Sortino Ratio
The Sortino ratio is a risk-adjusted measure equal to the portfolio's excess return over a target divided by the downside deviation, the standard deviation of only those returns falling below the target, so it penalises harmful volatility while ignoring upside swings.
Quick Answer
The Sortino ratio refines Sharpe by dividing excess return only by downside deviation, so a strategy is not punished for large upside moves. It suits asymmetric strategies where big gains are welcome and only losses count as risk. Read it with the target return in mind, since the figure shifts with that threshold.
Definition: Sortino Ratio
Sortino Ratio is a risk-adjusted return measure equal to excess return over a target divided by downside deviation, counting only returns below that target, so upside volatility is not penalised as risk.
Key takeaways: Sortino Ratio
- Sortino divides excess return by downside deviation, ignoring upside volatility
- It suits skewed strategies that Sharpe unfairly penalises for large gains
- It depends on the chosen minimum acceptable return, which must be disclosed
- Its downside sample is smaller, so the estimate is noisier than Sharpe's
- It shares Sharpe's blind spot: a catastrophic tail the data never sampled
Sortino Ratio at a glance
| Family | Risk-adjusted return |
|---|---|
| Formula | (Rp − T) ÷ downside deviation |
| Measures | Excess return per unit of downside volatility only |
| Versus Sharpe | Does not penalise upside volatility |
| Blind spot | Sensitive to the target and to few downside observations |
Sortino Ratio in simple words
The Sortino ratio is the Sharpe ratio's more forgiving cousin: it measures return per unit of bad volatility only, treating big gains as a good thing rather than as risk. It answers how much reward you earned for each unit of downside pain. This suits strategies with lumpy, asymmetric returns, such as trend following, that Sharpe unfairly penalises for their large winning moves.
What Sortino Ratio is for
The Sortino ratio exists because most investors do not fear upside volatility, only losses; by measuring dispersion below a target rather than around the mean, it aligns the risk term with what actually hurts.
Sortino Ratio — professional explanation
Downside deviation, the key ingredient
Downside deviation is computed by taking only the returns that fall below a chosen target, measuring how far each falls short, squaring those shortfalls, averaging over the sample, and taking the square root. Returns above the target contribute zero to the risk term rather than being ignored entirely, so the average is over the full sample count in the common convention. This is what makes Sortino asymmetric: a burst of large gains raises the numerator without inflating the denominator, unlike Sharpe where those same gains increase standard deviation.
The minimum acceptable return and why it matters
The target, often called the minimum acceptable return or MAR, defines what counts as a shortfall. It might be zero (any loss is bad), the risk-free rate, or a required hurdle. The choice materially changes the ratio: a higher MAR classifies more returns as shortfalls and raises downside deviation, lowering Sortino. Because there is no universal MAR, two Sortino ratios are only comparable if they use the same target, and the target must always be disclosed.
When Sortino beats Sharpe, and when it does not
Sortino is more informative than Sharpe when returns are skewed, because it stops punishing the desirable right tail. For a positively skewed trend strategy it can reveal quality that Sharpe obscures. But Sortino shares Sharpe's deeper blind spot: it is still a variance-style measure estimated from the sampled returns, so it cannot see a catastrophic loss the backtest never experienced. A strategy with huge hidden left-tail risk that simply has not fired yet can post a high Sortino just as it can a high Sharpe.
Estimation noise and small downside samples
Downside deviation is estimated from only the subset of returns below the target, so it is based on fewer data points than the full standard deviation. For a strategy with few losing periods, the downside sample can be very small, making the estimate unstable and the Sortino ratio erratic. Paradoxically, the better a strategy looks (the fewer its losses), the noisier its Sortino, which is a genuine statistical hazard when ranking high-win-rate strategies.
Annualisation and consistency with Sharpe
Like Sharpe, a periodic Sortino is annualised by multiplying by the square root of the periods per year, under the same independence assumption. When reporting both Sharpe and Sortino, they must use consistent return frequency, risk-free or target definitions, and annualisation, or the pair becomes internally inconsistent. A common presentation reports Sharpe, Sortino and their difference, since a Sortino much higher than Sharpe signals positive skew and one close to Sharpe signals roughly symmetric returns.
Formula
Sortino = (Rp − T) ÷ DD , where DD = √( mean( min(0, Ri − T)² ) )
Rp = portfolio mean return, T = the target or minimum acceptable return (MAR), Ri = the return in period i, DD = downside deviation. Only returns below T contribute to DD; shortfalls (Ri − T) are squared, averaged over the sample, and square-rooted. Annualise by multiplying by √p, with p = periods per year. The MAR choice changes the ratio and must be stated.
Worked example: Sortino Ratio
Illustrative example (Indian market)
Consider a Nifty trend-following strategy with monthly returns whose mean is 1.2 percent and a target (MAR) of 0 percent. Suppose the months below zero have shortfalls whose squared values, averaged over all 60 months in the sample, give a mean of 0.000225, so downside deviation = √0.000225 = 0.015, or 1.5 percent monthly. Monthly Sortino = (0.012 − 0) ÷ 0.015 = 0.8. Annualising by √12 (about 3.46) gives an annualised Sortino of about 2.77. If this strategy's Sharpe were only 1.6, the much higher Sortino would reveal that its volatility is dominated by favourable upside months, which Sharpe was penalising.
