What is implied volatility?

An option pricing model, the standard one being Black-Scholes-Merton, is a function. It takes five things you can observe or agree on, the spot price of the underlying, the strike, the time left to expiry, the risk-free interest rate, and a volatility, and it returns one number: a theoretical price for the option. Feed it a volatility, receive a price. That is the model running forward, and it is how a textbook introduces it.

The live market hands you the opposite problem. Four of those five inputs are already known, and the fifth thing you can read off a screen is not the volatility but the price, the premium the option is actually trading at. Volatility is the single quantity nobody can observe directly. So you turn the function around: given the traded price and the four known inputs, what one volatility must go in to make the model print exactly that price? You invert the model and solve for it. That solved value is the implied volatility, the volatility the market's price implies. There is no tidy formula that isolates it, so software searches numerically, nudging the figure up or down until the model's price lands on the market price.

It is worth being precise about why only volatility has to be solved for. The spot price comes straight off the tape. The strike and the time to expiry are written into the contract. The interest rate is read from short-term money markets, and dividends, where they matter, come from a known schedule. Every one of those is either observed or agreed before the option trades. Volatility alone concerns the future and cannot be looked up anywhere, which is why it is the single input the market is forced to imply rather than supply, and why it carries all the information the others do not.

Picture the search and the point becomes concrete. The model's price rises smoothly and without kinks as you raise the volatility you put in, so recovering IV is simple homing in: guess a volatility, price the option, compare with the market price, adjust, repeat. Because the curve only ever climbs, every traded price corresponds to exactly one volatility, and the same call priced at 20, 30 and 40 percent volatility gives three different premiums. Read that relationship the other way and it is the whole argument in one line: hand the model a price and it returns a volatility. No step treats volatility as a known cause of the price. The price is the target; the volatility is the unknown being hunted.

To make that loop concrete, take the call in the figure, a 30-day at-the-money option on a stock at ₹1,000 that is trading at about ₹36.9. Guess 25 percent volatility and the model returns about ₹31.3, too cheap. Guess 35 percent and it returns about ₹42.6, too rich. Move back toward 30 percent and the model prints about ₹36.9, the traded price, so 30 percent is the implied volatility. A computer does this in a handful of iterations and to many decimal places, but the logic is exactly that: adjust the one unknown until the model agrees with the market, then read off the value you had to use.

This is why IV moves the way it does. When traders bid options up, premiums rise, and the volatility the model needs to justify those richer premiums rises with them, so IV climbs. When demand cools and premiums soften, the implied figure falls. Causation runs from premium to IV, not the other way, and every later section depends on holding that arrow the right way round. It also dissolves a puzzle: two options on the same stock can carry different IVs at the same instant, because IV is a property of each option's own price, not one dial on the underlying.

Once you accept that IV is solved for, its real nature falls into place, and so do the misreadings. It is a consensus: inverted out of the prices that thousands of buyers and sellers have agreed on, so it is not any one analyst's forecast but the number the whole market is currently willing to transact at. It is forward looking: it concerns the movement still to come over the option's remaining life, which is why a fresh news shock can reprice it in minutes even though nothing in the stock's past has changed. And it is symmetric: because more expected movement makes both a rise and a fall more likely, IV prices the magnitude of the move and stays silent on its direction. A high IV never says up or down; it says far.

One more property is quietly assumed everywhere the number appears: IV is quoted as an annualised percentage. A 30 percent IV is the size of move, as a standard deviation of returns, that the market is pricing over a year, not over the option's remaining life and not over a single day. Turning that annual figure into an expected move for a specific expiry is a separate scaling step, but the number on the screen is always the annual one. That is why the same 30 percent can look mild on a long-dated option and severe on one expiring this week, and why any comparison between two options begins by putting their IVs on the same annual footing before anything else is said.

Implied volatility is not a measurement of anything that has happened. It is a price, quoted in volatility units, for movement that has not happened yet.

The single most common error is to read a high IV as a prediction that a big move is coming, and therefore as a reason to buy. It is not a prediction and not a reason. It is the market telling you that movement is expensive to own right now. That is information about the cost of the trade, not about its outcome, and the two are easy to confuse precisely because the number is quoted as a percentage that looks like a forecast. Treat IV as the price tag on movement and the rest of this guide reads cleanly. Treat it as a crystal ball and you will consistently pay the most for options exactly when they are most likely to disappoint.

