What is drawdown review?
Drawdown review is a performance review method focused on declines from a recent high point. In practice, it evaluates how much performance fell (depth) and how long it took to move away from that low point (duration), based on a time series such as an equity curve or account value history.
The core idea is that two performance records can look similar in total return, yet differ strongly in the size and timing of their worst falls. Drawdown review aims to describe those worst falls in a structured, comparable way.
How drawdown review works
1) Choose the performance series and definition
Most drawdown review begins with a chosen series (for example, account equity over time). Then you apply a consistent definition of “peak” and “trough.” A common approach uses the maximum value reached up to each point, and measures the distance down to the current or next low point.
Because different tools and communities use slightly different conventions (for example, whether drawdown is calculated on raw equity versus net of fees, or how missing data is handled), consistent definitions matter. If the goal is comparison, the same definition must be used across periods, strategies, or accounts.
2) Compute drawdown depth and track the timeline
After defining the calculation, you identify drawdown events: periods where the series moves away from its most recent high. For each event, drawdown review typically records:
- Depth: how far the series drops from the prior peak (often expressed as a percentage).
- Duration: the time from the start of the drawdown to recovery milestones.
- Recovery context: what happens after the low (for example, whether recovery is quick or prolonged).
Duration and recovery are important because a shallow but long drawdown can feel different from a deep but short one.
3) Review “what was happening” during drawdowns
A review does more than compute numbers. It links drawdown events to context that could plausibly affect performance, such as changes in trading activity, execution quality, or shifts in volatility regime. The goal is not to prove causality, but to create a verifiable checklist of factors you can later confirm with data.
If you are reviewing multiple periods, drawdown review can also support structured comparisons, such as whether the worst drawdowns tend to occur during certain conditions (for example, higher volatility), or after specific process changes.
4) Compare alternatives using the same metrics
A practical “factual comparison” approach is to define several criteria—like maximum drawdown depth, average drawdown depth across events, total time spent in drawdowns, and recovery speed—then compute them consistently for each alternative you want to evaluate.
Over time, this can help you distinguish between improvements in return and improvements in downside behavior. Two systems that both return money can still differ materially in how much drawdown they create.
Key limitations, risks, and what can be verified
Drawdown review measures downside behavior, not future safety
Drawdown review is descriptive of past behavior. It does not guarantee that future drawdowns will be smaller, shorter, or easier to recover from. Markets and conditions can change, and a curve that looks stable in one period can behave differently in another.
Because of this, drawdown review is best treated as evidence about historical downside characteristics, not as a predictor with certainty.
Results depend on data quality and consistent definitions
If the input series is incomplete (missing timestamps, inconsistent equity logging) or if fees, funding, or execution effects are handled differently across periods, drawdown metrics can be misleading. Even small convention differences (how peaks are tracked, how currency conversions are treated, how outliers are handled) can alter drawdown depth and duration.
To verify claims about drawdown behavior, you need reproducible inputs: the same data, the same calculation method, and a clear audit trail for any preprocessing.
Drawdown does not uniquely identify causes
A large drawdown can come from many sources: strategy dynamics, execution slippage, liquidity changes, risk exposure shifts, or random sequences of outcomes. Drawdown review can flag “when the pain happened,” but it cannot, by itself, establish which factor caused it.
A safe interpretation is that drawdown review helps generate hypotheses about contributing factors that must be tested with additional, independently verifiable data (for example, execution logs or risk exposure measurements).
Statistical uncertainty increases with shorter histories
If you review a short backtest window or limited historical data, drawdown metrics can be unstable. Rare events may not appear, and “maximum drawdown” can be dominated by a few observations.
This does not make drawdown review useless, but it does mean that comparisons based on small samples should be interpreted cautiously.
How to use drawdown review in a research process
Drawdown review is most useful when it sits inside a broader, verifiable research workflow:
- Use consistent definitions and document them.
- Recompute metrics under the same conventions for each alternative.
- Cross-check drawdown periods with independent logs or measurable context.
- Treat the output as evidence for downside behavior, not as a certainty about outcomes.
If your goal is to understand performance risk, drawdown review provides a structured lens for downside behavior. Its main value is clarity about “how bad it was” and “how it evolved,” along with an uncertainty-aware way to compare alternatives.