New firm finance and the feasibility of bank lending

This dissertation is concerned with the role of borrower reputation in bank lending. We show that a firm's reputation for being honest allows it to obtain a bank loan with limited collateral, while a reputation for being capable of creating safe investment opportunities requires substantial collateral to make bank lending feasible. These results hinge on the weakness of the reputation effects described in Diamond (1991) and on the assumption that banks base lending decisions on publicly available information. If these reputation effects are strong or if banks obtain proprietary information about firms over time, a borrower's reputation for being honest is not relevant.; Using The National Survey of Small Business Finance we find evidence suggesting these reputation effects are not important in lending relationships. We also find evidence that banks base lending decisions on publicly available information, they do not obtain proprietary information about firms over time.; We then conclude that if a lender is sufficiently pessimistic about a firm's honesty, the firm will never be able to build a sufficient reputation for being honest in order to obtain a relatively unsecured loan--i.e. the firm is caught in a leverage trap. This finding suggests how subtle forms of lending bias may affect the flow of credit into communities characterized by a high degree of asymmetric information. The importance of Diamond-type reputation effects also has implications for the efficacy of community lending programs. If these reputation effects are weak, we find that programs using community pressure to induce repayment inhibit reputation building necessary for borrowers to obtain unsecured loans.