Calmar Ratio: Formula, Calculator and Drawdown Explained

For your own data, enter one complete month per line: 2025-01;2.5 means January 2025 returned +2.5%. A tab can replace the semicolon; decimal commas are accepted. Use consecutive months and returns already adjusted for deposits, withdrawals and trading costs. Do not replace missing months with an assumed zero. Select Calculate after changing inputs or the window.

Two consecutive −20% months leave 0.8 × 0.8 = 0.64 of the preceding peak: a 36% drawdown. In the other history, a +25% month between the losses restores the previous peak, keeping the deepest fall at 20%.

Reordering the same return factors preserves their product and therefore the final growth. Their order changes the path and its drawdown. This is why two accounts with the same starting and ending values can have different Calmar ratios.

History A repeats +25%, −20%, +25%, −20%, +25%, then seven zero months. That annual block multiplies the index by 1.25. Repeating it for January 2023–December 2025 takes 100 × 1.25³ to 195.3125: a total gain of 95.3125%. These are deliberately constructed teaching returns.

There are 36 returns and 37 index points, including the initial 100. CAGR = (195.3125 / 100)^(12 / 36) − 1 = 25%. With the 20% drawdown, Calmar is 25% / 20% = 1.25. History B has the same CAGR but a 36% drawdown, giving about 0.69. Using total return instead would produce 95.3125% / 20% ≈ 4.77 for A, a different measure.

Do not annualize the arithmetic average of monthly returns. A +25% month followed by −20% averages +2.5%, yet 1.25 × 0.8 = 1: compound growth is zero.

The PerformanceAnalytics convention uses geometric annualized return. Other definitions use annualized excess return above a reference rate, as in Molyboga and L’Ahelec’s study. This calculator makes no risk-free-rate deduction; match that convention before comparing providers.

Maximum drawdown is the largest percentage fall from a previous high to a later low. The order matters: subtracting the table’s smallest value from its largest, regardless of dates, can identify a decline that never occurred. The worst individual month is also not necessarily the worst drawdown.

An index that moves from 100 to 125 and back to 100 has a 20% drawdown: (125 − 100) / 125. The recovery needed to move from 100 to 125 is 25%, because it starts from a different base. The calculator uses percentage drawdown from the running peak, including the initial index value.

The ratio does not measure how long the account stayed below its peak. Two histories can share both CAGR and maximum drawdown while taking very different amounts of time to recover.

A trailing three-year period is traditional, but the label “Calmar” does not guarantee 36 months. Record the actual dates. A calculation based on less than 12 months is flagged here: annualizing six months expresses a rate, not an observed one-year result.

Use a shared calendar period when comparing strategies. Shortening a history can remove an earlier loss episode. Each selected window here restarts the index at 100 before its first return; a peak outside the window is excluded, so this is not an all-time-high drawdown.

Only month-end values are observed. A fall and recovery between two month ends can disappear from this series. An intraday equity report may therefore show a larger drawdown; different sampling does not automatically mean either calculation is wrong.

A balance rising from 10,000 to 12,000 solely because of a 2,000 deposit has earned no investment return. A withdrawal can likewise reduce the balance without a trading loss. Use returns adjusted for those external flows, rather than percentage changes in raw balances.

The timing of each flow matters. Use a suitable periodic-return method or a correctly constructed time-weighted index; subtracting all deposits from the final balance cannot reconstruct the path. The GIPS Handbook explains flow treatment and linked returns. This tool does not certify a report or claim GIPS compliance.

Include applicable commissions, financing and other trading charges before calculation, and avoid deducting costs already included in the series. State how taxes and account fees are treated. The trading-cost guide distinguishes spread and commission; it does not supply the missing cash-flow accounting.

With positive CAGR and positive drawdown, a higher Calmar means more annualized growth relative to the worst observed fall. There is no universal pass mark: this historical relationship cannot establish future safety or the suitability of a strategy.

A very small drawdown can make the ratio unusually large. Check for a short sample, sparse observations, stale or smoothed valuations, and omitted costs before treating the number as persuasive.

Keep the sign of a negative CAGR. If the loss is unchanged, a larger drawdown makes a negative quotient closer to zero. That is why ranking losing histories by “higher Calmar” can reward the wrong feature; inspect return and drawdown separately.

Sharpe and Sortino use different risk measures. MAR ratio often uses the full record. The abbreviation MAR in Sortino calculations instead means minimum acceptable return: a target, not that ratio. Check each provider’s window and formula.

Use the Profit Factor guide to examine trade-level profit and loss totals. The trailing drawdown guide explains moving account limits and their breach conditions.

• Keep the original returns and identify whether they come from live trading or a backtest.
• Save the selected dates, sampling frequency, cost basis and cash-flow method alongside CAGR and maximum drawdown.
• Inspect the worst decline and the time below the peak before ranking strategies by one score.

The equity curve simulator explores hypothetical paths under stated assumptions. Use it for scenario analysis; use this calculator to measure a supplied return history.

Why is my platform’s Calmar different?

Check its dates, initial-value treatment, observation frequency, equity versus closed-trade balance, costs and cash-flow adjustments. Then compare the annualization method and whether a reference rate is deducted. Rounding is only one possible explanation.

Why does the calculator reject −100% or less?

The model requires positive wealth. A −100% month takes the index to zero; a lower return implies negative wealth. The calculator stops rather than deleting that observation, taking its absolute value or restarting the account. Examine such a record separately.

Does the inverse of Calmar show years needed to recover?

No. Dividing a historical drawdown by an annualized return does not describe the subsequent return path. Recovery depends on the future sequence of gains and losses; the ratio contains neither that sequence nor a measured recovery duration.