Sortino ratio
What is Sortino ratio?
The Sortino ratio measures return above a target or risk-free rate per unit of downside deviation below a chosen minimum acceptable return.
Why the Sortino ratio matters
Unlike the Sharpe ratio, which treats upside and downside volatility equally, the Sortino ratio penalizes only returns below a selected threshold. It can better match investors who welcome upside variability but care about shortfall. The result remains highly dependent on the target, period, frequency, and quality of the return history.
How it is calculated
Subtract the minimum acceptable return from portfolio return and divide by downside deviation. Downside deviation is based on observations below the threshold rather than all return variation. Annualization must use consistent return and downside-frequency conventions. Some implementations use arithmetic average excess return while others use compounded return, so methodology should be disclosed.
Example
Two portfolios earn 8% above a 2% target. Portfolio A has 6% annualized downside deviation and Portfolio B has 3%. Their simplified Sortino ratios are 1.0 and 2.0, respectively. Portfolio B delivered the same target-relative return with less measured downside variation, although tail loss and liquidity still require separate analysis.
How to interpret it
Higher is generally better when comparing compatible strategies and conventions. Confirm the minimum acceptable return, which might be zero, cash, inflation, or a required return. Investigate whether a high ratio reflects genuine downside control, a short calm sample, infrequent valuation, or an option strategy that has not yet experienced its rare loss.
Limitations
The ratio can be unstable when few observations fall below the threshold and can become undefined when downside deviation is zero. It does not show loss magnitude beyond the dispersion measure, maximum drawdown, skew, or liquidity. Rankings can reverse when the target changes. It should not be compared across different frequencies or methodologies without normalization.
Practical checklist
Set the threshold before analyzing performance, use enough observations to cover varied markets, and disclose return frequency and annualization. Compare Sortino with Sharpe, maximum drawdown, expected shortfall, and the underlying return distribution. Inspect the actual downside observations and valuation process. Use identical targets and dates for portfolio and benchmark comparisons, and avoid presenting excessive decimal precision.