Consumer information can be an important source of competitive advantage, especially in selection markets. How can we measure such information advantage? How do policies that limit this advantage affect market structure? We develop a method to analyze competitive equilibrium when firms have different products, cost structures, and information. Using a decade of market-wide micro-data from the Italian auto insurance industry, we document large differences in how accurately major firms assessed risk: the best outperformed the worst by 74%. A counterfactual policy that removes information advantage would curb information rent but concentrate the market towards low-cost insurers. An algorithmic transparency mandate that augments every insurer’s information precision decreases the cost to insure consumers by 4 euros per contract but hurts high-risk consumers.