Guide
Sharpe vs Sortino vs Calmar ratio
Sharpe divides return above cash by all of a portfolio's swings, Sortino by its downward swings only, and Calmar divides yearly growth by the single worst fall. Because they count risk differently they can crown different winners: since 20 October 2011, a 60/40 mix of SPY and bonds scored highest on Sharpe at 0.93, while QQQ came out on top on Calmar.
The three ratios in plain words
All three divide a return by a measure of risk, so you can compare a calm portfolio with a racy one on equal terms. They differ in what they count as risk.
| Ratio | Return on top | Risk underneath | What it punishes | What it misses |
|---|---|---|---|---|
| Sharpe | Yearly return minus the cash rate | Volatility: how much returns swing, up and down | Any bumpiness, even sudden gains | Whether the bumps were gains or losses |
| Sortino | Yearly return minus the cash rate | Downside volatility: only the swings below zero | Falls, not gains | How deep the single worst fall went |
| Calmar | CAGR, the average yearly growth rate | The maximum drawdown: the biggest fall from a high to a low | The worst crash in the period | Everything except that one crash |
The cash rate is what money would have earned in safe short-term bills; the ratio only credits return earned above it. A higher number is better on all three. On their own they mean little: they are for comparing portfolios measured the same way, over the same dates.
The follow-along example
We'll score four portfolios that most investors know, with dividends reinvested, from 20 October 2011 to 20 November 2026, the longest period all of them share.
Follow along: the four portfolios
Portfolio What it is SPY The S&P 500 in one fund, our benchmark QQQ The 100 largest non-financial Nasdaq companies, tech-heavy SCHD US dividend stocks SPY 60 / BND 40 A 60/40 mix: three-fifths in SPY, two-fifths in BND (US bonds), reset to target once a year Here is the full scorecard, with Sharpe in the table. The steps below add Sortino and Calmar, and the three don't agree on which portfolio did best.
| Basket | Yearly return | Volatility | Sharpe | Worst drop | $10,000 became |
|---|---|---|---|---|---|
| SPY 60 / BND 40 | 9.4% | 10.2% | 0.93 | 21.0% | $38,677 |
| QQQ | 18.1% | 20.7% | 0.91 | 35.1% | $122,393 |
| SCHD | 13.3% | 15.4% | 0.89 | 33.4% | $65,524 |
| SPY | 14.1% | 16.7% | 0.88 | 33.7% | $73,626 |
Full period: Indices Lab backtest, total return, mixes rebalanced yearly, single funds held throughout, 20 October 2011 to 20 November 2026. Hypothetical results: past performance does not predict future returns. Sharpe uses a 0% risk-free rate.
Step 1: Sharpe, return per unit of bumpiness
Sharpe takes the yearly return above cash and divides it by volatility. It asks: how much return did each unit of swinging around buy you? It is the most quoted ratio and a fair first look, because volatility captures most of what makes a portfolio hard to hold.
The trap is reading Sharpe as "better". A high Sharpe means a smooth ride for the return you got, not more money.
Follow along: the smoothest ride isn't the biggest pot
CAGR Volatility Sharpe SPY 60 / BND 40 9.4% 10.2% 0.93 QQQ 18.1% 20.7% 0.91 SCHD 13.3% 15.4% 0.89 SPY 14.1% 16.7% 0.88 The 60/40 mix had the highest Sharpe and the lowest return. Its bonds cut the swings by more than they cut the return. In money, the gap is large:
- SPY 60 / BND 40 $38,677
- QQQ $122,393
- SCHD $65,524
- SPY $73,626
Full period: Indices Lab backtest, total return, mixes rebalanced yearly, single funds held throughout, 20 October 2011 to 20 November 2026. Hypothetical results: past performance does not predict future returns.
$10,000 in the 60/40 mix grew to $38,677; in QQQ, to $122,393. Sharpe says the 60/40 mix used its risk best. It doesn't say which you'd rather have owned: that depends on how much swinging you can stand.
