Investment standard deviation measures how widely a set of periodic returns varies around its average. It is commonly used as a measure of historical volatility: a higher figure means the observed returns were more dispersed, while a lower figure means they clustered more closely around the mean.
That makes standard deviation useful for comparing investments measured over the same period and at the same frequency. It does not predict a return, guarantee a range of outcomes, or distinguish favorable upside from harmful downside. Use it as one input in a broader risk review.
What investment standard deviation tells you
Standard deviation is a measure of spread, calculated as the square root of variance. If two funds had the same average monthly return, the one with the larger monthly standard deviation experienced greater variation during the measured period.
The unit matches the input. If you use monthly percentage returns, the result is a monthly standard deviation expressed in percentage points. If you use annual returns, the result is annual. Never compare a daily figure directly with a monthly or annual figure.
The formula
For returns r1 through rn with mean return r̄:
Population standard deviation = square root of [sum of (ri − r̄)² divided by n]
Sample standard deviation = square root of [sum of (ri − r̄)² divided by (n − 1)]
The denominator is the key difference. Use the population version when the observations are the entire set you intend to describe. Use the sample version when the observations are treated as a sample used to estimate a broader population’s variability. NIST’s Engineering Statistics Handbook shows the sample formula and explains that standard deviation is the square root of variance.
Worked example
Assume an investment produced five annual returns: 5%, 10%, 7%, 2%, and 6%.
- Find the mean. Add the returns and divide by five: (5 + 10 + 7 + 2 + 6) ÷ 5 = 6%.
- Find each deviation. The differences from 6% are −1, 4, 1, −4, and 0 percentage points.
- Square the deviations. The squared values are 1, 16, 1, 16, and 0.
- Add them. The sum of squared deviations is 34.
- Divide and take the square root. For a population, √(34 ÷ 5) = 2.61%. For a sample, √(34 ÷ 4) = 2.92%.
The calculation says the five observed annual returns were typically dispersed by roughly 2.61 percentage points around their 6% average when treated as the full population. It does not say future returns will stay within that distance.
How to calculate standard deviation in Excel or Google Sheets
Put periodic returns in one column, using consistent periods and decimal values. For example, enter 5% as 0.05, not 5.
- Use
=STDEV.P(A2:A6)when the range is the complete population you want to describe. - Use
=STDEV.S(A2:A6)when the observations are a sample used to estimate broader variability.
For investment analysis, the choice depends on the question rather than a universal rule. Analysts often use the sample function for a finite historical return series intended to estimate an underlying return process. State which function you used so another reader can reproduce the result.
Use return data, not price levels
Calculating the standard deviation of share prices answers a different and usually less useful question. For comparable investment volatility, first calculate periodic total returns that include distributions where relevant. A simple return is:
(Ending value − beginning value + distributions) ÷ beginning value
Use adjusted-price or total-return data when available. Keep the frequency consistent, avoid mixing calendar and trading periods, and check how missing observations are handled.
Annualizing a standard deviation
A common convention annualizes a periodic standard deviation by multiplying it by the square root of the number of periods per year. A monthly figure is multiplied by √12; a daily figure may be multiplied by √252.
This shortcut assumes returns are sufficiently independent and similarly distributed through time. Volatility clustering, autocorrelation, changing market regimes, and inconsistent trading calendars can make the estimate misleading. Always label the frequency and method.
How to interpret the result
Standard deviation is most informative in a like-for-like comparison. Use the same:
- return frequency;
- measurement window;
- currency and distribution treatment;
- sample or population convention; and
- annualization method.
A higher standard deviation is not automatically bad. An investor might accept more variability in exchange for a different expected return or portfolio role. Nor is a low figure automatically safe: an illiquid asset with infrequent appraisals can report smooth returns while still carrying valuation, leverage, or liquidity risk.
Key limitations
It treats upside and downside alike
A return far above the mean increases standard deviation just as a large loss does. If your concern is loss, also examine maximum drawdown, downside deviation, value at risk with its assumptions, and stress scenarios.
It is backward-looking
The figure summarizes the chosen dataset. It can change sharply when a calm observation drops out of the window or a crisis enters it. Different lookback periods can produce different answers.
Returns may not be normally distributed
Rules that translate one or two standard deviations into fixed probabilities rely on a distributional assumption. Real investment returns can be skewed, have fat tails, and change behavior during stress. Do not interpret a historical standard deviation as a loss boundary.
It ignores sequence and liquidity
Two investments can share an average return and standard deviation but experience losses in a different order. Sequence matters for an investor making withdrawals. Standard deviation also does not reveal whether an asset can be sold quickly at a reasonable price.
Investor.gov’s broader investment-risk overview describes business, volatility, inflation, interest-rate, and liquidity risks. Combine quantitative measures with an understanding of the asset and your financial situation.
Frequently asked questions
What does investment standard deviation measure?
It measures how widely a set of periodic investment returns varies around its average. The result describes historical dispersion for the selected data, frequency, and calculation method.
Is higher standard deviation always worse?
No. Higher standard deviation means greater observed variability, not necessarily a worse investment. The appropriate level depends on expected return, portfolio role, time horizon, liquidity needs, and capacity for loss.
Should I use STDEV.S or STDEV.P?
Use STDEV.P when your observations are the entire population you want to describe. Use STDEV.S when the observations are a sample used to estimate broader variability. State the choice and apply it consistently.
Can standard deviation predict future losses?
No. It summarizes variability in a selected dataset. Future markets, correlations, liquidity, and return distributions can differ, so the measure cannot set a reliable limit on future losses.
How often should I calculate it?
Recalculate it on a consistent schedule that matches your review process, and also after material portfolio or market changes. Keep the return frequency and lookback window consistent when comparing results over time.
