Direct answer
Maximum drawdown is a simple statistic: it summarizes the largest percentage decline from a prior peak to a later trough within a specified measurement window. Its limitations come from that simplification. The number can change depending on how you define the window and returns, and it often hides timing details that matter for decision-making under real uncertainty.
Mechanism and definition (what it measures)
Maximum drawdown is usually defined from an equity curve (account value over time). Conceptually:
- Identify the highest equity value (“peak”) before a later low.
- Compute the percentage decline from that peak to the lowest subsequent equity (“trough”).
- Take the most negative such decline across the window.
This works cleanly when you have a consistent, correctly measured equity series and when the measurement uses the same units and accounting rules throughout (for example: whether equity includes fees and spreads, and how missing trading days or reporting gaps are handled). If any of those mechanics change, the resulting maximum drawdown can change even if underlying trading behavior does not.
Evidence or example (why the same strategy can show different drawdowns)
Consider a hypothetical equity curve with two losing phases. In one case, the decline happens slowly with partial recoveries before the final trough; in the other case, it drops quickly and recovers later. Both could produce the same peak-to-trough percentage, so maximum drawdown would match, yet the lived experience differs: rapid drawdowns can stress risk limits sooner, even if the eventual peak-to-trough size is identical.
Now change the calculation inputs. If you measure drawdown from gross returns (before costs) versus net returns (after fees, commissions, and financing), the maximum drawdown can differ because costs reduce equity. Similarly, using different start/end dates can create or remove peaks and troughs. Without stating these assumptions, the number is not fully comparable across studies or providers.
Limitations and risks (failure modes and uncertainty)
The key limitations are:
-
Dependence on measurement choices: maximum drawdown is sensitive to the chosen evaluation window, the starting equity, and how the equity curve is constructed (including reporting frequency and handling of gaps). Two calculations on the same underlying behavior can diverge.
-
Loss path is compressed into one value: the statistic shows the largest decline magnitude but not the sequence of intermediate recoveries or the speed of decline. This can matter because risk exposure often depends on timing.
-
Accounting and cost effects: in practice, real outcomes depend on transaction costs, bid/ask differences, slippage, and other execution frictions. Since maximum drawdown is computed from measured equity, any mismatch between modeled/net assumptions and observed execution can change the result.
-
Non-transferability to the future: historical maximum drawdown does not establish a dependable pattern for future drawdowns. Market regimes, leverage use, behavior under stress, and structural conditions can change, breaking historical relationships.
Verification or next question
To independently verify a claimed maximum drawdown, check at least the following: the exact definition used (peak-to-trough on an equity curve), the measurement window boundaries, whether figures are based on net or gross returns, and how fees/costs and missing data are treated. A useful next question is how many different drawdown scenarios could produce the same maximum drawdown, because that reveals how much “information” the single number is leaving out.