Standard deviation, and its blind spot
Standard deviation measures how much returns varied around their average. Higher means a bumpier ride.
It is genuinely useful and it has one structural flaw: it treats upside and downside identically. A month that surprised on the way up raises standard deviation exactly as much as an equally sized fall.
That is not how anyone experiences a portfolio. Nobody has ever been distressed by an unexpectedly good quarter. So a fund penalised for delivering occasional large gains scores as “riskier” than one that grinds along and then falls sharply once — which inverts the thing you were trying to measure.
Use it for what it is: a measure of variability, not of danger.
Sortino fixes exactly that
The Sortino ratio is the same idea as Sharpe with one change: it divides excess return by downside deviation only, ignoring upside variability entirely.
Which makes it the more honest measure for an investor, because it asks the question you actually care about — how much return did this fund produce per unit of the movement that hurt?
| Ratio | Divides excess return by | Reads as |
|---|---|---|
| Sharpe | Total variability | Return per unit of bumpiness, good and bad |
| Sortino | Downside variability only | Return per unit of pain |
A fund with a mediocre Sharpe and a strong Sortino is one whose variability is mostly to the upside. That is a meaningfully different product from one where the two are similar, and comparing only Sharpe would have missed it.
One caution that applies to both: they are ratios, so they can be high because the numerator was good or because the denominator was small. Always read the underlying return alongside them.
Beta measures relationship, not risk
Beta measures how much a fund moved relative to its benchmark. Above one means it amplified the index; below one means it moved less.
Two things it does not tell you. It says nothing about absolute risk — a low-beta fund tracking a volatile index can still be volatile. And it says nothing about direction: a beta of 1.2 means bigger moves both ways, which is desirable in a rise and unpleasant in a fall.
Beta is also only meaningful against the right benchmark. A small cap fund's beta against a large cap index is close to noise. Check what the beta was computed against before reading anything into it.
Capture ratios are the ones worth reading
The most intuitive measures published, and the least used.
Up-capture is the share of the benchmark's rise the fund captured. Down-capture is the share of its fall the fund suffered. Both as percentages.
| Up / Down | What it means |
|---|---|
| 110 / 110 | An amplified version of the index, both directions |
| 90 / 70 | Gives up some of the rise, avoids much of the fall — a defensive profile |
| 105 / 95 | The genuinely good combination, and rare |
| 85 / 105 | Less of the rise and more of the fall — the profile to avoid |
These two numbers describe the shape of a fund better than any single ratio. Two funds can post the same return over a period with completely different capture profiles, and the difference determines whether an investor stayed — a 90/70 fund is a far easier thing to hold through a bad year than a 110/110 fund, even where the end result matches.
Since the returns you actually receive depend on remaining invested, the shape is not a secondary consideration. It is frequently the deciding one.
Maximum drawdown is the one that predicts behaviour
Not a ratio at all: maximum drawdown is simply the largest peak-to-trough fall the fund has experienced, and the time it took to recover.
It is the most directly useful figure in this article, because it is the number an investor actually lives through. A fund with an excellent Sharpe and a 55% drawdown is a fund most people would have sold at the bottom — and someone who sold received none of the return the Sharpe describes.
Read drawdown before any ratio. The ratios describe the fund; the drawdown describes whether you would still be holding it.
What all of these share
Three honest limitations, and they apply to every measure above.
They are backward-looking. Computed from history, and the regime that produced them does not renew itself on request.
They depend on the period. A ratio computed over three years that excluded a crash describes a fund that has not been tested. Always check the window.
They can be distorted by overlapping data. Statistics computed on overlapping rolling windows are systematically overstated, because the arithmetic assumes an independence the data does not have — the point made in rolling returns. A suspiciously high ratio is more often a construction artefact than skill.
Which suggests the working order: drawdown first, capture ratios second, Sortino third, and everything else as context.
Reading them together
No single number describes a fund. Drawdown says how bad it got, capture ratios say what shape the ride had, and the ratios say what you were paid for it.
FNOTrader's Mutual Funds app computes these from the full AMFI NAV history — around 34 million NAV rows — alongside rolling returns across every start date, so the measures can be read over several windows rather than the one a factsheet chose. Formulas are published in the user guide.
Common questions
What does standard deviation measure in a mutual fund?
How much returns varied around their average. Its structural flaw is that it treats upside and downside identically, so a fund is penalised for occasional large gains exactly as much as for equivalent falls — which is not how anyone experiences a portfolio.
What is the difference between the Sharpe and Sortino ratios?
Sharpe divides excess return by total variability; Sortino divides it by downside variability only. Sortino answers the question you actually care about — return per unit of the movement that hurt.
Is beta a measure of risk?
It measures how much a fund moved relative to its benchmark, not absolute risk. A low-beta fund tracking a volatile index can still be volatile, and beta is only meaningful against the right benchmark.
What are capture ratios?
Up-capture is the share of the benchmark's rise a fund captured; down-capture is the share of its fall it suffered. Together they describe the shape of a fund better than any single ratio — 90/70 is a defensive profile, 110/110 is an amplified index.
Why do capture ratios matter more than the return?
Because the returns you receive depend on staying invested. Two funds can post the same return with completely different profiles, and a 90/70 fund is far easier to hold through a bad year than a 110/110 one — so the shape often decides the outcome.
Which risk measure should I look at first?
Maximum drawdown, because it is the number an investor actually lives through. A fund with an excellent Sharpe and a 55% drawdown is one most people would have sold at the bottom, receiving none of the return the ratio describes.
What are the limits of these measures?
They are backward-looking, they depend heavily on the period chosen — a three-year window excluding a crash describes an untested fund — and statistics built on overlapping rolling windows are systematically overstated.
Continue reading
More in Mutual Funds · App: Mutual Funds · Definitions: glossary · Free tools: calculators · All: every article