Análise dos valores de mercado e dos prêmios de seguro de veículos em uma região do interior paulista

Abstract

This study examined the relationship between the average market values of popular vehicles and their corresponding automobile insurance premiums in the Greater Campinas region from 2016.1 to 2021.2. The research began with the consolidation and standardization of datasets from SUSEP and the Fipe Table (via Kaggle), organized in Excel to ensure consistency across vehicle models and semesters. Six widely marketed vehicles in Brazil were analyzed: Ford Ka, Hyundai HB20, Nissan March, Renault Sandero, Toyota Etios, and Volkswagen Gol. The descriptive analysis revealed substantial variation in both market prices and insurance premiums across models and over time, indicating that these variables do not evolve proportionally. These findings supported the use of two complementary statistical approaches. The Beta regression model, applied to the premium-to-value index, showed a downward trend in several semesters and highlighted differences between vehicle models and policyholder profiles. The Gamma regression model, used to analyze premiums in absolute terms, confirmed the influence of vehicle value, average claim cost, and exposure on insurance pricing. Diagnostic and calibration analyses indicated satisfactory fit for both models. Overall, the results show that insurance premiums and vehicle market values do not follow parallel trajectories over time. While some models exhibited significant appreciation, the relative cost of insurance decreased, suggesting structural changes in portfolio behavior and in the insurance market. The combined use of relative and absolute modeling approaches provided a comprehensive understanding of how multiple factors interact in the formation of automobile insurance premia.