Volatility and Standard Deviation Explained: What They Really Measure

Volatility is the standard deviation of an investment’s returns. In plain language, it measures how widely the returns have scattered around their own average: a high volatility means returns have swung a lot from one period to the next, and a low volatility means they have stayed close to typical.

It is the most widely quoted risk statistic in finance, and it is also the most widely misread. Volatility measures variability. Risk, as most investors experience it, is the chance of losing money you cannot afford to lose. Those two ideas overlap, but they are not the same, and the gap between them explains most of the arguments about this number.

What it measures

Standard deviation is a general statistical idea, not a finance-specific one. Applied to returns, the recipe in words is:

1. Take the return for each period, say each day or each month.
2. Work out the average return across all those periods.
3. For each period, find the difference between its return and that average.
4. Square each difference, so that overshoots and undershoots both count as distance rather than cancelling out.
5. Average those squared differences. That average is the variance.
6. Take the square root of the variance. That is the standard deviation, and in finance we call it volatility.

The squaring step is worth pausing on, because it explains two properties. First, it is why volatility treats a period that was much better than average exactly like a period that was much worse: both are just distance from the middle. Second, it is why large deviations dominate the result. A single extreme period contributes far more than several mild ones, because the distance is squared before it is averaged.

Annualising it

Volatility is almost always quoted as an annual figure, even when it is computed from daily or monthly returns. The convention is the square root of time rule: multiply the standard deviation of daily returns by the square root of the number of trading days in a year, or multiply monthly volatility by the square root of twelve.

That rule assumes each period’s return is independent of the last. Real markets show streaks and mean reversion, so the annualised figure is a serviceable convention rather than a precise translation. It is still the right one to use, because it is what everybody else uses and comparability matters more here than false precision.

How to read it

Volatility is only comparable on identical terms. The frequency of the underlying returns, the length of the window, and the annualisation convention all change the printed number. Comparing a three year monthly volatility with a one year daily volatility tells you very little.

Judge it against a relevant benchmark and peer set. An equity portfolio and a short duration debt fund will naturally sit at very different levels, and neither figure is impressive or alarming on its own. What is informative is the level relative to a fair comparison over the same dates.

Expect it to move. Volatility is not a constant of nature. It clusters: quiet periods tend to follow quiet periods and turbulent periods tend to follow turbulent ones. A single trailing figure is a summary of a window that has already closed, so a rolling chart of volatility usually tells you more than one number.

Split it into up and down. Because standard deviation is symmetric, it treats a strong rally as risk. Most investors do not. Separating the variability of losing periods from the variability of gaining periods gives a more intuitive picture, which is exactly what downside deviation does.

Use it as a denominator, not a verdict. Volatility earns most of its keep as the divisor in risk adjusted return measures, where it converts a raw return into return per unit of variability. That is the role it plays in the Sharpe ratio.

Remember it says nothing about direction. Two portfolios can post the same volatility while one compounded steadily upward and the other went nowhere. Volatility is the width of the scatter, not its centre.

What it does not tell you

It does not distinguish good surprises from bad ones. A portfolio that occasionally jumps sharply higher will register as more volatile than one that grinds along. To an investor, those are not the same experience, but standard deviation cannot tell them apart. This is the single most common complaint about the measure and the reason downside-only alternatives exist.

It assumes a tidy distribution that markets do not always follow. Standard deviation is at its most meaningful when returns are roughly bell shaped. Market returns tend to have fatter tails than a bell curve, meaning extreme moves happen more often than the model implies, and they are often skewed, meaning the shape is not symmetric. So a volatility figure can understate how frequently the truly unpleasant days occur.

It says nothing about the order returns arrived in. Volatility is computed from an unordered bag of returns. Shuffle the same returns into a different sequence and the volatility is unchanged, yet the maximum drawdown can be completely different, because drawdown depends entirely on path. Two portfolios with identical volatility can put an investor through very different ordeals.

It does not measure the risk of permanent loss. A holding whose price barely moves can still be a business in structural decline, and a volatile holding can be a sound business in a jumpy market. Price variability and impairment of underlying value are different questions, and only one of them is answered by a return series.

It is invisible where prices are not marked. Illiquid or infrequently priced assets can appear to have low volatility simply because their reported values do not move often. Smooth reported prices are not the same as a smooth economic reality, and a naive comparison across liquid and illiquid holdings will systematically favour the latter.

It is an estimate, with error. Computed from a limited sample, volatility carries uncertainty. Short windows are noisy and one dramatic week can dominate the result. Extending the window reduces the noise but blends together market conditions that may have little to do with each other.

It does not survive a careless backtest. A simulated volatility calculated on prices that were never adjusted for splits and bonuses, or on a universe that quietly excludes companies which stopped existing, will not describe what an investor would have lived through. The measurement is only as good as the price history and the point in time discipline behind it.

The fair summary is that volatility is a good measure of what it actually measures. It is a precise, comparable, well understood description of how much returns have varied. It becomes a poor measure only when it is asked to stand in for everything the word “risk” carries in ordinary speech. Read it as one dimension, put drawdown and downside measures beside it, and it does honest work.

Related reading

• Portfolio metrics explained: the hub connecting every risk and return measure.
• What is maximum drawdown: the path-dependent risk measure volatility cannot capture.
• Downside deviation explained: measuring only the variability investors actually mind.
• What is the Sharpe ratio: the best known use of volatility as a denominator.
• What is beta in investing: variability relative to the market rather than in absolute terms.

This article is educational. Altys Labs is not a registered research analyst or investment adviser, and nothing here is investment advice or a recommendation to buy, sell, or hold any security.