Maximum Drawdown: How to Calculate It and Recovery Math
Maximum drawdown is the largest peak-to-trough fall in equity. See the calculation, a recovery-gain table, and why one backtest shows one ordering.
Put this into practice with a watchlist
Build a watchlist, then review each signal’s entry, stop, target, and reasoning. Broker access is optional.
Public subscriptions are paper-only; separately approved private beta accounts may submit live broker orders.
Maximum drawdown: Drawdown measures an equity decline relative to an earlier peak over a chosen observation frequency.
What Is Maximum Drawdown?
Maximum drawdown is the largest percentage decline from a peak in account equity to the lowest point that followed it, within a chosen period. If an account rises to $11,500, falls to $9,800 and later makes a new high, the drawdown for that episode is (11,500 - 9,800) ÷ 11,500, or 14.8%, and the maximum drawdown for the period is the largest such episode. It describes how deep a particular history got, which makes it a common yardstick for how much loss a strategy or an account has had to absorb.
Key Takeaways
- The gain needed to recover is larger than the loss. Recovery needed = drawdown ÷ (1 - drawdown). A 20% drawdown needs a 25% gain; a 50% drawdown needs a 100% gain.
- Maximum drawdown depends on how it is measured. Closed trades versus marked-to-market equity, daily versus intraday observations and the length of the history all change the number.
- A backtest shows one ordering of the trades. In a hypothetical shuffle of 100 identical trades, the maximum drawdown ranged from about 6% to about 17% across the 5th to 99th percentile, with the same final result.
- The worst drawdown has not necessarily happened yet. A longer history or a different market regime can produce a deeper one.
- Limits are rules you set in advance. A drawdown rule decided before a losing run is different from a decision made in the middle of one.
How to Calculate Maximum Drawdown
- Record account or strategy equity at regular points (end of each day, or after each closed trade).
- At each point, track the running peak: the highest equity reached so far.
- Drawdown at each point = (running peak - current equity) ÷ running peak.
- Maximum drawdown is the largest value of step 3 over the period.
Hypothetical example. Monthly equity for a hypothetical account:
| Month | Equity | Running peak | Drawdown from peak |
|---|---|---|---|
| 0 | $10,000 | $10,000 | 0.0% |
| 1 | $10,800 | $10,800 | 0.0% |
| 2 | $11,500 | $11,500 | 0.0% |
| 3 | $10,900 | $11,500 | 5.2% |
| 4 | $10,200 | $11,500 | 11.3% |
| 5 | $9,800 | $11,500 | 14.8% |
| 6 | $10,600 | $11,500 | 7.8% |
| 7 | $11,900 | $11,900 | 0.0% |
| 8 | $11,300 | $11,900 | 5.0% |
| 9 | $10,500 | $11,900 | 11.8% |
| 10 | $12,400 | $12,400 | 0.0% |
| 11 | $12,000 | $12,400 | 3.2% |
The maximum drawdown is 14.8%, from the $11,500 peak in month 2 to the $9,800 trough in month 5. It took five months to get from the month 2 peak to a new high in month 7. That span is the drawdown duration, which matters as much as depth for the person living through it. The second drawdown (months 8 to 9) was shallower at 11.8% and ended with a new high in month 10.
In a spreadsheet, add a running-maximum column (the highest equity so far) and a drawdown column, then take the maximum of the drawdown column.
Why a Loss Needs a Larger Gain to Recover
A loss is measured against the peak, but the recovery is measured against the smaller balance that remains. Recovery needed = d ÷ (1 - d), where d is the drawdown as a decimal.
| Drawdown | Gain needed to get back to the peak | Months at a hypothetical steady 1% per month |
|---|---|---|
| 5% | 5.26% | about 5 |
| 10% | 11.11% | about 11 |
| 15% | 17.65% | about 16 |
| 20% | 25.00% | about 22 |
| 25% | 33.33% | about 29 |
| 30% | 42.86% | about 36 |
| 40% | 66.67% | about 51 |
| 50% | 100.00% | about 70 |
| 60% | 150.00% | about 92 |
| 75% | 300.00% | about 139 |
The last column assumes a constant 1% monthly gain, which no trading account is guaranteed to deliver. It is only meant to show how quickly the time to recover lengthens. This asymmetry is why many risk plans set limits well before the deep end of the table. See how risk of ruin works for the probability side of the same idea.
