Sortino Ratio and Calmar Ratio: Formula, Meaning and Worked Examples
Sortino ratio: return above a target divided by downside deviation. Learn it and the Calmar ratio with F&O worked examples, good values and fund use.
The Sortino ratio tells you how much return you earned for each unit of bad risk. The formula is: (average return minus a target return) divided by downside deviation. Downside deviation measures only the periods that fell below your target, so big winning days do not pull the score down. A higher Sortino ratio is better.
The Calmar ratio asks a different question: was the return worth the worst fall? It divides the yearly return by the maximum drawdown (the biggest drop from a high point to a later low). Its creator used 36 months of data. Both are measures of risk adjusted return, which means return judged against the risk taken to get it.
Most pages on these ratios are written for mutual fund investors. This guide also works them out on a trader's own account. It uses a made-up 10-day F&O P&L list and shows every step. It also explains why the answer can mislead when you have only a few weeks of trades.
Key Takeaways
- 1.Sortino ratio = (average return minus target) divided by downside deviation. Only losing or below-target periods count as risk.
- 2.Calmar ratio = yearly return divided by maximum drawdown. The original version uses the last 36 months. The MAR ratio uses all data since the start.
- 3.When you work out downside deviation, divide by all periods, including the good ones. Dividing only by the losing periods gives a different, wrong number.
- 4.Fund guides from Nippon India and Aditya Birla Sun Life call a Sortino ratio of 1 to 2 good. Bajaj Finserv's guide says 2 and above. These are rules of thumb for yearly fund data, not targets for a 10-day trading record.
- 5.With only a few weeks of data, a yearly Sortino or Calmar figure can look huge. Read it next to drawdown, profit factor and expectancy, and wait for months of data.
What risk adjusted return means
Profit alone hides how you got it. Two traders can end the month with the same profit. One had calm days. The other had a few terrible days that nearly wiped out the account. The first result is easier to repeat and easier to live with.
A risk adjusted return ratio puts a number on this. It divides some measure of return by some measure of risk. The three you will see most are:
- Sharpe ratio: return above a risk-free rate, divided by the standard deviation of all returns (how much returns swing up and down around their average).
- Sortino ratio: return above a target, divided by downside deviation (swings below the target only).
- Calmar ratio: yearly return divided by the maximum drawdown in the same period.
Each one picks a different idea of "risk". That is why they can rank the same two traders or funds in different orders. The rest of this guide shows when that happens.
Sortino ratio: formula and meaning
The formula in plain words
The usual definition is S = (R minus T) / DR. Here R is the average return for the period, T is the target return and DR is the downside deviation. In early work the target was called the MAR (minimum acceptable return). Downside deviation is also called target semi-deviation.
The ratio is named after Frank Sortino. A Red Rock Capital paper on the ratio says its first mention was in Financial Executive Magazine in August 1980. It says the first calculation appeared in the Journal of Risk Management in September 1981. Sortino and Price later described it in the Journal of Investing in 1994.
How do you choose the target? A trader checking an own account usually uses zero: any losing day counts as bad. A fund reader often uses a risk-free rate, such as a fixed deposit rate. Groww's fund example does exactly that. The Red Rock paper adds one firm rule: use the same target every time you compare.
Why bother with a second ratio when Sharpe exists? Sharpe treats a big winning day as "risk" in the same way as a big losing day. Most traders do not see a big win as a problem. Sortino ignores the upside swings and punishes only the shortfalls.
How to work out downside deviation, step by step
Most top pages skip this part, yet it is where most mistakes happen. Follow these steps for any list of daily, weekly or monthly results:
- Choose the target. For your own trading, zero is the simple choice.
- Go through every period. If the result is below the target, write down the shortfall (result minus target). If it is at or above the target, write zero.
- Square each number from step 2. Squaring makes every value positive and makes big shortfalls weigh more than small ones.
- Add up the squares and divide by the total number of periods, counting the zero rows too.
- Take the square root. That is the downside deviation.
