Next, we present evidence that the firm-level real effects of universal banking derive from informational economies of scope across divisions, not from higher bank revenues from cross-selling. To demonstrate this, we ex- ploit mergers among commercial and investment banks as a source of variation in the resulting universal banks’ in- formation about borrower firms. We consider firms that received both a loan from a commercial bank and an un- derwriting product from an investment bank, and we then contrast two groups: (1) those firms whose lender and underwriter merged with each other and (2) those firms whose lender and underwriter did not merge with each other, but instead merged with other banks to form uni- versal banks.