Implied vs realized volatility: what IV actually is, and why it is not a forecast
Implied volatility is not a prediction the market makes. It is a price, expressed in volatility units, backed out of what options are trading for. Getting that backwards is the single most common error in options analysis.
Almost every explanation of implied volatility describes it as "the market's expectation of future movement." That description is close enough to be useful and wrong in a way that causes real errors.
Implied volatility is a price. It is what you get when you take the option's market price, put it into a pricing model, and solve backwards for the one input you did not already know.
Everything confusing about IV becomes simple once you have that the right way round.
Realized volatility: the measurable one
Realized volatility, also called historical volatility, is a plain statistic. Take a series of past returns, compute the standard deviation, annualise it.
Annualising is why the numbers look large. Volatility scales with the square root of time, so a daily standard deviation is multiplied by the square root of the number of trading days in a year:
`` annualised vol = daily standard deviation x sqrt(252) ``
A stock that typically moves 1% a day has an annualised realized volatility of about 16%. Nothing about that is a forecast. It is a description of what already happened, and it is completely determined by the data and your choice of lookback window.
That last clause matters. 30-day and 90-day realized volatility are different numbers about the same stock, and neither is more correct than the other.
Implied volatility: the one solved backwards
An option pricing model takes several inputs (spot price, strike, time to expiry, interest rate, volatility) and produces a theoretical price.
Every one of those is observable except volatility. So in practice the process runs in reverse. You observe the price the option is actually trading at, and you ask: what volatility number would make this model output that price?
That number is the implied volatility.
This is why IV is better understood as a quoting convention than a prediction. Traders express option prices in volatility units for the same reason bond traders quote yields instead of dollar prices: it strips out the mechanical effects of strike, spot and time, and leaves a number you can compare across contracts.
An option is not expensive because IV is high. IV is high because the option is expensive. The causation runs from price to IV, not the other way.
The consequence that trips people up
Since IV is a price, it behaves like one. It responds to supply and demand.
If a fund needs to buy a large quantity of downside protection, the price of those puts rises, and their implied volatility rises with it. Nothing has been predicted. Someone wanted insurance and paid for it.
This is also why IV rises ahead of scheduled events like earnings. It is not that the market foresees a big move. It is that nobody will sell you optionality across a known binary event at the same price they would sell it on a quiet Tuesday, so the price goes up, and the price expressed in volatility units goes up with it.
Comparing the two
The interesting question is the relationship between them: how does implied volatility compare to the realized volatility that subsequently occurs?
Two things are worth knowing.
They are measured over different periods. Implied volatility is forward-looking over the remaining life of the option. Realized volatility is backward-looking over whatever window you chose. Comparing "IV 30" against "30-day historical volatility" compares the next 30 days against the last 30, which is a legitimate comparison but not a like-for-like one.
On average, implied exceeds subsequent realized. This gap is persistent and well documented across markets, and it has a name: the variance risk premium. Options buyers, on average and over long periods, pay slightly more than the movement that follows justifies. That is what you would expect of an insurance market, and it is not a free lunch: the premium compensates the seller for taking a risk that occasionally arrives all at once.
Four practical cautions
IV is annualised even on a one-day option. A 0DTE contract quoted at 40% volatility is not going to move 40% today. The convention is annual, always, so you must scale down by the square root of the time remaining to get anything intuitive.
A single symbol's IV number is not one number. Different strikes and different expiries have different implied volatilities on the same underlying, which is the volatility surface. When a screen shows "IV: 32%" it has chosen a convention, usually something near at-the-money, and that choice is a summary that hides real structure.
Raw IV is not comparable across symbols. A biotech at 60% is calm. A utility at 60% is in crisis. Comparing the raw number across names tells you about the kind of company, not about whether options are currently cheap or dear. That is the entire reason IV rank and IV percentile exist.
Watch for scale. Some systems store implied volatility as a decimal fraction (0.32) and others as a percentage (32). Both conventions are common and both are defensible. Mixing them silently produces numbers that are wrong by a factor of one hundred while still looking plausible, which is worse than an error that crashes.
The summary worth keeping
Realized volatility is a measurement of the past. Implied volatility is a
price for the future, quoted in the units of a measurement.
The moment you treat IV as a forecast, you start reading supply and demand as prophecy. The moment you treat it as a price, questions like "is this expensive relative to its own history" become the obvious ones to ask.
Next: IV rank vs IV percentile, which is how you answer exactly that.