Volatility & Options Intermediate Sensitivity of premium to a 1-point IV change Forward-looking

IV and Vega

Vega tells you how much you are exposed to the level of implied volatility — and it lives almost entirely in the at-the-money strikes.

Quick answer: IV and vega are connected because vega is the rupees of premium an option gains or loses for a one-percentage-point change in implied volatility, and its size — a hump centred at the money, near zero in both wings, and growing with the square root of time to expiry — tells you exactly where on the option chain a position is exposed to the level of IV.

In simple words

Vega answers a single question: if implied volatility rises by one percentage point, how many rupees does this option's premium change? For the 30-day at-the-money NIFTY 24,000 call, vega is about ₹27, so a move in IV from 12.8% to 13.8% adds roughly ₹27 to the premium. Move away from the money and vega shrinks fast: the 23,000 call carries about ₹11 of vega and the 26,000 call only about ₹4. A far out-of-the-money option has almost nothing for volatility to work on, because there is barely any time value there for a change in IV to inflate or deflate.

Vega also depends heavily on how much time is left. Because an option's exposure to volatility builds with the square root of time, a long-dated option carries far more vega than a short-dated one at the same strike. The 7-day at-the-money NIFTY call has about ₹13 of vega; the 180-day one has about ₹62 — nearly five times as much, close to the √(180/7) ≈ 5.1 the square-root rule predicts. So if you want a position that gains a lot when the general level of implied volatility rises, you buy long-dated at-the-money options. If you want a position that reacts violently to the underlying actually moving, you buy short-dated ones — they carry little vega but enormous gamma.

Not to be confused with: Gamma, which measures sensitivity to the underlying's movement rather than to the level of implied volatility. Vega and gamma pull in opposite directions across tenors: long-dated options are rich in vega and poor in gamma, short-dated options are the reverse. A trader who wants exposure to the LEVEL of volatility and a trader who wants exposure to MOVEMENT want opposite ends of the term structure, and confusing vega with gamma means buying the wrong tenor for your actual view.

Vega as a function of implied volatility

How the premium's IV-sensitivity itself shifts with IV

Vega of a 30-day at-the-money NIFTY 24,000 call as implied volatility is varied, spot fixed at 24,000.

Vega is quoted as the rupees gained per one-percentage-point rise in IV, and for an at-the-money option it is remarkably stable across the range of implied volatilities that matter — which is why traders treat ATM vega as an almost fixed rupee figure per point. The gentle curvature that remains is vomma: the sensitivity of vega itself to volatility, and the reason a large volatility spike does not move the premium in a perfectly straight line.

Professional explanation

Vega is the rupee price of a one-point move in IV, and it is a hump

Vega measures how much an option's premium changes when implied volatility moves by one percentage point — by convention, always one point, not one unit. Because implied volatility acts only on time value, vega is largest where there is the most time value to act on, which is at the money, and it fades to almost nothing where time value is scarce, which is deep in or deep out of the money. On the 30-day NIFTY chain the at-the-money 24,000 call carries about ₹27 of vega, the 23,000 and 25,000 strikes about ₹11 and ₹18, and the 22,000 and 26,000 wings only about ₹1 and ₹4. Plotted across strikes it is a hump centred at the money. A far out-of-the-money option has almost nothing for volatility to work on — its premium is tiny and a point of IV barely moves it — which is the quantitative reason cheap wing options are a poor way to express a view on the level of volatility.

Vega grows with the square root of time

An option's sensitivity to volatility scales with the square root of its remaining life, so a longer-dated option carries proportionally more vega at the same strike. On the at-the-money NIFTY call, vega runs about ₹13 at 7 days, ₹27 at 30 days, ₹46 at 90 days and ₹62 at 180 days. The ratio of the 180-day to the 7-day vega is about 4.7, close to the √(180/7) ≈ 5.1 that the square-root rule predicts — the small gap is the interest-rate and moneyness drift over the tenors. This √T growth is the single most important fact about where vega lives: if you want a large exposure to the general level of implied volatility, you buy time, because time is what vega is built from.