Because IV is solved out of the premium, the two move together, and that gives the level of IV a direct meaning for anyone taking a position. A higher IV widens the range of prices the underlying might reach by expiry, which raises the chance the option finishes with real value, which is worth more today. Wider expected range, richer premium. So the practical translation of a high IV is blunt: the option is expensive. You pay more to buy it and you receive more to sell it, for the same strike and the same expiry.

For a buyer, an expensive option raises the bar. The underlying now has to travel further, in the right direction, before the position even recovers its cost. The figure below prices the same 30-day call at three volatilities and marks where its breakeven lands. At a low IV the breakeven sits close to spot; at a high IV it is pushed well away, so the buyer has to be right by more to make the same money. The seller of that same high-IV option is on the other side of the trade: paid a fat premium up front, and profitable across a wider band of outcomes, in return for carrying the risk of the large move.

Read down the last column. Nothing about the stock has changed between the rows, not the spot, not the strike, not the time to expiry. The only thing that moved is the price the market puts on movement, and it quadrupled the size of the move a buyer needs, from under two percent to nearly six. This is the sense in which trading an option is trading its IV: the level you buy at sets the hurdle you must clear, entirely before the underlying does anything. It is also why the same directional view can be a fair trade at a low IV and a poor one at a high IV. The view did not change; the price of expressing it did. Options sit alongside futures as a way to take a position, but only the option makes you pay this separate, shifting price for volatility on top of your directional call.

Implied volatility has a backward-looking twin: realised volatility, also called historical volatility. Realised volatility is a fact. You take the underlying's past returns over some window and compute how much they actually varied, and it tells you what already happened. Implied volatility is an expectation inverted out of today's option prices, and it tells you what the market is currently paying for the movement it thinks lies ahead. The two answer different questions pointed in opposite directions in time, and the relationship between them is one of the most important facts in options.

Realised volatility itself is a plain computation: take the underlying's daily returns over a chosen window, say the past 20 or 30 sessions, measure their standard deviation, and scale that up to an annual figure so it can be set beside implied on the same footing. It is entirely backward looking and it updates slowly, one new session at a time. That steadiness is the whole point of the contrast. The realised number crawls, while the implied number, being a live price, can leap the instant the crowd's expectation of movement changes, which is why the two can diverge sharply for weeks at a time.

The gap between the two is not noise; it is a structural feature. Implied volatility has tended to sit above the volatility the same underlying later realises, and that persistent premium of implied over subsequent realised is the variance risk premium. The figure below plots two years of one stock's implied and realised volatility, and the priced-in line mostly rides above the actual one: an average of about 19 percent implied against about 15 percent realised, a gap near five volatility points a year.

That gap is the price of insurance. An option seller is writing cover against movement, and on average the buyer pays for more volatility than turns up, which is exactly why the seller is compensated for carrying the risk that, sometimes, far more volatility arrives than was priced. The single coral month in the figure is that tail: realised volatility spikes above implied, and the seller pays. This is the honest shape of the trade. Selling volatility earns a small, steady premium most of the time and suffers the occasional large loss; buying it pays that premium away most of the time in exchange for the rare large win. It is the same structure as any insurance and hedging arrangement, and it explains why systematically buying options tends to bleed. Which side of that premium a position sits on, and whether the price paid is fair for the movement expected, is exactly the upstream judgement that the method we teach is built to make deliberately rather than by accident.

The sensitivity of an option's price to implied volatility has a name. Vega is the change in an option's price for a one percentage-point change in IV, with everything else held fixed. If an option's vega is a certain number of rupees, a one-point rise in IV lifts its price by roughly that many rupees and a one-point fall lowers it by the same. Vega is largest for at-the-money options and for longer-dated ones, because those have the most time value exposed to shifting expectations, and it shrinks toward expiry as that time value drains away. Vega is the bridge between the abstract IV number and the concrete rupees in your position.

