Maximum drawdown in one definition
Maximum Drawdown (MDD) is the largest observed drop from a prior peak to a later trough in a measured value series (often an equity or balance curve). Mechanics are simple: pick a starting peak value, track the subsequent lowest value until the next recovery, then record the maximum peak-to-trough percentage (or absolute) decline seen over the chosen period.
This definition matters because the MDD number is not a property of the market alone. It depends on what is measured (equity vs. balance), which period is used, and how the series is constructed.
How risks show up: operational, market, counterparty, and interpretation
Operational risks (how the measurement meets reality)
In practice, the value series behind MDD can be affected by operational factors. Examples include slippage, delayed execution, incorrect order handling, or system downtime. Even if a backtest-like curve looks coherent, real trading may execute differently, producing a deeper trough than the estimate implied.
A material failure mode is “measurement mismatch”: MDD calculated from gross performance (without realistic costs and execution) can underestimate what actually happens after costs and fills are included. Conversely, overly conservative cost assumptions may exaggerate expected drawdowns.
Market risks (path dependency and changing conditions)
MDD is path-dependent: two strategies with the same average result can have different worst declines. Sudden volatility expansions, gaps in tradable prices, or sustained trend shifts can extend the time to recover from a trough, worsening the observed MDD.
Also, markets are not stationary. Relationships that existed during the period used to compute MDD may weaken later. That does not mean MDD is “wrong,” but it means the future distribution of drawdowns can change.
Counterparty risks (where losses are not purely market-driven)
If your measured value depends on access to venues or services (for example, through a provider or trading service), counterparty-related issues can influence the equity path. Operational disruptions, withdrawal restrictions, or settlement problems are not the same as market moves, but they can still affect the measured series and the ability to manage risk during a downturn.
Because MDD is computed from observed values, non-market disruptions can create troughs that reflect access and reliability problems, not only trading performance.
Interpretation risks (what the number does—and does not—tell you)
MDD is often treated as a standalone “risk rating,” but it has limitations. Key interpretation risks include:
- Time-window sensitivity: changing the lookback period can change the worst trough.
- Benchmark and definition choices: measuring equity versus balance, or using different cost treatments, changes the curve.
- Scale ambiguity: absolute drawdowns and percentage drawdowns communicate different information.
A related limitation is that MDD summarizes the worst point but does not fully describe recovery speed, volatility between peak and trough, or how often drawdowns occur.
Realistic scenario impact (with explicit assumptions)
Assume you compute MDD from a daily equity curve over 12 months, using gross returns (no spread, no commission, no slippage) and that orders fill at requested prices. In a later live period, assume slippage occurs during volatile sessions and execution delays lead to worse fills.
Even if the same “strategy logic” is used, the realized equity curve can drop more than the gross curve predicted. The observable impact is a deeper peak-to-trough decline, i.e., a larger realized MDD. This scenario highlights that MDD is not only about market behavior; it is also about how the curve is produced and whether the assumptions behind it match reality.
Limitations and how to verify independently
At least one limitation: MDD is historical and selection-dependent
A core limitation is that MDD is computed from the available history. It cannot assure future outcomes. If future conditions differ, the future maximum drawdown could be larger or smaller than the historical worst decline.
Another limitation is selection dependency: choosing a different start date, measurement method, or cost model can change the result.
Verification and next questions
To verify claims about MDD for any system, independently check:
- The exact series definition: what values were used for “peak” and “trough. ”
- The period and frequency: daily, intraday, or monthly data can materially change the worst drawdown. - Cost and execution treatment: whether spreads, commissions, and realistic slippage were included.