What Is Average recovery time?

Average recovery time is the mean number of periods it takes a trading strategy or portfolio to climb from the trough of a drawdown back to its previous equity curve peak, averaged across all drawdown episodes in the measurement window. While the longest drawdown duration captures the single worst-case underwater spell, average recovery time provides a more representative view of how quickly a strategy typically heals after losses. Together with maximum drawdown depth and longest duration, it forms a three-dimensional picture of drawdown risk: how deep, how long at worst, and how long on average.

The Triple Penance Rule by Bailey and López de Prado (2014) offers a rough guide: recovery from a drawdown tends to take 2–3× as long as the drawdown’s formation. A strategy that declines over 5 trades should expect roughly 10–15 trades to recover. This asymmetry arises because recovering from a −20% loss requires a +25% gain — losses and gains are not symmetric in percentage terms. Strategies with high volatility may recover faster in calendar time but can also form deeper troughs, while low-volatility strategies recover slowly even from shallow dips, leading to surprisingly long average recovery times.

For allocators, average recovery time is a practical due-diligence metric. A strategy that claims a low maximum drawdown but has a long average recovery time may be frustrating to hold in practice: capital sits idle or slowly bleeds, opportunity cost mounts, and investor patience erodes. Comparing average recovery time across strategies with similar return profiles reveals which ones bounce back efficiently versus which ones spend most of their life grinding back to break-even. It is especially useful alongside exposure-adjusted CAGR, since a strategy with low time-in-market may show clean returns but glacially slow recoveries when it is invested.

Pros

  • Provides a more representative view of recovery behaviour than worst-case duration alone

  • Helps identify strategies that grind sideways even with moderate drawdown depth

  • Pairs well with longest drawdown duration for a full duration risk profile

  • Intuitive for allocators: directly answers “how long will I typically wait to get back to even?”

Cons

  • Averaging can mask bimodal recovery patterns (quick recoveries mixed with very slow ones)

  • Sensitive to how drawdown episodes are defined (threshold, frequency of returns)

  • A short backtest may not contain enough episodes for a stable average

  • Does not capture the path of recovery — a V-shaped bounce and a slow grind both count equally

See also