Set vega beside the other sensitivities and IV's peculiarity stands out. The table lists what moves a premium and the Greek that measures each force. Notice the highlighted row: a rise in IV lifts both a call and a put, whereas price and strike push calls and puts in opposite directions. That is the signature of a volatility input as opposed to a directional one, and it is why a buyer of options is long volatility and a seller is short volatility, entirely apart from any view on which way the underlying goes.

It helps to put a rupee number on vega. For the near-dated call we will follow through an event two sections from here, vega is around ₹0.55 for every one point of implied volatility near the money. So a 28-point collapse, from 54 percent down to 26 percent, is worth roughly ₹15 off the premium on that count alone, which is essentially the ₹14.8 vega loss the later decomposition books. Vega is not an abstraction; it is the exchange rate between the IV number and the rupees in your position, and around scheduled events that exchange rate does most of the damage while delta, the directional term, is left looking small.

Vega is the term that quietly dominates around events, and it is the mechanism behind the two behaviours that fill the rest of this guide. When IV mean-reverts, it is vega that turns the falling IV into a falling premium. When IV crushes after an event, it is vega that converts the collapse into a loss on the option even when the underlying has moved your way. Hold on to the highlighted row and the next two sections are just its consequences.

A stock price can trend a long way in one direction and keep going. Implied volatility does not behave like that. It is bounded and regime-like: it rests near a floor during calm stretches, spikes when fear arrives, and then decays back toward a long-run level as the calm returns. Extreme highs and extreme lows both tend to pull back over time. The figure below shows one stock's IV over two years, resting near a 14 to 16 percent floor and jumping to 34 and then 46 percent at two fear events before subsiding each time.

This mean-reverting shape is why a raw IV number means almost nothing on its own. Is an implied volatility of 18 percent high or low? The only honest answer is: for which underlying, and compared with what? A steady large-cap index and a jumpy small-cap live in entirely different volatility neighbourhoods, so the figure has to be read against the same underlying's own history, usually the trailing year. Two measures do this. IV rank asks where today's IV sits between its lowest and highest readings over the window, from zero at the bottom to a hundred at the top. IV percentile asks a different question: on what share of days over the window was IV below today's level? The difference bites after one violent spike, which stretches the high and can pin IV rank low for months, while IV percentile, weighing every day equally, barely notices it and is usually the steadier gauge.

There is a second regularity underneath the mean reversion, and it is worth naming: volatility clusters. Calm days tend to follow calm days and violent days follow violent ones, so IV does not flicker at random around its mean but moves in persistent regimes, a quiet stretch that can run for months and then a burst that takes weeks to subside. Mean reversion and clustering are not in tension. Together they say that IV will return to its long-run level, but not at once, and that a spike is both likely to fade and likely to stay elevated for a while first. For a position that is short volatility, that pairing is the whole risk: the reversion is the edge, the persistence is the drawdown you have to survive to collect it.

The practical warning writes itself. Because IV mean-reverts, buying options when IV is already near the top of its range often means buying near a peak that is likely to fall, and the falling IV then works against the position through vega even if the direction is right. In the figure, the worst moment to have paid up for volatility is the 46 percent panic peak, an IV rank near a hundred, where the only way left for volatility to travel is down. The best directional idea in the world is a poor options trade if it is expressed by buying volatility at its most expensive. Reading the level in its own range is how you avoid doing exactly that.

The most expensive lesson in implied volatility waits around a scheduled event: a results announcement, a central-bank decision, a policy or data release with a known date. Before the event nobody knows the outcome, so uncertainty is at its peak and the market pays up for options that straddle it. IV rises into the date and premiums swell, and much of that swelling is pure expectation of the move to come, not the move itself. Then the news breaks. The instant the outcome is known, the uncertainty that justified the rich premiums evaporates, IV collapses, and because vega links IV to price, every option on that underlying sheds extrinsic value at once. This is the IV crush, and its cruelty is that it is indifferent to whether you called the direction correctly.

The effect bites hardest on exactly the options a retail buyer reaches for first: short-dated, at-the-money contracts on a single stock going into its results. Those carry the most vega relative to their price and the least intrinsic cushion, so they inflate the most as IV climbs before the event and deflate the most as it collapses after. A longer-dated option feels a gentler crush, because much of its value is time rather than event premium, and a deep in-the-money option is anchored by intrinsic value the crush cannot reach. The cheap-looking weekly at-the-money call is the very instrument the crush is built to punish.

