Term Structure Intermediate IV across listed expiries Forward-looking

Expiry Structure

The market lists expiries; it does not list the smooth curve everyone draws through them.

Quick answer: Expiry structure is the implied volatility read off each individual listed expiry of an index, and on NIFTY it forms a saw-tooth — every weekly and the monthly carrying its own event content — rather than the smooth upward curve that interpolation pretends exists between them.

In simple words

An option chain does not give you one expiry. NIFTY lists a weekly expiry every Thursday plus a monthly, and each of those expiries has its own at-the-money implied volatility. If you write those numbers down in order — this Thursday 12.9%, next Thursday 13.4%, the monthly 13.8%, the following month 13.6% — they do not lie on a smooth line. They zig-zag. Each expiry is priced for the specific calendar days it contains, and different weeks contain different amounts of scheduled news, so their implied volatilities step up and down rather than sliding smoothly.

The mistake almost everyone makes is to draw a clean curve through those dots and read a value off it for some date in between. There is no option expiring on that in-between date, so the value you read off is a fiction. Worse, it is a fiction that systematically underprices the expiries that happen to straddle an RBI meeting or the Budget, because a smooth curve averages the event away instead of parking it on the one expiry that actually carries it.

Not to be confused with: The volatility term structure as a textbook object — a smooth function of continuous time to expiry. That idealised curve is a modelling convenience. The expiry structure is what the market actually quotes: a finite set of dated points, unevenly spaced, each with its own event load. India VIX interpolates to a constant 30 days precisely because no NIFTY expiry sits exactly 30 days out on most days.

The picture

The listed expiries do not lie on a line

At-the-money implied volatility of each successive NIFTY weekly and monthly expiry, spot 24,000.

The points step up and down instead of curving smoothly, and the monthly expiry sits proud of the weeklies bracketing it. That saw-tooth is the whole point: it proves the market prices each expiry for its own contents, so any smooth line drawn through these dots is inventing implied volatilities for dates on which nothing trades.

Professional explanation

A term structure is a set of points, not a curve

The phrase suggests a continuous line, but the market never quotes a continuum. It quotes the specific expiries it has chosen to list: on NIFTY, a weekly expiry every Thursday out to a few weeks, then monthly expiries on the last Thursday of each month, then quarterly and half-yearly points that barely trade. Each of those is a separate auction with its own supply and demand, its own liquidity, and its own bundle of calendar days. The implied volatility of the 11-day expiry and the implied volatility of the 18-day expiry are not two samples from one underlying function — they are two different prices for two different baskets of risk that happen to share an underlying. Treating them as points on a smooth curve is a choice, and it is a choice that throws away exactly the information the saw-tooth was carrying.

The monthly expiry is a liquidity magnet, and it shows in the vol

On the NSE, the monthly expiry is where hedgers, position traders and institutions concentrate. It is the most liquid contract, it is the one that index option open interest builds up in, and it is the expiry that absorbs the month's scheduled events — the RBI meeting, the monthly derivatives roll, the bulk of the systematic option-selling flow. Because it carries more event content and more of the market's real positioning, its at-the-money implied volatility typically sits slightly above the weeklies immediately around it, producing a visible bump rather than a smooth handover. This matters practically: a calendar spread that sells the monthly and buys a weekly is not a clean bet on the slope of a curve, it is a bet on a specific liquidity and event bump that the smooth-curve picture hides entirely.

0DTE and expiry-day options live in their own regime

An option in its final session is not a small version of a 30-day option — it is a different instrument. With hours left, the at-the-money premium is almost entirely gamma and almost no vega, time decay is measured in minutes rather than days, and the distribution of the terminal price is dominated by whether a single move happens or does not. Implied volatility computed on such an option is a division by a premium that has nearly decayed to zero, so a one-tick change in price swings the implied number wildly. This is why the last expiry on the chain routinely prints an implied volatility that looks absurd next to the rest of the structure: it is not part of the same continuum, and pretending it is corrupts any curve you fit through it.