Why the Number Depends on How You Measure
- Closed-trade versus marked-to-market. Equity measured only when trades close can hide the open-position loss in between. Marked-to-market equity is usually the more complete measure, and it is deeper.
- Frequency. Month-end equity can miss a mid-month trough that daily or intraday data show. Maximum adverse excursion measures the worst point inside each trade.
- Sample length. The longer the history, the more chances for a deep decline. Comparing a six-month figure with a ten-year figure compares different things.
- Position size. Maximum drawdown in percent scales with risk per trade. A strategy tested at one size does not carry its drawdown to another size without recomputing.
- Costs and slippage. A frictionless backtest hides losses that a real account pays on every trade.
A Backtest Is One Ordering of the Trades
A backtest's equity curve is the result of trades arriving in the order they happened. The same trades in a different order produce different drawdowns. Reshuffling the trades and recomputing the maximum drawdown many times is a simple Monte Carlo method that shows this spread.
Hypothetical simulation. Take 100 trades from a strategy: 40 winners that each add 2% of equity and 60 losers that each lose 1% of equity, compounded. Every ordering ends at the same place, about +20.8%. Shuffling the order 100,000 times and measuring the maximum drawdown gave:
| Percentile of the shuffled sequences | Maximum drawdown |
|---|---|
| 5th | 6.0% |
| 25th | 7.8% |
| 50th (median) | 9.0% |
| 75th | 11.4% |
| 95th | 14.2% |
| 99th | 16.9% |
| Worst possible ordering (all 60 losers last) | 45.3% |
The result of the strategy is the same in every case; the experience of it is not. A plan that assumed a 6% drawdown because that is what the backtest showed would have been unprepared for the middle of this distribution. The method keeps the same assumption as the loss-streak tables in the risk of ruin guide: trades are treated as exchangeable. Real losses cluster, so real drawdowns can be deeper than a reshuffle suggests.
Drawdown Rules, and What They Cannot Do
Traders and funds sometimes pre-commit to rules tied to drawdown. The following structure is a hypothetical illustration of the form such a rule takes, not a recommendation for any account:
| Equity drawdown from peak | Pre-committed action (illustration) |
|---|---|
| 5% | Review the log for rule breaks and reduce size by a stated fraction |
| 10% | Trade only the highest-priority setups, paper-trade the rest |
| 15% | Stop live trading and review the plan before resuming |
Whatever the thresholds, three points apply. A rule only helps if it was written before the losing run. A gap can carry equity through a threshold before any action is possible. And cutting size during a drawdown also slows recovery, which is a trade-off to understand in advance. The daily loss limit works the same way over a single session, and the algorithmic trading risk management guide shows how automated systems apply such limits.
Related Ratios
- Calmar ratio: annualized return divided by maximum drawdown. For example, a hypothetical 24% annual return with a 15% maximum drawdown gives a ratio of 1.6. It inherits all of the measurement issues above. See the Calmar ratio entry.
- Drawdown duration: time from a peak to the next new high, which can be long even when the depth is modest.
- Expectancy: average result per trade. It says nothing about drawdown without the variability of results. See the expectancy formula guide.
Put the setup on a watchlist first
Use the rules in this guide to evaluate a signal’s entry, stop, target, and reasoning before deciding what, if anything, to do.
Public subscriptions are paper-only; separately approved private beta accounts may submit live broker orders.
Tracking Drawdown When You Review Signals
Paper tracking is a low-stakes way to see what a drawdown looks like for a rule set. Tradewink is a research and signals app: each signal lists an entry, stop, target and thesis, so you can log the planned risk and record the realized result, then compute your equity curve, running peak and drawdown from the log. Compare the observed maximum drawdown with the reshuffled range, since your log is one ordering of the trades. A signal's confidence score is a ranking aid, not a win probability. Tradewink is not a broker and not a registered investment adviser. Public subscriptions are paper-only; separately approved private beta accounts may submit live broker orders. Create a free account to paper-track signals.
Maximum drawdown: practical checks
Include open-position marks, cash flows and shared capital consistently. Closing equity can conceal a larger intraday trough; compare the same frequency and reporting window when ranking two drawdown numbers.
