Term Structure Advanced The extra variance a scheduled event injects Forward-looking

Event Premium

The honest way to answer how much the market is pricing for Budget day.

Quick answer: Event premium is the extra variance that a single scheduled event — an RBI decision, the Union Budget, election counting — injects into an option's total implied variance, and separating it out lets you back out the one-day move the option market is actually pricing for that event day.

In simple words

An option that spans a scheduled event costs more than an otherwise identical option that does not, because one of the days it has to survive is not an ordinary day. Think of the option's total priced movement as a sum of daily pieces. Most days contribute an ordinary amount; the event day contributes an extraordinary amount. The event premium is that extra chunk — the difference between what the option charges and what it would charge if every day were ordinary. Once you isolate it, you can translate it into a plain-English number: the size of the move the market is pricing for the event day itself.

The reason this is worth doing is that it answers a question everyone asks badly. People eyeball a high implied volatility before the Budget and say the market is nervous. The event decomposition replaces that vague feeling with an actual number — for example, the market is pricing roughly a 1.4% move in NIFTY on the event day. That is a checkable, tradeable statement, and it comes straight from comparing two option prices you can see on the screen.

Not to be confused with: IV crush, which is what happens to the event premium after the event. The event premium is the extra volatility priced in before the event is known; IV crush is that same premium collapsing to zero the instant the outcome is public. They are two views of one thing — the premium is the charge, the crush is the refund of the part that was never used because the uncertainty is now resolved.

The picture

The event lifts one expiry, then decays out of the rest

At-the-money implied volatility by expiry around a scheduled RBI decision, spot 24,000.

The first expiry that closes after the event carries the full event variance and prints a pronounced bump; expiries that reach further past it dilute the same fixed chunk of variance over more ordinary days, so the bump decays with tenor. This proves the event is not a level shift you can interpolate away — it is a spike in one expiry that longer expiries average down.

Professional explanation

Variance is additive, which is what makes the decomposition legal

The whole method rests on one property: variance adds across independent time periods, while volatility does not. An option's total implied variance over its life is the sum of the variance expected on each day it spans. That means you can split the total into a part contributed by the ordinary, diffusive days and a part contributed by the one scheduled event day. Volatility — the square root of variance — cannot be split this way, which is exactly why traders who reason in volatility rather than variance get event pricing wrong. You cannot subtract a clean expiry's volatility from a dirty expiry's volatility and call the difference the event. You have to move into variance, subtract there, and only then take a square root. The event decomposition is not a heuristic; it is the direct consequence of variance being additive and volatility not being additive.

The arithmetic, symbol by symbol

Write the total implied variance of the event-containing expiry as σ²_total × T, where T is its life in years. Split its T days into (T − t_e) ordinary days and one event day of length t_e = 1/365. The ordinary days are assumed to behave like a clean expiry with no event in it, at diffusive volatility σ_diffusive. Then σ²_total × T = σ²_diffusive × (T − t_e) + σ²_event × t_e. Everything here is observable except σ_event: σ_total is the dirty expiry's implied volatility, σ_diffusive is the clean expiry's, T and t_e are calendar arithmetic. Rearranged, σ²_event × t_e = σ²_total × T − σ²_diffusive × (T − t_e). The left side is precisely the event-day variance, and its square root is the one-standard-deviation move the market is pricing for the event day. This is the calculation, and it is genuinely under-taught.

A spike in one expiry, not a bump you can interpolate away

The event variance is a fixed chunk. The first expiry that closes after the event absorbs all of it, so that expiry's implied volatility jumps. An expiry that reaches further past the event still contains the same fixed chunk of event variance, but now spread over more ordinary days, so per-day it is diluted and the expiry's overall implied volatility is lifted less. Push the expiry far enough out and the event is a rounding error. This is why the term structure around an event is a spike that decays with tenor, not a smooth bump you can draw a line through. And it is why interpolating implied volatility across the event — treating the elevated dirty expiry as if it lay on a smooth curve — misprices everything: the event does not belong to the curve, it belongs to one point on it.

The number you back out is a lower bound — the uncomfortable part

Here is the sentence a marketing department would cut. The event-day move you extract is not the whole story, and it is biased low. The decomposition assumes the diffusive volatility of the ordinary days is unchanged by the event. In reality, a big event often raises volatility for days or weeks afterwards — an RBI surprise or a shock Budget does not just move the market once and then return to calm, it changes the diffusive regime that follows. If the ordinary days after the event are actually more volatile than the clean pre-event expiry you used for σ_diffusive, then part of what you attributed to the single event day was really elevated diffusive volatility, and the true event-day move is smaller — or, read the other way, your σ_diffusive was too low and the event-day figure you computed is a floor, not a point estimate. Either way, the honest label on the output is at least this much, not exactly this much.

