Home / Performance Metrics / Average Drawdown
Risk metricIntermediate

Average Drawdown

Average drawdown is the mean depth of the declines below the high-water mark across an equity curve, describing the typical pain a strategy inflicts rather than the single worst loss captured by maximum drawdown.

Quick answer: Average drawdown is the mean depth of the declines below the high-water mark across an equity curve, describing the typical pain a strategy inflicts rather than the single worst loss captured by maximum drawdown.

In simple words

Where maximum drawdown reports the one worst fall, average drawdown describes the everyday experience: on a typical dip below your previous high, how deep does it usually go. A strategy might have a scary 30 percent maximum drawdown but an average drawdown of only 6 percent, telling you that big fall was a rare event and most declines were mild. It captures the routine discomfort of living with a strategy.

Purpose

Average drawdown exists because the single worst drawdown is a fragile outlier; averaging over all drawdown episodes gives a more stable, representative picture of the risk a trader actually experiences most of the time.

Professional explanation

Two common definitions

Average drawdown is defined in two ways that must not be confused. The first averages the depth of each distinct drawdown episode, where an episode runs from a peak to the trough before the next new high; this answers how deep a typical dip goes. The second averages the drawdown value at every point in time, including the many points at zero drawdown when equity is at a new high; this is closer to a time-weighted measure of how far underwater the curve is on average. The two give very different numbers, so the definition in use must always be stated.

Why averaging stabilises the estimate

Maximum drawdown rests on one episode and is therefore statistically fragile and prone to worsening with more data. Average drawdown pools information across all episodes, so it is a more stable and repeatable statistic that is less at the mercy of a single crisis. This makes it useful for comparing the routine risk profile of two strategies, and for detecting a strategy that is uncomfortable to hold day to day even if its worst-case is not extreme.

What it hides that maximum drawdown reveals

The cost of averaging is that it deliberately dilutes the tail. A strategy with a benign average drawdown can still harbour a catastrophic maximum drawdown that occurs rarely but ruins an account when it does, and average drawdown will not warn you about it. This is the mirror image of maximum drawdown's weakness: one hides the tail, the other is defined by it. They are complements, and reporting only one gives a distorted risk picture.

Average drawdown connects to several related constructs. The average of the N largest drawdowns underlies the Sterling ratio, offering a middle ground between the single worst and the full average. The time-weighted average drawdown relates to the pain index, and its root-mean-square cousin is the ulcer index, which squares drawdowns before averaging to punish deep and prolonged declines more heavily. Choosing among these is really choosing how much to weight depth versus frequency versus duration.

Frequency and duration as companions

Average drawdown depth is most informative when read together with how often drawdowns occur and how long they last. A strategy that spends 70 percent of its time underwater with a modest average depth can be more psychologically punishing than one with a deeper but rarer average drawdown that quickly recovers. Reporting average drawdown depth, the average drawdown duration, and the proportion of time underwater together gives a rounded view of the routine experience of a strategy that no single number provides.

Formula

Average drawdown = ( 1 ÷ N ) × Σ DD_i

Under the per-episode definition, DD_i = the depth (as a positive fraction) of drawdown episode i, N = the number of drawdown episodes, and the sum runs over all episodes. Under the per-point definition, DD_i is the drawdown at each time point and N is the number of time points (including zeros at new highs). The two definitions differ substantially, so state which is used.

Average drawdown vs Maximum drawdown

AspectAverage drawdownMaximum drawdown
CapturesTypical dip depthSingle worst fall
Statistical stabilityMore stable, pools episodesFragile, one episode
Sees the tailNo, dilutes itYes, defined by it
Best forRoutine risk comparisonWorst-case survival
Should be readWith frequency and durationWith duration and underwater time

Practical example

Illustrative example (Indian market)

Suppose a Nifty swing strategy had five distinct drawdown episodes over its backtest with depths of 4 percent, 6 percent, 5 percent, 3 percent and 30 percent. The per-episode average drawdown is (4 + 6 + 5 + 3 + 30) ÷ 5 = 48 ÷ 5 = 9.6 percent, while the maximum drawdown is 30 percent. The gap tells the story: four of the five dips were mild single-digit falls, and the 30 percent was a rare outlier. A trader reading only the maximum would overestimate the routine discomfort, while one reading only the average would be blindsided by the tail.

For an NSE options-selling strategy, average drawdown often looks reassuringly small because most months book steady premium and only shallow dips, but this is precisely where relying on the average is dangerous: the occasional volatility spike that produces the maximum drawdown is the event that matters, and the low average drawdown quietly hides it.

Advantages

  • More stable and repeatable than the single-episode maximum drawdown
  • Describes the routine, day-to-day pain of holding a strategy
  • Pools information across all drawdown episodes
  • Useful for comparing the everyday risk of two strategies
  • Less distorted by whether the window happened to contain one crisis

Limitations

  • Its blind spot: it dilutes and hides the catastrophic tail that maximum drawdown reveals
  • Highly definition-dependent (per-episode versus per-point) and non-comparable across definitions
  • Can make a strategy with rare ruinous drawdowns look safe
  • Depth alone omits frequency and duration of drawdowns
  • Sensitive to how a drawdown episode is delimited
  • Not a survival metric on its own

Why it matters in practice

  • It complements maximum drawdown, together describing both routine and worst-case pain
  • It is a more reliable input than the maximum for comparing everyday risk

