The Calmar ratio is a measure of investment performance that compares the return earned with the worst loss suffered along the way. It divides the annualized rate of return over a period by the maximum drawdown in that period, the largest fall from a peak in value to the following low. The standard period is the most recent 36 months. A higher ratio means more return for each unit of drawdown.
The ratio was introduced in 1991 by Terry W. Young, a fund manager in California, and takes its name from his firm's newsletter, California Managed Accounts Reports. It is used mainly to assess hedge funds, managed futures programmes run by a commodity trading advisor, and systematic trading strategies, where investors care more about the depth of losing streaks than about day-to-day volatility. It belongs to the family of measures of risk-adjusted return.
Formula and example
The calculation is:
- is the compound annualized return over the period, the same quantity as the compound annual growth rate.
- is the maximum drawdown over the same period, expressed as a percentage of the peak value and taken as a positive number.
Suppose a fund grows from 100 to 140.49 over three years. Its annualized return is percent. During those years its value at one point fell from a peak of 125 to a low of 100, a drawdown of percent, and no other fall was larger. The Calmar ratio is .
A second fund with the same 12 percent return but a worst fall of 8 percent has a ratio of 1.5. The two earned the same amount, but an investor in the second had a far less painful experience, and that difference is what the ratio is designed to show. The original version used monthly data and was recalculated each month on a rolling basis.
Interpretation
There is no universal standard for a good value. In the managed futures industry, ratios above 1 over three years are often described as good and ratios above 3 as exceptional, but the figure depends heavily on the period measured. A period that happens to contain a market crash will produce low ratios for nearly every strategy exposed to that market, and a calm period will flatter them.
Because drawdown and return both scale roughly in proportion when leverage is applied, the ratio is fairly insensitive to how much leverage a manager uses, which helps when comparing programmes run at different risk levels. Investors use it to rank candidates, to set expectations of the losses that accompany a given return, and to monitor whether a live strategy is behaving as it did in backtesting.
Comparison with other measures
The Sharpe ratio divides excess return by the standard deviation of returns, treating upward and downward swings alike. The Sortino ratio counts only downward deviations. Both describe the typical variability of returns. The Calmar ratio describes a single extreme event, the worst cumulative loss, which depends on the order in which returns arrive as well as their size. A run of small losses in succession produces a deep drawdown that a volatility measure would not reveal.
Close relatives differ mainly in the period used. The MAR ratio, named after the Managed Account Reports newsletter, applies the same division over the whole life of a fund instead of three years. The Sterling ratio uses an average of annual drawdowns, in some versions with a fixed amount added to the denominator. Because these definitions vary between sources, published figures should be compared only when the method is known to be the same.
Limitations
- One observation. The denominator rests on a single episode. It ignores how often losses occurred and how long recovery took, and two funds with the same maximum drawdown may have very different histories.
- Dependence on length. The longer the record, the greater the chance of having met a severe drawdown, so ratios over different spans are not comparable. A fixed 36-month window addresses this but discards older evidence.
- Abrupt changes. When a large drawdown passes out of the three-year window, the ratio can jump although nothing about the strategy has changed.
- Data frequency. Month-end values understate falls that occur and reverse within a month. Daily data give larger drawdowns and lower ratios.
- Hidden tail risk. A strategy that has not yet suffered its characteristic loss, such as one that sells options, shows a small drawdown and a high ratio until the loss occurs.
- No risk-free rate. Unlike the Sharpe ratio, the usual formula does not deduct the return on cash, so ratios rise with the general level of interest rates.
For these reasons the ratio is normally reported with others. A hedge fund or commodity pool will typically show its annualized return, volatility, worst drawdown and time to recovery side by side, with the cumulative return charted so that the pattern of gains and losses can be seen directly. The approach suits strategies such as trend following, whose returns are uneven and whose investors judge them largely by the losses they must sit through.