How do traders interpolate across the volatility surface when there's no quoted option at the exact strike and maturity they need?

Question

AcadiFi · Accepted Answer

Volatility surface interpolation is a critical practical skill that sits at the intersection of derivatives pricing and quantitative methods. The surface is sparse (options only trade at discrete strikes and expirations), but pricing requires continuous values. **The Volatility Surface:** Implied volatility varies along two dimensions: - **Strike dimension (skew/smile)**: IV changes with moneyness (K/S) - **Term structure dimension**: IV changes with time to expiration The surface combines both into a 3D landscape: IV = f(K, T) **Strike-Dimension Interpolation (Fixed Maturity):** **Method 1: Linear in Delta Space** Convert strikes to delta (option delta), then linearly interpolate IV between observed points. This is natural because delta captures moneyness more uniformly than strike. Example: You need IV at 30-delta but only have 25-delta (IV=32%) and 35-delta (IV=28%). IV(30) = 32% + [(30-25)/(35-25)] x (28% - 32%) = 32% - 2% = **30%** **Method 2: Cubic Spline** Fit a smooth cubic spline through observed IV points. This preserves the shape of the smile better than linear interpolation and avoids artificial kinks. **Method 3: SABR Model** A parametric model (Stochastic Alpha Beta Rho) that fits the entire smile with four parameters. Popular in interest rate derivatives. **Term-Dimension Interpolation (Fixed Strike):** Volatility across time is typically interpolated in **variance space** rather than volatility space: Total Variance = IV^2 x T Linear interpolation in total variance ensures no calendar arbitrage (negative forward variance). Example: 1-month IV = 20% (T=1/12), 3-month IV = 24% (T=3/12). Total var at 1M: 0.04/12 = 0.003333 Total var at 3M: 0.0576/4 = 0.014400 2-month total var: 0.003333 + [(2-1)/(3-1)] x (0.014400-0.003333) = 0.003333 + 0.005534 = 0.008867 2-month IV: sqrt(0.008867 x 6) = sqrt(0.053200) = **23.07%** ```mermaid graph TD A[Volatility Surface
Interpolation] --> B[Strike Dimension] A --> C[Time Dimension] B --> D[Linear in Delta
Simple, widely used] B --> E[Cubic Spline
Smooth, no kinks] B --> F[SABR Model
Parametric, 4 params] C --> G[Linear in Variance
Prevents calendar arb] C --> H[Square-root of Time
Crude approximation] style G fill:#c9a84c,color:#000 style D fill:#c9a84c,color:#000 ``` **Arbitrage Constraints:** The interpolated surface must satisfy: 1. **No negative forward variance**: d(IV^2 x T)/dT >= 0 (variance must increase with time) 2. **No butterfly arbitrage**: The smile can't be too steep (option prices must be positive for all strikes) 3. **Put-call parity**: Consistent across the surface **Why This Matters:** - Exotic option pricing requires IV at non-standard strikes/maturities - Risk management (VaR, greeks) needs a full, smooth surface - Mark-to-market for illiquid positions requires interpolated IVs **Exam Tip:** Know that time interpolation should be done in variance space (not volatility space), understand why (prevents calendar arbitrage), and be able to perform a basic interpolation calculation in both dimensions. Study advanced derivatives in our CFA Level II course.

How do traders interpolate across the volatility surface when there's no quoted option at the exact strike and maturity they need?

Master Level II with our CFA Course

Related Questions

Practice Questions