SVI and SSVI
Parametrise implied total variance slices and surfaces, with conditions for avoiding static arbitrage.
How it works
SVI represents implied total variance as a function of log-forward moneyness for one expiry. SSVI links slices with a specific surface parametrisation. Fitting total variance gives direct control over the shape of a smile, but price convexity and consistency across expiries must also be imposed or checked.
Worked example
At fixed forward moneyness, consider an annualised implied volatility of 20% at one year and 18% at two years. Total variances are 0.0400 and 0.0648 respectively. Volatility declines as expiry rises but total variance increases. A calendar screen driven directly by the volatility numbers would misread this example.
Validation and limitations
Fit in consistent forward and discount conventions, check the permitted parameter conditions and calculate butterfly and calendar diagnostics on a dense grid. Sparse data can produce smooth fitted slices with unreliable wing slopes. Compare with the price-wing construction in No-Arbitrage Smile Extrapolation.