Three essays in financial institutions and investments.

摘要

Essay one examines constraints that borrowers and syndicate arrangers sometimes employ that restrict a loan's liquidity. We analyze two types of such constraints on the resales of syndicated loans: (1) a prior consent constraint implemented by the borrower or lead loan arranger and (2) a minimum denomination requirement for loan sales. We hypothesize that constraints could be mechanisms for fostering relationships and/or facilitating the resolution of financial distress. We find that constraints are more likely when borrowers are small and have relatively poor credit ratings. We also find that loans with any type of constraint have higher all-in-spreads than unconstrained loans and the marginal cost of constraining liquidity is high.; Essay two examines commercial-bank participation in venture capital activity that culminates in an initial public offering. Following Hellmann, Lindsey, and Puri (2004), we examine the "future prospects" hypothesis that contends banks target their venture capital activity in order to create synergies between their investment and lending practices. Using firm-specific data, we find no support for this notion. We also examine the impact of commercial-bank venture capital involvement on the initial and short-run returns of IPOs. We find no significant relation between bank involvement and initial returns. We do find that during the "bubble" period of 1999-2000, IPOs with bank involvement in their venture capital syndicates have 30-day returns over 30% greater than those issues without bank involvement. We conclude that commercial banks have much the same motivation and impact as any other venture capitalist, at least in the case of those that are associated with IPOs.; Essay three examines the long-run underperformance of initial public offerings (IPOs) relative to seasoned issues. Although this underperformance is heavily documented, researchers have been unable to determine with confidence which IPO characteristics affect underperformance. Previous studies examining IPO long-run returns are plagued by biases associated with matched samples and overlapping time periods. We use a unique approach that concentrates on pairs of IPOs issued on the same day, thereby avoiding many of the biases associated with previous studies. In contrast to Brav and Gompers (1997), we find no relation between venture capital (VC) involvement and long-run returns. We also find no support for the proposed extended commitment hypothesis, as commercial bank involvement in VC activity has no relation to returns. We do find issues with lower 30-day returns and higher quality underwriters tend to have larger long-run returns.