Home / Volatility Metrics / Volatility Risk Premium
Volatility Metrics Advanced Implied volatility minus subsequently realised volatility Forward-looking (implied) against outcome (realised)

Volatility Risk Premium VRP

Small, positive and frequent. The losses are large, negative and rare. That asymmetry is the whole story.

Quick answer: The volatility risk premium is the systematic gap by which implied volatility exceeds the volatility that subsequently realises, typically one to four volatility points on liquid index options — a small, frequent, positive edge that exists precisely because it is occasionally overwhelmed by a large, rare, negative one.

In simple words

When you buy insurance, you pay a little more than the expected cost of the thing you are insuring against — that markup is how the insurer stays in business and gets paid for bearing your risk. Options are insurance on price movement, and they carry the same markup. The volatility risk premium is that markup, measured in volatility points: implied volatility, the price, sits a little above the volatility that actually shows up afterward, the cost. On liquid index options like NIFTY it averages perhaps 1 to 4 volatility points. So most of the time, the person who sold the option collected slightly more than the movement that arrived was worth.

Here is the part that ruins most people who discover this. "Most of the time" is doing enormous work in that sentence. The premium is small, positive and frequent — it shows up in the large majority of months, which makes selling options feel like a reliable income. But the losses are large, negative and rare: every so often a shock arrives, realised volatility explodes past what was implied, and a single month erases years of collected premium. The average across everything is mildly positive, and the distribution around that average is brutal. A trader who reads "implied volatility exceeds realised most of the time" as a reason to sell options has understood the statistic perfectly and missed its entire point.

Not to be confused with: The variance risk premium, which is measured on variance (volatility squared) rather than on volatility itself. Variance swaps pay on σ², which weights the tail far more heavily than volatility does, so the variance risk premium is larger and even more concentrated in extreme moves. The two are related but not interchangeable, and conflating them understates how much of the compensation is really payment for tail risk.

The picture

Frequent shallow gains, rare deep losses

Implied volatility minus subsequently realised volatility, period by period, shaded above and below zero.

-5%0%+5%20d70d120d170d220dpremium positive: option sellers are being paid to carry riskpremium negative: sellers underpriced the moveTrading dayImplied volatility − realised volatility
The premium is positive in most periods and shallow — a thin band of steady gains above zero. Then, occasionally, realised volatility overwhelms implied and the bar plunges far below zero. The area above the line is wide and thin; the area below is narrow and deep. The mean is positive and the picture is why that fact is dangerous.

Professional explanation

The asymmetry is the entire subject

Everything about the volatility risk premium reduces to one shape: a stream of small positive numbers punctuated by occasional large negative ones, whose average is mildly positive. This is not a footnote to the premium; it is the premium. Any description that emphasises the positive average without the shape of the distribution around it is not simplifying, it is lying by omission. The premium looks like income and behaves like the sale of insurance, because it is the sale of insurance. An insurer who quoted its expected annual profit without mentioning that a single catastrophe year could exceed a decade of premiums would be committing fraud; a volatility seller who thinks of the premium as steady income is committing the same error against themselves. The correct mental image is not a yield. It is a short position in disasters that pays you to hold it right up until it does not.

Two honest interpretations, and practitioners genuinely disagree

Why does the premium exist at all? There are two serious answers and the profession has not settled between them. The first is that it is fair compensation: selling options means bearing crash risk that cannot be diversified away, and it means posting capital that could be earning elsewhere, so the premium is simply the market price of underwriting insurance — no more an anomaly than the profit margin of an insurance company. The second is that it is a behavioural artefact: investors are persistently, structurally over-eager to buy protection, especially downside puts, and that chronic excess demand bids implied volatility above its fair level, leaving a premium for whoever will sell. Both stories fit the data. The first says the premium is earned and roughly fair; the second says it is partly an unearned windfall left by other people's fear — and neither camp can decisively refute the other, which is exactly why serious desks hedge their conviction as well as their positions.