Step 2: Sortino, counting only the falls
Sortino keeps Sharpe's top line but replaces volatility with downside volatility, which counts only the days that lost money. The idea is sound: nobody minds a sudden jump up. A portfolio that rises in sharp bursts and falls gently scores far better on Sortino than on Sharpe.
Follow along: same order, higher numbers
Sharpe Sortino SPY 60 / BND 40 0.93 1.31 QQQ 0.91 1.29 SCHD 0.89 1.28 SPY 0.88 1.24 Every Sortino is higher than its Sharpe, because only part of the swinging counts, and the ranking is exactly the same. That is the usual result for broad funds: their good and bad days are roughly mirror images, so leaving out the good ones changes the scale, not the order. Our example can't show the case where Sortino matters, which is lopsided returns: option-income funds that earn small steady gains and then drop hard, or trend-following funds that do the reverse.
The lesson: compare Sortino with Sortino, never with someone else's Sharpe. A Sortino looks impressive next to a Sharpe only because it's built to be bigger.
Step 3: Calmar, judged by the worst crash
Calmar divides CAGR by the maximum drawdown. It ignores everyday swings and asks one blunt question: how much growth did you get for the worst fall you had to sit through? That fall is the moment real people sell, so Calmar is closest to what investing feels like.
Follow along: Calmar changes the winner
CAGR Maximum drawdown SPY 60 / BND 40 9.4% 21.0% QQQ 18.1% 35.1% SCHD 13.3% 33.4% SPY 14.1% 33.7% Divide the first column by the second and QQQ comes out on top, ahead of the 60/40 mix that led on Sharpe and Sortino. QQQ's worst fall was only a little deeper than SPY's, while its growth was much faster. SCHD and SPY are level on Sharpe at two decimals, but SCHD falls clearly below SPY on Calmar.
- SPY 60 / BND 40 worst fall 21.0% (19 February 2020 to 23 March 2020), longest time below a previous high 2.1 years
- QQQ worst fall 35.1% (27 December 2021 to 3 November 2022), longest time below a previous high 2.0 years
- SCHD worst fall 33.4% (23 January 2020 to 23 March 2020), longest time below a previous high 2.0 years
- SPY worst fall 33.7% (19 February 2020 to 23 March 2020), longest time below a previous high 1.9 years
Full period: Indices Lab backtest, total return, mixes rebalanced yearly, single funds held throughout, 20 October 2011 to 20 November 2026. Hypothetical results: past performance does not predict future returns.
The chart shows the weakness. Calmar rests on one event, and the events differ: the other three hit bottom in the Covid crash of 2020, QQQ in the 2022 rate shock. A single fall decides each fund's score for the whole period.
Calmar has a second trap: there is no fixed window. Many fund reports use the last 36 months; backtests often use the whole history. Always check which one you're looking at.
Step 4: Change the period and watch the ranks move
Every ratio is a fact about a portfolio over particular dates, not a property of the portfolio. The test that catches this is to split the history in two and score each half.
Follow along: first half of the period
| Basket | Yearly return | Volatility | Sharpe | Worst drop | $10,000 became |
|---|---|---|---|---|---|
| SPY 60 / BND 40 | 9.7% | 7.9% | 1.22 | 11.7% | $20,149 |
| QQQ | 18.4% | 16.3% | 1.12 | 22.8% | $35,745 |
| SCHD | 13.6% | 11.9% | 1.13 | 17.1% | $26,090 |
| SPY | 14.4% | 13.3% | 1.08 | 19.3% | $27,545 |
First half: Indices Lab backtest, total return, mixes rebalanced yearly, single funds held throughout, 20 October 2011 to 7 May 2019. Hypothetical results: past performance does not predict future returns. Sharpe uses a 0% risk-free rate.
Second half of the period
| Basket | Yearly return | Volatility | Sharpe | Worst drop | $10,000 became |
|---|---|---|---|---|---|
| SPY 60 / BND 40 | 9.1% | 12.2% | 0.77 | 21.0% | $19,218 |
| QQQ | 17.8% | 24.4% | 0.79 | 35.1% | $34,327 |
| SCHD | 13.0% | 18.2% | 0.77 | 33.4% | $25,167 |
| SPY | 14.0% | 19.5% | 0.77 | 33.7% | $26,767 |
Second half: Indices Lab backtest, total return, mixes rebalanced yearly, single funds held throughout, 8 May 2019 to 20 November 2026. Hypothetical results: past performance does not predict future returns. Sharpe uses a 0% risk-free rate.
