Two equity curves can suffer the same maximum drawdown and make very different demands on a trader. Drawdown duration measures the missing part: how long each curve stayed below its previous high.
That does not turn duration into a pass-or-fail prop rule. It does stop a backtest from presenting a short, violent setback and a long, stagnant recovery as if they were the same experience.
Drawdown duration, defined
Drawdown duration is the time from an equity peak until the equity makes a new high. It begins when the curve drops below its prior peak and ends only when that peak is recovered.
Maximum drawdown answers, “How far did equity fall?” Drawdown duration answers, “How long was the strategy underwater?” Both are properties of the same equity path, but they describe different risks.
| Measure | The question it answers | What it can hide on its own |
|---|---|---|
| Maximum drawdown | How deep was the worst peak-to-trough fall? | Whether recovery was quick or painfully slow |
| Drawdown duration | How long did equity remain below a prior high? | How much capital was at risk during that wait |
| Daily loss limit | Did the account breach its rule during one day? | The longer recovery path after a compliant loss |
For a prop backtest, use the same equity series that the relevant loss rule observes. A balance-only curve can delay or conceal a floating loss. The distinction matters enough to warrant checking balance versus equity drawdown before treating either measure as settled.
Why the recovery clock belongs in a prop backtest
Drawdown duration is not a prediction of how long a future recovery will take. The historical path is one path, and a backtest is not a live record. It is still useful evidence about the path the strategy actually produced under its stated assumptions.
That evidence matters because prop trading is constrained by more than a final return. An account can remain within a daily and overall loss rule yet spend a long period with no new equity high. During that period, the trader may be tempted to change size, abandon rules, or replace the system because the recovery feels too slow. Those decisions are outside the backtest, but the backtest should not hide the condition that provokes them.
The metric also exposes a common comparison error. A strategy with a lower maximum drawdown is not necessarily easier to hold if it recovers much more slowly. Conversely, a deeper decline that is recovered quickly is not automatically safer; depth still matters to a firm’s loss limits. The point is not to crown a single winner. It is to inspect both dimensions before calling a curve “smooth.”
How to measure it without changing the story
The calculation is simple. The discipline is in defining the series before seeing the result.
- Use an equity series with the same costs, position sizing, and execution assumptions as the reported backtest.
- Mark each new running equity high.
- Start a drawdown episode when equity falls below that high.
- End the episode only when equity exceeds the old high; record an unrecovered episode separately at the test end.
- Report the longest duration alongside maximum drawdown, and inspect the full list rather than only the average.
The last point is where tidy summaries mislead. An average can make one unusually long recovery disappear inside many brief ones. A current, unrecovered drawdown should not be silently treated as recovered either: its duration is at least as long as the period observed, not a completed recovery time.
Frequency matters too. Repeated short drawdowns and one long drawdown can produce a similar average while creating different operational pressure. A trade list, an equity chart, and a dated list of peak-to-recovery episodes make that difference inspectable.
The measurement choices that change the answer
Drawdown duration has no useful meaning until its clock and curve are specified. Calendar time and trading time answer different questions; closing equity and intrabar floating equity can do the same.
For a system that holds positions through news or over a weekend, a calendar-based duration records the time actually spent below the high. For a system evaluated only on completed trading sessions, a session-based duration may be more useful operationally. Neither choice is universally correct. State it, then keep it unchanged across strategy comparisons.
The same caution applies to a prop-firm rule set. The rule is the firm’s, can change, and may define losses differently from a research chart. Check the firm’s current terms rather than importing a generic drawdown label into the test. The funding model is useful context for why a recovery path is not the same thing as a funding promise.
What a credible report should show
A credible report does not need a theatrical dashboard. It needs enough information for a reader to see what was measured and what was not.
- The equity basis: balance, closing equity, or floating equity.
- The time basis: calendar time, bars, or trading sessions.
- The longest completed drawdown duration and any unfinished episode at the sample end.
- Maximum drawdown from the same series.
- Costs and execution assumptions that generated the curve.
At realbacktesting, published cBot figures are cTrader-native backtests on cTrader broker M1 data from 2021-2026 for the FTMO line. The stated methodology uses intrabar M1 execution, a fixed 2-pip spread, charged commission, swap, and a 30% out-of-sample hold-out for the drawdown check. That is a method to reproduce and challenge, not a claim about future recovery time; see the methodology and proof page.
Frequently asked
Is drawdown duration the same as recovery time?
In a backtest, drawdown duration is commonly the peak-to-new-high interval, so it includes the recovery. “Recovery time” is sometimes used more loosely, which is why the report should define the start and end points.
Can a strategy have low maximum drawdown and long drawdown duration?
Yes. A curve can decline only modestly and then remain below its prior peak for an extended period. Maximum drawdown alone will not show that waiting time.
Does drawdown duration tell me what will happen live?
No. It describes the tested historical path under the test’s assumptions. Future paths, costs, execution, and prop-firm terms can differ.
The number that keeps the curve honest
Maximum drawdown tells you how deep the hole was. Drawdown duration tells you how long the strategy lived in it. A prop backtest that reports only one has left half the experience off the chart.