On the application of efficient hybrid heuristic algorithms – An insurance industry example

This paper proposes an optimization approach for generating an investment strategy for multi-period asset-liability management of long-term with-profit life insurance policies. Our approach uses models to simulate the processes insurance companies employ when determining multi-period investment strategies over a given planning horizon. The approach utilizes an enhanced heuristic algorithm to determine optimal multi-period investment strategies. Simulation models take into account asset numbers, objective functions, and asset allocation frequency. Strategy performance is evaluated by applying three single-period investment strategies to the simulation models. Computational results not only verify the efficiency and robustness of the algorithm, but also demonstrate the effectiveness of frequent asset reallocation, and dispute the suitability of traditional top-down investment strategies in maximizing investment returns of with-profit insurance policies.

Figure optionsDownload as PowerPoint slideHighlight
► An efficient hybrid algorithm combining evolution strategies and the Levenberg–Marquardt algorithm has been developed.
► The optimization approach is used to generate an investment strategy for multi-period asset-liability management of long-term with-profit insurance policies.
► Our experiment results show improved returns based on increased asset reallocation frequency.
► Our study disputes the traditional notion of the top-down investment approach to long-term with-profit liability.