Information design in competitive insurance markets

This paper characterizes the optimal information structure in competitive insurance markets with adverse selection. We consider a regulator that assigns ratings to individuals according to their expected costs. Insurers observe these ratings and compete as in Akerlof (1970). The optimal rating system minimizes ex-ante risk subject to participation constraints. We prove that in any such market there exists a unique optimal system under which all individuals trade and the ratings match low-cost types with high-cost types negative assortatively. We provide a simple algorithm that yields the optimal system and examine implications for government regulations of insurance markets.