Kristin Elizabeth Knox
Business Economics PhD
Dissertation Chair: Prof. J. Stein
Essays in Financial Economics: How Does the Presence of Collateral Affect Banks' Views of Takeover Defenses?; How Does Collateral Affect Loan Spreads?; Earnings Preannouncements (joint with Randolph Cohen and Tuomo Vuolteenaho)
This dissertation consists of three essays in financial economics. The first examines the impact of one aspect of the market for corporate control - takeover defenses - on a firm's cost of debt financing. Specifically, I examine the relationship between the strength of a borrower's takeover defenses, the presence of collateral, and the price paid for bank loans. I find that borrowers with strong takeover defenses pay more for their secured loans relative to borrowers with weak takeover defenses.
The second essay seeks to measure the effect of collateral on a firm's cost of bank financing. Estimation is complicated by the fact that whether or not a loan is secured is endogenous - there is unobserved, borrower-specific, time-varying risk that affects both the price a firm must pay to borrow and whether that firm borrows on a secured or unsecured basis. I use the amount a firm paid on its last loan of the same type (secured or unsecured) as a proxy for firm-specific risk. This greatly reduces, but does not eliminate, the endogeneity bias. I then lay out the econometric methods that would solve the endogeneity bias if a valid instrument could be found. To instrument for the presence of collateral, I use across-state and across-industry variation in the procedures banks must follow to perfect their security interests. I find that the instrument I consider is not effective in reducing the endogeneity bias, despite its economic plausibility.
The third essay, which is joint work with Randolph Cohen and Tuomo Vuolteenaho, seeks to determine whether firms now release their earnings news through preannouncements rather than actual earnings announcements. We find that firms release earnings news through both preannouncements and actual announcements, but from the perspectives of both analysts and investors, more information is conveyed on preannouncement days. Analysts and investors also both act as if bad news preannouncements are more informative than good news preannouncements, which is consistent with the fact that bad news preannouncements tend to be farther away from expectations.
