See exactly how much room your account gives you
Explore calculatorsThe Calmar ratio divides a strategy's annualised return by its maximum drawdown over the same period. It answers one question: how much return did you earn for each 1% of your worst peak-to-trough loss? A strategy that made 14% a year with a 12% maximum drawdown has a Calmar ratio of about 1.17.
Most performance measures treat risk as the wobble in returns. The Calmar ratio treats risk as the deepest hole the account fell into, which is the number that breaches account limits. This guide covers the formula, a worked calculation, the 36-month convention, the comparison with Sharpe, Sortino and MAR, and how to compute it from your own trade log.
Quick answer: The Calmar ratio is annualised return divided by maximum drawdown, normally measured over the last 36 months. Terry W. Young introduced it in 1991. A higher value means more return per unit of worst-case loss. There is no official threshold, but a ratio below 1 means the worst drawdown was larger than a typical year's return.
Highlights of this article
- Calmar ratio = annualised return / maximum drawdown, both taken over the same window, usually 36 months
- Terry W. Young published it in 1991, and the name comes from California Managed Accounts Reports
- It measures return against the single worst peak-to-trough loss, not against day-to-day volatility
- The ratio barely changes when you scale position size up or down, so it compares strategies fairly at any risk level
- One bad event can dominate the result, so read it next to Sharpe, Sortino and the raw drawdown figure
What is the Calmar ratio?
The Calmar ratio is a risk-adjusted performance measure. The numerator is the annual rate of return. The denominator is the maximum drawdown, the largest percentage fall from an equity peak to the following low.
According to Investopedia, Terry W. Young, a California-based fund manager, developed and introduced the ratio in 1991. The name is an acronym of California Managed Accounts Reports, his firm's newsletter. Young first published it in the trade journal Futures in an article titled "Calmar Ratio: A Smoother Tool".
Young's definition was specific: the average annual rate of return for the last 36 months, divided by the maximum drawdown for the last 36 months. He recalculated it every month. That rolling monthly update is why he called it smoother than the Sterling ratio it was adapted from.
It suits any trader who works under a hard loss limit, because the denominator is the same kind of number as the limit.
What is the Calmar ratio formula?
Calmar ratio = annualised return / maximum drawdown
Both inputs are percentages over the same period. The drawdown is entered as a positive number, so a strategy with a 20% drawdown uses 20, not -20.
2 details vary between sources:
- Average or compound return. Young used the average annual return. Many platforms now use the compound annual growth rate instead. The two are close for steady strategies and drift apart for volatile ones, so check which one your software reports.
- Risk-free rate. Some versions subtract a risk-free rate from the return first, which the Corporate Finance Institute describes in its formula notes. The original does not.
A worked calculation
The figures below are illustrative. An account starts at $100,000 and stands at $148,000 after 36 months.
- Total growth: $148,000 / $100,000 = 1.48.
- Annualised return: 1.48 raised to the power of 1/3, minus 1. That is 1.14 minus 1, or 14.0% a year.
- Find the worst drawdown: the equity peaked at $126,000 in month 14 and fell to $110,880 in month 19 before recovering. No other fall was deeper.
- Maximum drawdown: ($126,000 - $110,880) / $126,000 = $15,120 / $126,000 = 12.0%.
- Calmar ratio: 14.0 / 12.0 = 1.17.
Over these 3 years the account earned 1.17% of annual return for every 1% of its worst loss. Had the worst fall been 28%, the ratio would be 14.0 / 28.0 = 0.50.

Why is the standard window 36 months?
Young chose 36 months, and that convention stuck. The Corporate Finance Institute notes the timeframe is usually set at 36 months, and Investopedia calls it the standard three-year period.
The window is a compromise. A short one, such as 6 months, often contains no serious losing period, which makes the drawdown tiny and the ratio unrealistically high. A much longer one can be dominated by a loss from years ago.
The Corporate Finance Institute also points out that the ratio cannot be adjusted between time horizons, so 2 strategies should be compared over the same dates.
With fewer than 36 months of history, compute it anyway, label the window, for example "Calmar, 14 months", and treat it as provisional.
What is a good Calmar ratio?