Common mistakes

  • Reporting average drawdown without stating the per-episode or per-point definition
  • Using a low average drawdown to argue a strategy is safe when its tail is severe
  • Comparing average drawdowns computed under different definitions
  • Omitting the frequency and duration that give average depth its meaning
  • Treating average drawdown as a substitute for, rather than a complement to, maximum drawdown
  • Ignoring how the choice of episode boundaries changes the average

Professional usage

Experienced researchers report average drawdown alongside maximum drawdown precisely because the pair separates routine discomfort from tail catastrophe, and they state clearly which definition they use. They read average depth together with drawdown frequency, average duration and the fraction of time underwater, knowing that a strategy which is rarely at new highs can be harder to hold than a deeper but quicker one. They never let a soothing average drawdown obscure a ruinous but rare maximum, treating the two as inseparable.

Key takeaways

  • Average drawdown is the mean depth of an equity curve's declines
  • It describes typical pain, where maximum drawdown describes the worst
  • It is more stable than the maximum but deliberately dilutes the tail
  • Its per-episode and per-point definitions differ and must be stated
  • Read it with drawdown frequency and duration, and always beside the maximum

Frequently asked questions

What is average drawdown?
Average drawdown is the mean depth of an equity curve's declines below its high-water mark. It describes the typical dip a strategy produces, in contrast to maximum drawdown, which reports only the single worst fall.
How is average drawdown different from maximum drawdown?
Maximum drawdown is the deepest single peak-to-trough loss, while average drawdown pools all drawdown episodes into a representative depth. The maximum captures the tail; the average captures the routine experience and dilutes the tail.
What are the two definitions of average drawdown?
One averages the depth of each distinct drawdown episode, answering how deep a typical dip goes. The other averages the drawdown at every point in time, including zeros at new highs, giving a time-weighted underwater measure. They differ substantially and must be distinguished.
Why is average drawdown more stable than maximum drawdown?
Because it pools information across many drawdown episodes rather than resting on one worst case. This makes it less fragile and less prone to worsening simply because the window happened to include a crisis.
What is the danger of relying on average drawdown?
It dilutes the tail, so a strategy with a benign average can still hide a catastrophic maximum drawdown that ruins an account when it occurs. Average drawdown will not warn you about that rare event.
How does average drawdown relate to the ulcer index?
The ulcer index is a root-mean-square of drawdowns, squaring them before averaging so deep and prolonged declines are punished more heavily. Average drawdown is the simpler linear mean, weighting all drawdowns equally.
Should I report average or maximum drawdown?
Both. They are complements: the average describes routine pain and the maximum describes worst-case survival. Reporting only one gives a distorted risk picture.
What is the Sterling ratio's link to average drawdown?
The Sterling ratio divides return by an average of the largest drawdowns, a middle ground between the single worst and the full average. It is less hostage to one extreme episode than Calmar but still weighted toward the deeper falls.
Does average drawdown capture how long drawdowns last?
Not by itself. Depth and duration are separate dimensions, so average drawdown depth should be read together with average drawdown duration and the fraction of time spent underwater.
How do episode boundaries affect average drawdown?
A drawdown episode conventionally runs from a peak to the trough before the next new high, so how you delimit episodes changes both their count and their depths. Different conventions can move the average, which is another reason to state the method.
Can average drawdown be zero?
Effectively only for a curve that never declines below a prior peak, which in practice indicates too short or too idealised a backtest rather than a genuinely riskless strategy.
Is a lower average drawdown always better?
For routine comfort, generally yes, but a low average can coexist with a ruinous maximum. A strategy is only preferable if both its average and its worst-case drawdown are acceptable for your capital and tolerance.
Why does average drawdown suit high-win-rate strategies poorly on its own?
Because such strategies produce many shallow dips and a rare deep one, so their average drawdown looks reassuring while the rare deep drawdown is the event that actually matters. The average hides exactly the risk that ruins the account.
How does average drawdown help compare two strategies?
It offers a stable, representative measure of routine risk that is less distorted by a single crisis than the maximum. Comparing average drawdowns, alongside their maximums, shows which strategy is calmer day to day and which merely avoided a crisis in-sample.

Voice search & related questions

Natural-language questions people ask about Average Drawdown.

What is average drawdown in simple terms?
It is how deep your typical dip below a previous high tends to be, rather than the single worst fall you ever had.
How is it different from maximum drawdown?
Maximum drawdown is your one worst drop, while average drawdown is the everyday depth of your dips, so it describes routine pain.
Is a low average drawdown always safe?
No, because a low average can still hide a rare, huge drawdown, so you must also check the maximum.
Which drawdown number should I trust?
Use both together, since the average shows routine discomfort and the maximum shows whether you could survive the worst case.
Why does average drawdown matter for comfort?
Because it reflects how the strategy feels to hold most of the time, which is what actually tests your patience.
Does average drawdown show how long dips last?
No, depth and duration are separate, so you should also look at how long the drawdowns last and how much time you spend underwater.

Sources & references

    Last reviewed 11 July 2026. Educational content only — not investment advice. Markets and rules change; verify current conventions with SEBI, NSE/BSE and your broker.

    ← Previous
    Maximum Drawdown
    Educational content only — not investment advice. Examples use illustrative numbers and simplified models. Backtested results are hypothetical and trading derivatives involves substantial risk. See our Risk Disclosure and SEBI Disclaimer.