Same 10 points of excess, a smaller denominator because only losses count
The numerator is identical to the Sharpe ratio: 12% return less a 2% risk-free rate is 10 points of excess. The difference is entirely in the denominator. Sharpe divides by total volatility of 15%; Sortino divides by downside deviation of 10%, which counts only the returns that fell below the threshold.
Sortino = (Rp − Rf) / downside deviation
The gap between the two ratios is a measurement in itself. A Sortino well above the Sharpe says most of the volatility was upside — the strategy moved a lot, but mostly in the direction you wanted. The two converging says the movement was symmetric, and the distinction stops mattering.
Returns above the threshold are set to zero, not excluded
The standard calculation takes every period return, replaces anything above the target with zero, squares the shortfalls, averages across ALL periods, and takes the square root. Averaging across all periods rather than only the bad ones is the part most often got wrong, and it matters: dividing by the count of losing periods instead inflates the deviation and understates the ratio.
The target itself is a choice. Using zero measures deviation below break-even; using the risk-free rate measures it below the safe alternative; using a required return measures it below what you actually need. Each produces a different ratio, so a Sortino figure without its target stated is not comparable to another.
Asymmetric strategies need it; symmetric ones do not
Sortino earns its keep on strategies with genuinely asymmetric return distributions — trend following, long-volatility positions, anything with occasional large gains. On those, Sharpe systematically understates quality by treating the good tail as risk.
For a broadly diversified equity portfolio whose returns are roughly symmetric, the two ratios rank funds almost identically and Sharpe is simpler and more widely reported. Sortino also has a practical weakness: because it uses only the downside observations, it is estimated from fewer data points and is noisier on short histories.
Depth and duration of loss, which drawdown measures instead
Downside deviation summarises how far returns fell below a threshold on average. It says nothing about whether those falls were consecutive. A strategy losing 2% a month for a year and one losing 24% in a single month can produce similar downside deviation and completely different experiences.
Maximum drawdown answers that question directly — the largest peak-to-trough decline, and how long recovery took. Read a Sortino ratio alongside a drawdown figure, because the ratio describes the distribution while the drawdown describes the worst path actually travelled.
Frequently Asked Questions
How is the Sortino ratio different from the Sharpe ratio?
The Sharpe ratio divides excess return by total standard deviation, counting all volatility. The Sortino ratio divides by downside deviation only, so big upward swings do not lower the score. The Sortino ratio is usually the higher of the two for the same portfolio.
What is downside deviation?
Downside deviation is like standard deviation but computed using only returns that fall below a target, such as the risk-free rate or zero. It captures the variability of bad outcomes while ignoring good ones.
What is a good Sortino ratio?
Norms vary by strategy, so treat any threshold as a guide rather than a rule. Many investors view a ratio above 1 as solid and a negative ratio as a portfolio that did not beat a safe asset, but comparison against similar portfolios is what matters.
Why might I prefer the Sortino ratio?
If you only care about avoiding losses and are comfortable with upside volatility, the Sortino ratio fits better than the Sharpe ratio. It rewards strategies that are choppy on the way up but protect against downside moves.
Sources & References
Figures on this page are checked against primary, authoritative sources. Links open in a new tab.
Downside deviation is measured against a fixed target (often the risk-free rate or zero) over the same period as the return.
Only below-target returns count as risk; upside volatility is treated as harmless.
Figures are estimates; investment returns are not guaranteed and vary with market conditions.
Where it stops
Downside deviation needs enough negative observations to be reliable; sparse data makes it noisy.
If the downside deviation is zero the ratio is undefined, so the tool returns 0 in that case.
A single ratio reflects only your inputs and period and does not predict future performance.
Related Calculators
Sharpe RatioReturn above the risk-free rate per unit of total volatility, with the excess return shown separately.
Standard DeviationSample and population standard deviation, plus variance, mean, median, quartiles, z-scores, outliers, and confidence intervals.
Maximum DrawdownThe largest peak-to-trough fall in a value series, and the gain required to recover from it.
Expected ReturnProbability-weighted return across scenarios, with the best and worst cases and a check that probabilities sum to 100%.
Holding Period ReturnTotal return over a holding period, split into the capital gain and the income yield that produced it.
Annualized ReturnThe compound annual rate behind a total return, with the growth multiple alongside it.
CAGRThe compound annual growth rate between two values, or the end value a rate reaches and the years it needs, with the path year by year.
Rule of 72Years to double from a rate, or the rate to double in a set time, by the rule of 72 — with the exact answer and the error beside it, plus the rules of 114 and 144.
Returns are assumptions, not guarantees. Actual results may vary because of market performance, taxes, fees, inflation, and timing. This is an educational projection, not investment advice.
Published the calculator with its formula, worked example, assumptions, limitations and a bespoke guide, and added an automated formula test suite covering it.
Tested that the Sortino ratio exceeds the Sharpe ratio whenever downside deviation is below total volatility, which is the entire reason the measure exists.
Frequently Asked Questions now show on the page: 4 answers, checked in a page-by-page audit.
Add this calculator to your site
Responsive embed — and private: nothing your visitors type leaves their browser.