On valuing participating life insurance contracts with conditional heteroscedasticity

T.K. Siu, John Lau, H. Yang

    Research output: Contribution to journalArticle

    2 Citations (Scopus)

    Abstract

    In this paper, we consider a novel approach for the fair valuation of a participating life insurance policy when the dynamics of the reference portfolio underlying the policy are governed by an Asymmetric Power GARCH (APGARCH) model with innovations having a general parametric distribution. The APGARCH model provides a flexible way to incorporate the effect of conditional heteroscedasticity or time-varying conditional volatility and nests a number of important symmetric or asymmetric ARCH-type models in the literature. It also provides a flexible way to capture both the memory effect of the conditional volatility and the asymmetric effects of past positive and negative returns on the current conditional volatility, called the leverage effect. The key valuation tool here is the conditional Esscher transform of Bühlmann et al. (1996, 1998). The conditional Esscher transform provides a convenient and flexible way for the fair valuation under different specifications of the conditional heteroscedastic models. We illustrate the practical implementation of the model using the S&P 500 index as a proxy for the reference portfolio. We also conduct sensitivity analysis of the fair value of the policy with respect to the parameters in the APGARCH model to document the impacts of different conditional volatility models nested in the APGARCH model and the leverage effect on the fair value. The results of the analysis reveal that the memory effect of the conditional volatility has more significant impact on the fair value of the policy than the leverage effect.
    Original languageEnglish
    Pages (from-to)255-275
    JournalAsia-Pacific Financial Markets
    Volume14
    Issue number3
    DOIs
    Publication statusPublished - 2007

    Fingerprint Dive into the research topics of 'On valuing participating life insurance contracts with conditional heteroscedasticity'. Together they form a unique fingerprint.

    Cite this