Follow the buyer in the figure. They buy the call the session before results, when IV has already climbed to 54 percent, paying about ₹30.4. Results come in and the stock rises 1.5 percent, which is the right direction for a call, so on a naive reading the trade should be a winner. It is not. IV crushes from 54 percent back to 26 percent, and the option is worth about ₹20.5 the next session, a loss of roughly a third. The decomposition below shows why, rupee by rupee: the favourable move was genuinely worth about ₹8.4, but the collapse in IV handed back about ₹14.8 through vega, and a few sessions of time decay took another ₹3.5. The direction was right and the trade still lost, because the position was long volatility into the one event guaranteed to destroy it.

This is not a fringe outcome, and in India it is a mass one. Per SEBI's study of the equity derivatives segment, about 93% of individual traders in equity derivatives made net losses over FY22 to FY24, aggregate net losses exceeding ₹1.8 lakh crore (SEBI, September 2024). No single figure has one cause, but a large share of buy-side option losses share this exact shape: a directional bet placed by buying a rich, high-IV option into an event, undone by the crush and the decay rather than by the direction. The lesson is not that IV crush is avoidable. It is that the IV level is part of the trade, and ignoring it is how a correct view becomes a losing position.

Pull the threads together and implied volatility stops being a technicality and becomes the thing you are actually trading. It is solved out of the price, so it is a consensus rather than a forecast. It is a price, so its level sets the hurdle a buyer must clear and the premium a seller collects. It mean-reverts, so its level only means something against its own range. It usually exceeds realised volatility, so buyers pay a premium that sellers earn. And it feeds the option through vega, so a change in IV is a change in your position whether or not the underlying moves. Every one of those is a statement about the same number read correctly.

So the honest close is the one that keeps you out of trouble. When you buy or sell an option you take two positions at once, one on direction and one on volatility, and the second is the one beginners ignore. A high IV is not a signal to buy and not a promise of a move; it is a statement that movement is expensive right now, which cuts against buyers and toward sellers before either has taken a view on direction. Read IV as the cost of the thing rather than a prediction of the thing, check where that cost sits in its own range before you pay it, and the most common way that directionally-right traders still lose money simply stops happening to you. The index-level version of all this, and how an annualised figure converts into an expected move, is the subject of the separate India VIX guide; the way IV varies across strikes at one expiry, the volatility smile, is covered in reading an option chain.

• The pricing model that IV inverts. The Black-Scholes-Merton framework (Black and Scholes, 1973; Merton, 1973) defines the forward map from spot, strike, time, rate and volatility to an option price; implied volatility is the volatility that map requires to reproduce the observed market price, found numerically because no elementary closed form isolates it. The premium, breakeven and IV-crush figures on this page are computed from this standard model and labelled illustrative.
• The variance risk premium. Implied volatility has historically sat above the volatility the same underlying later realises, on average, which is the compensation option sellers earn for bearing movement risk. The two-year series shown is illustrative and authored to reflect this well-documented average relationship, including the tail months in which realised volatility exceeds implied.
• SEBI on individual derivatives traders. About 93% of individual traders in equity derivatives made net losses over FY22 to FY24, with aggregate net losses exceeding ₹1.8 lakh crore (SEBI, Analysis of Profit and Loss of Individual Traders Dealing in Equity Futures and Options Segment, September 2024). As of 17 July 2026; verify at source. sebi.gov.in
• India VIX. The NSE's volatility index, introduced in 2008 and adapted from the CBOE VIX methodology, expresses the expected 30-day annualised volatility of the Nifty 50, computed from the best bid and ask quotes of near and next-month Nifty option contracts. As of 17 July 2026; verify at source. nseindia.com
• Vega, IV rank and percentile, and IV crush. Vega is the change in option price per one percentage-point change in IV, largest for at-the-money and longer-dated options; IV rank and percentile are standard context measures over a trailing window; IV crush is the well-documented post-event collapse in IV, in which a vega loss can exceed a favourable directional move. These reflect standard options-market definitions.