A drawdown is the decline from an equity high to a later low before a new high is reached. Maximum drawdown is the largest such decline in the period being measured. It describes how much an account, strategy, or R-based equity curve fell from peak to trough.
Profit tells you where a sequence ended. Drawdown tells you what had to be endured along the way. Two strategies can finish with the same return and profit factor, yet one may lose 8% from its peak while the other loses 35%.
How maximum drawdown works
A drawdown starts after the equity curve reaches a high-water mark. It deepens while equity remains below that peak and ends only when equity reaches a new high. A temporary bounce that stays below the old peak does not finish the drawdown.
- Track cumulative account equity or cumulative R after every closed trade.
- Record the highest equity reached so far.
- Measure each later decline from that high-water mark.
- Reset the peak only when a new equity high occurs.
- Select the largest peak-to-trough decline as maximum drawdown.
Maximum drawdown example
Suppose an account rises from ₹5,00,000 to a peak of ₹6,00,000, then falls to ₹4,80,000 before recovering. The decline is ₹1,20,000. Maximum drawdown for that event is ₹1,20,000 ÷ ₹6,00,000 = 20%.
The starting deposit is not the relevant peak once a higher high has formed. Measuring from ₹5,00,000 would produce the wrong answer. Drawdown always references the previous high-water mark.
Drawdown in money, percentage, and R
- Money drawdown shows the actual currency decline, such as ₹1,20,000.
- Percentage drawdown adjusts for account size and supports comparison.
- R drawdown shows the decline in units of planned trade risk.
- Duration drawdown measures time or trades spent below the prior peak.
Each view answers a different question. Money expresses financial impact. Percentage expresses account severity. R helps compare execution across changing position sizes. Duration shows how long confidence and capital remain underwater. Learn how R-multiples normalize trade results.
Maximum drawdown versus a losing streak
A losing streak counts consecutive losing trades. Drawdown includes the entire path below the peak, including partial recoveries and additional losses. A strategy can remain in drawdown despite several winners if they do not restore the previous high.
For example, −1R, −1R, +0.5R, −1R is not a four-trade losing streak, but it creates a 2.5R drawdown from the starting peak. Both statistics matter because streak length affects behavior while drawdown measures cumulative damage.
Why recovery requires a larger percentage gain
Percentage losses and gains are asymmetric because recovery starts from a smaller base. After a 20% drawdown, equity is 80% of the peak. Returning from 80 to 100 requires a 25% gain—not 20%.
- A 5% drawdown requires about a 5.3% gain to recover.
- A 10% drawdown requires about an 11.1% gain.
- A 20% drawdown requires a 25% gain.
- A 30% drawdown requires about a 42.9% gain.
- A 50% drawdown requires a 100% gain.
What is an acceptable maximum drawdown?
There is no universal acceptable drawdown. It depends on trading capital, risk tolerance, strategy variance, leverage, income needs, investor constraints, and whether the observed result is worse than the tested range.
A 10% drawdown may be intolerable for one trader and ordinary for another strategy. The useful limit is defined before trading and connected to an action: reduce size, pause, investigate, or stop. It should never be invented while the account is falling.
- Personal limit: the decline you can financially and psychologically tolerate.
- Strategy expectation: drawdowns observed across robust tests and regimes.
- Operational limit: the level that triggers reduced size or a review.
- Hard risk limit: the level at which trading stops regardless of opinion.
Historical drawdown is not a worst-case guarantee
Maximum drawdown is backward-looking. The worst future decline can exceed the worst decline in a backtest or journal. Short histories may not contain difficult regimes, long losing clusters, liquidity shocks, or execution failures.
If only 30 trades were tested, the absence of a deep drawdown is weak evidence. The strategy sample-size guide explains why high-variance systems need more observations and broader regime coverage.
Why backtests often understate drawdown
- Entries and exits assume fills that were unavailable in real time.
- Spread, fees, funding, and slippage are omitted or understated.
- Rules are optimized after observing historical losses.
- Delisted instruments and failed setups disappear from the data.
