Sortino Ratio

Definition

The Sortino ratio is a measure of *risk-adjusted return* that counts only the bad kind of volatility. It is a direct refinement of the Sharpe ratio, and the two formulas look almost identical:

Sortino Ratio = (Portfolio Return − Minimum Acceptable Return) / Downside Deviation

There are three pieces:

• Portfolio return — the total return of the fund or portfolio over a period, including price change and reinvested distributions.
• Minimum acceptable return (MAR) — the threshold below which you consider a return a "bad" outcome. Often this is 0, sometimes the risk-free rate, and sometimes a target like your required income yield.
• Downside deviation — how much returns bounced around, but measured using *only the periods that fell below the MAR*. Returns above the threshold are treated as zero risk.

The one crucial difference from the Sharpe ratio is the denominator. The Sharpe ratio divides by standard deviation, which measures *all* volatility — a big gain counts against a fund just as much as a big loss. The Sortino ratio divides by downside deviation, which ignores upside swings entirely and penalizes a fund only when it drops below your minimum acceptable return. A higher Sortino ratio means more reward for each unit of *harmful* volatility.

Why It Matters

Most investors do not lie awake worrying that their fund went *up* too much. Yet standard deviation — and therefore the Sharpe ratio — treats a surprise 8% monthly gain as exactly as "risky" as a surprise 8% monthly loss. That is mathematically tidy but emotionally and practically backwards. The Sortino ratio fixes this by separating the volatility you fear from the volatility you welcome.

For income and ETF investors this distinction is especially valuable. Many high-yield and dividend-growth strategies are asymmetric: they grind out steady distributions and occasionally jump on a strong month. Under the Sharpe ratio those upside jumps inflate the standard deviation and *lower* the score, making a perfectly pleasant fund look riskier than it feels to own. The Sortino ratio only reacts to the drawdowns that actually threaten your principal and your income stream, which is what a retiree drawing a paycheck from the portfolio genuinely cares about.

It also complements the other risk numbers rather than replacing them. The Sharpe ratio tells you about efficiency against *total* variability; beta tells you how sensitive a fund is to the broad market; and the distribution rate tells you the headline income. The Sortino ratio adds the missing piece: reward per unit of *downside* risk. Reading them together gives a far fuller picture than any single figure.

Formula

The Sortino ratio is written almost exactly like the Sharpe ratio, with one substitution in the denominator:

Sortino = (Rp − MAR) / σdownside

  Rp        = the portfolio or fund return over the period,
              including price change and reinvested distributions
  MAR       = minimum acceptable return — the target you treat as
              "good enough" (often 0, sometimes the risk-free rate,
              sometimes a required income yield)
  σdownside = downside deviation — the spread of only those returns
              that landed BELOW the MAR; every period at or above it
              contributes nothing

The numerator, Rp − MAR, is the *excess* return over your threshold: how much the fund beat, or missed, the bar you set. That mirrors the Sharpe ratio's Rp − Rf exactly. The denominator is where the two part ways. The Sharpe ratio divides by standard deviation, which squares every period's distance from the average — so a surprise gain inflates it just as much as an equal-sized loss. Downside deviation instead squares only the *shortfalls* below the MAR and treats every above-target period as zero:

σdownside = sqrt( (1 / N) × Σ [ min(0, Ri − MAR) ]² )

  N   = total number of periods
  Ri  = the return in period i

A tiny worked calculation. Take six monthly returns — 3%, 2%, −1%, 4%, −3%, 1% — with a MAR of 0%. Only the two negative months fall below the target, so only they contribute to the sum:

• Shortfalls, min(0, Ri − MAR): 0, 0, −1%, 0, −3%, 0
• Squared: 0, 0, 1, 0, 9, 0 — which add up to 10
• Divide by N = 6 (≈ 1.67), then take the square root: ≈ 1.29%

The ordinary standard deviation of those same six months is about 2.38%, because it also counts how far the four *good* months sat from the average. Downside deviation is barely half as large here — and that gap is the entire point: it strips out the upside volatility that was quietly padding the Sharpe ratio's denominator.

Example

Consider two funds income investors frequently weigh against each other: SCHD, a low-cost dividend-growth ETF, and JEPI, an equity premium income (covered-call) fund. Covered-call funds like JEPI and its Nasdaq-100 counterpart QQQI sell call options, which *caps* their upside — they hand away the biggest up-months in exchange for option premium. That capped upside is exactly where Sharpe and Sortino diverge.

Suppose over a given year the return and volatility figures work out like this, with a minimum acceptable return of 4%:

• SCHD returns 10%, with total standard deviation of 14% but downside deviation of only 8% — its volatility is lopsided toward good months.
• JEPI returns 9%, with total standard deviation of 9% and downside deviation of 8% — its swings are small, but because the covered-call overlay caps the upside, nearly all of the volatility that remains is downside volatility: its downside deviation (8%) sits close to its total deviation (9%).
• QQQI returns 12%, with total standard deviation of 16% but downside deviation of only 10% — the Nasdaq-100 underlying is far more volatile, yet most of that movement shows up as good months rather than deep drawdowns.