The last hour of expiry day is not describable by any implied volatility — and settlement changes the endgame

This is the sentence a marketing department would cut. In the final hour of a NIFTY weekly expiry, the concept of an implied volatility has broken down. The remaining time is so short that the lognormal-diffusion assumption underneath Black–Scholes no longer approximates anything: the price will not diffuse, it will either jump or sit still, and no single σ describes that. On top of this, NSE settles index options at the weighted average of the underlying over the last 30 minutes of trading, not the closing tick. That averaging deliberately smooths the terminal value, which narrows the effective distribution of the settlement price relative to a last-price settlement and takes some of the knife-edge out of pin risk. So the terminal distribution an expiring option pays off against is not even the one its Black–Scholes implied volatility was describing — it is a time-averaged quantity that the model never had a symbol for.

Formula

Reading an expected move off a single listed expiry

ExpectedMove_expiry = S × σ_expiry × √(D / 365)

There is no single formula for the whole structure, because the structure is a list of independent points. What you can do is convert each listed expiry's own at-the-money implied volatility into the one-standard-deviation move that expiry is pricing, using that expiry's own days to expiry. Do not interpolate σ between expiries first and then apply this — invert the order and you will misprice the expiry that carries the event.

ExpectedMove_expiryThe one-standard-deviation move, in points of the underlying, that this particular expiry is pricing over its own life.
SSpot price of the underlying — 24,000 for NIFTY in every example on this site.
σ_expiryThe at-the-money implied volatility of that specific listed expiry, annualised, as a decimal (0.135 = 13.5%).
DCalendar days from today to that expiry. Calendar, not trading — the option decays over weekends even though the spot does not move.

Forward variance between two adjacent expiries

σ²_forward × (D₂ − D₁) = σ²₂ × D₂ − σ²₁ × D₁

Variance is additive over time, so the volatility priced for the window between an earlier expiry (D₁) and a later one (D₂) is backed out by differencing total variance. When this forward variance comes out higher than either expiry's own variance, an event is sitting in that window — the arithmetic localises the event to a segment instead of smearing it across a curve. If it comes out negative, one of the two quotes is stale or arbitrageable.

How to read the expiry structure without inventing points

List every expiry the exchange actually quotes with its at-the-money implied volatility and its calendar days to expiry. Do not add dates that are not listed.
Mark which expiries contain a scheduled event — an RBI MPC decision, the Union Budget, election counting day, a major results cluster. That mark, not the smoothness of the dots, explains most of the zig-zag.
Plot implied volatility against days to expiry as discrete points, and resist the urge to join them with a smooth line. If you must connect them, use straight segments so the saw-tooth stays visible.
For any expiry you care about, apply the expected-move formula with that expiry's own σ and its own days — never an interpolated σ.
To ask what the market prices for a window between two expiries, difference the total variance using the forward-variance identity rather than reading a curve.
Treat the final listed expiry with suspicion: its implied volatility is dividing by an almost-vanished premium, and in the last hour it is not a meaningful number at all.

Practical example

NIFTY worked example

Take three NIFTY expiries with spot at 24,000. This Thursday's weekly is 7 days out at σ = 13.0%; next Thursday's is 14 days out at σ = 13.4%; the monthly is 28 days out at σ = 13.8%. The one-standard-deviation moves are 24,000 × 0.130 × √(7/365) ≈ 432 points, 24,000 × 0.134 × √(14/365) ≈ 630 points, and 24,000 × 0.138 × √(28/365) ≈ 918 points. Notice the implied volatilities step up — 13.0, 13.4, 13.8 — which looks like a tidy rising curve until you learn that the RBI MPC decision falls between the second and third of these expiries. Difference the variance across that window: σ²_forward × (28 − 14) = 0.138² × 28 − 0.134² × 14, which is 0.53318 − 0.25135 = 0.28183, so σ_forward = √(0.28183 / 14) ≈ 14.2% — higher than either expiry's own number. The event did not lift the whole curve; it lifted the one segment that contains it, and only the differencing reveals that.

BANKNIFTY worked example

BANKNIFTY teaches a different lesson about the same structure: liquidity concentration. With BANKNIFTY at 52,000, its weekly at-the-money implied volatility might read 16.0% seven days out, pricing a one-standard-deviation move of 52,000 × 0.160 × √(7/365) ≈ 1,152 points. BANKNIFTY's structure is choppier than NIFTY's because the index is a narrow basket of lenders that all react to the same rate and asset-quality news, so its weeklies swing more between a quiet week and a policy week. More importantly, because open interest and turnover pile into the nearest expiries, the far weeklies on BANKNIFTY are thin — their quoted implied volatilities come from wide markets and should be treated as indicative, not as clean points you can difference against. A saw-tooth built partly from illiquid quotes has teeth that are quote noise, not information, and knowing which is which is the actual skill.

