Calmar and MAR Ratio: Return Divided by the Worst Thing That Happened

Sharpe and Sortino both measure the same abstract thing: the wobble in your return series. Neither measures the number that actually ends trading careers.

Nobody quits because their standard deviation was elevated. They quit because they were down 42% from the high, the account no longer felt survivable, and the next drawdown would have taken the mortgage with it. Investors do not redeem on volatility either. They redeem on drawdown.

Calmar and MAR are the two ratios that put drawdown in the denominator, where a trader would have put it in the first place.

The formulas, and the single difference between them

That is genuinely the whole distinction. Same arithmetic, different window. Calmar is a rolling three-year view, conventionally updated monthly. MAR covers the entire track record and therefore only gets harder as the record lengthens, because the denominator can rise but can never fall.

The naming is a piece of industry trivia worth knowing so you can decode the jargon. MAR takes its name from Managed Accounts Reports, the newsletter that popularised the measure among managed-futures allocators. Calmar was introduced by Terry W. Young in 1991 and is a contraction of California Managed Account Reports, his own publication. Two newsletters, two ratios, one idea.

What the denominator actually is

Maximum drawdown is the largest peak-to-trough decline in the equity curve, measured in percent, using the running high-water mark. Not the worst month. Not the worst trade. The deepest hole between one all-time high and the next.

Here is a twelve-month equity curve, starting at 100:

Gold bar is the high-water mark at 113.58. Red bar is the subsequent trough at 109.15. That gap is the maximum drawdown: 3.90%.

The year finished at 111.78, a compound return of 11.78%. So:

A ratio of 3.02 would be world-class. It is also completely meaningless, and understanding why is the most useful thing in this article.

Why one bad month defines you forever

Every other metric in trading averages something. Win rate averages outcomes. Expectancy averages R-multiples. Sharpe averages squared deviations. Averages get more reliable as the sample grows.

Maximum drawdown does not average anything. It is a single observation: the worst thing that happened once. That makes it the most sample-dependent statistic in finance, and it behaves in ways averages do not.

Our 3.02 above came from a single clean year in which nothing went wrong. Extend that trader to a decade and the first real crisis will roughly triple the denominator. That is why Calmar uses a three-year window as a minimum and why MAR figures on short records should be ignored entirely.

Why trend followers built their reputation on this number

Managed futures traders adopted MAR long before the rest of the industry, and for a reason that is worth understanding even if you never trade a futures contract.

A trend-following return distribution is the worst possible input for a Sharpe ratio. Long flat stretches, many small losses, then a handful of enormous winners that arrive when a market finally trends. That shape produces high standard deviation almost entirely from the upside, which Sharpe punishes, as covered in the Sharpe ratio article.

MAR does not care about the shape. It asks two questions a trend follower can answer proudly: what did you compound at, and how deep did the hole get. The metric happens to match the strategy’s actual risk profile, which is the only defensible reason to prefer any metric over any other.

Where the four ratios differ

What the numbers conventionally mean

For a track record of at least several years, the ranges allocators generally work with:

Notice how much lower these thresholds are than the Sharpe and Sortino thresholds. That is not a coincidence. Max drawdown is a much larger number than annualised volatility for almost every real strategy, so the ratio it produces is smaller. Never compare a MAR to a Sharpe and conclude anything.

Why this is the metric prop traders should care about most

Here is where the institutional metric becomes directly, unavoidably relevant to a retail trader.

A funded account does not fail when your Sharpe drops. It fails when you touch the drawdown limit. The rule is written in exactly the units MAR uses: peak-to-trough decline, expressed as a percent, measured against a high-water mark. The evaluation is, quite literally, a MAR ratio test with a hard floor and a profit target.

Which reframes the whole exercise. Passing a challenge is not about maximising return. It is about maximising return divided by the deepest hole you dig on the way, because the hole is the only thing that can disqualify you. That is why the trader who grinds a modest edge with tight loss limits passes more often than the trader with the better strategy and no drawdown discipline, and it is what a well-built drawdown protocol is actually optimising.

It also explains something we found in the data. Across a sample of prop accounts run by the same trader on the same strategy, half passed and half breached. The difference was never the edge. It was the depth of the hole.

Run your own numbers

Also available on its own page: the full trading performance metrics calculator.

Frequently asked questions about Calmar and MAR

What is a good Calmar ratio?

Above 0.5 over a three-year window is respectable and above 1.0 is strong. Anything above 2.0 usually means the window contained no meaningful stress, so check what the market did during those 36 months before drawing conclusions. A ratio earned in 2017 and a ratio earned across 2020 are not the same achievement.

What is the difference between the Calmar ratio and the MAR ratio?

Only the measurement window. Calmar uses the trailing 36 months. MAR uses the entire track record since inception. On a fund exactly three years old the two are identical. On a twenty-year record they can differ enormously, because MAR still carries the worst drawdown from year four while Calmar has long since rolled past it.

Should I use maximum drawdown or average drawdown?

Both, for different purposes. Max drawdown answers “what is the worst this has done”, which is the survival question and the one the ratio is built on. Average drawdown answers “what does a normal bad patch look like”, which is more useful for setting expectations and for knowing when to reduce size. The single-observation weakness of max drawdown is exactly what average drawdown fixes.

Can I compute a MAR ratio on my own trading account?

Yes, and it is one of the more useful things you can do with a journal. Track your equity curve daily, record the running high-water mark, and log the largest percentage decline from it. Divide your annualised return by that figure. Just do not compare the result to a fund’s published MAR unless your record is comparably long.

Why is my MAR ratio getting worse even though I am profitable?

Because the denominator ratchets. Once you record a 20% drawdown it stays in the calculation forever, while the numerator is an annualised return that has to be earned continuously. A MAR ratio drifting down over years of profitable trading is normal and expected. It is not a sign that your edge is decaying, which is why Calmar’s rolling window exists.