What Is the Variance Risk Premium?
Variance risk premium is the difference between the volatility implied by option prices and the volatility subsequently realised. It is usually positive, meaning option buyers on average pay more than the eventual movement justified, and that excess compensates sellers for bearing the risk of a large move.
Options generally imply more movement than subsequently occurs. That persistent gap is the variance risk premium, and it is compensation for a risk rather than a mispricing.
How it works
Option prices contain an expectation of future movement. Back that out and you have implied volatility — the market’s forecast of how much the underlying will move.
Then measure what actually happened. The difference between the two, averaged over many periods, is persistently positive across markets and decades.
Which means the average option buyer loses. Not on every trade, and on average across many of them, which is the same statistical statement as an insurance buyer losing on average.
Why it is not free money
The seller’s payoff is asymmetric. Steady small gains, then an occasional loss far larger than any individual gain — the shape of every insurance business ever run.
And the losses cluster. Large moves happen when everything else is going wrong, so the loss arrives alongside losses in whatever else the seller holds.
Which is exactly why the premium exists. Nobody would sell that payoff without being paid extra, and the extra is what the measurement picks up.
A worked example
Options imply 20% annualised volatility. Realised volatility over the following period turns out to be 16%.
The four-point gap is the premium, collected by whoever sold the options and paid by whoever bought them.
Repeat that across many periods and the seller accumulates steadily. Most periods look like this one, which is what makes the strategy so appealing on a track record.
Then one period realises 60% against an implied 20%. The loss on that single period can exceed several years of accumulated premium, and it arrives with no warning in the data.
Why the premium persists
Demand for protection is structural. Institutions hold equities and must limit drawdowns, so they buy puts regardless of whether the price is attractive — a constant one-directional demand.
And supplying that protection requires capacity. Selling options needs capital, risk tolerance and the ability to survive a bad period, which limits how much supply meets the demand.
So the imbalance is not an error being corrected. It is a persistent feature of who needs what, and it has survived decades of being documented.
Which makes it different from most published anomalies. There is an identifiable economic reason for it rather than a statistical pattern in search of one, and that reason has not gone away.
The original data
On this site’s shared series the median bar range is 0.493, the ninetieth percentile 1.101 and the largest single bar 2.338. ATR14 has a median of 0.5994 and a ninetieth percentile of 0.7954.
The gap between those percentiles is the whole trade. A seller pricing off the median collects consistently and is destroyed by the maximum; the premium exists because the distance between them is large and nobody knows when the tail arrives.
And costs erode the collection: a round trip costs 0.0098, about 2% of the median bar range. Harvesting this premium requires frequent trading, so transaction costs consume a meaningful share of what is collected.
How it is actually harvested
Variance swaps pay the difference between realised and implied variance directly, which is the cleanest expression and is institutional.
Selling straddles or strangles is the retail version, and it carries directional exposure unless it is delta-hedged continuously.
Short volatility exchange-traded products package the exposure and have a documented history of losing the great majority of their value in single sessions.
All three have the same underlying payoff. The wrapper changes the operational details and not the shape — steady collection and a rare severe loss — and any description of one of these as an income strategy has described half of it.
How it behaves through a cycle
It widens in calm periods. Implied volatility falls more slowly than realised volatility does, so the gap between them grows while nothing is happening.
It inverts during a crisis. Realised volatility spikes above implied, so the premium is briefly and sharply negative, which is when the accumulated losses occur.
And it is widest immediately afterwards. Implied volatility stays elevated after realised volatility has subsided, which historically has been the most profitable moment to sell and the moment fewest people are willing to.
That pattern is the whole difficulty. The best risk-adjusted entry follows the worst drawdown, and any process that reduces size after losses systematically misses it.
When it fails
The characteristic failure is sizing the position on the collection period. A seller observes years of steady returns and low measured volatility in the strategy itself, concludes it is low-risk, and increases size accordingly. The measured volatility was low because the tail had not occurred, and the position that looked conservative against that history is enormous against the event it was always exposed to. The strategy did not change and the leverage did, and it was raised precisely because the risk had not yet appeared in the data.
A second failure is treating a positive average as a reliable income. The average includes the losses that have not happened yet.
A third is ignoring that the losses correlate with everything else in the portfolio.
A fourth is using leveraged short-volatility products, which can lose most of their value in one session.
And a fifth is calling it an anomaly. It is compensation for a real and occasionally severe risk, and treating it as free money is the error the payoff punishes.
Related
Volatility risk premium covers the closely related measure in volatility rather than variance terms. Volatility risk covers the exposure being compensated. And tail risk covers the events that make the premium necessary.
This is one of the most robustly documented effects in finance and one of the most dangerous to trade. The premium is real, it pays consistently, and the losses when it stops paying are large enough to erase years of collection — which is precisely why it exists.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.