VVIX is the Cboe "VIX of VIX" — a model-free, risk-neutral estimate of the expected 30-day volatility of the VIX index itself, derived from the prices of VIX options the same way VIX is derived from S&P 500 options. Where VIX measures the market's expected magnitude of spot moves, VVIX measures the expected magnitude of moves in volatility — a second-order, "vol-of-vol" reading. Its core tension is that VVIX is one layer more abstract than almost everything else traders look at: it can carry genuine forward information about tail-hedge demand and the convexity of the volatility surface, but that information is subtle, easily over-interpreted, and only legible to people who already understand the VIX options market it sits on top of.

How it's calculated / formed

VVIX uses the same generalized variance-swap replication formula as the VIX (the Demeterfi-Derman-Kamal-Zou framework), but applied to VIX options instead of SPX options. The engine integrates a strip of out-of-the-money VIX puts and calls across strikes, weighting each option price by 1/K², to extract the risk-neutral expected variance of the VIX over a 30-day horizon. Because VIX options expiring exactly 30 days out rarely exist, Cboe computes the measure from the two nearest VIX expirations straddling 30 days and interpolates between them, then annualizes — yielding VVIX in the same percentage-point units as VIX (Cboe VVIX white paper; the calculation mirrors the VIX methodology PDF).

Cboe launched VVIX on March 14, 2012, and backfilled the series to 2007 using historical VIX option quotes. Two consequences follow from the construction. First, VVIX inherits all the frailties of the VIX options market — wide bid-ask spreads, thin OTM strikes, and the fact that VIX options are priced off VIX futures, not spot VIX. Second, VVIX is structurally tied to the skew and convexity of VIX options: because traders systematically buy upside VIX calls as crash insurance, the right tail of the VIX option strip is usually rich, which keeps VVIX elevated relative to where a symmetric model would put it.

How it's used in practice

Practitioners use VVIX in three main ways.

• Level / regime gauge. VVIX's long-run mean is commonly cited near the mid-80s to low-90s, with a "normal" band roughly 80–110 (Cboe dashboard data; commonly cited ranges from volatility-trading commentary). Readings persistently below ~80 are read as complacency / cheap convexity; readings above ~120 signal that VIX-option buyers are paying up for large VIX moves — typically during stress or ahead of a known event.
• The VVIX/VIX divergence. The most-discussed application is watching VVIX rise while spot VIX stays low. This pattern says hedgers are accumulating VIX calls and OTM SPX puts even though realized calm persists — convexity demand building under a quiet surface. The frequently-cited example is the run-up to the February 5, 2018 "Volmageddon," when spot VIX sat in the low-to-mid teens while VVIX pushed into the ~100–115 area (Cboe data; widely reported in post-mortems). Some commentators formalize this as a VVIX/VIX ratio (e.g. ratio elevated with VIX subdued), though specific thresholds are heuristic, not validated rules.
• Pricing and timing vol trades. Vol-of-vol is a direct input to anything with VIX-option optionality: VIX call spreads, S&P put-spread collars, and dispersion/convexity books. A high VVIX means VIX-option premium is expensive — favoring sellers of that convexity (with tail risk) and penalizing buyers; a low VVIX cheapens tail hedges.

Adoption, debate & evidence

VVIX is a niche, professional indicator, not a retail mainstay — it lives in the world of volatility desks, tail-risk funds, and dealer-flow analysts. The strongest evidence for it being more than noise is academic. Yang-Ho Park's 2013 Federal Reserve FEDS paper, Volatility of Volatility and Tail Risk Premiums, found that VVIX has forecasting power for the returns of tail-risk hedges: when VVIX rises, it raises current prices of S&P 500 puts and VIX calls and thereby lowers their subsequent returns over the next ~3–4 weeks, consistent with rare-disaster theory. Park attributes the predictability mainly to integrated vol-of-vol and its risk premium rather than to jumps. Follow-on work (e.g. the ScienceDirect study on vol-of-vol and tail-risk hedging returns) reaches similar conclusions.

The honest caveat: this evidence is about the cost/return of hedges, not a clean "VVIX up → stocks down next week" timing signal. The folklore version — that a VVIX/VIX divergence reliably predicts crashes — is much weaker. Crashes are rare, divergences are not, and the famous cases (2018, 2020) are a handful of in-sample anecdotes. No widely accepted study establishes a tradeable equity-direction edge from VVIX divergence with out-of-sample base rates. Treat divergence as a conditioning variable (hedges are getting bid; convexity is in demand), not a trigger.

Strengths & limitations

VVIX works best as a second-order context layer: it surfaces stress that hasn't yet shown up in spot vol, and it correctly prices when tail insurance is cheap vs. expensive — a real, repeatedly-documented edge for hedge valuation. It is genuinely additive information beyond VIX alone.

It fails when used as a standalone directional or market-timing signal. VVIX is noisy, mean-reverting, and can stay elevated for long stretches without a crash; it can also spike on idiosyncratic VIX-option flow. The #1 misuse is treating a high VVIX or a VVIX/VIX divergence as a crash prediction and sizing equity bets on it — the empirical hit rate doesn't support that. A secondary trap is forgetting that VVIX is built on VIX futures-priced options, so it can move for term-structure/microstructure reasons unrelated to fundamental risk.

Sources

• Cboe — VIX & volatility-index methodology (variance-swap replication; VVIX construction): VIX Methodology PDF; Cboe VVIX white paper "Double the Fun with Cboe's VVIX Index."
• Cboe Global Indices — VVIX Dashboard (live levels, history).
• Yang-Ho Park (2013), Federal Reserve FEDS — Volatility of Volatility and Tail Risk Premiums (predictive power for tail-hedge returns; primary academic source).
• ScienceDirect — Volatility-of-volatility and tail risk hedging returns.
• Range/interpretation commentary (commonly-cited bands, divergence heuristics): SpotGamma, Convex/ConvexTrade glossary, Volatility Box. Flagged dispute: specific VVIX/VIX ratio thresholds and "divergence predicts crashes" claims are practitioner heuristics, not academically validated.