Max Drawdown (MDD) Explained: Meaning, Math and Recovery
Maximum drawdown (MDD) is the largest fall in an account's or strategy's value from a peak to a later low. This guide shows how to calculate it, why losses need bigger gains to recover, how leverage can push a drawdown past 100% of the money allocated, and how to size a position so you can live with it.
What maximum drawdown means
A drawdown is any decline from a running peak in equity. The maximum drawdown is the deepest one over the period measured:
MDD = (lowest value after the peak - peak value) / peak value
A hypothetical example: an account goes 1,000 → 1,400 → 980 → 1,500 USDT. The peak before the fall is 1,400 and the low is 980, so the drawdown is (980 - 1,400) / 1,400 = -30%. The later climb to 1,500 doesn't erase it; MDD records the worst stretch you had to sit through.
Two companion numbers matter as much:
- Time underwater: how long the account stayed below its previous peak.
- Frequency: one deep drawdown in five years is different from one every few months.
Peak-based vs start-based drawdown
In the example above, the account at 980 USDT is only 2% below its starting 1,000. Measured from the start, the drawdown looks tiny; measured from the peak, it is 30%. Peak-based drawdown is the standard measure, and it matches what you feel: you watched 420 USDT of gains disappear.
Two more things to check in any report:
- Closed-trade vs open-position drawdown. A curve built only from closed trades hides how far open positions fell before closing; for strategies that hold through swings or add to positions, the mark-to-market drawdown is deeper.
- Capital base. Is the drawdown a share of the whole account or of the amount allocated to the strategy? With leverage, the difference is large.
The recovery math: why losses hurt more than gains help
After a loss, recovery takes a larger percentage gain, because the gain is earned on a smaller balance:
Gain needed to recover = 1 / (1 - drawdown) - 1
| Drawdown | Gain needed to get back to the peak |
|---|---|
| 10% | 11.1% |
| 20% | 25% |
| 30% | 42.9% |
| 50% | 100% |
| 70% | 233% |
| 90% | 900% |
The curve is gentle at first and then brutal. A 30% drawdown needs a 43% gain; a 50% drawdown needs the account to double. That asymmetry is why limiting deep drawdowns usually matters more than squeezing out extra gains.
How leverage pushes a drawdown past 100%
In spot trading without leverage, the worst case is the coin going to zero: you lose what you put in, and no more.
With futures, profit and loss are calculated on the full position size, not on the margin you post. A hypothetical example: you allocate 1,000 USDT to a strategy that trades a 3,000 USDT position (3x). On paper, a 35% move against it loses 1,050 USDT, more than the entire allocation. Add-on entries raise the position size further, and a series of losing trades can stack on top. Measured against the allocation, the peak-to-trough drawdown can exceed 100%.
Live, with isolated margin, the position is liquidated once its margin runs out; with cross margin, the rest of your futures balance is used to keep it open. So a backtest drawdown above 100% of the allocation tells you that, live, you would have been liquidated or paid the excess from your other funds. The mechanics are explained in liquidation, margin and funding.
An honest example from our own testing. In Pullwave's SOL futures backtest, the maximum peak-to-trough drawdown exceeded 100% of the allocated amount. The backtest doesn't simulate liquidation or margin, so it can't show what would have happened to a live account sized the same way. That's why we suggest keeping the allocated amount well below your balance and starting small. The full backtest and every trade are on the performance page.
How to size a position around drawdown
Work backward from the loss you can accept, not forward from the gain you hope for.
- Start with the worst historical drawdown, as a share of the allocated amount.
- Assume the future can be worse. Backtests tuned on past data are in-sample, and the next bad stretch may look nothing like the last one (see bull, sideways and bear markets). A common rule of thumb is to plan for one and a half to two times the historical worst.
- Decide the most you could lose from the whole account without it changing your life or your behavior.
- Divide: allocation = acceptable loss / planned drawdown.
A hypothetical example: you can accept losing 300 USDT, and you plan for a drawdown of 150% of the allocation. Allocation = 300 / 1.5 = 200 USDT. With a planned drawdown above 100%, the allocation must be smaller than the loss you can bear.
Then check each trade: the risk from entry to exit, including add-ons, and whether the liquidation price sits well beyond the planned exit.
Can you actually sit through it?
A drawdown on a chart feels very different from one in your own account. Ask yourself:
- If the account fell this far over several weeks, would you stick to the rules, or quit near the bottom?
- How many months could you stay below your previous peak without abandoning the plan?
- Is this money you might need within a year?
- After a losing streak, would you raise your size to win it back?
If any answer worries you, lower the allocation. Most people overestimate their tolerance until they have lived through a real drawdown, so start small and scale up only after you have. To judge a published drawdown in context, see how to read a backtest.
FAQ
Q. Does a lower max drawdown mean a better strategy?
Not on its own. A low drawdown from a short test may only mean the strategy hasn't met a bad market yet. Compare it with the test length, the number of trades and the market conditions covered.
Q. Can a drawdown be more than 100%?
Not for an unleveraged spot account, where the worst case is losing what you put in. With leverage and add-ons, losses measured against an allocated amount can exceed it in a backtest. Live, the position would be liquidated or the extra loss would come out of your other funds.
Q. What is the difference between a loss and a drawdown?
A loss is the result of one trade. A drawdown is the fall in total equity from a peak across trades and open positions, so several small losses in a row can add up to a deep drawdown.
Q. How long does it take to recover from a drawdown?
It depends on how deep the drawdown is and how quickly the strategy gains, and the recovery table shows that deeper drawdowns need disproportionately larger gains. Check the longest time underwater in any backtest to see how long past recoveries took.
Note: This article is general education, not investment advice. Backtest results describe the past only and do not guarantee future results. Futures can be liquidated and you can lose your capital. Start small.