IMPLEMENTATION OF SAFETY FIRST RATIO IN PORTFOLIO OPTIMIZATION USING GENETIC ALGORITHM FOR INSURANCE INDUSTRY

• Main Author: Tan Putra, Howie
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/81500
• Institution: Institut Teknologi Bandung

In premium pricing, insurance companies typically rely on various assumptions, including the expected return from investments. Consequently, insurance companies must construct an investment portfolio that meets or exceeds the assumed return level while minimizing investment risk. Investments made are expected to be efficient, thereby providing the largest excess return ratio per unit of investment risk taken. To achieve these goals, the safety-first ratio can be employed. This ratio helps mitigate the risk of the portfolio falling below the assumed return level and optimizes investment allocation. The investment activities of insurance companies in Indonesia are regulated by the Otoritas Jasa Keuangan (OJK). Consequently, when constructing an investment portfolio, insurance companies must adhere to the constraints established by the OJK, in addition to other common constraints like buy-in threshold and round lot constraints. The inclusion of these constraints and the variety of assets in portfolio determination introduce complexity in finding the optimal portfolio. This complex optimization problem can be solved using a metaheuristic method. One of the metaheuristic methods that is applicable is genetic algorithm. The application of safety first ratio using genetic algorithm on varying return rate produces an efficient frontier similar to the results obtained from applying the Markowitz model to the same portfolio optimization problem. This research provides optimal asset allocation for the insurance industry in Indonesia, aiming to maximize the safety first ratio for a desired level of return.