Level versus movement: the term structure is a choice of exposure

The √T growth of vega and the 1/√T growth of gamma point traders in opposite directions depending on what they actually want. A trader who wants exposure to the LEVEL of implied volatility — who thinks IV across the surface is too low and will rise — buys long-dated at-the-money options, because those carry the most vega and the least gamma, so the position gains on a rise in IV without depending much on the underlying moving day to day. A trader who wants exposure to MOVEMENT — who thinks the underlying will realise more than the market expects — buys short-dated at-the-money options, which carry little vega but enormous gamma, so the position gains from the underlying actually moving rather than from the quoted IV re-rating. These are different trades with different Greeks, and the term structure is the dial between them.

Vega does not add up the way traders assume

Two facts about vega ruin more risk reports than any other. First, vega is quoted per one percentage point of IV, so a book showing '₹50,000 of vega' means it makes ₹50,000 for a one-point rise in the relevant IV — but which IV? Vega across different strikes and expiries is exposure to different points on the surface, and those points do not move one-for-one, so summing them into a single number pretends a parallel shift that rarely happens. Second, and more dangerous, vega is not additive across underlyings. A book that is 'vega neutral' because its positive NIFTY vega cancels its negative BANKNIFTY vega is not hedged at all, because NIFTY IV and BANKNIFTY IV do not move together one-for-one — BANKNIFTY realises and re-rates more, so a stress that lifts both IVs by different amounts leaves the supposedly neutral book with a real, uncancelled loss. And because vega itself changes as IV moves — that second-order sensitivity is called vomma — even a single-underlying vega hedge drifts as volatility moves, which is the uncomfortable detail most retail risk tools quietly ignore.

Vega is a hump at the money and nothing in the wings

Vega across every strike of a 30-day NIFTY expiry, spot 24,000, IV 12.8%.

The profile peaks sharply at the at-the-money strike and falls to almost nothing at both wings. A far out-of-the-money option has hardly any time value for a change in implied volatility to act on, so its vega is negligible — which is why a book that looks 'long volatility' because it holds a stack of cheap far-OTM options is often carrying almost no vega at all, and is really long tail gamma instead.

Formula

Vega — the premium's sensitivity to a one-point change in IV

ν = ( S · φ(d₁) · √T ) / 100, d₁ = [ ln(S/K) + (r + σ²/2)·T ] / ( σ·√T )

The division by 100 makes vega the rupees gained per ONE PERCENTAGE POINT of implied volatility, which is the trading convention. Vega is proportional to √T (why long-dated options carry more) and is maximised near the money, where φ(d₁) — the standard normal density — is largest. It falls to almost zero in the wings, where d₁ is far from zero and φ(d₁) is tiny.

νVega — the change in the option's premium, in rupees, for a one-percentage-point change in implied volatility. Identical for a call and a put at the same strike and expiry.
SSpot price of the underlying — 24,000 for NIFTY throughout this site.
φ(d₁)The standard normal probability density evaluated at d₁. Largest when d₁ is near zero, i.e. at the money — which is why vega humps there.
d₁The Black–Scholes d₁ term, measuring how far in or out of the money the option is in standardised units.
√TSquare root of time to expiry in years. Vega grows with √T, so a 180-day option carries several times the vega of a 7-day one at the same strike.
KStrike price of the option contract.
rRisk-free interest rate, 6.5% as a rupee proxy.
σImplied volatility, annualised, as a decimal (0.128 = 12.8%).

Vomma — how vega itself changes when IV changes

vomma = ν · ( d₁ · d₂ ) / σ

Vega is not constant: as implied volatility moves, vega moves too, and vomma measures that second-order effect. It is why a large IV spike does not change the premium in a perfectly straight line and why a vega hedge drifts as volatility moves — a detail that matters most for options away from the money, where d₁·d₂ is large.

How to choose a strike and tenor for the volatility exposure you actually want

Decide what you are betting on: the LEVEL of implied volatility across the surface rising or falling, or the underlying MOVING more or less than expected. These need different Greeks.
For a view on the level of IV, target vega. Buy at-the-money options, where vega humps, and buy time, because vega grows with √T — a 180-day ATM option carries several times the vega of a 7-day one.
For a view on movement, target gamma. Buy short-dated at-the-money options, which carry little vega but enormous gamma, so the position gains from realised movement rather than from IV re-rating.
Avoid far out-of-the-money options if your view is on the level of IV — their vega is negligible because there is almost no time value for a point of IV to act on.
Read vega as rupees per one percentage point, and size the position by the IV move you expect times the vega, not by the premium.
Do not net vega across NIFTY and BANKNIFTY into a single figure — their implied volatilities do not move one-for-one, so the cancellation is fictional.
For a large expected IV move, check vomma: vega itself will change as IV moves, so the linear vega figure understates the response to a big volatility spike.

