The Sortino Ratio and Downside Deviation: A Fairer Risk Number for Traders in 2026
The Sortino ratio exists because the Sharpe ratio treats a great day and a terrible day as the same thing. Sharpe measures risk as the standard deviation of all returns, and standard deviation does not care about sign. A month with one unusually large winning day scores as more volatile, and therefore worse, than a month of small grinding losses.
That is a strange thing to tell a trader. Nobody has ever ended a session upset that their gains were too irregular. What ends accounts is the downside, and the Sortino ratio is the version of the calculation that only measures the downside.
This guide covers what the Sortino ratio is, how downside deviation is actually computed and the step most people get wrong, how it compares with Sharpe on the same data, what the number does and does not tell you, and how it fits alongside the rule-based metrics that govern a funded account.
Key takeaways
- Sortino divides by downside deviation. Only returns below your chosen threshold contribute to the risk figure, so good periods stop being penalized.
- The threshold is a choice you must state. The minimum acceptable return is usually zero for traders, and a ratio quoted without it is not comparable to anything.
- Above-threshold returns become zero, not missing. Dropping them instead of zeroing them inflates the ratio, and it is the most common calculation error.
- Sortino is almost always higher than Sharpe. That is arithmetic, not evidence of skill, so never compare one strategy's Sortino to another's Sharpe.
- It cannot see a risk that has not happened yet. Many small gains and one rare large loss produce a strong ratio right up until the rare loss arrives.
What this guide covers
What the Sortino ratio measures
The Sortino ratio is a risk-adjusted return measure that divides excess return by downside deviation rather than by total standard deviation. In plain terms, it asks how much return you produced per unit of the variability that actually hurt.
The formula is straightforward: take your average return over the period, subtract the minimum acceptable return, and divide by the downside deviation of the same return series. The numerator is the same shape as Sharpe's. The denominator is where everything changes.
The minimum acceptable return
The threshold in the calculation is called the minimum acceptable return, and it is the return level below which you consider an outcome a shortfall. Most traders use zero, which makes the ratio measure the variability of losing periods specifically. Some analysts use the risk-free rate. A funded trader can reasonably use a threshold tied to their own account requirements.
All three are defensible. What is not defensible is quoting a ratio without saying which one you used, because the choice changes the number materially. If you switch thresholds midway through tracking your own performance, your series stops being comparable to itself.
Why the asymmetry matters to a trader
Standard deviation is a symmetric measure. It treats a return three points above the mean and a return three points below the mean as equally informative about risk. For a portfolio manager reporting to a committee, that symmetry has a defensible logic. For a trader operating under a daily loss limit and a drawdown allowance, it does not describe the thing that ends accounts.
What ends accounts is a run of shortfalls. The Sortino ratio is the version of the statistic that measures those and nothing else. The SEC's investor material on managing risk is a reasonable neutral grounding on why risk measurement is not a single number.
How downside deviation is calculated
Downside deviation is the standard deviation of only the returns that fell below the threshold, with above-threshold returns entered as zero rather than removed from the sample. That final clause is the step most people get wrong, and getting it wrong inflates the ratio.
The four steps
Work through a return series one period at a time.
- Step one. For each period, compute the difference between the return and the minimum acceptable return.
- Step two. If that difference is positive or zero, record zero. If it is negative, record the difference.
- Step three. Square each recorded value, and average across the total number of periods, including the ones you recorded as zero.
- Step four. Take the square root. That is downside deviation.
The critical detail is in step three: you divide by the full period count, not by the count of losing periods only. Dividing by the losing count alone treats a strategy with rare losses as if it were as volatile as one with frequent losses of the same depth, which is backwards, and it produces a smaller denominator and a flattering ratio.
A worked illustration
Suppose a hypothetical trader records ten daily returns, in percent: +1.2, +0.6, minus 0.9, +2.1, minus 0.4, +0.3, minus 1.6, +0.8, +0.5, minus 0.2. The average is +0.24 percent. With a minimum acceptable return of zero, the four negative days contribute their squared values, the six other days contribute zero, and the sum of squares is divided by ten before taking the square root. The result is a downside deviation of roughly 0.60 percent, against a total standard deviation of roughly 1.01 percent for the same series.
Same ten days. Two different risk figures, and the Sortino denominator is roughly forty percent smaller because the plus 2.1 day no longer counts against the trader. These figures are illustrative and are used only to show the mechanics of the calculation.
Standard deviation treats a large winning day as evidence of risk. Downside deviation ignores it. Illustrative example built from the ten hypothetical returns in the text; not drawn from any account.
Teal bars sit at or above the threshold and are entered as zero in the downside calculation. Crimson bars are the only ones that contribute to downside deviation, and each one is still divided by the full ten-period count.
