I study the coexistence of formal and informal finance in underdeveloped credit markets. While weak institutions constrain formal banks, shallow pockets hamper informal lenders. In such economies, informal finance has two effects. By increasing the investment return it decreases borrowers’ relative payoff following default, inducing banks to lend more liberally (disciplinary effect). By channeling bank capital it reduces banks’ agency costs from lending directly to borrowers, limiting banks’ extension of borrower credit (rent-extraction effect). Among other things, the model shows that informal interest rates are higher, borrower welfare lower, and informal finance more prevalent when the rent-extraction effect prevails, consistent with stylized facts in poor societies.