There is no official cut-off. Neither Investopedia nor the Corporate Finance Institute publishes a numeric benchmark. Both say only that a higher ratio means better risk-adjusted performance. The bands below are a practitioner rule of thumb, not a standard.
| 36-month Calmar ratio | How to read it |
|---|---|
| Below 0.5 | The worst loss was more than twice a typical year's return. |
| 0.5 to 1.0 | A year's return roughly matches the worst loss. |
| 1.0 to 3.0 | Return clearly exceeds the worst loss. A solid result if the window includes a hard period. |
| Above 3.0 | Rare over a full 36 months. Check the window and the sample size. |
Context matters more than the band. A ratio of 4 from a 5-month backtest tells you less than 1.2 from 3 years of live trading.
Calmar vs Sharpe vs Sortino vs MAR ratio
All 4 ratios divide return by risk. They disagree on what risk is.
| Ratio | Return used | Risk used | Typical window | Best at showing |
|---|---|---|---|---|
| Calmar | Annualised return | Maximum drawdown | Last 36 months | Return against the worst loss actually suffered |
| MAR | Compound annual return since inception | Maximum drawdown since inception | Whole track record | Lifetime return against lifetime worst loss |
| Sharpe | Return above the risk-free rate | Standard deviation of all returns | Any, often annualised | How smooth the path was overall |
| Sortino | Return above a target or risk-free rate | Downside deviation only | Any, often annualised | Smoothness, counting only harmful volatility |
Calmar vs MAR. The difference is the window. MAR, popularised by the Managed Account Reports newsletter, uses all data since inception. Calmar uses the last 36 months.
Calmar vs Sharpe. Sharpe penalises every swing, up or down. A strategy that is smooth for 35 months and loses 30% in the last one can have a fair Sharpe and a poor Calmar. Investopedia notes that Calmar ignores overall volatility, which makes it a narrower view of risk.
Calmar vs Sortino. Sortino counts only downside moves but still averages them. Calmar looks at their worst cumulative run.
Used together, Sharpe or Sortino describe the ride, and Calmar describes the worst moment of it.
How to calculate the Calmar ratio from a trade log
You need a dated equity curve from a trade log or backtest report. The steps work in any spreadsheet.
- Build the equity series. One row per day or per month, with the closing account value. Use equity that includes open positions if you have it.
- Track the running peak. In the next column, take the highest equity seen so far. In a spreadsheet:
=MAX($B$2:B2). - Compute the drawdown at each row. Equity divided by the running peak, minus 1:
=B2/C2-1. The result is 0 at a new high and negative below it. - Take the maximum drawdown. The most negative value in that column:
=MIN(D:D). Drop the sign. - Annualise the return. (Ending equity / starting equity) raised to the power of (12 / number of months), minus 1.
- Divide. Annualised return over maximum drawdown.
A short illustrative log shows the mechanics.
| Month | Equity | Running peak | Drawdown |
|---|---|---|---|
| 0 | $100,000 | $100,000 | 0.0% |
| 1 | $104,000 | $104,000 | 0.0% |
| 2 | $101,400 | $104,000 | -2.5% |
| 3 | $98,800 | $104,000 | -5.0% |
| 4 | $103,000 | $104,000 | -1.0% |
| 5 | $108,000 | $108,000 | 0.0% |
| 6 | $106,000 | $108,000 | -1.9% |
The maximum drawdown is 5.0%. The account grew 6% in 6 months, which annualises to 1.06 squared minus 1, or 12.4%. The ratio is 12.4 / 5.0 = 2.47.
That 2.47 is flattering, because 6 months holds only 1 losing stretch. As rows are added, the maximum drawdown can only stay the same or grow.
Daily data finds deeper drawdowns than monthly data, because month-end values skip the lows in between. A trading journal that records equity each day makes this a 5-minute job. Any serious backtesting report prints both inputs, and quant trading desks track the ratio on a rolling basis.
2 worked examples with different strategies
Both examples are illustrative and use 36 months of results.
Example 1: a trend-following strategy
A daily-chart trend following system wins about 35% of its trades and relies on a few large winners.
- Annualised return: 24%
- Maximum drawdown: 30%, during a 7-month sideways market
- Calmar ratio: 24 / 30 = 0.80
The price of that return was a loss larger than a full year of gains.
Example 2: a mean-reversion strategy
An intraday mean reversion system wins about 62% of its trades with small targets.
- Annualised return: 12%
- Maximum drawdown: 6%
- Calmar ratio: 12 / 6 = 2.00
Half the return, one fifth of the drawdown.