Practical example

NIFTY worked example

NIFTY at 24,000, 30 days, r = 6.5%, IV 12.8%. Vega across strikes forms a hump: the 22,000 call carries about ₹1.10 of vega, the 23,000 about ₹11.43, the at-the-money 24,000 about ₹27.08, the 25,000 about ₹17.51 and the 26,000 about ₹3.59. So a one-point rise in IV, from 12.8% to 13.8%, adds about ₹27 to the ATM premium but only about ₹4 to the 26,000 wing. Now hold the strike at the money and vary the tenor: vega runs about ₹13.22 at 7 days, ₹27.08 at 30 days, ₹45.67 at 90 days and ₹62.03 at 180 days. The 180-day to 7-day ratio is 62.03 ÷ 13.22 ≈ 4.7, against √(180/7) ≈ 5.1 — the square-root rule, confirmed. Interpret it: if you believe the whole NIFTY volatility surface is too cheap and will re-rate upward, the 180-day at-the-money option gives you almost five times the rupee exposure per IV point of the weekly, and it barely cares whether NIFTY moves in the meantime. The weekly would give you a fraction of the vega and a mountain of gamma you did not ask for.

BANKNIFTY worked example

BANKNIFTY at 52,000, lot 30, 30 days, IV 15%. The at-the-money 52,000 call carries about ₹58.85 of vega — more than double the NIFTY ATM call's ₹27.08 — while across strikes it still humps: about ₹7.94 at the 48,000 strike, ₹58.85 at 52,000 and ₹17.14 at 56,000. The BANKNIFTY lesson is the one about additivity. Suppose a book is long ₹58,850 of BANKNIFTY vega per lot and short an offsetting amount of NIFTY vega, and the risk system reports it 'vega neutral'. It is not hedged. BANKNIFTY implied volatility is more volatile than NIFTY's and does not move one-for-one with it; in a stress episode BANKNIFTY IV can jump several points while NIFTY IV rises one, and the 'neutral' book takes a real loss on the difference. Netting vega across two underlyings assumes their volatilities are the same variable. They are not, and the risk report that says otherwise is quietly wrong.

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. A stack of cheap far-out-of-the-money options is often sold as a way to be 'long volatility' for very little money. The vega profile says otherwise: wing options carry almost no vega, so such a book barely gains when the general level of implied volatility rises. What it actually holds is tail gamma — it pays off on a large, fast move, not on a re-rating of IV. Marketing the position as long volatility hides that it is long a specific, rare kind of movement, and that it bleeds theta every day the move does not come.

Advantages & limitations

What it is good for

Vega gives a single, tradeable rupee figure for a position's exposure to the level of implied volatility: rupees gained per one-percentage-point rise in IV, which is exactly what you size a volatility view on.
Its hump shape tells you precisely where on the strike ladder your IV exposure lives — at the money — and where it does not, so you never waste premium buying wing options to express a view on the level of volatility.
Its √T growth lets you dial exposure up or down by choosing tenor: buy time to buy vega, sell time to shed it, with the square-root rule giving a quick estimate of how much.
Separating vega from gamma across the term structure lets a trader isolate a view on the level of IV from a view on movement — long-dated for level, short-dated for movement — instead of tangling the two.
Because ATM vega is nearly constant across the practical range of IV, it behaves almost like a fixed rupee-per-point figure, which makes at-the-money volatility positions easy to size and monitor.