And Sortino for each half:
Sortino, first half Sortino, second half SPY 60 / BND 40 1.76 1.08 QQQ 1.61 1.12 SCHD 1.62 1.10 SPY 1.54 1.08 For Calmar, divide each yearly return by its worst drop. In the first half the 60/40 mix led on all three ratios, with QQQ and SCHD level on Sharpe behind it. In the second half QQQ led on all three, and SCHD fell to last on Sharpe and Calmar. Nothing about the funds' rules changed; the markets did.
Step 5: Know what "good" really means
You'll read that a Sharpe above 1 is good. That rule of thumb comes from comparing trading strategies, and for long-term portfolios it depends mostly on the dates.
Follow along: same funds, two verdicts
Sharpe, last 3 years Sharpe, full period SPY 60 / BND 40 1.16 0.93 QQQ 0.83 0.91 SCHD 1.24 0.89 SPY 1.05 0.88 Over the last 3 years all four cleared 1. Over the full period, none did. The funds didn't get worse; the full period includes 2020 and 2022.
Three more things move the number without changing the portfolio:
- The cash rate. The figures on this page subtract nothing for cash, as the table notes say. A tool that subtracts the going rate on short-term bills shows lower Sharpe and Sortino for the same fund, so check the rate before comparing figures from two sources.
- Daily or monthly returns. The same history measured monthly usually gives a different Sharpe than measured daily.
- Small gaps. Over the full period the four Sharpes run from 0.88 to 0.93. Gaps that size can reverse with a few weeks' more data. Treat portfolios that close as tied.
How to use the three together
Reading the three ratios together
- Same dates and data for every option
- Worst fall first, in money
- Sharpe for the overall ride
- Sortino when gains and losses are lopsided
- Calmar for what the worst crash cost
- Scored over more than one period
- Small gaps treated as ties
Start with the worst fall in money: could you have sat through it? Then use Sharpe for the overall ride, Sortino when a portfolio's gains and losses are lopsided, and Calmar to see what the worst crash cost. Score every option over the same dates and data, check more than one period, and pick on the plan you can live with, not on the second decimal.
The step-by-step backtest guide shows how to set up a test like this one, and our methodology gives the exact definitions we use. To see two of these funds side by side, compare QQQ vs VOO or SCHD vs VOO. The portfolio optimizer can search for weights that maximise one of these ratios, which makes Step 4 essential.
Questions
Which is better, the Sharpe ratio or the Sortino ratio?
Neither is better in general. Sharpe treats every swing as risk; Sortino counts only the falls. For broad stock and bond funds they usually rank portfolios the same way. Sortino earns its place when a portfolio's gains and losses are lopsided.
What is a good Sharpe ratio?
There is no fixed line. Above 1 is often called good, but for long-term portfolios the figure depends mostly on the period: the same funds can clear 1 in a calm stretch and fall below it over a period that includes a crash. Compare portfolios over the same dates instead.
What is a good Calmar ratio?
Like Sharpe, it depends on the window. Check whether it covers the last 36 months or the whole history, and remember it rests on one fall. Use it to compare portfolios over the same period, not as a pass mark.
Why is the Sharpe ratio on one site different from another?
Usually the period, the cash rate subtracted, or whether daily or monthly returns were used. Change any of them and the number changes, even for the same fund.
Can a portfolio have a high Sharpe ratio and still lose a lot?
Yes. Sharpe averages the ride over the whole period, so one deep fall can hide behind many calm years. Read the maximum drawdown, or the Calmar ratio, alongside it.
About this page. Figures on this page come from Indices Lab’s own data, powered by EODHD, and every number is recalculated after each market close; the text is reviewed by a person before it is published. Indices Lab is a research and educational tool: everything here is hypothetical and informational, not investment advice and not a recommendation to buy or sell any security. Past performance does not predict future results. See our methodology.