Variance risk premium is not the volatility risk premium

The premium changes character depending on whether it is measured on volatility or on variance. A variance swap pays on realised σ², not σ, and squaring is not a cosmetic difference — it weights large moves vastly more heavily, because the square of a big number is enormous. So the variance risk premium loads even more of its compensation onto the extreme tail than the volatility risk premium does, and it is correspondingly larger and more violently negative when a shock hits. This matters practically: much of the academic literature quotes the variance risk premium (it is cleaner to define from option prices via the VIX-style replication), while a trader selling straddles is closer to short volatility than short variance. Reading a variance-premium result as if it described a volatility-selling position overstates the edge and understates how much of it is really payment for the tail.

It lives in the wings, not spread evenly across strikes

The premium is not distributed uniformly across an option chain — it is concentrated in the out-of-the-money puts, the wings. This is where the chronic demand for crash protection sits, so it is where implied volatility is bid furthest above any reasonable estimate of realised, and it is why the volatility skew on an equity index is as steep as it is. An at-the-money straddle captures only a modest slice of the total premium; the bulk of it is in the low-delta puts that almost always expire worthless and occasionally pay out catastrophically for whoever sold them. This has an uncomfortable implication that most retail premium-selling ignores: the strikes that offer the richest-looking premium are the very strikes whose rare payout is largest, so the apparent generosity of the wings is not a bonus, it is the market pricing the shape of the disaster correctly. Selling the wing is not harvesting an inefficiency; it is being paid the going rate to stand in front of the worst outcomes.

What was priced against what arrived

Implied volatility as priced at the start of each period against the volatility that actually realised over it.

5%10%15%20%25%30%20d70d120d170dIV is biased high in the calm……and catastrophically low right before the shockDay the forecast was madeAnnualised volatilityimplied volatility (the forecast, made today)realised volatility over the NEXT 20 days (the outcome)
Line up what the option market charged against what the underlying then delivered. Implied sits above realised in the calm majority and far below it in the rare shock. The premium is the average vertical gap — positive — but the trade lives and dies on the handful of periods where the gap goes the wrong way.

Formula

Volatility risk premium — implied minus subsequently realised

VRP = E[σ_implied] − E[σ_realised]

The premium is the expected gap between the implied volatility priced at the start of a period and the volatility that actually realises over it, averaged across many periods. It is typically +1 to +4 volatility points on liquid index options. The expectation matters: in any single period the realised term can overwhelm the implied one, so the premium is a property of the long-run distribution, not of any one month.

  • VRPThe volatility risk premium in volatility points. Small and positive on average; the sign in any single period is unknown in advance.
  • E[σ_implied]The expected implied volatility priced into options at the start of each period — the forward-looking price, known when the position is opened.
  • E[σ_realised]The expected volatility that subsequently realises over the period — the outcome, knowable only at the end. This is the number that does not exist yet when the trade is placed.

Variance risk premium (measured on σ², not σ)

VRP_var = E[σ²_implied] − E[σ²_realised]

Defined on variance rather than volatility. Because squaring weights large moves far more heavily, the variance risk premium concentrates even more of its value in the tail, is larger in magnitude, and turns more sharply negative in a shock. Variance swaps pay on this quantity; it is not interchangeable with the volatility risk premium a straddle seller is exposed to.

How to think about the volatility risk premium

  1. Start from the definition: implied volatility priced today minus the volatility that actually realises over the option's life, averaged over many periods — not a single month.
  2. Estimate it like-for-like: compare a 30-day implied volatility against the realised volatility over the following 30 days, annualised on the same 252-day convention.
  3. Read the sign in context. A positive average is the normal state and reflects the price of insurance; it is not evidence that options are mispriced.
  4. Look at the whole distribution, not the mean. The premium is small, positive and frequent; the losses are large, negative and rare, and the mean hides both facts.
  5. Distinguish volatility from variance. If your source quotes a variance risk premium, remember it loads more heavily on the tail and overstates the edge a volatility seller actually earns.
  6. Locate where the premium sits. Most of it is in the out-of-the-money puts, so selling the wings is being paid the market rate to bear the largest rare losses, not harvesting an inefficiency.
  7. Never size a short-volatility position as if the premium were income. It is compensation for a tail that will eventually arrive, and short options carry theoretically unlimited loss.