Where it breaks down

Vega assumes a parallel shift in the volatility used to price the option, but real surfaces twist and steepen rather than shift in parallel, so a single vega figure overstates how hedged a position across multiple strikes really is.
Vega is not additive across underlyings. Netting NIFTY vega against BANKNIFTY vega assumes their implied volatilities move one-for-one, which they do not, so a 'vega neutral' cross-underlying book can carry a large real exposure.
Vega itself changes as IV moves — the vomma effect — so the linear rupees-per-point figure understates the premium response to a large volatility spike and causes a static vega hedge to drift.
Vega is negligible in the wings, so it says almost nothing about the risk of a far out-of-the-money position, whose real exposure is to a large move (gamma and tail risk) rather than to the level of IV.
The √T growth means a long-dated vega position is exposed to the whole term structure re-rating, and the long end of the term structure can move quite independently of the front end, so long-dated vega is not simply 'more of the same' exposure as short-dated vega.

Common mistakes

Buying cheap far out-of-the-money options to be 'long volatility' and finding the position barely moves when IV rises — because wing vega is negligible and the book is really long tail gamma, not volatility level.
Netting vega across NIFTY and BANKNIFTY and believing the book is hedged. Their implied volatilities do not move one-for-one, so the cancellation is fictional and a stress event reveals a real loss.
Treating vega as a fixed number when planning for a large IV spike. Vomma means vega itself grows or shrinks as IV moves, so the premium response to a big spike is not the linear vega estimate.
Buying short-dated options to express a view that the level of implied volatility will rise. Short-dated options carry little vega and enormous gamma, so the position depends on the underlying moving, not on IV re-rating.
Buying long-dated options to profit from an expected big move over the next few days. Long-dated options carry lots of vega but little gamma, so they respond slowly to movement and the position underperforms if the move comes and goes quickly.
Quoting a book's vega without saying which IV it is exposure to. Vega on the front-month at-the-money strike and vega on a six-month wing are exposures to different points on the surface that do not move together.
Forgetting that vega is quoted per one percentage point, and mis-sizing a position by a factor of a hundred by treating a one-unit (one hundred percentage points) move as one point.

Professional usage

Volatility desks decompose every position into vega buckets along the term structure and across the strike ladder, precisely because a single vega number hides the twists that actually cause profit and loss. A desk running a relative-value trade might be long 90-day vega and short 30-day vega — a view that the term structure is too flat — and it manages that as a term-structure position, not as a net vega figure, because the two tenors re-rate independently. Across underlyings the desk never nets: NIFTY vega and BANKNIFTY vega sit in separate books with separate hedges, because their implied volatilities are correlated but far from identical, and a beta-weighted hedge between them is itself a position that has to be managed. The vega hump across strikes tells the desk where a customer's trade has loaded its exposure, and the desk re-strikes or spreads to move that exposure to where it wants it.

Dispersion and correlation desks live off the fact that index vega and single-stock vega do not add up. A dispersion trade is short index vega and long the vega of the index's constituents, a bet that the components will move more than the index because their correlation will fall — a position that a naive vega sum would report as roughly flat and that is in fact one of the largest volatility exposures a book can carry. And vomma is a traded quantity in its own right on structured-products desks, where long-dated options with large vomma are used to gain exposure to the volatility of volatility itself. In every case the professional treats 'vega' not as one number but as a vector across strike, tenor and underlying, and treats the naive scalar sum as a red flag rather than a hedge.

Key takeaways

Vega is the rupees an option's premium changes for a one-percentage-point move in implied volatility — quoted per one point, by convention.
Across strikes vega is a hump centred at the money and near zero in both wings, because a far out-of-the-money option has almost no time value for a change in IV to act on.
Across tenors vega grows with √T: the 180-day ATM NIFTY call carries about ₹62 of vega against about ₹13 for the 7-day, a ratio near the √(180/7) ≈ 5.1 the rule predicts.
For exposure to the LEVEL of IV, buy long-dated at-the-money options; for exposure to MOVEMENT, buy short-dated at-the-money options, which carry little vega but enormous gamma.
Vega is not additive across underlyings: a book 'vega neutral' across NIFTY and BANKNIFTY is not hedged, because their implied volatilities do not move one-for-one. Vomma — vega's own sensitivity to IV — makes even a single-underlying vega hedge drift.

