Realized volatility: calculate and interpret historical movement

Realized volatility summarizes past returns using a chosen method and window. Learn the calculation and what it leaves uncertain.

By OptionsKing. Publisher and editorial methods.

Define the data window and estimator before quoting a realized-volatility number. The close-to-close example turns historical prices into log returns and annualizes their dispersion. We will then examine how a different window or use of intraday data changes the measurement.

The computation, in four steps

The standard version, called close-to-close, is the one every platform means unless it says otherwise.

1. Take the daily closes. For a 30-day reading you need 31 of them.
2. Turn each pair into a log return: the natural log of today's close divided by yesterday's.
3. Take the sample standard deviation of those 30 numbers. On Nike that came to 0.01518, so about 1.52 percent.
4. Annualize by multiplying by the square root of 252. 0.01518 x 15.87 = 0.241, so 24.1 percent.

That is it. No model, no risk-free rate, no assumption about what anyone believes.

The window decides the answer

This is the part people skip and then get confused by. Realized volatility is not a property of a stock. It is a property of a stock and a window, and the window is your choice.

Nike's 30-day reading is 24.1 percent. Stretch to 90 days and you pull in the March selloff and the number rises. Shorten to 10 days and you get a reading dominated by whatever happened in the last two weeks, which on a quiet fortnight can print in the teens.

None of those is more correct. They answer different questions. What matters is that when you compare implied against realized, the windows have to match the thing you are trading. A 45-day option is not usefully compared against a 10-day realized reading.

Two traps in the annualization

Square-root scaling depends on assumptions. Multiplying daily standard deviation by the square root of 252 uses a trading-day convention and assumes variance accumulates proportionally with time. Serial dependence, changing volatility and concentrated event risk can break that approximation. State the day count and horizon when comparing values.

Log returns, not simple returns. With a $74 stock over 30 days the difference is in the third decimal and you will not see it. On something that halved and doubled it is large. Use logs and stop thinking about it.

What realized volatility is good for

A trailing estimate supplies context for an option price, not the future volatility against which the trade settles. Record the entry IV and later measure realized movement over the same holding period. Even then, an option's profit also depends on payoff shape and trading costs.

One job, mostly: it is the yardstick you hold implied volatility against. On its own it tells you what already happened, and the market has already priced whatever caused it.

The trap is treating it as a forecast in disguise. A stock that realized 24 percent last month is not thereby a 24 percent stock next month, and the times it is most obviously wrong are exactly the times it costs you. Realized volatility going into a merger announcement, an earnings print or a Fed meeting is a measurement of the calm before, and it will tell you the option is expensive right up until it turns out to have been cheap.

Volatility does cluster, which is the honest defence of the number. Quiet months tend to follow quiet months. But clustering is a tendency, not a mechanism, and it breaks on the scheduled dates you can see on a calendar.

Close-to-close throws away most of the day

Here is the objection that turns into a whole other page. Close-to-close uses exactly two prices out of every session and discards the high, the low and the open. A day that opened at $74, ran to $77, collapsed to $72 and closed at $74.05 registers as almost nothing.

That is a lot of information on the floor, and there are four well-known estimators that pick it up. On this same Nike window they disagree with each other by nearly six volatility points, which is larger than most of the edges anyone is trying to trade. That comparison has its own page, and it is the one that will change how you read a realized-volatility figure on any platform.

Common questions

Sources

• Parkinson (1980), The Journal of Business 53(1), 61 to 65
• OCC, Characteristics and Risks of Standardized Options

Rules and thresholds above were checked against these documents on August 4, 2026. Exchange and broker rules change. Confirm anything you are about to act on with your own broker.

Keep reading

• The forward-looking half of the pair
• The gap between them, and who gets paid for it
• Four estimators, six volatility points apart

Do the math on your own trade

• Implied volatility calculator solve for IV from a market price, and get the move it implies.
• IV rank calculator whether today is actually a high-premium day.
• Black-Scholes calculator price and Greeks from five inputs.

Every price in this article is an illustrative worked example, not a quote. Read Volatility for the rest of the series, and the disclaimer before you act on any of it. Selling options carries real risk of loss, and the loss can be far larger than the premium you collected.