Guide
Sharpe ratio vs. Sortino ratio: what's the difference?
Both numbers try to answer "how much return did I get for the risk I took" — but they disagree on what counts as risk. That disagreement is the whole difference, and it matters more than it looks.
Quick answer
Sharpe ratio divides return by total volatility — every up-and-down swing counts against it equally. Sortino ratio divides return only by downside volatility, so big upside moves don't hurt the score at all. A strategy with occasional large gains and otherwise calm losses will show a much higher Sortino than Sharpe — and that gap itself tells you something useful about the shape of its returns.
What each ratio actually punishes
Sharpe ratio uses standard deviation of all returns as its risk measure — a strategy that swings up 5% one week and down 5% the next gets penalized for both moves, even though only one of them lost money. This treats volatility as symmetric: good and bad surprises count the same.
Sortino ratio only counts returns that fall below a minimum acceptable threshold (usually zero, or sometimes the risk-free rate) when calculating its risk measure. A big upside week doesn't touch the denominator at all. The idea is simple: investors don't actually mind upside volatility, so it shouldn't be counted as risk.
A worked comparison
Take two strategies with the identical average return and identical total volatility — so their Sharpe ratios come out the same. Now look at how that volatility is distributed:
| Strategy | Return pattern | Sharpe | Sortino |
|---|---|---|---|
| A | Steady small gains, steady small losses | 1.2 | 1.3 |
| B | Frequent small losses, rare large gains | 1.2 | 2.1 |
Both strategies show the same Sharpe. But Strategy B's Sortino is nearly double — because most of its volatility comes from the upside, which Sortino doesn't penalize. Looking at Sharpe alone, these two strategies look identical. They are not.
Why the gap between them is informative
The size of the difference between a strategy's Sharpe and Sortino tells you how asymmetric its returns are, without needing to look at the raw return distribution at all:
- Sortino close to Sharpe: returns are roughly symmetric — upside and downside swings are similar in size and frequency.
- Sortino noticeably higher than Sharpe: the strategy's volatility is upside-heavy — occasional large wins alongside calmer, smaller losses.
- Sortino close to or below Sharpe: uncommon, but can flag a strategy whose losses are unusually large relative to its overall volatility — worth a closer look at the tail risk.
Which one should you actually use?
Not an either/or choice — look at both, together. Sharpe tells you how smooth the ride was overall. Sortino tells you specifically how painful the downside was. A strategy you're about to trade with real capital is worth judging on both, since a strategy that only "looks good" on one of them is showing you an incomplete picture of its actual behavior.
Or see both, calculated honestly, automatically
The Honest Backtest Engine reports Sharpe and Sortino side by side, out-of-sample and net of real costs — so you can see not just how good a strategy looks, but what shape its risk actually takes.
See how it worksFrequently asked questions
What is the main difference between Sharpe ratio and Sortino ratio?
Sharpe ratio divides return by total volatility, treating upside and downside swings as equally bad. Sortino ratio divides return only by downside volatility, so a strategy with big upside spikes but calm downside won't be penalized for the upside at all.
Is Sortino ratio always higher than Sharpe ratio?
Usually, since downside deviation is typically smaller than total deviation, which makes the denominator smaller and the ratio larger. The size of the gap tells you how asymmetric the strategy's returns are — a small gap means upside and downside swings are similar in size.
Which ratio should I use to judge a trading strategy?
Look at both. A strategy where Sortino is much higher than Sharpe has lumpy, asymmetric returns — occasional large gains — which is worth understanding before you trust it. Neither number replaces checking the result is out-of-sample and net of costs.
Can a strategy have a bad Sharpe ratio but a good Sortino ratio?
Yes. A strategy with frequent modest upside moves and rare, calm downside periods can look mediocre on Sharpe (which penalizes the upside volatility too) while looking strong on Sortino (which ignores that upside entirely).