For an NSE trend strategy that catches occasional large directional moves in Bank Nifty, Sortino often looks far better than Sharpe because the big winning months inflate Sharpe's standard deviation but not Sortino's downside deviation; still, the downside sample may contain very few months, so the Sortino estimate can swing noticeably when one more losing month is added.
Sortino vs Sharpe
| Aspect | Sortino ratio | Sharpe ratio |
|---|---|---|
| Denominator | Downside deviation below a target | Total standard deviation |
| Treats upside as risk | No | Yes |
| Reference point | Minimum acceptable return (MAR) | Risk-free rate |
| Best for | Skewed, asymmetric returns | Roughly symmetric returns |
| Estimation stability | Noisier (fewer downside points) | More stable |
Advantages of Sortino Ratio
- Penalises only harmful downside volatility, not favourable upside
- More informative than Sharpe for skewed or asymmetric return profiles
- Aligns the risk term with what investors actually fear, losses
- The gap between Sortino and Sharpe is a quick read on return skew
- Uses a tunable target that can reflect a real hurdle rate
Limitations of Sortino Ratio
- Downside deviation is estimated from fewer points, so it is noisier than Sharpe's denominator
- Its blind spot is the same as Sharpe's: a catastrophic tail the sample never contained
- The ratio depends heavily on the chosen MAR, hurting comparability
- Very few losing periods make the estimate unstable
- Can be gamed by a target choice that minimises apparent downside
- Still says nothing about maximum drawdown or capacity
Why Sortino Ratio matters in practice
- It is the natural risk-adjusted metric for asymmetric strategies Sharpe treats unfairly
- Comparing it against Sharpe exposes the skew of a return stream at a glance
How professionals treat Sortino Ratio
Quants use Sortino alongside Sharpe precisely because the pair is more informative than either alone: a Sortino far above Sharpe flags positive skew, while a Sortino near or below Sharpe warns of negative skew and hidden left-tail risk. They fix and disclose the MAR, check that the downside sample is large enough to be stable, and never treat a high Sortino as evidence a strategy is immune to a tail it has not yet met. In practice Sortino informs but does not replace drawdown and stress analysis.
Common misconceptions about Sortino Ratio
Misconception: A high Sortino means my strategy is safe.
Reality: It only reflects the downside the data has shown, so a hidden crash that has not happened yet stays invisible.
Common mistakes with Sortino Ratio
- Comparing two Sortino ratios computed with different targets
- Trusting a high Sortino built on a downside sample of only a handful of periods
- Assuming Sortino captures tail risk it has never actually observed
- Choosing a MAR that conveniently minimises the apparent downside
- Reporting Sortino without stating the target used
- Reading a high Sortino as proof of safety rather than of favourable skew
Sortino Ratio: frequently asked questions
What is the Sortino ratio?
The Sortino ratio is a strategy's excess return over a target divided by its downside deviation, the volatility of only the returns that fall below that target. It rewards return while penalising only harmful downside movement, unlike Sharpe, which penalises all volatility.
How is Sortino different from Sharpe?
Sharpe uses total standard deviation, counting upside and downside equally, while Sortino uses only downside deviation below a target. Sortino therefore does not punish a strategy for large favourable swings, making it fairer for asymmetric returns.
What is the MAR in the Sortino ratio?
The MAR, or minimum acceptable return, is the target below which a return counts as a shortfall. It might be zero, the risk-free rate or a required hurdle, and because it changes the ratio, it must always be disclosed for the number to be meaningful.
When should I prefer Sortino over Sharpe?
When returns are skewed or asymmetric, such as trend following, because Sharpe unfairly penalizes the large winning moves. For roughly symmetric returns, Sharpe and Sortino tend to agree and either is fine.
Does Sortino capture tail risk?
Not the way people hope. It only measures downside the sample actually contained, so a catastrophic loss the backtest never experienced is invisible to it, exactly as with Sharpe. A high Sortino is not proof of safety.
Why can the Sortino ratio be unstable?
Because downside deviation is estimated from only the subset of returns below the target, which can be a small sample for a high-win-rate strategy. Adding one more losing period can then move the ratio noticeably.
Voice search: how people ask about Sortino Ratio
Natural-language questions people ask about Sortino Ratio.
What is the Sortino ratio in simple terms?
It is like the Sharpe ratio but it only counts the bad, downside volatility, so big gains no longer count against you.
Why use Sortino instead of Sharpe?
Because Sharpe punishes big winning moves as if they were risk, while Sortino only measures the downside that actually hurts.
What is the target in the Sortino ratio?
It is the minimum acceptable return, the line below which a return is treated as a loss, and it must be stated for the number to mean anything.
Sources & references
- Sortino, F. A., & van der Meer, R. (1991). “Downside Risk.” Journal of Portfolio Management, 17(4), 27–31.
- Sortino, F. A., & Price, L. N. (1994). “Performance Measurement in a Downside Risk Framework.” Journal of Investing, 3(3), 59–64.
Published 11 July 2026. Educational content only — not investment advice. Markets and rules change; verify current conventions with SEBI, NSE/BSE and your broker.