Run the Sharpe ratios first:

1. SCHD: (10% − 4%) / 14% = 0.43
2. JEPI: (9% − 4%) / 9% = 0.56
3. QQQI: (12% − 4%) / 16% = 0.50

By Sharpe, JEPI wins — its low total volatility flatters it. Now run Sortino:

1. SCHD: (10% − 4%) / 8% = 0.75
2. JEPI: (9% − 4%) / 8% = 0.63
3. QQQI: (12% − 4%) / 10% = 0.80

Line the three funds up side by side:

The verdict flips. Once you stop punishing SCHD for its cheerful up-months and measure only the downside, SCHD delivered more return per unit of *harmful* risk. JEPI's advantage under Sharpe was partly an artifact of a covered-call strategy that never surprises you to the upside — so it has little "good volatility" for the Sharpe ratio to penalize in the first place. QQQI shows the pattern is about the *shape* of the volatility, not the label: even though it sells calls like JEPI, its far more volatile Nasdaq-100 underlying still throws off enough good months to open a wide gap between its Sharpe and Sortino scores. This is the classic pattern: capped-upside income funds with little upside surprise score relatively better on Sharpe than on Sortino, while funds with asymmetric, upside-skewed returns look better on Sortino than on Sharpe.

Common Mistakes

• Comparing Sortino ratios computed with different MARs. The minimum acceptable return is a free parameter. One fact sheet may use zero, another the T-bill rate, another a 6% income target. Change the MAR and the ratio changes — so two Sortino numbers are only comparable if they share the same threshold.
• Comparing across different time periods. Like every risk-adjusted metric, a Sortino ratio measured in a calm year looks far better than one measured through a drawdown. Only compare funds over the *same* window.
• Treating it as a standalone verdict. A high Sortino ratio does not tell you a fund's expense ratio, its tax treatment, how much of the distribution is return of capital, or whether it fits your goals. It is one input, not a buy signal.
• Assuming Sortino always beats Sharpe. It is not "the better ratio," just a different lens. For a symmetric, broadly diversified fund the two often tell nearly the same story; the gap only widens when returns are skewed.
• Trusting a ratio built from too little, or too coarse, data. Downside deviation needs enough below-target periods to mean anything. Estimate it from a handful of annual returns, or from a single calm year, and you will capture almost no drawdowns — starving the denominator and inflating the ratio. Monthly or daily data spanning a full market cycle gives a far more honest downside picture; it is the same sampling caveat that applies to the Sharpe ratio.
• Chasing tiny differences. A Sortino ratio of 0.75 versus 0.72 is noise. Use it to separate clearly efficient funds from clearly inefficient ones, not to split hairs.

FAQ

What is a good Sortino ratio?

As a rough guide, a Sortino ratio below 1.0 is considered sub-par, 1.0 to 2.0 is good, 2.0 to 3.0 is very good, and above 3.0 is excellent. Because it ignores upside volatility, a fund's Sortino ratio is usually *higher* than its Sharpe ratio, so do not compare the two numbers head-to-head against the same threshold. As always, compare a fund against its peers over the same time frame and using the same minimum acceptable return rather than against a fixed cutoff.

Sortino vs Sharpe — which is better?

Neither is universally better; they answer different questions. The Sharpe ratio rewards funds for low *total* volatility, which suits investors who dislike surprises in either direction. The Sortino ratio rewards funds for low *downside* volatility only, which suits income investors who are happy to keep upside surprises. For funds with lopsided, upside-skewed returns the Sortino ratio is often the fairer measure; for symmetric funds the two usually agree.

What is downside deviation?

Downside deviation is the statistic that makes the Sortino ratio different from the Sharpe ratio. It is a standard-deviation-style measure of volatility, but it is computed from *only* the periods whose return fell below your minimum acceptable return; every period at or above the target is set to zero and adds nothing. The result is a single number, in the same units as your returns, that captures how hard a fund tends to move against you. The Formula section above walks through the exact calculation with a worked six-month example.

What MAR should I use?

The minimum acceptable return is a choice, and the best one depends on the question you are asking. Set it to 0% if you simply want to penalize losing money. Use the risk-free rate — a short T-bill yield — if you want a true "excess return" number comparable to the Sharpe ratio. Use your required income yield or spending target if you are a retiree who cares whether the fund clears the bar you actually live on. None of these is "correct" in the abstract, but whichever you choose, apply the *same* MAR to every fund you compare — changing the threshold changes the ratio.

Can the Sortino ratio be negative?

Yes. A negative Sortino ratio means the fund's return fell *below* your minimum acceptable return over the period — the numerator is negative. It does not necessarily mean the fund lost money in absolute terms; it can simply mean the return failed to clear the threshold you set, whether that threshold is zero, the risk-free rate, or an income target.

Why is downside deviation different from standard deviation?

Standard deviation measures how far *all* returns spread around their average, so a big gain and a big loss both increase it. Downside deviation only measures the spread of returns that fall *below* the minimum acceptable return, treating every above-threshold period as zero risk. That is why two funds can share the same standard deviation but have very different downside deviations if one's volatility is concentrated in good months rather than bad ones.