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. Calendar and diagonal spreads are routinely sold as low-risk ways to harvest term-structure slope. They are not low-risk. A calendar spread is short the near expiry and long the far one, and it makes its money only if realised movement stays inside the near expiry's priced range while the structure behaves. An unscheduled shock that lands in the near expiry's life can move against the position faster than the far leg's vega compensates, and the loss is real. Selling the front of a saw-tooth because it looks rich is a short-gamma position wearing a term-structure costume.

Advantages & limitations

What it is good for

It shows the market's own timeline of expected risk, expiry by expiry, so you can see which specific weeks the market is charging extra for rather than a smeared average.
It localises scheduled events. Because each expiry is priced for its own days, an RBI meeting or the Budget shows up as a bump on the expiries that contain it, and forward-variance differencing pins the event to a segment.
It is the honest input for expected-move and range calculations, because it uses the σ of the expiry you are actually trading rather than an interpolated fiction.
It exposes calendar-spread and roll opportunities that a smooth curve hides, since the tradeable edge lives in the bumps between adjacent listed expiries, not in a fitted line.
It is directly observable and needs no model to construct — you are reading quoted implied volatilities off dated contracts, not estimating a latent function.

Where it breaks down

It is defined only at the dates the exchange lists. The moment you need a value for an unlisted date, you are interpolating, and interpolation reintroduces exactly the smoothing that the saw-tooth was warning you against.
Its far points are built from illiquid expiries. Beyond the front couple of weeklies and the near monthly, NIFTY and especially BANKNIFTY quotes come from wide markets, so distant teeth of the saw-tooth may be quote noise rather than priced information.
The final listed expiry corrupts any fitted curve. In its last sessions its implied volatility is dividing by a near-zero premium, and in the last hour no σ describes its terminal distribution at all.
It says nothing about direction or about the shape across strikes. It is one point — the at-the-money — per expiry; the skew of each expiry is a separate object the structure does not capture.
Settlement mechanics break the model at the boundary. NSE's 30-minute weighted-average settlement means an expiring option pays off against a time-averaged price, not the lognormal terminal the option's own implied volatility assumed.

Common mistakes

Drawing a smooth curve through the listed expiries and reading an implied volatility off it for a date with no option. That number is invented, and it is biased low on any segment that contains an event.
Interpolating σ first and then computing an expected move, instead of computing the move from each expiry's own σ. The two orders give different answers, and only the second is a price the market actually quotes.
Treating the monthly expiry's bump as an anomaly to be smoothed away. That bump is the market's liquidity and event concentration; erasing it erases the information.
Building an IV Rank, a screener alert or a signal off the last listed expiry's implied volatility. In its final sessions that figure is an artefact of a premium that has almost fully decayed.
Selling a front-week calendar because the near expiry looks rich relative to the far one, without noticing the near expiry is rich because a shock-prone week sits inside it. That is a short-gamma bet, not a term-structure arbitrage.
Comparing the raw expiry structures of NIFTY and BANKNIFTY tooth for tooth. They have different liquidity profiles and different event sensitivities, so a bump on one is not the same signal as a bump on the other.

Professional usage

Volatility desks do not consume the expiry structure as a curve; they consume it as a set of forward variances between adjacent listed expiries, because the tradeable quantity is the volatility priced for the window between two dates, not the spot-to-expiry number. A market maker fits a surface but keeps the listed expiries as the anchoring nodes and treats anything between them as a modelled interpolation to be hedged, not trusted. When a desk wants to isolate an event, it does exactly the differencing shown on this page: it takes the total variance of the first expiry after the event and the total variance of the clean expiry before it, and backs the event variance out of the difference rather than reading a bump off a chart.

Calendar and diagonal traders live inside the saw-tooth. Their book is a series of long-and-short positions across adjacent expiries, and their profit and loss is driven by how the relative implied volatilities of those specific listed expiries move, plus the gamma and theta of the near leg. Risk managers, meanwhile, use the expiry structure to see where the market's priced risk is concentrated in time — a steep front driven by an imminent event tells them the near-dated book is the one to stress first. None of these users would accept a smoothed curve, because the thing they are managing is precisely the unevenness the smoothing destroys.