Practical example

NIFTY worked example

Over the past twelve months, NIFTY's 30-day at-the-money implied volatility averaged about 13%, while the volatility NIFTY actually realised over each subsequent 30-day window averaged about 10.5%. The volatility risk premium is therefore roughly 13 − 10.5 = 2.5 volatility points, squarely in the typical +1 to +4 band for a liquid index. Read naively, that says an option seller collected about 2.5 points of implied volatility more than the movement that arrived, month after month — which sounds like a reliable stream. But unpack the average. In perhaps ten of those twelve months, realised volatility came in below the implied 13% and the seller kept a small premium; in one or two, a risk-off episode drove realised volatility to 22% against the 13% that was priced, and a single such month lost far more than the calm months made. The +2.5 is real and it is positive, and it is the average of a wide, left-skewed distribution whose rare tail is what the seller was actually paid to hold.

BANKNIFTY worked example

BANKNIFTY makes the wing concentration concrete. Suppose its at-the-money 30-day implied volatility runs 17% and realises 14% on average — a 3-point volatility risk premium, a touch richer than NIFTY's, as befits a more volatile index. Now look at a low-delta out-of-the-money BANKNIFTY put, several thousand points below spot. Its implied volatility might be 24% because of the steep skew, while the realised volatility relevant to that strike being reached is far lower in the vast majority of months — so the premium embedded in that put looks enormous, far more than the 3 points at the money. That is not a gift. Those are exactly the strikes that pay out most violently in a banking-sector shock, when BANKNIFTY can gap through several strikes in a session. The rich premium in the wing is the market pricing a rare, large, sector-specific disaster correctly; the seller of that put is being paid the going rate to be the counterparty to it, not finding an edge the market overlooked.

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 volatility risk premium is the single most dangerous statistic in options because it is true and it is positive. "Implied volatility exceeds realised volatility most of the time" is an accurate base rate about the market, and it is routinely used to justify selling options as though the positive average were an edge. It is not an edge in any usable sense: the premium is small, positive and frequent, and the losses that offset it are large, negative and rare, so a short-volatility position collects steady income right up until a single shock returns years of it. Short options carry theoretically unlimited loss, the premium is concentrated in exactly the strikes whose rare payout is largest, and no amount of favourable base rate changes the shape of that tail. A high frequency of small wins is not a measure of edge, and reading it as one is how leverage accumulates in short-volatility positions across the whole market before a coordinated unwind.

Advantages & limitations

What it is good for

  • It is a genuine, persistent feature of option markets, documented across underlyings and decades, so it is a real economic phenomenon rather than a fitting artefact.
  • It gives a coherent, forward-looking definition of whether options are richly priced — implied against subsequently realised — that the naive HV-vs-IV comparison only approximates.
  • It explains why option-selling strategies have a structural tailwind, and therefore why they attract capital, which is essential context for understanding market positioning.
  • It is measurable and decomposable: you can separate the at-the-money premium from the wing premium, and the volatility premium from the variance premium, which sharpens any analysis of where compensation actually comes from.
  • Understanding it correctly is the strongest available inoculation against the most common way retail traders are ruined — mistaking a high win rate for an edge.

Where it breaks down

  • It is a long-run average of a violently left-skewed distribution, so it tells you almost nothing about any single period, and a position sized to the average can be destroyed by one realisation.
  • The realised term is knowable only at expiry, so the premium is never observable in advance — every real-time estimate substitutes a forecast or trailing figure for the number that decides the outcome.
  • Its size and even its sign depend on the estimation window and the sample: a sample that happens to exclude a crash overstates it, and a sample that includes one can show it negative.
  • The variance version and the volatility version differ materially, so quoting one while trading the other overstates the edge and understates the tail exposure.
  • It is concentrated in the wings, so the premium available at the money is far smaller than the headline figure, and capturing the full premium means selling exactly the strikes with the largest rare payouts.
  • Its two explanations — fair compensation versus behavioural anomaly — have different implications for whether it will persist, and the data cannot currently decide between them, so any claim that it is durable free-standing edge is unsupported.

