Leveraging liquid and long-term volatility data for effective calibration of economic scenario generators

The calibration of models for insurance liabilities presents a significant challenge due to the mismatch between long-term horizons and the limited availability of market instruments. Insurers must manage long-term guarantees while relying on short-term market data. Traditionally, the calibration process has struggled to reconcile these long-term liabilities with available financial instruments, leading to potential mispricing and regulatory scrutiny.

This paper explores recent S&P Global Market Intelligence and Milliman research, featuring insights from Enrico Piccin, Pierre-Edouard Arrouy and Paul Bonnefoy. The study focuses on calibrating economic scenario generators using swaption volatility matrices informed by liquidity analysis. Key objectives include assessing the impact of incorporating very long-term volatilities (up to 70 years, without extrapolation) on best estimates, capital calculations, and asset-liability management (ALM) outcomes. The findings emphasise that a well-structured calibration process, combined with high-quality data, enables insurers to achieve accurate results without distorting ALM projections, ultimately enhancing financial stability.