Vega is the Greek that tells you where a position is exposed to the level of implied volatility, and its two shapes are the whole story: a hump across strikes that says the exposure lives at the money, and a √T rise across tenors that says the exposure is bought with time. Get those two shapes into your hands and you stop making the classic errors — buying wing options to be long volatility, buying weeklies to bet on IV rising, netting NIFTY and BANKNIFTY vega into a comforting zero. The uncomfortable part is that vega is the most abused number on a risk report: it does not add across strikes the way a parallel shift assumes, it does not add across underlyings at all, and it changes under your feet as IV moves. A single vega figure is a starting point for a conversation, never the end of one.

Frequently asked questions

What is vega in options?

Vega is the change in an option's premium, in rupees, for a one-percentage-point change in implied volatility. For a 30-day at-the-money NIFTY 24,000 call it is about ₹27, so a move in IV from 12.8% to 13.8% adds roughly ₹27 to the premium. It is quoted per one point, by convention.

Why is vega highest for at-the-money options?

Because implied volatility acts only on time value, and at-the-money options carry the most time value for a change in IV to inflate or deflate. Vega is a hump centred at the money — about ₹27 on the NIFTY 24,000 strike — and it falls to almost nothing in the wings, where there is little time value to work on.

Why do far out-of-the-money options have low vega?

Because they have almost no time value for a change in implied volatility to act on. The 26,000 call on a 30-day NIFTY chain carries only about ₹4 of vega against the ATM strike's ₹27, so a point of IV barely moves its premium. This is why cheap wing options are a poor way to express a view on the level of volatility.

How does vega change with time to expiry?

Vega grows with the square root of time to expiry. On the at-the-money NIFTY call it runs about ₹13 at 7 days, ₹27 at 30 days, ₹46 at 90 days and ₹62 at 180 days. The 180-day to 7-day ratio, about 4.7, is close to √(180/7) ≈ 5.1 — the square-root rule.

Which options should I buy to bet on rising implied volatility?

Long-dated at-the-money options, because they carry the most vega and the least gamma. A 180-day ATM NIFTY call gives you nearly five times the rupees-per-IV-point of the weekly and barely depends on whether the underlying moves in the meantime. Nothing here is investment advice.

Which options should I buy to bet on a big move?

Short-dated at-the-money options, which carry little vega but enormous gamma, so the position gains from the underlying actually moving rather than from implied volatility re-rating. The trade-off is severe time decay if the move does not arrive quickly.

Is vega quoted per one point or one unit of IV?

Per one percentage point, by universal convention — a vega of ₹27 means ₹27 of premium change for a move from, say, 12.8% to 13.8%. Treating it as a one-unit (hundred-point) sensitivity mis-sizes a position by a factor of a hundred, a common and expensive beginner error.

Is vega the same for a call and a put?

Yes, exactly, for a European call and put at the same strike and expiry. Vega depends on the standard normal density at d₁, which is identical for the two, so a straddle's vega is simply twice a single leg's. This is why straddles are the natural instrument for a pure volatility position.

Can I add up vega across different strikes?

You can, but the sum assumes a parallel shift in the whole volatility curve, which rarely happens — real surfaces twist and steepen. A net vega figure across many strikes overstates how hedged the position is, because the strikes are exposure to different points on the surface that move independently.

Is a vega-neutral book across NIFTY and BANKNIFTY hedged?

No. Vega is not additive across underlyings, because NIFTY implied volatility and BANKNIFTY implied volatility do not move one-for-one. BANKNIFTY re-rates more, so a stress that lifts both IVs by different amounts leaves the supposedly neutral book with a real, uncancelled loss.

What is vomma?

Vomma is the sensitivity of vega itself to a change in implied volatility — a second-order Greek. It means vega is not constant: as IV rises or falls, vega moves too, so a large volatility spike does not change the premium in a straight line and a static vega hedge drifts. It is largest for options away from the money.

Why does BANKNIFTY have higher vega than NIFTY?

Largely because of its higher spot: vega is proportional to the spot price, and BANKNIFTY at 52,000 is more than double NIFTY at 24,000, so the at-the-money 30-day BANKNIFTY call carries about ₹59 of vega against NIFTY's ₹27. The higher rupee vega does not by itself mean BANKNIFTY options are a better volatility bet.

Does vega stay constant as IV changes?