Common mistakes

  • Reading "IV exceeds realised most of the time" as a reason to sell options. The high frequency of small wins is not an edge; it is the visible half of an insurance sale whose rare large losses are the other half.
  • Sizing a short-volatility position to the average premium. The average is drawn from a left-skewed distribution, and a single realisation in the tail can exceed years of the average, which is how short-vol books blow up.
  • Confusing the variance risk premium with the volatility risk premium. The variance version is larger and more tail-heavy because it is measured on σ², so treating a variance-premium figure as the edge on a straddle overstates it.
  • Assuming the premium is spread evenly across strikes and selling at-the-money to "capture the VRP". Most of the premium is in the out-of-the-money puts, and the at-the-money straddle captures only a modest slice of it.
  • Selling the wings because the premium there looks richest. Those strikes offer the largest premium precisely because their rare payout is largest; the richness is the market pricing the disaster correctly, not an oversight.
  • Treating the premium as guaranteed to persist. Whether it is fair compensation or a behavioural anomaly is unsettled, and if it is the latter it can erode as protection-buying behaviour changes — so no version of it is a durable free edge.
  • Estimating the premium from a sample that excludes a crash and concluding it is larger and safer than it is. The tail that defines the premium may simply not have shown up in your window yet.

Professional usage

Volatility desks treat the premium as the raw material of their business, not as a signal to act on blindly. A systematic short-volatility book sells the premium deliberately and then spends most of its risk budget on managing the tail it is being paid to hold — buying cheaper far-wing protection, capping position size relative to a stress scenario rather than to the average premium, and delta-hedging so the residual exposure is to the volatility gap rather than to direction. The desk's edge, if it has one, is not the existence of the premium — everyone knows it exists — but surviving the realisations that periodically try to return it, which is a risk-management problem before it is a trading one.

Relative-value and dispersion desks trade the premium's structure rather than its level: they compare the premium across indices, across single stocks versus the index, and across the wings versus the at-the-money, selling where it is richest relative to the tail risk taken and buying where it is thin. Some also trade the variance-minus-volatility spread directly, expressing a view on how much of the compensation is tail-driven. Risk managers, meanwhile, watch aggregate short-volatility positioning across the market as a systemic gauge, because the premium's very reliability lures capital into the same trade, and a crowded short-volatility market is one where a shock forces a coordinated unwind that amplifies the realised volatility everyone was short.

Key takeaways

  • The volatility risk premium is the gap by which implied volatility exceeds subsequently realised volatility, typically +1 to +4 points on liquid index options.
  • The premium is small, positive and frequent; the offsetting losses are large, negative and rare. That asymmetry, not the positive average, is the entire subject.
  • Two honest interpretations coexist and practitioners disagree: fair compensation for bearing crash risk and capital costs, or a behavioural anomaly driven by chronic over-demand for protection.
  • The variance risk premium is measured on σ² and is not the same thing — it weights the tail far more heavily and overstates the edge a volatility seller earns.
  • The premium is concentrated in the out-of-the-money puts, so selling the wings is being paid the market rate to bear the largest rare losses, not harvesting an inefficiency.

The volatility risk premium is real, it is positive, and it is the most dangerous fact in options precisely because both of those things are true. It is the price of insurance, and insurance is sold above its expected cost so that the seller is paid to bear a rare catastrophe — which means the steady stream of small gains is not income but the visible half of a short position in disasters. Whether it is fair compensation or a behavioural anomaly is genuinely unsettled, and it lives disproportionately in the wings, where the richest-looking premium guards the largest rare losses. Understand it and you understand why option-selling attracts capital and why it periodically destroys it; misread the positive average as an edge, and you have understood the statistic and missed the point of it entirely.

