A1442
Title: Sustainable health insurance: An application of GAMLSS for claims expenditure prediction
Authors: Fabio Baione - Sapienza University of Rome (Italy) [presenting]
Davide Biancalana - la sapienza Roma (Italy)
Abstract: The Principles for Sustainable Insurance represents the milestone for insurance industries to manage and reduce Environmental, Social, and Governance (ESG) risks. To this aim, the interest in developing the so-called green insurance solutions represents a challenge for life and non-life insurance companies. Within the offer of significant social value covers, we can refer to products that supplement the public health service, designed to help manage the costs of treatment and assistance, as well as the reduction in earnings of customers in the event of serious illnesses or the loss of self-sufficiency. We apply Generalized additive models for location, scale, and shape (GAMLSS) to price health insurance which provides coverage for several categories of medical expenditure. It is worth noting that the coverage exhibits high repeatability. Hence the number of claims for each policyholder per year can be much greater than one, with a high probability. Moreover, depending on the claims category, we observe low or high frequencies and highly skewed claim size distributions. GAMLSS is a flexible class of regression models for analyzing data allowing for a relevant extension of distributions assumed for the response variable even in the presence of truncated or censored data. Given their flexibility and thriftiness, they can represent a valuable tool for solving actuarial problems even in the presence of big data.