Formula

Decomposing implied variance into diffusive and event components

σ²_total × T = σ²_diffusive × (T − t_e) + σ²_event × t_e

Rearranged to solve for the event: σ²_event × t_e = σ²_total × T − σ²_diffusive × (T − t_e). The left-hand side is the event-day variance; its square root is the one-standard-deviation move the market prices for the event day. Because variance is additive over time and volatility is not, the subtraction must be done in variance and the square root taken last.

σ_totalImplied volatility of the expiry that CONTAINS the event (the dirty expiry), annualised as a decimal (0.146 = 14.6%).
σ_diffusiveThe ordinary, event-free volatility of the non-event days, taken from a clean expiry just before the event (the diffusive component).
σ_eventThe annualised volatility attributed to the single event day — the unknown being solved for.
TTime to expiry of the dirty expiry in years, calendar days ÷ 365.
t_eThe length of the event in years — one calendar day, so t_e = 1/365 ≈ 0.00274.

The one-day event move you actually report

EventMove_1day = √( σ²_event × t_e ) = √( σ²_total × T − σ²_diffusive × (T − t_e) )

The event-day one-standard-deviation move as a fraction of spot. It is the square root of the event-day variance directly, so it does not require you to compute σ_event first. Multiply by spot to express it in points: on NIFTY at 24,000, a 1.40% event move is about 335 points.

How to back out the implied event-day move

Identify the first expiry whose life contains the event (the dirty expiry) and read its at-the-money implied volatility — this is σ_total.
Find a clean expiry just before the event with no scheduled event in its life, and read its at-the-money implied volatility — this is σ_diffusive.
Convert both to calendar-day arithmetic: T is the dirty expiry's days over 365, and t_e = 1/365 for a single event day.
Compute total variance σ²_total × T and diffusive variance σ²_diffusive × (T − t_e); subtract the second from the first to get the event-day variance.
Take the square root of the event-day variance — that is the one-standard-deviation move the market prices for the event day, as a fraction of spot.
Multiply by spot to state the move in points, and, if you want it, divide the event variance by t_e and take the root to express σ_event as an annualised volatility.
Label the result a lower bound, and check it against the clean-expiry assumption: if the event likely raises volatility afterwards, your diffusive input was too low and the true event move is smaller.

Practical example

NIFTY worked example

NIFTY is at 24,000. A weekly expiry seven days out has an RBI MPC decision inside its life and its at-the-money implied volatility prints 14.6%; the clean weekly just before the meeting, with no event in it, prints 11.4%. Treat the dirty expiry as six ordinary days at the clean 11.4% plus one event day. In variance-days (the 1/365 factors cancel): total is 0.146² × 7 = 0.149212, diffusive is 0.114² × 6 = 0.077976, so the event-day variance-days figure is 0.149212 − 0.077976 = 0.071236. The event-day one-standard-deviation move is √(0.071236 / 365) = √0.00019517 ≈ 0.01397, or about 1.40% — roughly 335 NIFTY points. Expressed as an annualised volatility, σ_event = √(0.071236) ≈ 26.7%. So the option market is pricing a move of about 1.4% in NIFTY on RBI day, far larger than an ordinary session, and that number came straight from two implied volatilities you can read off the chain. Remember it is a lower bound: if the RBI decision also lifts volatility in the days after, the clean 11.4% understated the diffusive part and the true event move is a touch smaller.

BANKNIFTY worked example

BANKNIFTY makes the same event feel larger, and shows why. With BANKNIFTY at 52,000, an RBI decision sits inside a seven-day expiry printing 18.2% while the clean weekly before prints 14.0%. In variance-days: total 0.182² × 7 = 0.231868, diffusive 0.140² × 6 = 0.117600, event 0.114268, so the event-day move is √(0.114268 / 365) ≈ √0.000313 ≈ 0.01769, about 1.77% — roughly 920 BANKNIFTY points. The event move is bigger than NIFTY's not because the RBI decision is a different event but because BANKNIFTY is a leveraged bet on lenders, and a rate decision hits their earnings directly. The lesson is that the same event injects a different amount of variance into different underlyings, and the decomposition quantifies exactly how much more — you cannot assume the NIFTY event move and the BANKNIFTY event move scale with anything simple like their overall volatility ratio.

