كنفرانس بين المللی مهندسی صنايع و سيستم ها

صفحه اصلی / دوازدهمین كنفرانس بين المللی مهندسی صنايع و سيستم ها

Dynamic Insurance Pricing Optimization under Price-Dependent Risk Selection: An Auto Insurance Case Study

نویسندگان :

Melika Rezaei¹ Hadi Mosadegh²

1- دانشگاه صنعتی امیرکبیر (پلی‌تکنیک تهران) 2- دانشگاه صنعتیِ امیرکبیر (پلی‌تکنیک تهران)

کلمات کلیدی :

Dynamic pricing،Automobile insurance،Insurance optimization،Risk selection،Mixed-integer programming،CVaR

چکیده :

This paper develops a multi-period optimization model for dynamic insurance pricing under price-dependent risk selection. Conventional demand-based insurance pricing usually allows premiums to affect the number of retained contracts while treating the loss distribution of the retained portfolio as fixed. This assumption is restrictive when policyholders differ in both expected claim cost and price sensitivity. The proposed model divides an automobile insurance portfolio into heterogeneous risk segments and optimizes segment-level premiums over a finite horizon. Premiums determine renewal probabilities, renewal probabilities determine the retained portfolio, and the retained portfolio determines future premium revenue and aggregate claim losses. A scenario-based mixed-integer linear programming formulation is developed with practical constraints on premium variation and minimum portfolio retention. Downside underwriting performance is evaluated through Conditional Value-at-Risk. A semi-synthetic automobile insurance experiment shows that modeling price-dependent risk selection improves expected profit and lower-tail performance compared with actuarial markup and aggregate-retention benchmarks.