Drawdown Explained: The Number That Decides If You Quit

The measurements that matter

Drawdown

The percentage decline from a previous equity peak to the current value. It is always measured from a peak, so a new high resets it to zero.

Maximum drawdown

The largest such decline over a period. The most quoted risk statistic and the most misunderstood, because it describes one historical path.

Drawdown duration

The time from the peak to the trough. Measures how long conditions were deteriorating.

Time underwater

The time from the peak until a new peak is reached. Usually far longer than the duration and the figure that determines whether you stay with a strategy.

Ulcer index

A measure combining depth and duration, giving more weight to long, deep declines than to brief sharp ones.

Recovery factor

Total return divided by maximum drawdown. A rough measure of return earned per unit of pain.

Calmar ratio

Annualised return divided by maximum drawdown, commonly used to compare systematic strategies.

Why duration matters more than depth

A sharp 25 percent decline that recovers in three months is unpleasant. A 15 percent decline that grinds sideways for two years is far more corrosive, because every month brings fresh evidence that the strategy might be broken, and there is no way to distinguish a normal flat period from a genuine failure in real time.

Estimating the drawdown you will actually face

1. 1

   Start with the backtested maximum

   Record both depth and time underwater. This is your baseline, not your expectation.

2. 2

   Resample the trade sequence

   Run a Monte Carlo simulation that reshuffles your historical trades thousands of times. The distribution of maximum drawdowns will contain many outcomes worse than the historical one, because history produced only one ordering.

3. 3

   Take the 95th percentile, not the median

   Plan around the bad draws. If the 95th percentile drawdown is 34 percent, that is the figure your position sizing and your emotional preparation should assume.

4. 4

   Add a margin for edge decay

   Live performance is typically worse than backtested performance. A common rule of thumb is to expect drawdowns 1.5 times the historical maximum.

5. 5

   Check whether that figure is tolerable

   If not, reduce position size until it is. Halving risk per trade roughly halves the drawdown, at the cost of roughly halving the return.

6. 6

   Write the number down

   Record the expected worst case in your trading plan. When it occurs, the plan will say that this was anticipated, which is exactly the information you will need.

Responding to a drawdown

There are only three legitimate explanations for a drawdown, and each has a different response. The purpose of a prewritten schedule is to work out which applies without making the decision under stress.

Structural ways to reduce drawdown

• Reduce position size. The most direct lever, and it works proportionally. Half the risk per trade produces roughly half the drawdown.
• Add uncorrelated strategies. Combining families that fail in different conditions reduces aggregate drawdown more reliably than improving any single strategy.
• [Volatility targeting](/library/risk-management/volatility-targeting). Reducing exposure when volatility rises cuts the depth of crisis drawdowns, since they occur in high-volatility regimes.
• Drawdown-based de-risking. Halving size after a defined decline extends survival, at the cost of slower recovery.
• Trend filters. Exiting when a long-term trend breaks avoids the extended portion of sustained declines, at the cost of whipsaws.
• Diversify across markets, not tickers. Ten equity positions are one market. Adding rates, commodities, and currencies genuinely reduces the shared exposure.