Key takeaways

The market lists a finite set of dated expiries, each independently priced for its own calendar days; the smooth term-structure curve is something an observer draws, not something the market quotes.
The NIFTY structure is a saw-tooth: weeklies step up and down with their event content and the liquid monthly sits proud of the weeklies around it.
Interpolating implied volatility between listed expiries systematically underprices the expiries that carry scheduled events, because it averages the event away instead of localising it.
0DTE and expiry-day options are a separate regime; in the last hour no single implied volatility describes the terminal distribution, and NSE's 30-minute weighted-average settlement changes that distribution again.
To ask what the market prices for a window between two expiries, difference total variance rather than reading a value off a fitted line.

Stop drawing the curve. The expiry structure is a saw-tooth because the market prices each listed expiry for the specific days it owns, and every tooth is telling you something — this week is quiet, that week has the RBI, the monthly is where the liquidity and the events pool. The smooth line an interpolator draws through those teeth is not a better version of the data; it is the data with its most useful feature deleted. Read the points as points, difference the variance when you want a window, and distrust the last one on the chain.

Frequently asked questions

What is expiry structure?

Expiry structure is the at-the-money implied volatility of each listed expiry of an index, read as a set of dated points. On NIFTY those points form a saw-tooth rather than a smooth curve, because every weekly and the monthly is priced for the specific calendar days it contains.

Why is the NIFTY term structure a saw-tooth instead of a smooth curve?

Because each listed expiry is an independent auction over a different bundle of calendar days. A week with an RBI meeting or heavy results is priced higher than a quiet week around it, so the implied volatilities step up and down rather than sliding smoothly from one expiry to the next.

Why does the monthly expiry usually show higher implied volatility than the weeklies around it?

Because the monthly is the liquidity magnet. Open interest, hedging flow and the month's scheduled events concentrate in it, so it carries more event content and more real positioning than the weeklies bracketing it, which shows up as a visible bump in the structure.

What is wrong with interpolating implied volatility between listed expiries?

Interpolation assumes the curve is smooth, so it averages an event across the whole span instead of parking it on the one expiry that carries it. That systematically underprices the expiries straddling an RBI decision or the Budget and overprices the clean ones next to them.

How do I find the implied volatility for a date with no listed option?

You cannot read it directly, because no option expires on that date. Any value you produce is an interpolation between neighbouring listed expiries, and you should treat it as a modelled estimate — not a quoted price — and never difference against it as if it were real.

What is 0DTE and why is it a different regime?

0DTE means zero days to expiry — an option in its final session. It is dominated by gamma with almost no vega, its time decay is measured in minutes, and its payoff hinges on whether one move happens, so a single σ no longer describes it the way it describes a 30-day option.

Why does the last listed expiry print such a strange implied volatility?

Because solving for σ near expiry is dividing by a premium that has almost decayed to zero. A one-tick change in a nearly worthless option implies an enormous change in volatility, so the final expiry's number looks absurd next to the rest of the structure and should not be trusted.

How does NSE settle index options on expiry day?

NSE settles index options at the weighted average price of the underlying over the last 30 minutes of trading on expiry day, not the closing tick. That averaging deliberately smooths the terminal value and reduces the knife-edge of pin risk around the settlement.

How does the 30-minute settlement change the terminal distribution?

Averaging the last half hour narrows the effective distribution of the settlement price compared with settling on a single last price. So an expiring option pays off against a time-averaged quantity that is less dispersed than the lognormal terminal its Black–Scholes implied volatility assumed.

Can I compute an expected move from the expiry structure?

Yes — for each listed expiry, multiply spot by that expiry's own at-the-money implied volatility and by the square root of its days-to-expiry over 365. The key discipline is to use each expiry's own σ, never a σ interpolated to some in-between date.

What is forward variance in the context of expiry structure?

Forward variance is the volatility the market prices for the window between two adjacent expiries, backed out by differencing total variance: σ²_forward × (D₂ − D₁) = σ²₂ D₂ − σ²₁ D₁. It localises an event to a segment instead of smearing it across a curve.

What does it mean if forward variance between two expiries comes out negative?

It means the no-arbitrage relationship between the two total variances is violated, so at least one of the quotes is stale, illiquid, or genuinely arbitrageable. A calendar structure cannot price negative variance for a future window, so the number is telling you the inputs are bad.

