Maximum Drawdown Is the Number Investors Actually Feel
Volatility describes scatter around an average. Drawdown describes the worst stretch of losing money, which is the experience that causes people to abandon a strategy.
The Definition
Maximum drawdown is the largest percentage decline from a peak to a subsequent trough, before a new peak is reached.
A strategy rising to 100, falling to 62, and later recovering had a maximum drawdown of 38 percent. It does not matter how quickly the recovery came, or what the average return was over the whole period. The measure captures the worst experience of holding it.
Why It Complements Volatility
Standard deviation treats each period independently and discards the ordering. Two return series with identical volatility can have completely different drawdowns depending on whether the bad periods were scattered or consecutive.
Consecutive losses are what break people. A series of small declines spread across five years is tolerable. The same total decline compressed into four months is what causes redemptions, stop outs, and abandoned plans.
Volatility describes the statistical properties of a return series. Drawdown describes what it did to whoever was holding it, which determines whether they were still holding it at the recovery.
The Recovery Arithmetic
| Drawdown | Gain required to recover |
|---|---|
| 10 percent | 11 percent |
| 25 percent | 33 percent |
| 50 percent | 100 percent |
| 75 percent | 300 percent |
| 90 percent | 900 percent |
The asymmetry is severe and is the reason drawdown control matters more than it appears to. Avoiding a 50 percent decline is worth considerably more than capturing an extra few percentage points of return in good years, because recovering from it requires doubling.
The Companion Measure
Depth is only half the picture. Time to recovery, sometimes called the underwater period, measures how long the strategy remained below its previous peak.
A 30 percent decline recovered in eight months is a different experience from a 30 percent decline that takes six years to recover. The second tests conviction, institutional patience, and career survival in ways the first does not.
Major equity markets have experienced underwater periods measured in years and, in some historical cases, more than a decade. Any plan built on long run average returns is implicitly assuming the investor remains invested through those stretches.
The Limitations
Maximum drawdown is a single observation, drawn from whatever happened to occur in the sample. It carries no information about whether a worse decline was possible and simply did not happen.
It is also sensitive to the length of the period examined. A longer history has more opportunity to contain a severe decline, so comparing the maximum drawdown of a three year record with a twenty year one is not a fair comparison.
And a strategy that has never experienced a stress event reports an attractive drawdown figure that describes only the absence of a test.
The Derived Ratios
The Calmar ratio divides annualised return by maximum drawdown, producing a return per unit of worst case decline. It is popular in managed futures and hedge fund evaluation.
Like all such ratios it inherits the weaknesses of its inputs. A strategy that has not yet experienced its characteristic loss shows a small denominator and an excellent ratio, which is a description of the sample rather than the strategy.
Why It Matters for Plans
The practical use is honesty about tolerance. An investor who states they can accept a 40 percent decline and who has never experienced one is making a forecast about their own behaviour, and the evidence on such forecasts is poor.
Sizing a position so that its plausible drawdown is genuinely survivable, both financially and psychologically, is more important than optimising expected return, because a plan abandoned at the bottom returns nothing regardless of what the model projected.
The Bottom Line
Maximum drawdown measures the worst peak to trough decline and captures the path dependence that volatility discards. The recovery arithmetic makes deep declines disproportionately damaging, and the underwater period matters as much as the depth. It is a single historical observation rather than a bound, so treat a small drawdown on a short record as evidence of an untested strategy rather than a safe one.