Quant with Vahab
Quant Systems Lab · Control Systems for Quantitative Finance

Implied Volatility Smile and Surface

The smile shows how implied volatility varies with strike; the surface extends this across maturities under no-arbitrage constraints.

Explanation

For each maturity, implied volatility as a function of strike forms a smile or skew, reflecting non-lognormal returns and supply–demand for options.

Across maturities, implied vols form a surface σ_imp(K, T); traders think and quote in terms of this surface rather than a single volatility number.

Arbitrage-free interpolation requires call prices to be decreasing and convex in strike and total variance σ²T to be non-decreasing in maturity.

Parametric forms such as SVI or SABR-style fits are used to build smooth, arbitrage-consistent surfaces that drive pricing, hedging, and risk for vanilla and structured products.


implied volsmilesurfaceno-arbitragesvisabr
Interactive visualisation

The smile is σ(K) at fixed maturity. The surface extends it across maturities. The diagnostics below check basic no-arbitrage shapes: calls decreasing/convex in strike, and total variance non-decreasing in maturity.

Strike KImplied vol σF ≈ 100.00 (ATM reference)T = 1.00607692108124140
Surface view (σ(K,T)) — Hover the surface for a value.0.25y0.50y1.00y2.00y607692108124140Strike K (x-axis) · Maturity T (y-axis)
Maturities shown (toggle):
Numbers
Slice T=1.00y: ATM σ(T)≈ 30.00%
Range: σ ∈ [28.68%, 69.76%]
Diagnostics (selected slice + surface)
Call monotone in K: ok (violations: 0)
Call convex in K: ok (violations: 0)
Calendar (w non-decreasing in T): ok (violations: 0)
Interpretation

The smile reflects how implied volatility changes with strike, which is shorthand for non-lognormal returns and option supply–demand. The surface extends this across maturities.

The diagnostics are deliberately basic. They are not a full arbitrage-free construction. They act as a guardrail: if you bend the sliders until these fail, you are likely building a surface that cannot correspond to valid call prices.

Practical modelling step: instead of inventing σ(K,T), many desks parameterise total variance w=σ²T and enforce smoothness and calendar structure there.