Publicação

Price elasticity in auto insurance: impact of premium fluctuations on policyholder behavior

Ver documento

Detalhes bibliográficos
Resumo:This project aims to explore the price elasticity of demand in the auto insurance mar ket. It will use machine learning models, such as the Logistic Regression model and Gradient Boosting model, to predict how price increases can influence policyholder behavior, not disregarding other key variables such as policyholder demographics, vehicle characteristics, and payment frequency. The models were trained using historical data from an insurance company covering the years 2020 to 2022. The Gradient Boosting model, which performed better, was also tested using price increase simulations to evaluate its performance and how it could lead to policy cancellations and revenue loss. This test revealed a nonlinear relationship. Addressing consumer behavior when there’s a premium change will help insurance companies determine better strategies to retain their policyholders while staying competitive and profitable. The findings suggest that not only do price fluctuations strongly influence policy cancellations, but other variables such as policyholder demographics, vehicle characteristics, and payment frequency also play an essential role in assessing the reasons that lead to policy cancellations. This research is important to understand how insurance companies can adapt their premiums in order to not lose customers or profitability.
Autores principais:Vaz, Filipa Correia Inês e Correia
Assunto:Auto Insurance Price Elasticity Logistic Regression Model Gradient Boosting Model Policyholder Demographics Vehicle Characteristics Seguro Automóvel Elasticidade de Preço Regressão Logística Gradient Boosting Dados demográficos dos segurados Características dos veículos
Ano:2024
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso aberto
Instituição associada:Universidade de Lisboa
Idioma:inglês
Origem:Repositório da Universidade de Lisboa
Descrição
Resumo:This project aims to explore the price elasticity of demand in the auto insurance mar ket. It will use machine learning models, such as the Logistic Regression model and Gradient Boosting model, to predict how price increases can influence policyholder behavior, not disregarding other key variables such as policyholder demographics, vehicle characteristics, and payment frequency. The models were trained using historical data from an insurance company covering the years 2020 to 2022. The Gradient Boosting model, which performed better, was also tested using price increase simulations to evaluate its performance and how it could lead to policy cancellations and revenue loss. This test revealed a nonlinear relationship. Addressing consumer behavior when there’s a premium change will help insurance companies determine better strategies to retain their policyholders while staying competitive and profitable. The findings suggest that not only do price fluctuations strongly influence policy cancellations, but other variables such as policyholder demographics, vehicle characteristics, and payment frequency also play an essential role in assessing the reasons that lead to policy cancellations. This research is important to understand how insurance companies can adapt their premiums in order to not lose customers or profitability.