Frequently asked questions

What is the volatility risk premium?
The volatility risk premium is the systematic gap by which implied volatility exceeds the volatility that subsequently realises, typically 1 to 4 volatility points on liquid index options. It is the markup option buyers pay and sellers collect for insuring price movement, analogous to an insurance premium above expected cost.
Why does the volatility risk premium exist?
There are two serious explanations and practitioners disagree. One is that it is fair compensation for bearing crash risk and posting capital; the other is that it is a behavioural anomaly caused by chronic over-demand for protection. Both fit the data, and neither has decisively won.
Is the volatility risk premium a free lunch?
No. It is small, positive and frequent, but it is offset by losses that are large, negative and rare, so it behaves like selling insurance rather than earning income. A single shock can return years of collected premium, and short options carry theoretically unlimited loss.
Why is a high win rate on selling options not an edge?
Because the wins are small and the rare losses are large. "Implied exceeds realised most of the time" is an accurate base rate, but the frequency of small gains says nothing about the size of the infrequent loss that offsets them. Confusing the two is the classic way short-volatility books are ruined.
What is the difference between the volatility and variance risk premium?
The volatility risk premium is measured on volatility (σ); the variance risk premium is measured on variance (σ²). Squaring weights large moves far more heavily, so the variance premium is larger and more tail-concentrated. Variance swaps pay on σ², so the two are not interchangeable.
Why does squaring matter for the variance risk premium?
Because the square of a large move is enormously larger than the move itself, so σ² loads most of its value onto extreme outcomes. The variance risk premium therefore concentrates even more of its compensation in the tail and turns more violently negative in a shock than the volatility version.
Where in the option chain is the premium concentrated?
In the out-of-the-money puts — the wings — where chronic demand for crash protection bids implied volatility furthest above realised. That concentration is why the equity index skew is so steep, and an at-the-money straddle captures only a modest slice of the total premium.
Why is selling the wings not harvesting an inefficiency?
Because the wings offer the richest premium precisely because their rare payout is the largest. The market is pricing the shape of the disaster correctly, so selling a low-delta put is being paid the going rate to bear the worst outcomes, not exploiting a mistake.
How big is the volatility risk premium on NIFTY?
On liquid index options it is typically 1 to 4 volatility points — for example, a 30-day implied volatility of 13% against a subsequently realised 10.5% is about 2.5 points. The exact figure depends on the estimation window and whether the sample included a shock.
Can the volatility risk premium be negative?
Yes, in any single period. When realised volatility overwhelms what was implied — during a shock — the premium for that period is deeply negative. The premium is positive only as a long-run average, and a sample that includes a large crash can even show it negative overall.
How is the volatility risk premium measured?
By comparing the implied volatility priced at the start of a period against the volatility that actually realised over it, like-for-like on tenor and annualisation, averaged across many periods. The realised term is knowable only at the end, so real-time estimates use forecasts or trailing figures as proxies.
Why can't I know the volatility risk premium in advance?
Because it depends on the realised volatility over the option's life, which does not exist until expiry. When you open the position you know the implied side but not the realised side, so the premium you will actually earn is unknown at the moment it matters.
Does the volatility risk premium mean option sellers make money on average?
On average and over long horizons, sellers have historically collected a positive premium, but the average conceals a left-skewed distribution where rare losses are severe. "On average" is not "reliably", and sizing to the average ignores the tail that periodically returns years of gains.
Is the volatility risk premium the same as the HV-vs-IV gap?
It is the forward, averaged version of it. HV vs IV often compares implied against trailing realised volatility; the volatility risk premium compares implied against the realised volatility that subsequently arrives, averaged over many periods, which is the economically correct form.
Will the volatility risk premium persist in the future?
That depends on which explanation is right, and it is unsettled. If it is fair compensation for risk it should persist as long as the risk does; if it is a behavioural anomaly it could erode as protection-buying behaviour changes. No version of it is a guaranteed durable edge.
How do professionals trade the volatility risk premium?
They sell it deliberately and spend most of their risk budget managing the tail — buying cheaper far-wing protection, sizing to a stress scenario rather than the average, and delta-hedging so exposure is to the volatility gap rather than direction. The craft is surviving the realisations, not the premium's existence.
What is variance replication and why does it matter here?
Variance replication is the technique behind VIX-style indices, building a model-free implied variance from a strip of options across strikes. It matters because much of the academic literature defines the premium on that replicated variance, which is why quoted results are often variance premia, not volatility premia.
Does the volatility risk premium explain the volatility skew?
Partly, yes. Chronic demand for downside protection bids up the implied volatility of out-of-the-money puts, concentrating the premium in the wings and steepening the skew. The skew is, in effect, a map of where the premium is richest and the tail risk largest.
Why is the volatility risk premium dangerous for retail traders?
Because it is true and positive, which makes selling options feel like reliable income. The steady small gains lull traders into oversized positions, and the rare shock that returns the premium arrives without warning. The statistic is real; reading it as an edge is the error.
Can I capture the whole volatility risk premium by selling straddles?
No. An at-the-money straddle captures only part of it, because most of the premium sits in the out-of-the-money puts. Capturing the full premium means selling the wings, which is exactly where the rare losses are largest — so the fuller the capture, the fatter the tail.
How does crowding affect the volatility risk premium?
The premium's reliability draws capital into the same short-volatility trade, and a crowded market is one where a shock forces a coordinated unwind. That unwind amplifies the very realised volatility everyone was short, which is why systemic risk managers watch aggregate short-vol positioning closely.