Not exactly — that is what vomma measures. For an at-the-money option vega is remarkably stable across the practical range of IV, which is why traders treat ATM vega as nearly fixed. Away from the money, vega changes more as IV moves, and a large spike can shift it noticeably.

How do I size a volatility position using vega?

Multiply the vega by the number of IV points you expect the relevant implied volatility to move, then by the lot size. A position with ₹27 of vega per unit, on a NIFTY lot of 75, gains about ₹27 × 75 = ₹2,025 per lot for each one-point rise in IV — before any offsetting delta, theta or gamma effects.

Why is my long-dated option barely reacting to the underlying's moves?

Because a long-dated option is rich in vega and poor in gamma. It responds to changes in the level of implied volatility, not strongly to the underlying's day-to-day movement. If you wanted exposure to movement you needed a short-dated option, which carries the gamma.

What is the difference between vega and gamma?

Vega measures sensitivity to the level of implied volatility; gamma measures sensitivity to the underlying's movement. They pull opposite ways across tenors — long-dated options are rich in vega and poor in gamma, short-dated options the reverse — so the tenor you choose decides which exposure you take on.

Does India VIX rising help all my long options equally?

No. A rise in the general level of implied volatility helps positions with vega, and vega lives at the money and grows with tenor. Your long-dated at-the-money options gain the most; your cheap far out-of-the-money weeklies, which carry almost no vega, barely move on the VIX print alone.

Why is vega proportional to the square root of time?

Because an option's cumulative exposure to volatility over its life scales with √T, the same way the standard deviation of a random walk grows with the square root of the horizon. Twice the time gives √2 ≈ 1.41 times the vega, not twice — which is why the term-structure ratios never look linear.

Can vega be negative?

For a single long option, no — a long option always gains when IV rises, so its vega is positive. A position can be net short vega if it holds more short options than long, and a short option's vega is negative from the holder's point of view because a rise in IV increases the premium they owe.

Why do dispersion traders care that vega does not add across underlyings?

Because their whole trade depends on it. A dispersion position is short index vega and long the vega of the index's constituents — a bet that correlation falls so the components move more than the index. A naive vega sum would report the trade as roughly flat, when it is in fact one of the largest volatility exposures a book can hold.

Voice search & related questions

Natural-language questions people ask about iv and vega.

What does vega tell me about an option?

It tells you how many rupees the option's premium will move if implied volatility changes by one percentage point. A high vega means the option is very sensitive to the level of IV; a low vega, in the wings or on a short-dated option, means it barely reacts to IV and depends on other things.

Why don't cheap OTM options make money when volatility spikes?

Because far out-of-the-money options carry almost no vega — there is barely any time value for a change in IV to act on. A stack of them is really a bet on a large, fast move in the underlying, which is gamma, not a bet on the level of implied volatility rising.

Should I buy weekly or monthly options to trade volatility?

It depends on which volatility you mean. To trade the LEVEL of implied volatility, longer-dated options carry far more vega and are the right tool. To trade MOVEMENT — the underlying realising more than expected — short-dated options carry the gamma. Buying the wrong tenor gives you the wrong exposure. Nothing here is investment advice.

Why does a longer-dated option cost so much more in vega terms?

Because vega grows with the square root of time, so a 180-day option carries roughly √(180/7) ≈ 5 times the vega of a 7-day one at the same strike. You are buying more exposure to the level of implied volatility, and that exposure is literally built out of time.

Is my book really hedged if its vega nets to zero?

Only if all that vega is exposure to the same implied volatility. If you have netted NIFTY vega against BANKNIFTY vega, the answer is no — their IVs do not move together, so the zero is fictional and a stress event will show you a real loss. Vega across underlyings does not add.

What is vomma in plain terms?

Vomma is how much your vega itself changes when implied volatility moves. Vega is not a fixed number; on a big volatility spike your sensitivity to IV grows or shrinks, so the premium does not move in a straight line and a hedge you set up at one IV level slips as IV moves.

Why did India VIX jump but my options barely moved?

Most likely you were holding short-dated or far out-of-the-money options, which carry little vega. A VIX move is a move in the level of implied volatility, and only positions with real vega — at-the-money and longer-dated — respond to it strongly. Low-vega options need the underlying to actually move.

Sources & references

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