Lot sizes used above (NIFTY 75, BANKNIFTY 30) are those in force at the time of writing; NSE revises them periodically. Figures exclude brokerage, STT, exchange charges, stamp duty and GST. Examples are teaching scenarios built on round numbers — they are not historical quotes, not backtests and not trade calls.

Risk note. The event decomposition is often used to justify selling the event: the market is pricing a 1.4% move, I think it will be less, so I sell the straddle and collect the premium. The decomposition is honest arithmetic, but this use of it is not risk-free. The backed-out move is a lower bound on the event-day risk, the event day can print far more than the priced move, and a short straddle over an RBI decision or the Budget carries theoretically unlimited loss if the surprise is large. Knowing the priced move tells you the terms of the bet, not whether the bet is good.

Advantages & limitations

What it is good for

It converts a vague sense that implied volatility is high before an event into a specific, checkable number — the one-day move the market is pricing for the event day.
It uses only observable inputs: two at-the-money implied volatilities from the chain and simple calendar arithmetic, with no model beyond the additivity of variance.
It localises the event correctly. By assigning the event variance to a single day rather than smearing it across the expiry, it explains the spike-and-decay shape of the term structure around events.
It is directly comparable across events and underlyings. The same procedure on RBI, Budget or election expiries, or on NIFTY versus BANKNIFTY, produces numbers you can line up against each other.
It exposes when an expiry is rich for a good reason. An elevated dirty-expiry implied volatility is not a mispricing to fade if the event variance justifies it, and the decomposition shows how much of the richness the event explains.

Where it breaks down

It assumes the diffusive volatility of the ordinary days is unchanged by the event, so if the event raises volatility afterwards the backed-out event move is a lower bound rather than a point estimate.
It depends on a clean expiry existing just before the event. If every nearby expiry contains some scheduled item, there is no uncontaminated σ_diffusive to subtract, and the decomposition has no clean baseline.
It treats the event as a single calendar day, which breaks down for multi-day events like election counting or a Budget session whose effects bleed across sessions.
It uses at-the-money implied volatility and so ignores the skew: an event that is priced mostly as downside risk shows up differently across strikes than a symmetric one, and the single-number move hides that.
It is only as clean as the two quotes it uses. Near expiry or on an illiquid weekly, the implied volatilities feeding the subtraction are noisy, and small errors in each are magnified when you difference their variances.

Common mistakes

Subtracting the clean expiry's volatility from the dirty expiry's volatility and calling the difference the event. Volatility is not additive; you must subtract in variance and take the square root last, or the number is simply wrong.
Interpolating implied volatility across the event as if the dirty expiry lay on a smooth curve. The event belongs to one point, not the curve, and interpolating misprices every expiry around it.
Reporting the backed-out event move as an exact figure. It is a lower bound, because the method assumes the post-event diffusive volatility equals the pre-event clean expiry, which events routinely violate.
Assuming a longer expiry that also contains the event prices the same event move. The longer expiry dilutes the fixed event variance over more ordinary days, so its implied volatility is lifted less — the event has not shrunk, the average has.
Selling the event straddle because the priced move looks large, treating the decomposition as a signal that the market is wrong. The arithmetic tells you the terms of the bet; it says nothing about whether the realised move will be smaller, and the loss is unbounded.
Applying a NIFTY event move to BANKNIFTY. The same event injects different variance into different underlyings, and the decomposition must be run separately on each to get the right number.

Professional usage

Event desks and volatility arbitrage traders run this decomposition as standard pre-event work. Before an RBI decision, the Budget or election counting, they take the first expiry after the event and the clean expiry before it, difference the total variances, and back out the implied event-day move — then compare it against their own estimate of how large the move is likely to be, and structure the trade around the difference rather than around the raw implied volatility. Because the method isolates the event from the diffusive background, it lets a desk separate two questions that beginners conflate: is background volatility rich, and is the event specifically rich? A desk can be short the event and long the diffusive background at the same time, and the decomposition is what makes that distinction tradeable.