Voice search & related questions

Natural-language questions people ask about volatility risk premium.

If IV usually beats realised, why not just sell options all the time?
Because the usual small wins are paid for by rare huge losses, and you do not get to choose which month the huge loss lands in. Selling options is selling insurance — reliable income right until a catastrophe returns years of it, with theoretically unlimited loss on the short. The positive average is real and it is not an edge you can lean on. None of this is investment advice.
Is the volatility risk premium real or just theory?
It is real and well documented across markets and decades. What is unsettled is why it exists — whether it is fair pay for bearing crash risk or a behavioural quirk from everyone wanting protection. The existence is solid; the explanation is genuinely debated.
Why do people say the premium is in the wings?
Because demand for crash protection is heaviest in the out-of-the-money puts, so their implied volatility is bid furthest above what tends to realise. That is where the premium is richest — and also where the rare payout is biggest, so the richness is the market pricing the disaster, not a free lunch.
What's the difference between variance and volatility risk premium in plain terms?
Volatility premium is measured on the move itself; variance premium is measured on the move squared. Squaring makes big moves count for much more, so the variance premium is bigger and even more about the tail. If someone quotes a variance premium as the edge on selling straddles, they have overstated it.
How can the average be positive if the losses are so big?
Because the losses, though large, are rare, and the small gains are frequent enough that they add up to slightly more on average. It is exactly how an insurer is profitable across many years despite the occasional catastrophic payout. The average being positive is completely consistent with a brutal distribution around it.
Should the volatility risk premium make me confident about selling premium?
It should make you respectful of it, not confident. Understanding the premium mostly teaches you why the trade is dangerous — steady income that periodically detonates — rather than why it is safe. If anything, knowing the premium is real is the reason to size small and manage the tail, not the reason to lever up.
Does the volatility risk premium mean the option market is wrong?
Not necessarily. Under the fair-compensation view the market is pricing insurance correctly and the premium is the going rate for the risk. Only under the behavioural view is the market arguably over-pricing protection. Both are defensible, so calling the premium a market error takes a side the evidence has not settled.

Sources & references

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

← Previous
HV vs IV
Educational content only — not investment advice. Every diagram on this page is generated from the site's own model, using illustrative inputs rather than live quotes. Options and futures carry substantial risk, including loss exceeding your deposit on short-volatility positions. See our Risk Disclosure and SEBI Disclaimer.