Step 4 is the trap. The Red Rock paper says it has seen the ratio worked out wrongly "more often than not", in books and in trading software. The usual error is to throw away the zero rows and take the standard deviation of the losing periods only. The paper gives a simple case. A stream of shortfalls of 0, 0, 0 and 10 has far less downside risk than 10, 10, 10 and 10. If you drop the zeros, both look the same.
Worked example: Sortino on a 10-day F&O P&L list
Here is a made-up record of 10 trading days for an F&O trader. Each figure is the net P&L for the day, after charges. The target is zero, so every losing day counts as a shortfall. These numbers are examples only. They are not real results.
| Day | Net P&L (₹) | Below target? | Shortfall squared |
|---|---|---|---|
| 1 | 4,500 | No | 0 |
| 2 | -2,000 | Yes | 40,00,000 |
| 3 | 3,000 | No | 0 |
| 4 | -1,500 | Yes | 22,50,000 |
| 5 | 6,000 | No | 0 |
| 6 | -3,500 | Yes | 1,22,50,000 |
| 7 | 1,000 | No | 0 |
| 8 | 2,500 | No | 0 |
| 9 | -1,000 | Yes | 10,00,000 |
| 10 | 2,000 | No | 0 |
| Total | 11,000 | 4 losing days | 1,95,00,000 |
Now the arithmetic:
- Average daily P&L = ₹11,000 / 10 = ₹1,100.
- Sum of squared shortfalls = 1,95,00,000. Divide by all 10 days, not by the 4 losing days: 19,50,000.
- Downside deviation = square root of 19,50,000 = about ₹1,396.
- Daily Sortino ratio = 1,100 / 1,396 = about 0.79.
Ratios are often shown on a yearly basis. The common shortcut multiplies a daily ratio by the square root of the number of trading days in a year. A year is usually taken as about 252 trading days, and the square root of 252 is about 15.87. So the yearly Sortino here is 0.79 x 15.87 = about 12.5.
That 12.5 is not a forecast. It assumes the next 242 days look like these 10. They almost never do. A Sortino of 12.5 would be far above what fund pages call good. It only looks that high because the sample is tiny.
For comparison, the Sharpe ratio on the same 10 days is lower. The sample standard deviation of the daily P&L is about ₹3,053, so the daily Sharpe is about 0.36 and the yearly figure about 5.7. Sortino is higher than Sharpe here because the biggest swings were winning days, and Sortino does not count those as risk.
Sortino ratio in Excel or Google Sheets
You do not need a special calculator. Put one day's net P&L in each cell of column B, from B2 down to B11 for the 10-day example. Keep the no-trade days as zero. Then follow these steps:
- In C2, type =MIN(B2,0)^2 and copy it down to C11. It gives the squared shortfall for a losing day and zero for any other day.
- In E2, type =SQRT(SUM(C2:C11)/COUNT(B2:B11)). This is the downside deviation.
- In E3, type =AVERAGE(B2:B11)/E2. This is the daily Sortino ratio.
On the example list, E2 shows about 1,396 and E3 about 0.79, the same as the working above. COUNT counts every day, so the formula avoids the common mistake of dividing by losing days only. For a target other than zero, subtract the target from each day first in a new column, then use that column.
Same profit, different Sortino
This second example shows why Sortino can rank traders differently from win rate. Two made-up traders each make ₹6,000 over 10 days.
- Trader A wins ₹1,500 on 8 days and loses ₹3,000 on 2 days.
- Trader B wins ₹2,400 on 5 days and loses ₹1,200 on 5 days.
| Measure | Trader A | Trader B |
|---|---|---|
| Total net P&L | ₹6,000 | ₹6,000 |
| Win rate (winning days) | 80% | 50% |
| Profit factor (gross wins / gross losses) | 2 | 2 |
| Downside deviation (target 0) | about ₹1,342 | about ₹849 |
| Daily Sortino ratio | about 0.45 | about 0.71 |
| Daily Sharpe ratio | about 0.32 | about 0.32 |
Win rate says Trader A is better. Profit factor says they are equal. Even the Sharpe ratio cannot tell them apart, because their results swing by the same amount around the same average. Sortino prefers Trader B. B loses more often, but each loss is smaller. Trader A's two losses of ₹3,000 each are the kind that hurt most, and Sortino squares them.