What happens when you scale them
Position size moves return and drawdown together. Trade half the size and both roughly halve, so the ratio stays about the same. That lets you compare strategies at the same worst loss.
Suppose you will only accept a 6% maximum drawdown.
| Trend following | Mean reversion | |
|---|---|---|
| Original return and drawdown | 24% and 30% | 12% and 6% |
| Size needed for a 6% drawdown | One fifth of original | Unchanged |
| Expected annual return at that size | 24% x 0.2 = 4.8% | 12% |
| Calmar ratio | 0.80 | 2.00 |
At equal drawdown, the strategy with the lower headline return earns 2.5 times as much. The scaling is approximate, since compounding and costs do not scale perfectly, but the ordering holds.
Sizing rules such as the Kelly criterion decide where on that line you sit. The risk reward ratio describes 1 trade. The Calmar ratio describes what hundreds of those trades did to the account.
Why does drawdown deserve its own ratio?
Losses and recoveries are not symmetric. A deeper hole needs a disproportionately larger gain to climb out of.
A 10% drawdown needs an 11.1% gain to recover. A 50% drawdown needs 100%. A ratio with maximum drawdown in the denominator keeps that in view, which is why it belongs in any risk management in trading review.
What are the limits of the Calmar ratio?
- One event dominates. 2 strategies with the same worst loss get the same risk score, even if one came close to that loss 6 times and the other once.
- Window dependence. Move the start date a few months and a large drawdown can enter or leave the window. The ratio jumps although the strategy has not changed.
- It ignores the volatility of returns. It says nothing about how rough the path was outside the worst episode.
- It is backward looking. The largest drawdown in a sample is not a ceiling for future ones.
- Small samples inflate it. A backtest with 40 trades can show a tiny drawdown by luck.
- Duration is invisible. A 12% drawdown that lasted 3 weeks and one that lasted 14 months score the same.
A practical fix is to report 3 numbers together: the ratio, the maximum drawdown itself, and the longest time spent below a previous peak.

Using the Calmar ratio in a prop firm challenge
A prop evaluation is a Calmar problem in miniature: reach a set return before a set drawdown.
At Velotrade, a multi-asset prop trading firm offering simulated, educational evaluations on DXtrade, the maximum drawdown is static. The floor is fixed from the starting balance and does not move. The CLASSIC 2-Step plan has a 10% static maximum drawdown with profit targets of 10% and then 5%. The CLASSIC 1-Step plan has a 7% static maximum drawdown and a 10% target. The profit split is 80%.
2 points follow.
Scale the strategy to the limit, with a margin. If your backtest shows a 12% maximum drawdown and the limit is 10%, the strategy at that size has already failed once on paper. Cutting size so the historical worst case is about half the limit leaves room for a deeper one. Scaled to a 5% drawdown, a ratio of 2.00 implies about 10% a year and a ratio of 0.80 about 4%.
A static floor is not the same as peak-to-trough. The Calmar denominator measures the fall from the highest equity point. A static maximum drawdown measures the fall from the starting balance. A 6% dip after the account is up 8% leaves you above where you began. The same dip in week 1 uses most of the room, so early losses are the dangerous ones.
The prop trading drawdown calculator converts the percentages into dollar floors for any account size and plan.
Sources
- Investopedia, "The Calmar Ratio Explained: Assessing Hedge Fund Performance": Terry W. Young, 1991, the name origin, the standard three-year period and the note that the ratio ignores overall volatility.
- Corporate Finance Institute, "Calmar Ratio": the 36-month timeframe, how to read high and low values, and the limit on comparing different time horizons.
- Wikipedia, "Calmar ratio": Young's original definition, the 1991 Futures article "Calmar Ratio: A Smoother Tool", and how the MAR ratio differs.
Frequently Asked Questions
About the author

Gianluca Pizzituti
Chief Executive Officer
Formerly on the derivatives desk at Dresdner Kleinwort in London, then founded and ran a proprietary HFT firm in FX and equity indices out of Singapore.
Velotrade was founded in Hong Kong in 2016 by a team from JP Morgan, Dresdner Kleinwort and Bank of America, has facilitated over $2.5 billion in trade-finance payments, and has been featured in Bloomberg, the Financial Times, the Wall Street Journal and on Nasdaq. Its prop trading evaluations are simulated and educational.
View author page




