Sortino ratio
Risk-adjusted return that penalises only downside volatility, making it more suitable than Sharpe for skewed equity funds.
The Sortino ratio is a refinement of the Sharpe ratio. It answers the same core question — how much excess return per unit of risk? — but it redefines "risk" to mean only the bad kind: losses below a target return. A fund that swings upward frequently is not penalised; only the downward swings count.
What it measures
Sortino separates return volatility into two halves: upside deviations (gains above target) and downside deviations (losses below target). Standard deviation treats both equally. Sortino discards the upside portion and measures only the downside, producing a number that better reflects an investor's actual experience of discomfort.
How it is computed
Sortino = (R_p − T) / downside_σ
Where:
R_p= annualised portfolio return (3-year monthly, same as Sharpe)T= target return, typically the risk-free rate R_f (91-day T-bill)downside_σ= annualised standard deviation computed using only the months where return fell belowT
Computing downside_σ (monthly):
For each month i:
d_i = min(R_i − T_monthly, 0) // floor at zero for above-target months
downside_variance = Σ(d_i²) / n
downside_σ_monthly = √downside_variance
downside_σ_annual = downside_σ_monthly × √12
Example: Fund monthly returns over 36 months; 12 months fell below T. Monthly downside σ = 2.8% → annualised = 9.7%. Fund return = 15%, R_f = 6.8%. Sortino = (15 − 6.8) / 9.7 = 0.85.
For the same fund, Sharpe (using full σ) might be 0.62 — Sortino is higher because upside months are excluded from the denominator.
How to interpret
Sortino ratios are generally higher than Sharpe ratios for the same fund. A Sortino > 1.0 is considered good for equity funds; > 2.0 is excellent. The ratio is most useful comparatively within a peer group rather than in isolation. A fund with Sortino 1.2 vs. a peer at 0.8 is materially more efficient at managing the risk that matters to investors.
Limitations + caveats
When very few months fall below the target (e.g. during extended bull markets), downside_σ is estimated from a small sample, making Sortino unstable. During a prolonged bear market, Sortino converges toward Sharpe because most months are below target. Like Sharpe, it remains backward-looking and does not predict future drawdowns.
Related metrics
- Sharpe Ratio — the parent metric; Sortino is a more investor-friendly variant.
- Downside Deviation — the denominator of Sortino; measuring it in isolation shows the absolute magnitude of downside risk.
- Max Drawdown — a complementary tail-risk metric that captures the single worst episode.
Sources
Monthly NAV returns: AdvisorKhoj API. Risk-free rate: RBI 91-day T-bill, monthly. Sortino recomputed monthly after NAV refresh; minimum 24 months of history required.