Neither trader is "right". The table shows that a single number never tells the full story. If big single-day losses are your weak spot, Sortino will show it sooner than win rate.
What is a good Sortino ratio?
Nippon India Mutual Fund's guide says a Sortino ratio between 1 and 2 is generally considered good. It adds that some investors accept less than 1 and that the ratio can go above 2. A negative ratio means no reward for the risk taken. Aditya Birla Sun Life Mutual Fund's guide agrees and calls 1 to 2 ideal. It says the ratio can go up to 3, which it calls exceptional. Not every guide uses the same line. Bajaj Finserv Mutual Fund's guide says a ratio of 2 and above is generally considered good. Groww's guide says a higher Sortino ratio is typically considered better. OneTradeJournal's dashboard uses a similar rule of thumb: above 1 is good and above 2 is excellent.
Treat these as rough guides. They come from fund data measured over years against a risk-free rate. A trader's number depends on the period, the target, whether returns are in rupees or percent, and how many days are in the sample.
- Higher or lower? Higher is better. It means more return for each unit of downside risk.
- Is 0.5 good? On yearly fund data, it is below the 1 to 2 range fund pages call good. On a daily trading figure, 0.5 cannot be compared with that range at all until you annualise it.
- Is 1.5 good? On yearly fund data, it sits inside the 1 to 2 range Nippon calls generally good, but below the 2 that Bajaj Finserv's guide uses. Compare it only with funds in the same category over the same period.
- Negative? Your average return was below your target. With a target of zero, you lost money on average.
Calmar ratio: formula and meaning
The formula and where the name comes from
Terry W. Young created the Calmar ratio and first published it in 1991 in the trade magazine Futures. The name comes from his firm's newsletter, CALifornia Managed Accounts Reports. Young defined it as the average yearly return over the last 36 months divided by the maximum drawdown over the same 36 months. It was worked out monthly.
Calmar ratio = yearly return / maximum drawdown, both in percent for the same period.
Maximum drawdown is the largest fall from any high point in your account to a later low, before a new high. The drawdown glossary entry explains it in more detail, and the maximum drawdown calculator works it out from a list of balances.
A Calmar ratio of 1 means the yearly return was the same size as the worst fall. Above 1, the return was bigger than the worst fall. Below 1, the worst fall was bigger than a year's return. Wikipedia's Calmar entry says a higher ratio means more yearly return per unit of maximum drawdown.
Calmar ratio vs MAR ratio
People often treat these as the same thing. They are close cousins with one key difference: the time window.
- Calmar ratio: uses the last 36 months only.
- MAR ratio: uses all data from the start (inception). It was popularised by a rival newsletter, Managed Account Reports, founded in 1979 by Leon Rose. It divides the compound yearly return since inception by the maximum drawdown since inception.
Some later versions of the Calmar ratio subtract a risk-free rate from the return, which makes it look more like Sharpe. Whatever version you use, Wikipedia notes a weakness that matters a lot for traders: the ratio is sensitive to the chosen period. Compare two records only when they cover the same start and end dates.
Worked example: Calmar on the same 10 days
Go back to the 10-day list above. Say the trader started with ₹1,00,000 (an example figure) and added or withdrew nothing.
- The account rose to ₹1,10,000 after day 5. That was the peak before the worst fall.
- Day 6 lost ₹3,500, taking it down to ₹1,06,500.
- Maximum drawdown = ₹3,500, or 3,500 / 1,10,000 = about 3.18% of the peak.
- The account ended at ₹1,11,000, a return of 11% over 10 days.
Now try the textbook version. Turn 11% in 10 days into a yearly return by compounding: 1.11 raised to the power (252 / 10), minus 1. That gives about 1,287% a year. Divide by the 3.18% drawdown and the Calmar ratio comes out at about 405.
