CASE STUDY: RSP PROFITABILITY MAXIMIZATION

Context

• Large automobile insurer with important volume of business in Québec
• Losses incurred in the last years on policy transfers to Québec's risk-sharing pool (RSP), the Plan de Répartition des Risques managed by the Groupement des assureurs automobiles (GAA)

Problem

• Make RSP transactions profitable by ceding the most undercharged risks
• Replace the existing risk selection model with a better performing model for ceding risks
• Implement a solution so that cessions are performed on a regular basis, as new business and renewals are processed

Solution

• Using historical policy and claims data, ApSTAT developed and compared hundreds of risk estimating predictive models in order to choose the best performing (profitable) with respect to the particular RSP challenge. Monthly, the system sends a series of recommendations:
  • New business to be ceded
  • Upcoming renewals to be ceded at renewal assuming the policy remains in force
  • Risks currently ceded that, given recent policy changes, should be pulled back from the RSP into the insurer's portfolio

Benefits

• Increased profitability of 0.5% of the insurer's volume for its RSP transactions

To Know More

Read the white paper Data Mining Techniques for Optimizing Québec’s Automobile Risk Sharing Pool.

« Back to Case Studies