In this paper, we present a new approach to measuring interest rate risk for insurers within the Swiss Solvency Test, which overcomes the shortcomings of the standard model. The standard model of the Swiss Solvency Test is based on more interest rate risk factors than are actually needed to capture interest rate risk, it allows for significantly negative interest rates and it tends toward procyclical solvency capital requirements. Our new approach treats interest rate risk with direct reference to the underlying term structure model and interprets its parameters as a canonical choice of the relevant interest rate risk factors. In this way, the number of interest rate risk factors is substantially reduced and interest rate risk measurement is linked to the term structure model itself. The consideration of empirical interest rate data and the acceptance of the economical implausibility of persistently negative interest rates significantly below the cost of holding cash motivate the introduction of a truncated Gaussian process to simulate innovation in the future development of the parameters of the underlying term structure model. In a natural way this leads to mean-reverting interest rate behaviour and to countercyclical solvency capital requirements.
interest rate risk yield curve truncated Gaussian process Swiss Solvency Test