- The test covers only a favorable market regime.
- Several correlated positions are treated as independent trades.
- The trader’s missed entries and discretionary mistakes are excluded.
Use conservative execution assumptions, untouched validation data, and forward testing. Compare backtest and live drawdown as ranges rather than expecting an exact match.
Closed-equity versus intraday drawdown
A journal built from closed trades usually calculates closed-equity drawdown. It may not capture adverse movement while a position is open. Intraday or mark-to-market drawdown can therefore be deeper, especially with wide stops, leverage, or multiple simultaneous positions.
State which method you use. Do not compare a backtest based on closed trades with a broker risk report based on intraday equity and assume the figures describe the same thing.
How position sizing changes drawdown
Strategy outcomes determine the sequence in R; position sizing determines how that sequence affects account equity. Risking 3% per trade creates roughly three times the initial percentage impact of risking 1%, before compounding and correlation.
Large risk accelerates both growth and decline, but deep drawdowns make recovery progressively harder. If normal strategy variance can produce ten losing R, the risk per trade must allow the account and trader to survive that sequence plus a safety margin.
Correlation can create hidden drawdown risk
Four positions are not four independent risks when they respond to the same market factor. Long BTC, ETH, and other crypto assets during one broad selloff can behave like one concentrated position. The same applies to correlated currency pairs, indices, or stocks within one sector.
- Measure total open risk across correlated positions.
- Group trades by underlying thesis or market event.
- Review drawdown by day as well as by individual trade.
- Reduce combined exposure when several stops can be hit together.
How to reduce trading drawdown
1. Lower risk before the limit is breached
Choose size from the tested distribution and personal limit. Reducing risk after panic arrives is reactive; defining size and step-down rules before a drawdown is risk management.
2. Stop mixing unrelated or untested setups
A profitable core strategy can be dragged into a deeper decline by experiments. Separate strategies and use setup tags so you can identify which process creates the damage.
3. Control behavioral losses
Revenge trades, moved stops, and excess trade frequency add losses that were not part of the strategy distribution. Use a written post-loss protocol from the revenge trading guide, and review stop-loss discipline.
4. Respect correlated exposure
Cap risk at the portfolio or idea level, not only per ticket. Several individually reasonable trades can combine into one unacceptable event.
5. Revalidate when live behavior changes
A drawdown beyond the tested range may indicate ordinary variance, changed market conditions, execution decay, or a broken assumption. Pause long enough to diagnose instead of changing several rules mid-sample.
A practical drawdown response plan
- Before trading, define warning, reduction, pause, and hard-stop levels.
- At the warning level, verify data, costs, adherence, and open exposure.
- At the reduction level, lower risk by the prewritten amount.
- At the pause level, stop new trades and review the complete sample.
- Resume only with a specific diagnosis and objective conditions.
- Never increase size merely to recover the account faster.
Questions to ask during a drawdown review
- Is the current decline inside the strategy’s tested range?
- Were all trades valid according to the frozen rules?
- How much loss came from mistakes versus on-plan outcomes?
- Did fees, slippage, or liquidity materially change?
- Are losses concentrated in one setup, session, or market regime?
- Did correlated positions create more total risk than intended?
- Would the conclusion change if the largest winner or loss were removed?
- Is the sample large enough to distinguish variance from deterioration?
Do not judge only the last few trades. Compare the current path with the full distribution, strategy version, and trading expectancy. A negative short window can occur inside a positive long-term process.
Tracking drawdown in a trading journal
Reliable drawdown analysis needs complete, chronological results. Selectively recording dramatic wins or losses breaks the equity path and can materially understate the decline.
Traderizz keeps actual P&L and R-multiple history in order, with day views, calendars, tags, strategy context, and an annual P&L heatmap. This makes it easier to connect a drawdown to specific days, setups, execution mistakes, and clusters rather than seeing only one account-level number.
Users of Delta Exchange India and Shark Exchange can import broker trades to reduce missing-history bias. A drawdown metric is only as trustworthy as the sequence supplied to it.