Market makers use the same arithmetic in reverse to price expiries consistently across an event. Rather than quoting each expiry's implied volatility independently and hoping they hang together, they fit a diffusive term structure and add a fixed event-variance bump to every expiry that spans the event, which automatically produces the spike-and-decay shape and keeps the expiries arbitrage-free relative to each other. Risk managers, meanwhile, read the backed-out event move as the market's own stress scenario for the event day and feed it into margin and scenario models — it is a live, market-implied estimate of the event's one-day risk, which is more responsive than any historical average of past RBI or Budget days.

Key takeaways

Event premium is the extra variance a single scheduled event injects into an option's life, and it can be isolated because variance is additive over time while volatility is not.
The decomposition σ²_total·T = σ²_diffusive·(T − t_e) + σ²_event·t_e, with t_e = 1/365, lets you back out the one-day move the market is pricing for the event day from two observable implied volatilities.
In the worked NIFTY example — a dirty expiry at 14.6% and a clean expiry at 11.4% — the implied event-day move is about 1.40%, roughly 335 points on NIFTY at 24,000.
The event is a spike in the first expiry after it that decays with tenor, not a bump you can interpolate away, because longer expiries dilute the same fixed event variance over more ordinary days.
The backed-out move is a lower bound: if the event also raises diffusive volatility afterwards, the clean expiry understated the background and the true event move is smaller.

The event premium turns a feeling into a number. Instead of saying implied volatility looks high before the Budget, you take the dirty expiry and the clean one, subtract in variance, and read off that the market is pricing roughly a 1.4% move on the day — a statement you can check, compare and trade against. Do the subtraction in variance, not volatility; assign the event to one point, not a curve; and label the answer a lower bound, because the day that surprises the market also tends to leave it more volatile than the calm expiry you measured against.

Frequently asked questions

What is event premium in options?

Event premium is the extra variance a scheduled event — an RBI decision, the Budget, election counting — adds to an option's total implied variance beyond what ordinary days contribute. Isolating it lets you back out the one-day move the market is pricing specifically for the event day.

How do I calculate the implied move for an event day?

Take the total variance of the expiry that contains the event and subtract the diffusive variance of a clean expiry before it: σ²_event·t_e = σ²_total·T − σ²_diffusive·(T − t_e). The square root of the event-day variance is the one-standard-deviation move the market prices for the event day.

Why must the event decomposition be done in variance, not volatility?

Because variance is additive across independent time periods and volatility is not. You can add the ordinary-day variance and the event-day variance to get the total, but you cannot add or subtract volatilities directly. Subtract in variance, then take the square root last.

What is a worked NIFTY event premium example?

With NIFTY at 24,000, a dirty weekly containing an RBI decision at 14.6% and a clean weekly at 11.4%, the event-day variance in variance-days is 0.146²·7 − 0.114²·6 = 0.071236, so the event move is √(0.071236/365) ≈ 1.40%, about 335 points.

What one-day move does that NIFTY example imply?

About 1.40% of spot, which on NIFTY at 24,000 is roughly 335 points. That is the one-standard-deviation move the option market is pricing for the RBI event day — far larger than an ordinary session, and read straight off two implied volatilities.

Why is the backed-out event move a lower bound?

Because the method assumes the diffusive volatility of the ordinary days equals a clean pre-event expiry. If the event raises volatility afterwards, that clean number understated the background, so part of what you attributed to the event was elevated diffusive volatility, and the true event move is smaller.

Why does the first expiry after an event absorb all its variance?

Because the event variance is a fixed chunk tied to a single day, and the first expiry to close after that day is the shortest one that fully contains it. That expiry carries the whole event variance over the fewest ordinary days, so its implied volatility jumps the most.

Why do longer expiries show a smaller event bump?

Because they spread the same fixed event variance over more ordinary days. The event has not shrunk; the per-day average has, so a three-month expiry containing the event is lifted far less than the weekly that closes just after it. That is why the bump decays with tenor.

Can I interpolate implied volatility across an event?

No. The event belongs to one point on the term structure, not to a smooth curve through it. Interpolating treats the elevated dirty expiry as if it lay on a line, which misprices the event and every expiry around it.

What is t_e in the event premium formula?

t_e is the length of the event in years. For a single event day it is one calendar day, so t_e = 1/365 ≈ 0.00274. It is the fraction of the year over which the event-day variance is charged.

What is σ_diffusive in the decomposition?

σ_diffusive is the ordinary, event-free volatility of the non-event days, taken from a clean expiry just before the event. It represents what the market would charge if every day in the option's life were an ordinary session, and it is the baseline the event variance is measured against.

