Blog
Same period, same two strategies. Sharpe rewards B; Sortino and Calmar reward A.
Blog
Two traders close the year at +30%. The first climbed fairly steadily and never gave back more than 8% from a high. The second had doubled the account by summer, then handed almost all of it back in a brutal autumn, before a year end rally saved the number. On paper they tied. In reality one of them is far safer to give more capital to, and the single "return" percentage does not tell you which. That is the job of a risk adjusted return ratio, and the three you meet everywhere are Sharpe, Sortino and Calmar.
They are not three names for the same thing. Each divides your return by a different definition of risk, so each can crown a different winner from the exact same track record. Understanding which risk each one punishes is what lets you read your own numbers without fooling yourself.
In short: Sharpe divides return by total volatility, so it penalises good and bad swings equally. Sortino divides only by downside volatility, so it is fairer to asymmetric styles. Calmar divides annual return by maximum drawdown, so it focuses on the deepest hole. On the same track record they can reward different strategies: the honest move is to read all three together.
The Sharpe ratio is the oldest and most quoted of the three. It takes your average return above a risk free threshold and divides it by the standard deviation of all your returns. The formula is simply (R minus Rf) divided by sigma, where sigma is total volatility.
The catch is hidden in that word "total". Standard deviation treats a +6% day and a −6% day as equally guilty, because both are departures from the mean. So a strategy that occasionally lands an enormously positive week gets penalised: its volatility rises, its Sharpe falls. For a long only fund that wants smooth, boring growth that is fine. For a trader who cuts losses short on purpose and lets profits run, it quietly understates skill. That asymmetry is exactly what the next ratio was built to fix.
The Sortino ratio keeps the same shape but changes the denominator. Instead of total volatility it uses downside deviation: the volatility of only the returns that fell below your target (usually 0%). Positive days, however wild, are simply ignored. The reasoning is intuitive once you hear it, since no trader ever lost sleep over a day that went too well. Because it stops punishing the upside, Sortino reads asymmetric, trend following and breakout styles far more fairly than Sharpe. We work the arithmetic, including the mistake almost everyone makes, in the guide on how to calculate the Sortino ratio.
What Sortino still does not capture is the shape of the pain: ten small losses and one catastrophic one can produce a similar downside deviation, yet they are not the same risk. For that you need the third lens.
The Calmar ratio throws out volatility entirely and asks a blunter question: how much annual return did you earn per the worst beating you took along the way? It divides the compound annual growth rate by the maximum drawdown, the deepest peak to trough fall in the period (classically measured over 36 months). A Calmar of 2 means you made twice your worst drawdown in annual return.
It is the ratio that worries about the one number that truly ends careers and fails challenges: the deepest hole. Two strategies can have an identical Sortino and still have very different Calmars if one took its downside in a single cliff and the other in a steady drizzle. Where Sharpe and Sortino describe the texture of your returns, Calmar describes the worst moment you had to survive.
| Ratio | Divides by | What it punishes | Best for |
|---|---|---|---|
| Sharpe | Total volatility | Every swing, up and down | Diversified portfolios, smooth growth |
| Sortino | Downside deviation | Only losses below the target | Asymmetric styles, trend and breakout |
| Calmar | Maximum drawdown | The deepest hole | Leverage, prop firms, low crash tolerance |
A table makes them look tidy. The interesting part is what happens on real data, where they openly contradict each other. Picture two strategies measured over the same period. Strategy A is a trend follower: jagged and choppy on the way up, with several explosive winning runs, but it cuts losers fast so its worst drawdown stays shallow. Strategy B is a mean reversion system: smooth and steady almost all the time, which flatters its daily volatility, except for one deep crash when a trade went against it for a week.
Same period, same two strategies. Sharpe rewards B; Sortino and Calmar reward A.
Strategy B wins on Sharpe because its smooth days lower total volatility. Strategy A wins on Sortino and Calmar because it controls losses and its worst drawdown stays shallow. Notice what just happened: Strategy B looked safer by the most famous metric in finance, and it was the more dangerous one to size up. The lesson is not that one ratio is right and the others wrong. It is that each answers a different question, so the real skill is knowing which question matters for your situation.
If you run a balanced, diversified portfolio and simply want to know whether your returns justify the bumps, Sharpe is a great single number, and it is the one most outsiders will ask you for. If your style is asymmetric, you let profits run, you trade breakouts or momentum, then Sharpe will lie to you and Sortino is the grade to trust. And if your number one enemy is the deep hole, because you are leveraged, because you trade for a prop firm, or because a 40% drawdown means you stop trading, then Calmar is the ratio that speaks your language, and it reads best next to the shape of your equity curve.
Prop firm traders are the clearest case. A funded challenge does not fail you for volatility; it fails you the moment drawdown crosses a fixed line. A strategy with a glittering Sharpe but a habit of deep crashes is structurally unfit to pass, while a more modest strategy with a high Calmar sails through. That is why "maximise Sharpe" is bad advice for a challenge and "respect the drawdown" is good advice. The honest move is to read all three together: it takes seconds at a glance and each one covers the others’ blind spot.
You can compute them by hand once to understand them, and you should. Redoing it after every session is the kind of mechanical work that quietly stops happening. AlgoTech connects to your MetaTrader 4, MetaTrader 5 or cTrader account in read only mode (encrypted credentials, no orders, no fund movements) and recomputes Sharpe and Sortino on every imported trade, both annualised, alongside max drawdown and a recovery factor that covers the same return versus worst crash ground that Calmar measures. Each account lives in a separate environment with its own history, so every ratio is read on the right sample.
Common questions about the Sharpe, Sortino and Calmar ratios.
Sharpe, Sortino and Calmar are not competing for the title of best ratio: they answer different questions about risk. Sharpe asks whether the returns justify the bumps, Sortino whether the falls are controlled, Calmar how deep the worst one was. Reading all three is what stops you from sizing up the wrong strategy.
This article is for informational purposes only and does not constitute financial advice. The numerical examples are illustrative. Algotech Srl is not a financial intermediary.