Read the Bailey et al., The Probability of Backtest Overfitting for the underlying definitions and limits. For a next step, use Closing Equity Drawdown: A Reproducible Intraday Example and Trading Risk Management: Sizing, Stops and Portfolio Heat.
Frequently Asked Questions
What is a good maximum drawdown?
There is no universal figure. It depends on the strategy, the position size, the length of the history and how much loss the person or mandate can tolerate. A drawdown that is acceptable on paper can feel different when it is real money, so many traders set a limit in advance and size to it.
How do you calculate maximum drawdown in Excel?
Put equity in one column, create a running maximum column, compute (running maximum - equity) ÷ running maximum in a third column, and take the maximum of that column. Use the same observation frequency every time you compare.
Why does a 50% loss need a 100% gain to recover?
The loss is measured against the peak, but the recovery is measured against the remaining balance. From $10,000 down 50% to $5,000, a 100% gain is needed to return to $10,000. The needed gain is d ÷ (1 - d).
Is a smaller backtest drawdown proof a strategy is safe?
No. A backtest is one ordering of trades, one market history and one set of assumptions about costs. Reshuffling trades, testing other periods and including costs can show a wider range of drawdowns.
How is maximum drawdown different from risk of ruin?
Maximum drawdown is a measured historical decline. Risk of ruin is a modeled probability of reaching a defined loss threshold in future. See the risk of ruin guide.
How long does it take to recover from a drawdown?
It depends on the rate of return after the drawdown, which is uncertain. Deeper drawdowns need disproportionately larger gains: at a hypothetical steady 1% per month, a 20% drawdown takes about 22 months and a 50% drawdown about 70.
Risk Disclaimer
Trading involves substantial risk of loss and is not suitable for everyone. This article is educational and is not investment advice. All numbers marked hypothetical are illustrations, not results or forecasts. Tradewink is not a broker and not a registered investment adviser. Signal confidence is not a win probability. Simulated results don't guarantee future results.
Read next
Keep learning with a related guide before putting an idea on your watchlist.
Closing Equity Drawdown: A Reproducible Intraday Example
Closing equity drawdown: This synthetic example shows identical closing values and a 30% drawdown in the finer equity observations.
Trading Risk Management: Sizing, Stops and Portfolio Heat
Risk management trading: Trading risk management defines exposure and response rules; it cannot eliminate losses or create an edge.
Day Trading Risk Management: Position Sizing, Loss Limits, and the 1% Rule
day trading risk management with the 1% rule, position sizing formulas, daily loss limits, and the circuit breakers professional traders rely on.
Algorithmic Trading Risk Management: Position Sizing, Drawdown Limits, and Kill Switches (2026)
Algorithmic trading risk management covers position sizing, drawdown limits, and kill switches. Learn the safeguards every automated system needs to.
MFE and MAE: The Two Numbers That Reveal If Your Trading Strategy Actually Works
MFE and MAE reveal whether your strategy captures enough of each move. Learn how to track these metrics, calibrate stops and targets, and improve quality.
Risk of Ruin in Trading: Formula, Assumptions and Limits
Risk of ruin is the chance of reaching a loss threshold. See the gambler's ruin formula, losing-streak odds, simulated sizing results and its limits.
Trading Expectancy Formula: How to Calculate It, With Examples
Trading expectancy is the average gain or loss per trade. Learn the formula, R-multiple examples, a break-even win rate table, and common mistakes.
How Many Trades Do You Need to Know If a Strategy Works?
How many trades prove a strategy works? See confidence-interval tables for win rate and expectancy, and how many tested variations a backtest can survive.
Try AI signals on your watchlist
Send yourself a signal preview, then add tickers to see ranked entries, exits, and risk notes in Tradewink.
Sources cited in this guide
- Bailey et al., The Probability of Backtest Overfitting (davidhbailey.com)
Tradewink builds explainable market research for self-directed traders. Build a watchlist, inspect signal reasoning and risk context, and paper-track ideas before you decide. Public subscriptions are paper-only; separately approved private beta accounts may submit live broker orders.
How this guide is reviewed
Tradewink reviews educational content against its documented market-data sources, risk controls, and product methodology. See our data sources and evaluation methodology for the evidence and limitations behind the platform, and our editorial standards for how guides are sourced, dated, and corrected.