Generating Economics Scenarios for the Long-Term Solvency Assessment of Life Insurance Companies: The Orthogonal ARMA-GARCH Method

Constructing the models that can generate possible economic scenarios of the returns on major asset classes is essential for solvency assessment. The key issues in establishing a comprehensive ESG models include: how to deal with the large number of risk factors, how to model the dynamics of some chosen factors, and how to incorporate the relations among risk factors. We propose the orthogonal ARMA-GARCH method to tackle these issues. This method includes applying factor analysis to asset class so that modeling dimension can be significantly reduced. Furthermore, we may model the relations among the risk factors within an asset class by common factors in addition to using correlated random shocks. We may also enjoy great flexibilities in establishing the time-series models for individual factors since the retrieved common factors can be orthogonal to each other. Therefore, our O-GARCH modeling is computationally efficient (modeling fewer factors), econometrically appropriate (providing fitness statistics as well as using general time-series models), and economically sound (by using common factors in addition to random shocks).