and publishes it on a Merkle-anchored log, so PnL, RME, correlation to SPX, mean R:R and trade count over 30, 100 and 200-day windows are computed over everything that was committed, losses included. $15/month to keep a record; reading one is free.
Sources
- CFA Institute, Global Investment Performance Standards (GIPS) for Firms, 2020 edition read 2026-08-16
- Brown, Goetzmann, Ibbotson & Ross, 'Survivorship Bias in Performance Studies', Review of Financial Studies 5(4), 1992, 553–580 read 2026-08-16
- Lo, 'The Statistics of Sharpe Ratios', Financial Analysts Journal 58(4), 2002, 36–52 read 2026-08-16
Frequently asked questions
How much do you need to gain to recover a drawdown?
1/(1−d) − 1. A 20% drawdown needs 25%, a 30% drawdown needs 42.9%, a 50% drawdown needs 100%, and a 90% drawdown needs 900%.
What is an acceptable maximum drawdown?
It depends on the return earned for it and on what you can sit through. The more useful question is the duration: a 25% drawdown lasting six weeks is survivable, the same depth lasting fourteen months usually is not.
Is max drawdown a worst case?
No. It is the worst that has happened in the sample so far, which is a tail observation rather than a bound. Plan on the assumption that the future worst exceeds the past worst.
Should drawdown be measured on closed trades or marked equity?
Marked equity. Computing it on closed trades only excludes open losing positions, which is exactly where an unbooked loss is hiding.