This article explains what the Sortino ratio is and how to read it. It is educational and general, not personalised investment advice, and not a performance claim about any specific strategy. Where it refers to results, those are historical; past performance does not guarantee future results.
The Sharpe ratio gets all the attention, but it carries a quiet unfairness: it treats a sharp move up as just as much “risk” as a sharp move down. For a strategy that earns its keep from a handful of large winners, that is precisely the wrong way to keep score. The Sortino ratio fixes that one flaw by measuring only the volatility that actually hurts. Here is what it is, why it is the fairer lens for an asymmetric return shape, and what counts as a good one.
What the Sortino ratio is
The Sortino ratio is a risk-adjusted return measure, like the Sharpe ratio, but it divides a strategy’s excess return by its downside deviation alone, the volatility of returns that fall below a target (usually zero or the risk-free rate), instead of by total volatility. Because it ignores upside swings entirely, it does not penalise a strategy for making money in large, uneven jumps. In one line: where the Sharpe ratio penalises all volatility, up and down alike, the Sortino ratio penalises only the downside.
That single change is the whole idea. Both ratios put excess return on top. Both divide by a measure of variability. The Sharpe ratio uses the standard deviation of every return, so a string of big up-months inflates the denominator and drags the score down. The Sortino ratio throws those up-months out of the risk calculation and keeps only the disappointment, the returns below the target, on the bottom. It is the same question asked more carefully: not “how much did this wobble?” but “how much did it wobble in the direction I care about?”
Why the distinction matters
For a smooth, symmetric return stream, the two ratios largely agree, and the distinction is academic. It stops being academic the moment returns are skewed, and momentum strategies are about as skewed as it gets. A momentum book makes most of its money from a small number of large winners while the typical position does very little; the distribution has a long right tail. That is not a defect to be smoothed away. It is the engine of the return.
The Sharpe ratio mishandles exactly this shape. It sees those few outsized winners as volatility and marks the strategy down for them, as if a 30% month were a hazard to be avoided rather than the entire point of being in the trade. A strategy can be doing precisely what a momentum strategy is supposed to do, cutting the losers quickly, letting a handful of winners run very far, and post a merely respectable Sharpe, because the big upside is counted against it. The Sortino ratio removes that penalty. By measuring only downside variability, it judges the strategy on the risk that actually matters to a holder: the risk of losing, not the “risk” of winning unevenly.
This is why the choice of ratio is not a cosmetic preference. For a right-skewed book it changes the verdict. Read through the Sharpe ratio alone, a good momentum strategy can look ordinary; read through the Sortino ratio, the same track record reads closer to what it actually delivered. The reason a long momentum tilt produces that asymmetric shape in the first place is the subject of momentum investing, and the edge living in the unglamorous bulk of trades rather than the highlights is the boring middle.
So is the Sortino ratio just a better Sharpe ratio?
No, and it is worth being honest about that, because the Sortino ratio has limits of its own. It inherits most of the Sharpe ratio’s weaknesses and adds one. It is still period-dependent: a Sortino measured over a calm stretch is the calm stretch’s number, not the strategy’s. It is still blind to the worst-case path, a ratio built from deviations says nothing about the deepest peak-to-trough loss a holder would have had to sit through, which is why drawdown has to be read alongside it.
The weakness it adds is subtler: because downside deviation is estimated from fewer data points than total deviation, only the below-target returns count, it is a noisier statistic, especially on a short history. A strategy that simply hasn’t met many bad months yet will show a flattering Sortino on thin evidence. So the Sortino ratio is not a strictly better number; it is a different and fairer one for a particular, common-in-momentum return shape. Use it as a complement to the Sharpe ratio, not a replacement for it.
The Calmar ratio: a useful sibling, with a catch
A close relative is worth knowing, because it answers the question the Sortino ratio sidesteps. The Calmar ratio divides a strategy’s compound annual return by its maximum drawdown , the deepest peak-to-trough fall over the period. Where the Sharpe and Sortino ratios measure return against wobbliness, Calmar measures return against the single worst loss you would actually have lived through. It is wonderfully intuitive: roughly, how much annual return did this strategy earn per unit of maximum pain?
The catch is structural, and it is the one to watch for. Maximum drawdown is a single worst-case event, so on a short track record that simply hasn’t encountered its bad day yet, the Calmar ratio looks spectacular, not because the strategy is robust, but because the worst loss is still in the future. A young strategy with a shallow worst-drawdown-so-far can post a Calmar that a far better one, tested across a real crisis, could never match. Calmar only means something over a window long enough to contain genuine stress. The same instinct, that a low drawdown is a claim that has to survive a hard tape before you trust it, is exactly why a backtest run only over calm years flatters everything; see why backtests lie.
So, what is a good Sortino ratio?
As a working rule of thumb, and over a long, varied sample: anything below 1 is weak, the downside risk wasn’t adequately rewarded; around 1 is solid; around 2 is strong; and a sustained 3 or more is excellent and worth interrogating for a short sample or a hidden tail. If those bands sound like the Sharpe ratio’s, that is the point of the comparison, with one systematic twist.
For the same strategy, the Sortino ratio is almost always the higher number. The arithmetic guarantees it: downside deviation excludes the up-moves, so it is smaller than total deviation, and a smaller denominator lifts the ratio. The gap between the two is itself informative, a Sortino that towers over the Sharpe is the statistical fingerprint of a right-skewed book, one whose variability lives mostly on the winning side. A strategy where the two are nearly equal has a roughly symmetric return shape. So you cannot compare a Sortino to a Sharpe head-to-head and call the bigger one better; they are answering different questions on different scales.
This is why Shishin reports the Sortino and Calmar ratios next to the Sharpe ratio on its five-year track record, rather than leading with whichever flatters most. A regime-aware momentum system, the four guardians each handling a different market state, produces exactly the right-skewed return shape where the Sharpe ratio alone understates the result, and the Sortino ratio is the fairer reading of it. Publishing all three, and explaining why the skew makes the Sortino ratio the more honest lens, is more useful than quoting one number in isolation. No single ratio is the verdict; why the return, the risk-adjusted figure, and the worst loss only mean something together is the argument of the trinity. The Sortino ratio answers one question well, how efficiently was the return earned, counting only the losses against it. It is a sharper question than the Sharpe ratio asks. It is still not the only one.
Sources & further reading
- Sortino, F. A. & Price, L. N. (1994). “Performance Measurement in a Downside Risk Framework.” Journal of Investing, 3(3), 59 to 64.