Stats › Volatility and VIX › Variance risk premium
The variance risk premium
Options get priced as though the market will be more volatile than it usually turns out to be. Not occasionally, systematically. Over 1990 to 2024 the VIX averaged 19.59% while the volatility that actually followed averaged 15.50%, a gap of just over four percentage points. That gap is the variance risk premium, and the reason it exists isn't complicated. Selling an option is selling insurance, and nobody takes on that shape of risk for the expected value alone, so the price carries compensation for bearing it.
Which raises the obvious question, and that question is why this page exists. If option sellers are being overpaid by four points, why haven't the strategies built to collect it beaten the market? Cboe's own put writing index has returned 7.0% a year since 2007 against the S&P 500's 11.0%. Its covered call index has returned 8.5% since 1986 against 11.2%. Both cut volatility by roughly a third. Both trailed, and put writing didn't win on a Sharpe ratio basis either, at 0.51 against 0.61.
So the page carries the measurements, the academic work behind them, which establishes the premium as a separately priced risk rather than repackaged equity risk, what the writing indices actually delivered on return and on risk, and the story of an inverse volatility product that lost 96.3% in a single session in February 2018. Everything is sourced and dated at the bottom, including the index factsheets, which are short PDFs you can check in a minute.
Implied volatility is systematically higher than the volatility that follows. Over 1990 to 2024 the VIX averaged 19.59% against subsequent realised volatility of 15.50%, a gap of 4.09 percentage points. That gap is real, persistent and one of the better documented anomalies in finance. What it has not done is make money for the obvious strategies. Cboe's own put writing index has returned 7.0% a year since 2007 against the S&P 500's 11.0%, and its buy write index 8.5% against 11.2% since 1986. Both cut volatility by about a third and both trailed. And in February 2018 one inverse volatility product lost 96.3% in a single day.
Members only
The rest of this page is for members
Below this point there are 9 sections, 1 chart, 1 table and 5 named sources, roughly 1850 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.