Alternative Modeling of Surrender Risk in Life Insurance Using the Fuzzy Logic

Karima Lamsaddak,Driss Mentagui

Nowadays, the regulatory and prudential framework of the Moroccan insurance sector is undergoing significant changes in terms of regulation. With the proposed introduction of risk-based solvency, insurers must properly manage the future exposed risks including them in the technical provisions calculation. Also, with regard to the taxation of retirement savings products, Moroccan regulation has undergone since 2018 some changes, through the possibility to transfer the retirement savings without tax impact (Morocco finance Law, 2018). Given their impact on policyholders' behavior, these changes should be considered in the modeling of the future claims behavior. As a result, conventional modeling isn’t sufficient to predict claims experience marked by uncertainty, imprecision and insufficient information. It’s within this framework that modeling a surrender risk adopting the fuzzy logic approach constitutes a well-adapted alternative and it is the aim of this paper. This approach is based on three categories of factors: the characteristics of the insured, of the product and those related to the insured management. Each category is modeled separately by the fuzzy logic approach; their outputs will be injected as inputs in the final modeling which explains the loss experience of the surrender and classifies it into three fuzzy sets: low risky, medium and high risky.

Volume 11 | 05-Special Issue

Pages: 1781-1791