Nobody should trust that number. It assumes the trader repeats an 11% fortnight about 25 times in a row, and that no worse drawdown ever comes. This is the most common way the Calmar ratio misleads new traders: a short good run, scaled up to a year.
A more honest view for a short record is to skip the yearly step. Over these 10 days, the return was 11% and the worst fall was 3.18%, so the return was about 3.5 times the drawdown. Call it a "return to drawdown" figure for this period only. It is not the textbook Calmar ratio, and it should be read as one data point, not a verdict.
What is a good Calmar ratio?
There is no official scale. None of the primary sources checked for this guide gives a fixed "good" number. The one safe reading is the plain meaning: above 1, the yearly return beat the worst fall; below 1, it did not. OneTradeJournal's dashboard uses that same line: its note for a Calmar above 1 reads "Returns outpace worst drawdown". Beyond that, compare Calmar only between records with the same dates and the same length.
Sharpe ratio in short, and how the three compare
William F. Sharpe introduced his measure in 1966 for mutual funds and called it the "reward-to-variability ratio". In a 1994 paper he accepted the common name, the Sharpe ratio. It divides the return above a benchmark (often the risk-free rate) by the standard deviation of that return.
The Red Rock paper points out a strange side effect. For a strategy with rare big wins, removing the largest winning days can raise the Sharpe ratio. That makes little sense to a trader, and it is the main reason Sortino exists.
When do the two give the same answer? Wikipedia's Sortino entry says they give similar results when returns are spread evenly around the middle and the target sits near the middle value. They drift apart when returns are lopsided (skewed), for example many small losses and a few big wins. Option buying often looks like that.
Some fund pages also show the information ratio. It compares a fund with a benchmark index, such as the Nifty 50. It takes the fund's return minus the index return (the active return) and divides it by how much that gap swings. It tells you if a fund manager beat the index steadily. It does not tell you how deep your own losses were.
| Ratio | What it counts as risk | Useful for | Weak spot |
|---|---|---|---|
| Sharpe | All swings, up and down | Comparing funds or strategies with fairly even returns | Punishes big winning days the same as big losing days |
| Sortino | Only swings below the target | Traders with uneven results, such as option buyers with occasional big wins | Needs enough losing periods to be reliable; depends on the target chosen |
| Calmar | The single worst drawdown | Asking whether the return was worth the worst fall | One bad day can change it a lot; very sensitive to the period used |
| Information ratio | Swings in the gap between the fund and its benchmark index | Checking if a fund beat its index steadily | Needs a fitting index; does not separate upside from downside swings |
Do these ratios mean anything with few trades?
This is the question no top-ranking page answers for traders, and it matters more than the formulas. Short records produce numbers that look precise but are not.
- The textbook warns about it too. Wikipedia's Sortino entry says these measures do not give accurate results over the short term, such as asset trading. Use them on months of data, not days.
- Annualising magnifies noise. Sharpe's 1994 paper shows that, under simple assumptions, a ratio measured over T periods equals the one-period ratio times the square root of T. Multiplying a 10-day daily ratio by 15.87 also multiplies its error.
- Downside deviation needs losing periods. Groww's guide notes there must be enough "bad" observations for the result to mean much. If you had 2 losing days out of 10, the whole denominator rests on those 2 days.
- No losing day breaks the formula. If every day was positive, downside deviation is zero and the ratio has no finite value. OneTradeJournal shows "Inf" in that case.
- Calmar rests on one event. It uses only the single worst fall. One bad expiry day can halve it overnight, and a short record may not have met its worst day yet.
- The frequency matters. Wikipedia's Sortino entry notes that the simple discrete form needs monthly returns to get enough data points. Daily data gives more points, but each day is noisy.
So what should you do with a few weeks of trades? Work the ratios out, but do not act on them alone. First read the plain numbers: average win, average loss, win rate, profit factor and your largest single loss. Then watch whether Sortino and Calmar stay in a similar range as months are added. A ratio that holds steady over several months tells you far more than a large one from a single good month.
