TWO ESSAYS ON BORROWING FROM BANKS AND LENDING SYNDICATES

A loan deal is often composed of several components (for example, a 3-year revolving loan, a 10-year secured senior term loan, and a 5-year subordinated term loan). The division of a deal into two or more components, each with different risk characteristics, is called tranching. This study recognizes the importance of tranching and establishes tranching as an integral component of a syndicated loan structure. In the first essay, we present a model to explain the economic value of tranching and show that riskier firms are more likely to take loans with multiple tranches. Therefore, the average credit spread on syndicated loans with multiple tranches is higher than that on nontranched loans. However, after accounting for the risk characteristics of a tranched loan, we show that a given tranche of a multi-tranche loan, on average, has a lower credit spread than an otherwise similar loan that is not part of a multi-tranche loan. We also show that the benefits of tranching accrue primarily to borrowers with speculative debt ratings. Prior studies have found an abnormal stock return of 100 to 150 basis points for firms that announce they have borrowed funds from a bank. Despite some conflicting evidence (Peterson and Rajan, 2002; Thomas and Wang, 2004; Billett, Flannery and Garfinkel, 2006), the literature tends to interpret this positive bank loan announcement effect as the markets response to the mitigation of information asymmetry regarding the borrowing firm caused by the certification role of the lending banks who act as quasi-insiders. In the second essay, we document that a strong selection bias exists in prior studies. We show that less than a quarter of the loans made by banks are ever announced by borrowing firms and the loans that are announced are systematically different from loans that are never announced by the firms. Firms with low debt ratings, firms with zero or negative profits but positive interest expense, firms that take large loans in relation to their assets base, firms with little analyst following, and firms with high forecasted EPS growth are more likely to announce their loans. We show that while there was a positive announcement effect over the period 1987 to 1995, loan announcements elicited zero or negative returns in the last ten years as the mix of companies announcing loans changed over time