> Paste an equity curve and get maximum drawdown, the peak and trough behind it, time under water, and whether the series ever recovered. Working shown. - Canonical: https://hadalinstruments.com/tools/maximum-drawdown-calculator/ --- Tool # Maximum drawdown calculator Maximum drawdown is the deepest decline from a running peak anywhere in a record. Most calculators stop there. The number that ends accounts is usually the other one — how **long** you spent below the peak — because capital, patience and mandates expire on the clock rather than on the percentage. Your equity curve 10000, 10420, 10310, 9880, 9640, 10050, 10390, 10120, 9730, 10480, 10910 Balances in order, separated by commas, spaces or new lines. Paste a column straight from a spreadsheet if you like. Instrument EUR/USD GBP/USD USD/JPY XAU/USD (gold) US30 / Dow NAS100 Something else Value per pip, per lot Typical value — overwrite it with your broker's. Position size (lots) Drawdown allowance If you trade under one. Leave at zero to skip the comparison. Maximum drawdown — Depth in pips at your size — From peak — To trough — Periods under water — Against your allowance — ## How it is computed Walk the series from left to right holding the highest value seen so far. At every point the drawdown is that running peak minus the current value. The maximum drawdown is the largest of those differences, and the percentage is that difference divided by the peak it fell from — not by the starting balance, which is a common and flattering error. Duration is measured separately: from the peak that began a decline to the first later point that exceeds it. Where no later point exceeds it, the series never recovered and the drawdown is still open — the tool says so rather than reporting a recovery that has not happened. ## What the series has to be before the answer means anything The walk above is only as honest as the curve you paste into it, and the most consequential property of that curve is one most people set without noticing: how often it is sampled. A series built from closed trades moves only when a position is booked, so every excursion that happened while a trade was open — the float against you before it came back — is invisible to it. A series sampled at the end of each day catches more. A true equity series catches all of it. The same account can produce three different maximum drawdowns this way, all of them correctly computed. That gap matters most in the case this page is usually opened for. Where a floor is tested against equity rather than against booked results — and which of the two governs you is a property of your rule text, not of the word *drawdown* — a closed-trade curve can report a comfortable maximum drawdown for an account that was already breached, and it will do it without any error in the arithmetic. Only the record is narrow. Which anchor your own programme uses is the subject of [whether a drawdown limit is computed on balance or on equity](https://hadalinstruments.com/research/is-my-drawdown-on-balance-or-equity/); what a statement can and cannot establish about a path is the subject of [what a trading statement proves](https://hadalinstruments.com/research/what-can-a-trading-statement-prove/). Whichever series you use, use one throughout: mixing granularities inside a single curve produces a number that describes no account at all. ## Two things that decide whether the number means anything **Maximum drawdown grows with record length.** Even for a strategy whose behaviour never changes, the deepest excursion observed over ten years will in expectation exceed the deepest observed over two — more history is more opportunity for a bad sequence. Comparing the maximum drawdowns of records of different lengths, without adjustment, is not a comparison. **A backtested figure is biased shallow.** A configuration chosen from many trials was selected partly for having dodged the worst sequences in that particular data. The live distribution is systematically worse before anything about the market changes. If this series came from a backtest, treat the answer as a floor rather than an estimate. ## What this does not do The calculator computes no probability of anything. It does not tell you whether a drawdown of this size is acceptable, what position size to use, or whether the strategy behind the curve is sound — those depend on constraints this page cannot see, and a tool that answered them would be guessing on your behalf. What turns a drawdown into an ending is a barrier: a margin requirement, a prop-firm floor, a redemption trigger, a personal breaking point. That is the domain of [risk of ruin](https://hadalinstruments.com/glossary/risk-of-ruin/), and the barrier that moves is a [trailing drawdown](https://hadalinstruments.com/glossary/trailing-drawdown/). ## Depth in pips, and depth against an allowance A decline of 780 in account currency is not a quantity you can act on, because you do not hold account currency — you hold lots. The same 780 is 78 pips at one standard lot on a ten-per-pip instrument and 7.8 pips at ten, and only the second framing tells you whether the decline was a normal excursion at your size or an outlier. Set your instrument and position size above and the depth is restated in the unit you trade. The allowance field turns the descriptive question into the decisive one. A drawdown is only an ending when it crosses a barrier, so entering the allowance you trade under answers whether this excursion would have survived it. Read that answer narrowly: it compares against a **static** floor, the simplest case. Under a trailing rule the floor climbs behind every new high, so a decline that clears a static allowance comfortably can still breach a trailed one — the [trailing drawdown calculator](https://hadalinstruments.com/tools/trailing-drawdown-calculator/) is where that arithmetic lives. The term is defined at [drawdown](https://hadalinstruments.com/glossary/drawdown/). Where a floor moves with your equity rather than staying put, the [trailing drawdown calculator](https://hadalinstruments.com/tools/trailing-drawdown-calculator/) computes the distance that governs your next order; where the constraint is on the size of a single day rather than a decline, that is the [consistency rule calculator](https://hadalinstruments.com/tools/consistency-rule-calculator/). Everything here measures a decline that already happened; for the forward question — how likely a given sizing is to reach a floor at all — that is the [risk of ruin calculator](https://hadalinstruments.com/tools/risk-of-ruin-calculator/). The return side of the same record — and how wide its error bars actually are once the track record is short — is the [Sharpe ratio calculator](https://hadalinstruments.com/tools/sharpe-ratio-calculator/). All of them are listed under [calculators](https://hadalinstruments.com/tools/). A drawdown is only as honest as the curve it was measured on, and whether that curve is a faithful record is a different question with a different instrument — [Data Forensics](https://hadalinstruments.com/instruments/data-forensics/). What the same book would have done through a day the curve does not contain is the [Stress Harness](https://hadalinstruments.com/instruments/stress-harness/). And when the inputs cannot support an answer, the panel above refuses and says why rather than rendering a zero — [why an instrument refuses to answer](https://hadalinstruments.com/docs/insufficient-n/) is that design, written down.