Glossary

Calmar Ratio — Definition, Formula, and Drawdown Sensitivity

The Calmar ratio divides annualised return by maximum drawdown. Definition, formula, worked example, and why a single bad day can dominate the figure.

By NakedPnL Research·May 7, 2026·4 min read

TL;DR

Calmar is annualised return divided by the absolute value of maximum drawdown over a window.
The denominator is dominated by a single peak-to-trough event, so one bad day can halve the ratio.
NakedPnL doesn't display Calmar on profile pages because a single-event-driven number is a poor summary on a public registry.

Definition

The Calmar ratio, popularised by Terry W. Young in 1991 in his Calmar (California Managed Account Reports) newsletter, is the ratio of a portfolio's annualised return to the absolute value of its maximum drawdown over the same measurement window. It is intended to express how many units of compounded annual return a manager generates per unit of worst peak-to-trough loss.

Formula

Calmar = AnnualisedReturn / |MaxDrawdown|

where:
  AnnualisedReturn = compounded annual rate over the window
  MaxDrawdown      = largest peak-to-trough decline in NAV over the same window

Convention: Young's original definition uses a trailing 36-month window.

Some vendors use a 12-month window or the entire history; the choice changes the result materially.

Worked example

A trader posts an annualised return of 24% over the last 36 months. The worst peak-to-trough decline in equity over that window was 12%, occurring during a single 9-day stretch. Calmar = 24 / 12 = 2.0. Now extend the window by one extra month that contains a 25% drawdown; the numerator changes only slightly, but the denominator more than doubles, dropping Calmar to roughly 0.96. The summary number more than halved despite identical underlying performance for 35 of the 37 months.

Why NakedPnL doesn't display this on profile pages

Calmar is, by construction, dominated by a single event in the price path. That makes it an unstable summary on any public profile: extending the lookback window by one observation can flip the figure from very strong to mediocre, and a trader who is aware of the visible window length can manage position size around it. On a public registry of verified performance, displaying a metric whose value is so concentrated in a single observation invites the wrong incentives. NakedPnL therefore restricts profile-level metrics to TWR, total PnL, and trade count, all of which integrate over the entire history rather than collapsing onto one peak-to-trough event. Allocators who specifically want to study tail behaviour can compute Calmar themselves from the daily NAV series exposed via the API, choosing a window that is appropriate for their underwriting question.

Maximum drawdown
MAR ratio — historical predecessor of Calmar
Sharpe ratio
Sortino ratio

Frequently asked questions

How is Calmar different from the MAR ratio?

The MAR ratio (Managed Account Reports ratio) is computed over the entire historical track record. Calmar is conventionally computed over a trailing 36-month window. They are otherwise identical in form: annualised return divided by maximum drawdown.

What is a 'good' Calmar?

Industry shorthand puts 1.0 at acceptable, 3.0 at strong, and above 5.0 at exceptional. These rules of thumb depend heavily on asset class, leverage, and the lookback window, so they should not be ported between strategies.

Does Calmar work for short track records?

It is unreliable for windows shorter than a full market cycle because the maximum drawdown denominator may not yet reflect the strategy's tail behaviour. A high Calmar over six months says very little about how the strategy will behave in the next adverse regime.

References

Calmar ratio — Wikipedia