The Sharpe ratio treats all volatility the same - including upside surges. The Sortino ratio asks a narrower question: how much return did you earn per unit of bad volatility?
Formula (intuition)
Sortino is similar to Sharpe, but the denominator uses downside deviation instead of total standard deviation:
Sortino ≈ (average return − target return) / downside deviation
Only returns below your target (often zero or the risk-free rate) count toward the risk measure. Large positive swings do not inflate the denominator.
When Sortino helps
- Strategies with positive skew (occasional big wins) can look unfairly punished by Sharpe.
- Income or option overlays where you accept smooth small gains and rare spikes.
- Comparing two strategies with similar returns but different upside noise.
When to stick with Sharpe
Sharpe is more comparable across studies and literature. If you report Sortino, report Sharpe alongside it so readers can cross-check.
Reading the number
Like Sharpe, there is no universal cutoff. Use Sortino to rank variants of the same strategy on the same data window, not as a standalone "buy" signal.
On Quantly
Tear sheets and performance views include ratio metrics next to drawdown and volatility. Pair Sortino with max drawdown and Sharpe for a fuller risk picture.
Run a demo backtest to see these metrics on your own rules.
