Essays on Financial Reporting, Contracting, and Regulation

This dissertation consists of empirical accounting essays in three different settings related to the interplay among financial reporting, debt contracting, and capital markets regulation: the reflection of earnings incentives of lenders on loan contracting, unintended consequences of banning short-selling on information quality, and the ex ante risk-limiting role of debt covenants.;The first essay examines earnings management in the financial industry via syndicated loan originations, under which lead arrangers of lending consortia can recognize a disproportionate fraction of front-end fees in the quarter of issuance. I find evidence that lenders initiate additional loans in the last month of fiscal quarters when their reported EPS just-meets-or-beats benchmarks. I also observe that this boost is not costless. These loans are offered at a discount of 15-20 basis points and are associated with questionable quality as the borrowers experience subsequent credit rating downgrades and CDS initiations. To further understand why lenders underwrite and participate in costly suspect loans, I find evidence that these lead banks concurrently engage in other methods of earnings management and that these loans involve less experienced syndicates.;In the second essay, I focus on short selling, which has been a controversial subject, especially in bear markets. Specifically, I investigate short sellers' net value added in firms' information environments by benefiting from a quasi-natural experiment provided by the Australian short selling bans, which were imposed during the recent financial crisis and lifted on different dates for financial and non-financial stocks. My evidence suggests that banning short selling is detrimental to information quality. First, I observe that the ban period is associated with conditionally poorer information quality relative to pre-ban and post-ban regimes. Second, and more importantly, non-optionable stocks, compared with optionable stocks, experience conditionally higher increases in bid-ask spreads and analyst forecast dispersions. In response, stocks without options traded are more likely to provide voluntary disclosures and repurchase shares for signaling purposes during the ban. These results, confirmed by dif-in-dif-in-diff and placebo specifications, are consistent with a reaction from various economic actors to deteriorated information quality, an unintended consequence of short sellers' exclusion from price discovery and information revelation.;In the third essay, co-authored with Stephen Karolyi, we ask whether financial covenants have an ex ante role in firms' operating decisions and financial stability. We answer this question by analyzing bankruptcies and showing a positive association between the covenant slack at loan origination and subsequent corporate failures. The observed relation is statistically significant only during the Great Recession. Economically, a one-standard-deviation increase in covenant slack corresponds to a marginal increase of 5.6% in the probability of bankruptcy. This finding is consistent with the ex ante restrictive role of covenants: Loose covenants provide an avenue for unconstrained risk taking, which leads to excess exposure to corporate failure risk in macroeconomic downturns. We eliminate several alternative explanations, including unobservable credit quality signals and relaxed lending practices around the middle of the last decade. Overall, we present new evidence on the efficacy of loan covenants beyond their role as a tripwire.