Implied volatility is a price, not a forecast
IV tells you what the market is charging for uncertainty today. Treating it as a prediction of what will happen is the most common misreading in options.
Implied volatility has an intimidating name and a simple definition.
Take an option pricing model. Feed it everything you can observe — the underlying, the strike, the time to expiry, rates. One input is left: how much the underlying is expected to move. Now adjust that one input until the model's output equals the price the option is actually trading at.
The number you arrive at is the implied volatility. It is the market's price of uncertainty, expressed as an annualised percentage. It is not a forecast, not a measurement, and not anybody's opinion in particular. It is what you are being charged.
That distinction does all the work in this chapter. A forecast can be right or wrong. A price can only be expensive or cheap — and whether it is expensive is a question you answer by comparison, not by looking at the number.
Implied against realised
There are two volatilities and traders constantly conflate them.
Realised volatility is how much the underlying actually moved, computed from past prices. A measurement, available only after the fact.
Implied volatility is how much the market expects it to move, extracted from option prices. A price, available right now.
The gap between them is where option sellers make their living and where option buyers get their bills. And the two behave completely differently around an event, in a way worth looking at directly.
This is the mechanism behind the losing trade in the previous chapter, drawn as a picture. And it explains the apparent paradox of an option losing value on the most volatile day of the month: realised volatility was high, but you do not get paid for realised volatility. You get paid for changes in the price of expected volatility, and that price fell.
How to tell whether IV is expensive
Never by its absolute level. An implied volatility of 22 is high for one instrument and low for another, high in one month and ordinary in the next.
Three comparisons that do work:
Against its own recent range. Where does today's implied volatility sit within the last six months of its own readings? Near the top of its range, or near the bottom? This is the single most useful question, and it is the one that survives across instruments, because each instrument is compared only to itself.
Against realised volatility over a comparable window. If implied is 22 and the underlying has actually been moving at 14, you are being asked to pay for considerably more movement than has lately occurred. That may be justified — something may be coming — but you now know what you are paying for and can ask whether you believe it.
Against the move it implies in rupees. This is the comparison that makes the abstraction concrete. Convert the implied volatility into an expected move over your holding period and ask whether you think the instrument will move more than that. If implied volatility says the index will move roughly 240 points this week and you think it will move 400, you have a reason to buy. If you merely think it will go up, you do not.
That third comparison is the one that turns "I have a view" into "I have a trade", and it is why the implied move — not the implied volatility — is the number worth having on screen.
Why "sell high IV" is not a strategy
The most repeated piece of options advice is to sell when implied volatility is high. There is something real behind it and it is incomplete in a way that costs money.
The real part: implied volatility has tended, over long periods and across markets, to sit above the volatility that subsequently materialises. Sellers of options are providing insurance, and insurance is generally sold above its expected cost — otherwise nobody would provide it. That tendency is the structural tailwind behind option selling.
The incomplete part: high implied volatility is often correctly high. It is elevated because something genuinely uncertain is happening, and sometimes the uncertainty resolves violently in the direction you are short. The premium exists to compensate for exactly that, and it is paid out on a schedule nobody controls — many quiet periods of small gains, punctuated by losses that are large enough to erase several of them.
So the advice is not wrong, it is a description of a payoff shape rather than a strategy. Selling high volatility is a trade with a positive mean, a high hit rate and a fat left tail. Whether that is a good trade for you depends entirely on whether your sizing can survive the tail — which is the reason the risk module comes before this one, and why its final chapter is about ruin rather than about returns.
The skew, briefly
If you look at implied volatility across strikes at the same expiry, it is not flat. Typically, out-of-the-money puts carry higher implied volatility than equivalently distant calls.
This is not an inefficiency and you should not trade it as one. It reflects a real asymmetry in how markets move and in what market participants need to hedge: large downward moves happen faster than large upward ones, and there is persistent demand for downside protection from people who hold the underlying. The skew is the price of that demand.
The practical use is as a reference point. A skew that is much steeper than usual tells you downside protection has become expensive — which is information about positioning and nervousness. A skew that has flattened tells you it has become cheap. Both are more informative than the level of implied volatility on its own, because they describe what the market is worried about rather than merely how much.
When an IV reading is not a reading
One caution that matters more on illiquid strikes than anything else in this chapter.
Implied volatility is computed from a price. If that price is stale — a last-traded value from forty minutes ago on a strike nobody has touched — then the implied volatility derived from it is not a reading of the market. It is an arithmetic consequence of an old number, carrying a precision it has not earned.
Worse is a strike with no trade at all. A displayed value of zero is not a volatility of zero; it is the absence of a measurement, and software that treats it as a number will produce greeks, probabilities and strategy recommendations built on nothing. The honest treatment of an untraded strike is to price it from the midpoint of its bid and ask if a two-sided quote exists, and to mark it unavailable if it does not — never to pass a zero down the chain as though it were data.
Before you act on any volatility reading from a far strike, check that something traded. The number is only as good as the transaction underneath it.
What to take from this chapter
- Implied volatility is the price of uncertainty, not a forecast. Prices are expensive or cheap; forecasts are right or wrong.
- Implied peaks before an event and collapses on resolution. Realised peaks after. Being out of phase is the trap.
- Judge IV against its own recent range, against realised volatility, and above all against the rupee move it implies.
- "Will it move more than the implied move?" is a trade. "I think it will go up" is not.
- Option selling has a real structural tailwind and a fat left tail. It is a payoff shape, not a strategy — whether it suits you is a sizing question.
- The put skew is the price of persistent demand for downside protection, not an inefficiency. Its changes are the information.
- An IV computed from a stale or absent price is not a reading. Zero is not a volatility; it is a missing measurement.