What Is the VIX and How Is It Calculated?

The variance-swap math behind the fear gauge.

The VIX is the CBOE's measure of how much volatility the market expects in the S&P 500 over the next 30 days, expressed as an annualized percentage. When traders ask how is VIX calculated, the short answer surprises most of them: it does not come from a single at-the-money option. It comes from a model-free variance-swap replication that aggregates the prices of an entire strip of out-of-the-money puts and calls, weighted by 1/K². That design makes it a cleaner read on aggregate demand for protection than any single strike could ever be. To see why that matters, it helps to understand how implied volatility works at a single strike — and then appreciate why the VIX deliberately avoids relying on one.

The core insight: variance swaps, not ATM options

A variance swap is a contract whose payoff is the difference between realized variance over a period and a fixed variance strike agreed at inception. It can be replicated — in theory, perfectly — by holding a continuous strip of options across all strikes, with each strike weighted by 1/K². The VIX methodology is a discretized version of exactly that replication.

The conceptual payoff is large: no model assumptions about the shape of the distribution are needed. The price of the variance swap is pinned by no-arbitrage, not by Black-Scholes or any other parametric model. That is what "model-free" means in this context. If the market bids up OTM puts — because portfolio managers are buying crash protection — those higher prices flow directly into the VIX through the sum, without any model translating them. The VIX therefore rises when fear rises, almost by construction.

The formula, term by term

The CBOE defines the VIX variance as:

σ² = (2/T) · Σ (ΔK_i / K_i²) · e^rT · Q(K_i) − (1/T) · (F/K₀ − 1)²

VIX itself is then σ · 100 (the square root of σ², annualized). Each symbol does specific work:

• T — time to expiration in years. Dividing by T scales the result to a per-year variance.
• ΔK_i — the interval between adjacent strikes around strike K_i, typically half the distance to the neighboring strikes on each side. It acts as the integration weight in a Riemann sum approximating a continuous integral.
• K_i² — the square of the strike in the denominator. This is the 1/K² weighting that comes directly from the variance-swap replication math.
• e^rT — discounting in reverse: it brings the option prices, which are already present values, back to forward space so the sum represents the forward variance.
• Q(K_i) — the midpoint of the bid-ask spread of the out-of-the-money option at strike K_i. Puts are used below the forward price F, calls above it, and at-the-money both are averaged.
• F/K₀ − 1 — the convexity correction. F is the forward price implied by put-call parity; K₀ is the nearest listed strike below F. Because listed strikes are discrete, F and K₀ rarely coincide, and this term removes the resulting bias.

When you use an options pricing calculator to back out implied vols at different strikes, you are computing the inputs that go into Q(K_i). The VIX machinery aggregates all of those Q values simultaneously.

Interpolating to a constant 30-day horizon

SPX options expire on fixed calendar dates, not on a rolling 30-day cycle. On any given day, the two nearest expiries straddle the 30-day mark. The CBOE computes σ² separately for the near-term expiry (T₁, typically the next Friday expiry) and the next-term expiry (T₂), then interpolates linearly in variance-time to pin down exactly 30 days:

VIX² = 100² · [ w₁ · σ₁² · (T₂ − 30/365) / (T₂ − T₁) + w₂ · σ₂² · (30/365 − T₁) / (T₂ − T₁) ]

where w₁ and w₂ are time-weighting factors that ensure the weighted average maturity lands on exactly N₃₀ = 30 × 1440 minutes. The CBOE actually measures T in minutes to avoid day-count ambiguity around weekends and holidays. The result is a single, consistent 30-day variance number regardless of where on the calendar you are.

Why OTM puts drive the VIX higher

Portfolio managers hedging tail risk buy OTM puts — say, SPX 5% or 10% below spot. When demand surges, put prices rise, and since those strikes enter the sum with positive weight, σ² increases and the VIX moves up. The 1/K² weighting means that lower-strike puts get progressively larger weight per dollar of notional, amplifying the effect of deep-strike put buying relative to near-ATM options. This is why a spike in demand for crash protection — even without any movement in the index itself — can push the VIX up sharply. The VIX is fundamentally a measure of the cost of protection, not just a statistical forecast of future swings.

Annualization and what VIX 20 actually means

σ² from the formula is already scaled to annual units via the 1/T and 2/T prefactors. Taking the square root and multiplying by 100 gives the annualized volatility in percentage points. A VIX reading of 20 means the market implies a daily move of roughly 20 / √252 ≈ 1.26% (one standard deviation) in the S&P 500. It does not mean a 20% move is expected over the next 30 days — that misreading is the single most common error in financial commentary.

The VIX is also forward-looking by construction. It tells you nothing directly about how volatile the market has been; it prices how volatile the options market thinks it will be. That divergence between implied and realized volatility is itself a tradeable quantity — the volatility risk premium — and historically the VIX has run above subsequent realized vol more often than not, which is why systematic short-volatility strategies have had structural positive carry.

Common misreadings to avoid

• VIX is not the probability of a crash. It is a price, not a probability. Converting it to a probability requires additional distributional assumptions.
• VIX does not predict direction. A high VIX is compatible with a rally — it measures the cost of optionality, not the direction of the underlying.
• VIX is not tradeable directly. VIX futures and options trade on a different term structure than the spot VIX, and they do not converge to spot VIX in a simple way. The roll cost in VIX futures products can be severe during contango.
• The 30-day horizon is an interpolated construct. No single option has a 30-day expiry on most days; the number is synthesized, so it inherits any illiquidity in the component strikes.

The VIX endures as a benchmark because the variance-swap framework is principled — it is grounded in no-arbitrage replication theory, not a heuristic. Understanding the math also clarifies its limits: it is the price of a specific kind of insurance, measured at a specific horizon, across a specific universe of listed strikes. Treat it as that, and it is genuinely informative.