Insurance Pricing Optimization via Off-Policy Evaluation

2026-05-27

AI Summary

Traditional insurance pricing often neglects policyholders’ price sensitivity, hindering revenue optimization. This work formulates pricing as a sequential decision-making problem and introduces an optimization framework that integrates off-policy evaluation with stochastic control. The proposed approach employs a kernelized inverse propensity score estimator that leverages the local structure of the action space to reduce variance and combines an interpretable Lasso model with neural networks for policy parameterization. Experimental results in a synthetic travel insurance environment demonstrate that the method significantly outperforms existing techniques, achieving higher pricing revenue while preserving policy interpretability.

Abstract

Traditional insurance pricing relies on risk-based principles that ensure actuarial fairness and solvency but do not explicitly account for policyholders' price sensitivity. We formulate insurance pricing as a decision-making problem and study it using tools from off-policy evaluation and stochastic control. We propose a kernelized inverse propensity score estimator that exploits local structure in the action space and yields variance reduction compared to the classical inverse propensity score estimator. Building on these value estimates, we investigate policy optimization and present two practical approaches for computing optimal pricing rules: an interpretable data-shared Lasso formulation and a flexible policy parameterization based on neural networks. Using a controlled synthetic travel insurance environment, we empirically confirm the theoretical results and show that neural networks outperform existing techniques for policy optimization.

Problem

• insurance pricing
• price sensitivity
• off-policy evaluation
• policy optimization
• actuarial fairness

Innovation

• off-policy evaluation
• kernelized inverse propensity score
• insurance pricing optimization
• neural network policy
• variance reduction