How Sortino and Calmar fit with drawdown, profit factor and expectancy
The two ratios summarise things you can also measure directly. Look at them side by side. In the 10-day example, the gross wins were ₹19,000 and the gross losses ₹8,000, so the profit factor was about 2.38. Six of 10 days were winning days, a 60% win rate.
| Measure | What it answers | Where to work it out |
|---|---|---|
| Maximum drawdown | How deep was the worst fall? | Maximum drawdown calculator |
| Profit factor | How many rupees won for each rupee lost? | Profit factor calculator |
| Expectancy | What does an average trade earn or lose? | How to calculate trading expectancy |
| Risk-reward ratio | How big are wins compared with losses? | Risk-reward ratio explained |
| Sortino ratio | How much return per unit of below-target swings? | This page, or your journal's analytics |
| Calmar ratio | Was the yearly return worth the worst fall? | This page, or your journal's analytics |
A good habit is to look for agreement. If profit factor, expectancy and Sortino all improve over a quarter, the change is more likely to be real. If Sortino jumps while expectancy falls, check whether one big winning day is hiding a weaker run.
Sortino and Calmar for mutual fund investors
Many people searching for the Sortino ratio are choosing a mutual fund. Guides from Groww and Nippon India Mutual Fund explain it with yearly returns and a risk-free rate, and that use is different from a trader's.
Here is a made-up fund example. Say a fund returned 12% in a year, the risk-free rate was 7% and the downside deviation was 5%. The Sortino ratio is (12 minus 7) / 5 = 1. If the fund's standard deviation was 10%, its Sharpe ratio would be (12 minus 7) / 10 = 0.5. The fund looks better on Sortino because much of its swing was on the upside. All these figures are examples, not data for any real fund.
If you read these ratios on a fund page, keep three rules in mind:
- Compare funds in the same category (for example, two large-cap funds), over the same period.
- Check how the figure was worked out: which period, which risk-free rate and whether it uses monthly or daily returns. Look for these details next to the ratio on the page or factsheet that shows it.
- Never compare a fund's Sortino with your own trading Sortino. The period, target, units and data frequency are all different.
These ratios describe the past. They do not predict what a fund or a trader will do next. This page does not recommend any fund.
How to track Sortino and Calmar on your own trades
To work these ratios out properly, a record needs more than a list of trades. It needs:
- Net P&L for every trading day, after brokerage and charges, because gross figures overstate both ratios.
- Your starting capital, plus every deposit and withdrawal, so returns can be measured in percent instead of rupees.
- Days with no trades, counted as zero. If you skip them, the average looks higher and the ratios look better than they are.
- Enough history. Months are better than weeks.
OneTradeJournal shows Sharpe, Sortino and Calmar, next to maximum drawdown and profit factor, in the Risk-Adjusted Performance card on its Trade Analysis screen. You log each trade by hand; there is no broker link. The app estimates charges per trade, so the figures use P&L after estimated charges. These are estimates, not contract notes.
Here is how the app works the ratios out, so you can match it with the method on this page:
- It builds a daily series from your first to your last trade day. Weekdays with no trades count as zero (every day counts for crypto).
- If you have logged starting capital, each day is a percent return on the account value at the start of that day. Otherwise it uses rupee P&L.
- Sortino uses a target of zero. Downside deviation divides by all days, the method shown above. The daily figure is then multiplied by the square root of 252 (365 for crypto and 260 for forex).
- Sharpe uses no risk-free rate and the sample standard deviation.
- Calmar, with capital logged, is the yearly time-weighted return (deposits and withdrawals removed) in percent, divided by maximum drawdown in percent. Without capital, it uses yearly average rupee P&L divided by maximum drawdown in rupees.
One difference from the textbook is worth knowing. The original Calmar uses 36 months. The app uses whatever period you have logged or filtered. On a short period you will see the same inflated numbers as the 10-day example, so read them with care. Because the app uses percent returns when capital is logged, its Sortino can differ a little from a rupee calculation on the same days.