Why are the far expiries in the structure less reliable?

Because beyond the front weeklies and the near monthly, NIFTY and especially BANKNIFTY expiries are thinly traded. Their quoted implied volatilities come from wide bid-ask markets, so the distant teeth of the saw-tooth may be quote noise rather than priced information.

Is expiry structure the same as India VIX?

No. India VIX is a single number interpolated to a constant 30-day horizon from the whole near-dated NIFTY chain. Expiry structure is the set of individual listed expiries at their own dates. India VIX interpolates precisely because no NIFTY expiry usually sits exactly 30 days out.

Does the expiry structure tell me market direction?

No. It is one at-the-money implied volatility per expiry, and volatility is sign-agnostic. A steep front means the market prices more movement in the near term, not that it expects that movement up or down; direction is not encoded in the structure.

Why does a policy week's weekly expiry cost more than a quiet week's?

Because that weekly's life contains a session in which NIFTY could move several times a normal day's range. Whoever sells that option must be compensated for the extra risk, so its implied volatility is higher than a neighbouring weekly whose life contains no scheduled event.

How is a calendar spread related to the expiry structure?

A calendar spread is long one listed expiry and short another, so its value depends on how those two specific expiries' implied volatilities move relative to each other. It is a bet on a bump in the saw-tooth, plus the near leg's gamma and theta — not a clean bet on a smooth slope.

Should I sell the front weekly when it looks rich versus the monthly?

Be careful — the front weekly is often rich because a shock-prone week sits inside its life. Selling it is a short-gamma position, and an unscheduled move inside that week can lose more than the term-structure slope was ever going to pay. Richness is not automatically an edge.

Why do NIFTY and BANKNIFTY expiry structures look different?

Because they have different liquidity and event sensitivities. BANKNIFTY is a narrow basket of lenders reacting to the same rate and asset-quality news, so its weeklies swing more between quiet and policy weeks, and its far expiries are thinner than NIFTY's.

Do weekends affect the expiry structure?

Yes, through the day count. Expected-move and variance calculations use calendar days over 365 because an option decays across weekends even though the spot does not move. Two expiries separated by a long weekend carry more calendar time than trading time, and the arithmetic must use calendar days.

Can the expiry structure be downward sloping?

Yes. When the market is currently stressed but expected to calm, near expiries can price higher implied volatility than far ones, giving a downward or humped structure. The saw-tooth still sits on top of whatever the overall slope is; slope and teeth are separate features.

Voice search & related questions

Natural-language questions people ask about expiry structure.

Why isn't the term structure just a smooth curve?

Because the market never quotes a smooth curve — it quotes a handful of dated expiries, each priced for its own days. When one week has the RBI meeting and the next does not, their implied volatilities step apart, and the picture zig-zags. The smooth curve is something an analyst draws on top; it is not the data.

What happens to an option in the last hour of expiry?

It stops being describable by an implied volatility. There is so little time left that the price will either jump or sit still rather than diffuse, so no single σ fits it. On top of that, NSE settles on the weighted average of the last 30 minutes, so the option pays off against an averaged price, not the closing tick.

Why does the monthly expiry stick up above the weeklies?

Because that is where the liquidity and the scheduled events pool. The monthly is the most traded index expiry on the NSE, it holds the bulk of open interest, and it absorbs the month's big events, so it is priced a touch richer than the weeklies on either side of it. That bump is information, not a glitch.

Can I read an IV for any date I want off the structure?

Only if an option actually expires on that date. For any other date you are interpolating between listed expiries, and that interpolated number is a modelled guess that averages events away. It is fine as a rough estimate, but never treat it as a price the market is really quoting.

How do I see what the market is pricing for RBI week specifically?

Difference the variance across the RBI window rather than eyeballing a bump. Take the total variance of the first expiry after the meeting and of the clean expiry before it, subtract, and divide by the days in the window. That forward volatility isolates the meeting instead of smearing it over a curve.

Is the far end of the structure trustworthy?

Not very. Past the front weeklies and the near monthly, the expiries are thinly traded, so their implied volatilities come from wide markets and wobble on little turnover. Use the near, liquid expiries for anything you are actually trading, and treat the far teeth as indicative at best.

Sources & references

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