Description: The purpose of this paper is to help practitioners and regulators more accurately quantify the potential impact of market risk on insurance products with equity-linked guarantees. To this end, this paper examines time series model selection and assessment based on residuals, with a focus on regime switching models. Also it discusses the difficulties in defining residuals for such processes and propose several possible alternatives. We determine that a stochastic approach to defining the residuals is the only way to generate residuals that are normally distributed under the model hypothesis.

Description: Regime switching models have become a popular tool in econometric time series modeling since their introduction in Hamilton [1989].These models have been widely applied for equity-returns for investment guarantees. A prominent risk management challenge is hedging these guarantees, given the equity regime switching process. This presentation develops some methodologies for regime-switching portfolio replication and use real world data to demonstrate and evaluate their effects. A Bayesian approach for regime switching model fitting is used, which allows for the examination of the robustness of each of the optimization methods when parameter uncertainty is present.

Description: A simple short-term disability model is developed, where the loss process is comprised of two processes: new claims process and continuing claims process. This results in a surplus function with stationary dependent increments. Three dimensional transitional probabilities are needed to model the transitions of this non-Markovian process. An approximating model with stationary independent increments is found. The approximating model leads to bounds for the finite probability of ruin. Numerical comparisons are made.

Description: This is the abstract on the paper on regime-switching portfolio replication.

Description: This presentation estimates several models of interest rates, including seven without and four with regime shifts.