The free plan lets you log one trade per day per market, and it does not expire. Paid plans add AI features, such as asking questions about your own logged trades. Its answers can be wrong, and it sees only the trades you entered. For a wider view of what to review each week, see how to track trading performance.
Common mistakes with Sortino and Calmar
- Dividing by the losing days only. Downside deviation divides by all periods. Dropping the zero rows changes the answer and hides how rare your losses were.
- Annualising a few weeks. A 10-day Calmar of 405 is a maths result, not a skill result.
- Changing the target between comparisons. A target of zero and a target of 7% give different answers. Keep one target.
- Using gross P&L. In F&O, charges on many small trades add up. Use net P&L after charges.
- Comparing different periods. A Calmar ratio over a calm quarter and one over a volatile year are not comparable.
- Comparing a fund with a trading account. Fund ratios use yearly or monthly data and a risk-free rate; your journal may use daily data and a zero target.
- Reading one ratio alone. Check it against expectancy, drawdown and profit factor before drawing any conclusion.
Sources
- Wikipedia: Sortino ratio (definition, target or MAR, downside deviation, discrete form needs monthly data), checked 10 Nov 2026.
- Wikipedia: Calmar ratio (Terry W. Young, Futures 1991, 36 months, MAR ratio, period sensitivity), checked 10 Nov 2026.
- Red Rock Capital: Sortino, A Sharper Ratio (history and target downside deviation method), checked 10 Nov 2026.
- William F. Sharpe: The Sharpe Ratio (1994), checked 10 Nov 2026.
- Nippon India Mutual Fund: Sortino ratio (what is a good Sortino ratio), checked 10 Nov 2026.
- Aditya Birla Sun Life Mutual Fund: Sortino ratio (1 to 2 ideal, up to 3 exceptional), checked 10 Nov 2026.
- Bajaj Finserv Mutual Fund: What is the mutual fund Sortino ratio (2 and above generally good), checked 10 Nov 2026.
- Groww: Sortino ratio (fixed deposit rate as risk-free, need for enough bad observations), checked 10 Nov 2026.
- Wikipedia: Information ratio (active return against a benchmark divided by its swings), checked 10 Nov 2026.
Related Topics
Related Articles
Risk Adjusted Return Calculator 2026: Sharpe, Sortino Ratio
Sharpe, Sortino and Calmar calculator from six manual inputs: formula, India risk-free rate, annualising, Excel steps and how the tool's shortcuts work.
Paper Trading App in India: What Each App Offers and What SEBI Says
Paper trading apps in India compared from their own pages: cost, virtual money and price delay. Plus SEBI's 2024 advisory and how to judge your practice trades.
Share Market Kaise Sikhe: A Step by Step Plan for Beginners in India
Share market kaise sikhe: a 12-week beginner plan, free NISM and NSE resources, real costs of a first trade, SEBI loss data and a journal habit. No promises.
TradingView Paper Trading in India: Steps, Limits and Nifty Rules
TradingView paper trading is free, but its help page lists Indian exchange symbols as non-tradable. Setup steps, NSE data delay, limits and a journal method.
Zerodha Console P&L: How to Read Your Report and What It Misses
Read your Zerodha Console P&L step by step: realised and unrealised P&L, charges, heat map, Tax P&L, Verified P&L, why numbers differ and what it can't show.
TradeZella Price in India: Plans in Dollars and Rupees (2026)
TradeZella price: Essential $35, Pro $59, Ultra $99 a month or $315 to $891 a year. Rupee cost, the July 2026 rise, trial facts and checks for Indian traders.
The trading journal for Nifty, Bank Nifty and Indian F&O. Track your trades, spot patterns, build discipline.
- Log one trade a day by hand, on purpose
- AI mentor finds your repeat mistakes
- Behavioural analytics catch tilt early
- Trading calendar with P&L heatmap
- Pre-trade checklist flags risks
Yearly ₹3,999 · No broker credentials