Last time, we looked at the Sharpe ratio — and flagged one thing about it that annoys a lot of investors once they notice it: Sharpe punishes all volatility equally, including the good kind.
A portfolio that suddenly jumps 12% in a month gets treated by Sharpe's maths exactly the same way as a portfolio that suddenly drops 12% in a month. Both are just "volatility." But nobody's calling their financial adviser in a panic because their portfolio went up too fast.
The Sortino ratio fixes that. It only counts the downside.
What it actually measures
Sortino asks almost the same question as Sharpe — return earned per unit of risk taken — but changes what "risk" means. Instead of measuring how much your returns bounce around in either direction, it only measures how much they bounce around on the bad side.
That "bad side only" measure is called downside deviation. It's calculated the same way as regular volatility, except it only looks at the returns that fell below some minimum you care about (usually zero, or the risk-free rate) — every good month is simply ignored in the risk calculation.
The formula, in plain English:
(Your return − the risk-free return) ÷ how much your portfolio dropped below what you'd accept
For the technically curious
Written as a formula rather than in plain English, the Sortino ratio is:
Where Rₚ is your portfolio's return, R_f is the risk-free rate, and σ_d is downside deviation — the standard deviation of only the returns that fell below a minimum acceptable return (typically 0% or R_f).
The numerator is identical to Sharpe's. The only thing that changes is the denominator — Sharpe uses total volatility (σₚ, every swing, up and down), Sortino uses downside volatility (σ_d, only the drops).
How to read the number
- Below 1.0 — not being well compensated for the downside risk you're carrying.
- Above 1.0 — generally good.
- 2.0 and above — excellent.
- Negative — you'd have done better in cash.
But here's the important part: Sortino will almost always come out higher than Sharpe, for the same portfolio. That's not a bug, and it doesn't mean Sortino is "more generous" in a misleading way — it's simply measuring a smaller slice of volatility (only the bad half), so the number it's dividing by is naturally smaller, which pushes the ratio up. Don't panic if Sortino reads 3.0 and Sharpe reads 1.5 on the same portfolio — that's completely normal and expected.
Why it matters more than Sharpe, for some investors
If you genuinely don't mind big upside swings — and let's be honest, almost nobody does — Sortino is arguably the more honest measure of the risk you actually care about. It answers a more specific, more useful question: "When things went wrong, how wrong did they go, and was I compensated for that?"
This matters especially for portfolios with a handful of high-conviction, high-growth positions — the kind that can have huge up-months. Sharpe would flag those up-months as "risk." Sortino correctly ignores them and focuses only on what actually hurt.
The one thing to watch out for
Because Sortino only looks at bad months, a portfolio with very few data points — a new portfolio, or one that simply hasn't hit a rough patch yet — can produce a Sortino ratio that looks artificially fantastic. A small sample of "bad" months (or none at all) means very little downside deviation to divide by, which can inflate the ratio in a way that doesn't reflect how the portfolio would actually behave over a full market cycle.
The fix isn't to distrust Sortino — it's to remember that any risk ratio is only as reliable as the amount of history behind it. A Sortino ratio calculated over three months means a lot less than one calculated over three years.
This is part 2 of a series on the risk numbers behind your portfolio. Next up: the Calmar ratio — the number that asks "what if things went about as badly as they've ever gone?"