How is event premium different from IV crush?

Event premium is the extra volatility priced in before the event is known; IV crush is that premium collapsing to zero the moment the outcome is public. They are two views of one thing — the charge before, and the refund of the unused part after the uncertainty is resolved.

Does the same event inject the same move into NIFTY and BANKNIFTY?

No. The same RBI decision produces a larger implied event move on BANKNIFTY than on NIFTY, because BANKNIFTY is a leveraged bet on lenders whose earnings react directly to rate decisions. Run the decomposition separately on each underlying to get the right number.

What events matter most for Indian index options?

The RBI Monetary Policy Committee decisions, the Union Budget, and election counting days are the main scheduled events on the Indian calendar. Each sits inside specific expiries and injects event variance into them, which the decomposition can isolate and price.

Can I use the event premium to decide whether to sell the straddle?

It tells you the terms of the bet — the move the market is pricing — but not whether the realised move will be smaller. Selling the event straddle because the priced move looks large treats the arithmetic as a signal it is not, and a short straddle over an event carries unbounded loss.

Why does event premium require a clean expiry to exist?

Because you need an uncontaminated σ_diffusive to subtract. If every nearby expiry contains some scheduled item, there is no event-free baseline, and the decomposition has nothing clean to measure the event variance against.

How do I express the event as an annualised volatility?

Divide the event-day variance by t_e and take the square root: σ_event = √(event-day variance / t_e). In the NIFTY example that is √(0.071236) ≈ 26.7% annualised, which is the diffusion-equivalent volatility of the single event day.

Does the event premium capture skew?

No. The decomposition uses at-the-money implied volatility, so it produces a symmetric one-day move and ignores the skew. An event priced mostly as downside risk shows up across strikes in a way the single at-the-money number cannot reveal.

Why are noisy quotes a problem for the decomposition?

Because you difference two variances, and small errors in each implied volatility are magnified in the difference. Near expiry or on an illiquid weekly the inputs are noisy, so the backed-out event move can be unreliable even though the arithmetic is exact.

What does an annualised event volatility of 26.7% mean intuitively?

It means the single event day carries as much variance as a day would under a market realising 26.7% annualised volatility — more than double the ordinary 11.4% diffusive level. It is a way of stating how much more violent the market expects the event day to be than a normal session.

Is the event premium a market inefficiency?

No. It is the option correctly charging more for an expiry that must survive an extraordinary day. The premium reflects genuine event risk, and its collapse after the event is not a mispricing being corrected but a risk that has simply ceased to exist.

Voice search & related questions

Natural-language questions people ask about event premium.

How much is the market pricing for Budget day?

You back it out by comparing two expiries. Take the first expiry after the Budget and a clean expiry before it, subtract their total variances, and the square root of the difference is the one-day move being priced. On the RBI example on this page it comes to about 1.4% for NIFTY — a real number, not a hunch.

Why can't I just subtract the two implied volatilities?

Because volatilities do not subtract — variances do. The dirty expiry's 14.6% minus the clean 11.4% is 3.2%, and that number means nothing. You have to square each into variance, weight by days, subtract there, and take the square root at the end. Do it in volatility and you simply get the wrong answer.

Why does the near expiry jump so much before RBI?

Because it is the shortest expiry that still has to survive the whole event day, so it packs the entire event variance into very few days. The event is a fixed chunk of extra variance, and the near expiry carries all of it over the least time, which is why its implied volatility spikes while longer ones barely move.

Is the event move I calculate exact?

No — treat it as a floor. The calculation assumes the days after the event are as calm as the clean expiry you used, but big events often leave the market more volatile for a while. If that happens, some of what you blamed on the event day was really elevated background volatility, so the true event-day move is a bit smaller.

Why is the BANKNIFTY event move bigger than NIFTY's?

Because BANKNIFTY is a concentrated bet on lenders, and an RBI rate decision hits their earnings directly. The same event injects more variance into BANKNIFTY than into the broader NIFTY, so when you run the decomposition on each, BANKNIFTY's implied event-day move comes out larger. You have to compute each separately.

What happens to the event premium right after the announcement?

It vanishes — that is IV crush. The moment the RBI decision or the Budget is public, the uncertainty the option was charging for no longer exists, so the event portion of the implied volatility collapses to zero. The premium was the charge before; the crush is the refund of the part that was never needed.

Sources & references

Last reviewed 10 July 2026. Educational content only — not investment advice.