How To Use Public-Private Databases In Insurance Risk Management: Geography, Climate And People In Motor Insurance

Abstract

This work focuses on the use of public information sources in the application of relational models in insurance companies, for a better understanding of risks and assisting decision-making in new sustainability environments. Firstly, we propose using Eurostat's degree of urbanization methodology to group motor claims or policies into potentially more homogeneous categories in the insurance sector (urban / suburban / rural) for segmentation and analysis. Secondly, we analyze how insurance companies can use local weather information in conjunction with the degree of urbanization to model the number of motor claims in a specific geographic area. Finally, we apply relational models to databases with anonymized information on passengers in traffic accidents provided by the Spanish General Traffic Directorate for the purpose of better defining the characteristics of the claim based on the profile of the people inside the vehicle. It is about knowing, for example, the profile of the passengers in vehicles driven by elderly people, also in conjunction with sex and the geographical area. Insurance companies know the enormous potential of data analytics and must focus on the search for relationships using information that may be dispersed in multiple databases, including those that are for public use and that can facilitate the homogenization and comparison of results, together to the design of preventive and risk management policies. We also include the R codes making them available to the insurance sector and academia for use.