The Sortino ratio is the Sharpe ratio’s fairer sibling: it measures return per unit of downside volatility only. A portfolio is not punished for months when it surges — only for months when it falls.
Because upside volatility is excluded, Sortino typically runs 1.2–1.5× the Sharpe ratio for the same portfolio.
Downside deviation = √(mean of squared negative monthly returns). Risk-free rate is taken as 0. Undefined when the history has no negative months.
Two portfolios both average +1% a month with 3% volatility. Portfolio A’s swings are mostly upward spikes; portfolio B’s are sharp sell-offs. Their Sharpe ratios are identical — but A’s Sortino might be 2.1 while B’s is 1.2, correctly flagging B’s pain as the real risk.
Reference bands on monthly data:
| Sortino | Reading |
|---|---|
| < 0 | poor — losing on average |
| 0–1.0 | weak |
| 1.0–2.0 | solid |
| 2.0–3.0 | very good |
| > 3.0 | excellent — or a short, lucky history |
The Sharpe/Sortino pair is more informative than either alone: Sortino far above Sharpe means volatility is mostly upside; Sortino close to Sharpe means drawdowns make up a meaningful share of the swings.
Most useful for concentrated or momentum-heavy portfolios whose upside spikes unfairly inflate plain volatility — Sortino separates the good turbulence from the bad.
With few negative months the downside sample is tiny and the ratio becomes unstable (and undefined with none at all). The same caveats as Sharpe apply: risk-free assumed 0, monthly granularity smooths risk, and short histories overstate quality.