Abstract:This work describes statistical modeling of detailed, microlevel automobile insurance records. We consider 1993–2001 data from a major insurance company in Singapore. By detailed microlevel records, we mean experience at the individual vehicle level, including vehicle and driver characteristics, insurance coverage, and claims experience, by year. The claims experience consists of detailed information on the type of insurance claim, such as whether the claim is due to injury to a third party, property damage to a third party, or claims for damage to the insured, as well as the corresponding claim amount. We propose a hierarchical model for three components, corresponding to the frequency, type, and severity of claims. The first model is a negative binomial regression model for assessing claim frequency. The driver’s gender, age, and no claims discount, as well as vehicle age and type, turn out to be important variables for predicting the event of a claim. The second is a multinomial logit model to predict the type of insurance claim, whether it is third-party injury, third-party property damage, insured’s own damage or some combination. Year, vehicle age, and vehicle type turn out to be important predictors for this component. Our third model is for the severity component. Here we use a generalized beta of the second kind of long-tailed distribution for claim amounts and also incorporate predictor variables. Year, vehicle age, and person’s age turn out to be important predictors for this component. Not surprisingly, we show a significant dependence among the different claim types; we use a t-copula to account for this dependence. The three-component model provides justification for assessing the importance of a rating variable. When taken together, the integrated model allows more efficient prediction of automobile claims compared with than traditional methods. Using simulation, we demonstrate this by developing predictive distributions and calculating premiums under alternative coverage limitations.

PREDICTIVE CLAIM SCORES FOR DYNAMIC MULTI-PRODUCT RISK CLASSIFICATION IN INSURANCE

A Claim Score for Dynamic Claim Counts Modeling

Customer Price Sensitivities in Competitive Automobile Insurance Markets

Frequency-severity experience rating based on latent Markovian risk profiles

Boosting insights in insurance tariff plans with tree-based machine learning methods

GAMLSS for Longitudinal Multivariate Claim Count Models

Individual Claims Reserving using Activation Patterns

A Posteriori Risk Classification and Ratemaking with Random Effects in the Mixture-of-Experts Model

Dynamic online prediction model and its application to automobile claim frequency data

Enhanced Gradient Boosting for Zero-Inflated Insurance Claims and Comparative Analysis of CatBoost, XGBoost, and LightGBM

Hierarchical Insurance Claims Modeling

Adaptive Pricing in Insurance: Generalized Linear Models and Gaussian Process Regression Approaches

A multi-task network approach for calculating discrimination-free insurance prices

A data science approach to risk assessment for automobile insurance policies

Three-step Risk Inference in Insurance Ratemaking

Bonus-Malus scale premiums for tweedie's compound poisson models

Generalized Additive Modelling of Dependent Frequency and Severity Distributions for Aggregate Claims

Machine Learning in Ratemaking, an Application in Commercial Auto Insurance

Bivariate Gamma Mixture of Experts Models for Joint Insurance Claims Modeling

Estimation of foreseeable and unforeseeable risks in motor insurance