A breakout trader closes three trades in a week: two small, controlled losses and one trade that runs far past target for an outsized win. Their Sharpe ratio takes a hit for that — the math treats the big winning swing as "risk" in exactly the same way it treats the losses. Sortino ratio was built to fix that blind spot, and for a certain kind of trader, it tells a much truer story than Sharpe does.
It answers a narrower, more useful question than Sharpe ratio: not "how much did my returns bounce around," but "how much did they bounce around on the downside" — because most traders don't actually want to be penalized for a winner that ran further than expected.
What Is the Sortino Ratio?
The Sortino ratio measures return per unit of downside risk, where downside risk is defined as downside deviation — the volatility of only the returns that fell below a minimum acceptable threshold (usually zero, or the risk-free rate). Unlike standard deviation, which treats every swing away from the average as risk, downside deviation ignores upside swings entirely.
Sortino Ratio = (Average Return − Minimum Acceptable Return) ÷ Downside Deviation
The "minimum acceptable return" is usually set to zero for trading purposes — meaning only returns that lost money get counted in the risk calculation. A trading period where every negative return was small and every positive return was large, even wildly so, produces a high Sortino ratio, because the formula never penalizes the size of the winners.
A Worked Example
Take two traders who both average a 2% monthly return over a year, with identical standard deviation of 4% — meaning they'd show the same Sharpe ratio. But the shape of their returns is different: Trader A's volatility comes mostly from a few outsized winning months, while Trader B's comes from a mix of large winning and large losing months.
| Trader | Avg Monthly Return | Std. Dev. (Sharpe input) | Downside Dev. (Sortino input) | Sortino Ratio |
|---|---|---|---|---|
| A | 2.0% | 4.0% | 1.2% | 1.67 |
| B | 2.0% | 4.0% | 4.0% | 0.50 |
Both traders would show the exact same Sharpe ratio, since Sharpe can't distinguish where the volatility came from. Sortino ratio splits them apart immediately: Trader A's swings are almost entirely upside, so the ratio barely dings them for it. Trader B's swings are genuinely two-sided, and the ratio reflects that real downside exposure. Sharpe would have called these two traders identical. Sortino shows they're not.
Why Standard Deviation Punishes Good Trades
Standard deviation, the input behind Sharpe ratio, treats every deviation from the average as equally undesirable — a month that's way better than average moves the number in the same direction as a month that's way worse than average. That's a reasonable assumption for a passive index fund, where a sudden spike often just means more volatility is coming in both directions. It's a much weaker assumption for a trading strategy, where a big winning trade is usually the entire point of the edge, not a warning sign.
- Breakout and trend-following strategies routinely produce a few outsized winners among many small, controlled losses — exactly the shape Sharpe penalizes and Sortino doesn't
- A strategy with strict stop losses but no cap on how far a winner can run will always look more "volatile" under Sharpe than it deserves to
- Sortino isolates the volatility that actually threatens the account — losing streaks and large drawdowns — from the volatility that grows it
- Comparing Sharpe and Sortino side by side reveals whether a strategy's variance comes from good asymmetry or genuine two-sided risk
What Counts as a Good Sortino Ratio?
Because Sortino ratio excludes upside variance from the denominator, it's almost always higher than Sharpe ratio for the same strategy — so the reference bands run higher too, calculated on an annualized basis:
- Below 0: negative — the strategy's average return didn't clear the minimum acceptable return at all
- 0–1.0: below average — meaningful downside volatility relative to the return produced
- 1.0–2.0: good — solid downside-risk-adjusted performance, roughly the professional-grade band
- 2.0–3.0: very good — losses are small and contained relative to the return generated
- Above 3.0: excellent, but check the sample — as with Sharpe, an extreme reading over a short track record often means too few losing periods to trust yet, not a flawless strategy
TradeEdge calculates downside deviation and Sortino ratio automatically from your logged trades, alongside Sharpe ratio, profit factor, and max drawdown — so you can see whether your edge's volatility is coming from winners running or losses piling up.
Sortino Ratio vs. Sharpe Ratio: When Each One Tells You More
Neither metric replaces the other — they answer related but distinct questions, and reading both together catches things one alone would miss:
- 1Sharpe ratio — return per unit of total volatility, treating upside and downside swings as equally risky; better for comparing strategies with symmetric, normally-distributed returns
- 2Sortino ratio — return per unit of downside volatility only; better for strategies with asymmetric payoffs, like trend-following, breakout, or any setup that lets winners run past a fixed target
- 3A large gap between the two — high Sortino, mediocre Sharpe — is itself informative: it tells you the strategy's variance is mostly good variance, which Sharpe alone would never surface
A strategy that shows a strong Sharpe ratio and an even stronger Sortino ratio is usually one where losses are genuinely well-controlled and the volatility comes from the winners doing their job — the exact profile most trend and breakout traders are actually trying to build.
Where Traders Get Sortino Ratio Wrong
The most common mistake is using it as the only risk metric and ignoring Sharpe and max drawdown entirely — Sortino says nothing about how large a single losing streak could get, only how volatile the losing periods have been on average. A second common error is setting the minimum acceptable return too high (some traders plug in their target return instead of zero), which inflates downside deviation and produces a ratio that understates how the strategy is actually performing relative to a simple breakeven bar.
Why This Needs a Journal, Not a Quick Mental Estimate
Downside deviation requires isolating every losing period from your full return history and calculating variance on that subset alone — not something to eyeball from a P&L chart. A trading journal that logs every closed trade can separate winning and losing periods automatically and compute rolling Sortino ratio as the sample grows, the same way it already tracks Sharpe ratio, profit factor, and max drawdown. Reading all four together is what separates a strategy with good asymmetry from one that just got a favorable calculation.