Standard deviation of all ten returns, including the best day. Roughly 1.01 percent in this illustration.
The strongest day makes this number larger, which makes the ratio smaller.
Downside deviation of the four shortfall days only. Roughly 0.60 percent in this illustration.
A smaller denominator is why Sortino is nearly always the higher of the two ratios.
SubtractReturn minus the minimum acceptable return, period by period.
Zero itPositive results become zero. Negative results are kept as they are.
Square, then averageDivide by every period, not only the losing ones.
Square rootThat result is downside deviation, the Sortino denominator.
tradefundrr.comIllustrative example only. Simulated trading environment. Figures demonstrate a calculation and are not a projection of any result.
Sortino versus Sharpe on the same data
For the same return series, the Sortino ratio is almost always higher than the Sharpe ratio, because downside deviation is almost always smaller than total standard deviation. This is arithmetic, not evidence of anything. It also means that comparing one strategy's Sortino to another strategy's Sharpe is meaningless, and it happens constantly in marketing material.
| Property | Sharpe ratio | Sortino ratio |
|---|---|---|
| Risk measure used | Standard deviation of all returns | Downside deviation below a threshold |
| Treats upside volatility as risk | Yes | No |
| Requires a stated threshold | No, uses the risk-free rate in the numerator | Yes, the minimum acceptable return |
| Typical relative value | Lower for the same series | Higher for the same series |
| Sensitive to a single large gain | Yes, it worsens the ratio | No, it improves the ratio |
| Sensitive to a single large loss | Yes | Yes, and more so proportionally |
| Sees path dependence and drawdown order | No | No |
Structural comparison of the two ratios. Neither measures the sequence in which returns arrived, which is the part that matters most under a drawdown rule.
When Sharpe is the more honest number
There is a case for Sharpe that deserves stating. If your strategy produces occasional very large gains, Sortino will reward you for exactly the behavior that makes your equity curve unpredictable. A trader whose results depend on a handful of outsized days has real fragility, and Sharpe at least registers that as variability while Sortino waves it through.
The reasonable position is to compute both, on the same series, with the settings written down. If the two diverge sharply, the divergence itself is telling you something about the shape of your returns. Our post on the Sharpe ratio for day traders covers the other side of the pair.
Where the ratio misleads
The Sortino ratio has three well-known failure modes: small samples, unrealized tail risk, and blindness to the order in which returns arrived. All three matter more to a funded trader than to a fund manager.
Small samples produce confident nonsense
The ratio depends on the variability of a subset of your returns. If you have thirty daily returns and eight of them fell below the threshold, your downside deviation is estimated from eight observations. That is not a measurement, it is an impression. Add one bad day and the number can move substantially.
Traders computing a Sortino ratio on a few weeks of a new strategy and treating the result as a finding are the most common misuse of this statistic. Treat any ratio built on a short series as a rough indication, and expect it to move as the sample grows.
It cannot see a risk that has not shown up
A strategy that produces many small gains and rare large losses posts an excellent Sortino ratio for as long as the rare loss stays away. Selling far out-of-the-money options is the textbook case. The statistic is not lying; it is describing the sample it was given, and the sample does not yet contain the event that defines the strategy.
This is a general property of ratio statistics and it is worth internalizing, because it is the mechanism behind a great many blown accounts. A number that summarizes past variability cannot warn you about a distribution's tail until the tail has arrived. Our post on risk of ruin approaches the same problem from the other direction.
It ignores the order of returns
Neither Sharpe nor Sortino cares whether your losing days were spread evenly or arrived consecutively. Shuffle your return series into any order and both ratios are identical. For an investor holding for years, that may be acceptable. For a trader under a drawdown rule, the order is the whole game: the same set of returns arriving in a bad sequence can end an account that the identical returns in a different sequence would not have touched.
Using it inside a funded account
In a funded account, the Sortino ratio is a diagnostic about the shape of your returns, and your rule metrics are what determine whether you keep the account. Those are different jobs and you need both.
What the ratio is genuinely good for
Tracked over time with a fixed threshold and a fixed return frequency, the Sortino ratio answers a question worth asking: are my losing periods getting shallower relative to what I make? A rising ratio across comparable periods usually means your losses are being contained better, which is exactly the improvement a funded program is asking for. That is a more informative signal than a rising equity curve, which can improve for a single lucky reason.
What it will not protect you from
A daily loss limit is path-dependent and absolute. A single breach ends a session, and on a hard daily loss rule it ends the account, regardless of how strong your long-run ratio was on the days before it. A drawdown allowance works the same way: it tracks the worst point of your path, not the average shape of your distribution.
