VIX vs realized volatility: the historical gap
VIX is a 30-day forecast of S&P 500 volatility, priced from options. Compare it to the volatility the S&P actually delivered over the following month and a clear pattern shows up: VIX is usually too high — by an average of about 3–4 points since 1993, and higher roughly 85% of the time.
The gap is the premium
That persistent overshoot is the volatility risk premium: option buyers pay for protection they usually don’t need, and the average outcome is that implied volatility was priced richer than what followed. It’s widest when markets are calm — precisely when insurance feels least necessary and is therefore under-demanded relative to the eventual risk.
Except when it isn’t
The relationship flips in a crisis. When a shock hits, daily moves blow past anything the options market had priced, realized volatility overshoots VIX, and the gap turns sharply negative. Every major drawdown in the sample — 2008, the February 2018 “volmageddon”, March 2020, 2022 — shows realized above implied for a stretch. The premium you collect in the calm is repayment for those episodes.
The full series, and where it is now
The S&P 500 page charts implied against subsequent realized month by month back to 1993, breaks the premium out by year and by volatility regime, and shows a “right now” gauge of the current gap versus its own history.
Open the S&P 500 page →Common questions
Is VIX usually higher than realized volatility?
Yes — historically about 85% of the time for the S&P 500. On average VIX has sat roughly 3–4 points above the realized volatility that followed over the next month. That persistent gap is the volatility risk premium.
When is realized volatility higher than VIX?
During and just before crashes. When a shock hits, actual daily swings blow past what options were pricing, so realized volatility overshoots implied and the gap goes sharply negative — 2008, February 2018, March 2020, and 2022 all show it.
Does a low VIX mean stocks are safe?
No. VIX reflects the market's current expectation, which is often lowest right before volatility spikes. The premium is actually largest in calm regimes and inverts when fear arrives.
Research and education only — not investment advice. Realized volatility here is close-to-close over the next 21 trading days, annualized.