So the practical structure is to track your rule metrics daily and your ratio metrics periodically. Worst day against the daily loss limit, current distance to the drawdown line, and number of soft-limit days used are the numbers that decide whether the account survives. The Sortino ratio is the number that tells you whether your process is improving. Confusing the two is how a trader with genuinely improving statistics still loses the account.
- State your minimum acceptable return and keep it fixed. Zero is the usual choice for traders.
- Enter above-threshold returns as zero, not as missing values, and divide by the full period count.
- Compute Sharpe on the same series so you can see how much of the difference is upside variability.
- Note the sample size next to the ratio, every time. A ratio without a period count is not a result.
- Compare your own ratio across periods with identical settings, never against a number from someone else's material.
- Track worst day, distance to the drawdown line, and soft-limit days used as separate daily metrics.
- Ask what event would ruin this ratio, and whether your sample contains one yet.
Where the honest limit sits
None of these ratios were designed for someone operating under a daily loss limit. They came from portfolio analysis, where the question is how a return stream behaves over years and where nobody gets shut off at four in the afternoon. Borrowing them is reasonable, and pretending they answer the funded trader's actual question is not.
The question a funded trader needs answered is closer to: given how my losses cluster, how likely is it that I hit a limit before I compound anything. That is a simulation question rather than a ratio question. The Sortino ratio is a useful input to it, because the shape of the downside is what you would be simulating. FINRA's material on day trading and the SEC's day trading guidance are worth reading alongside any performance statistic, because both make the same point from a different angle: short-term trading outcomes are dispersed, and a single summary number hides that dispersion.
Frequently asked questions
What is the Sortino ratio?
The Sortino ratio is a risk-adjusted return measure that divides excess return by downside deviation instead of by total standard deviation. It only counts the variability of returns that fell below a chosen threshold, so unusually good periods do not make the risk figure look worse. It answers how much return you produced per unit of the variability that actually hurt.
What is downside deviation?
Downside deviation is the standard deviation of only the returns that fell below a threshold, usually called the minimum acceptable return. Returns at or above the threshold are entered as zero rather than being dropped, which keeps the denominator of the calculation consistent. The average is taken across every period in the sample, not only across the losing ones.
How is the Sortino ratio different from the Sharpe ratio?
Sharpe divides excess return by the standard deviation of all returns, so a large winning day increases measured risk exactly as much as a large losing day. Sortino divides by downside deviation, so only shortfalls below the threshold count. Sortino is usually the higher of the two numbers for the same series, which is arithmetic rather than evidence of skill.
What is a good Sortino ratio for a trader?
There is no universal threshold, and any specific number quoted without the sample period, the return frequency and the minimum acceptable return behind it is not comparable. Daily returns and monthly returns produce different ratios for the same strategy. The useful comparison is your own ratio across periods with identical settings, not your number against someone else's.
How many trades do I need before the Sortino ratio means anything?
More than most traders assume. The ratio depends on the variability of a subset of your returns, so a small sample gives you very few observations below the threshold. Treat a ratio computed on a few weeks of daily returns as a rough indication rather than a measurement, and expect it to move meaningfully as the sample grows.
Does the Sortino ratio work for a funded account with a daily loss limit?
Partly. It measures the shape of your return distribution, which is useful, but a funded account is governed by path-dependent rules: a single daily loss limit breach can end a session or an account regardless of a strong long-run ratio. Use it alongside your rule metrics such as worst day and distance to the drawdown line, not in place of them.
What minimum acceptable return should I use?
Zero is the most common choice for traders, which makes the ratio measure the variability of losing periods specifically. Some use the risk-free rate, and a funded trader can reasonably use a threshold tied to their own account requirements. Whatever you pick, keep it fixed and state it, or your own series stops being comparable to itself.
Can the Sortino ratio be gamed?
Yes, in the same way most ratio statistics can. A strategy that produces many small gains and rare large losses can post a strong ratio until the rare loss arrives. The ratio describes the sample you fed it, and it cannot see a risk that has not shown up in that sample yet. Always ask what event would ruin the number, and whether your data contains one.
The number is a mirror, not a verdict
The Sortino ratio is worth adding to your review because it removes an obvious unfairness from the Sharpe calculation and because tracking it forces you to look at your losing periods as a group rather than as individual bad days. Those are real benefits.
What it will not do is tell you whether you are going to keep a funded account. That is decided by a daily loss limit, a drawdown allowance and a position limit, all of which are written down in your account terms before you place a trade. A simulated funded account is a reasonable place to build both habits at once: track the ratio to see whether your process is improving, and track the rules because those are what end the account. The 80/20 split applies on all programs if you reach a payout by following them.
Measure the downside against a published number
TradeFundrr publishes the daily loss limit, drawdown allowance, position rules and 80/20 split for every simulated program, so the threshold you measure against is written down before you start.
Get Funded →