Dissertation/Thesis Abstract

Essays in Accounting Theory: Corporate Earnings Management in a Dynamic Setting and Public Disclosure in the Financial Services Industry
by Du, Kai, Ph.D., Yale University, 2012, 144; 3535385
Abstract (Summary)

This dissertation consists of three essays on the interactions between economic fundamentals and accounting information in three different settings: an infinite-horizon financial reporting problem, a coordination game with trading in the secondary market, and a bank which provides risk sharing among demand depositors. In the first essay, I propose a dynamic model of corporate earnings management in which investors have different expectations schemes. I find that while earnings management may exist when investors have rational expectations or misspecified Bayesian beliefs, it disappears in the long run of an adaptive learning process. The model also offers ample predictions on the time-series properties of asset prices and return predictabilities. The second essay studies the role of public disclosure by a distressed firm whose creditors engage in a coordination game with trading. I find that conditioned on the private information environment and equilibrium selection, better public disclosure could lead to higher probability of creditor run and lower expected welfare. The third essay introduces loan loss provision to a bank that transforms illiquid assets into demand deposits, and shows that the contingency in demand-deposit contracts induced by loan loss provision reduces risk sharing.

Indexing (document details)
Advisor: Sunder, Shyam
Commitee: Antle, Rick, Thomas, Jacob, Tsyvinski, Aleh, Versano, Tsahi
School: Yale University
Department: Accounting
School Location: United States -- Connecticut
Source: DAI-A 74/05(E), Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Accounting
Keywords: Corporate earnings, Earnings management, Financial services industry, Public disclosures
Publication Number: 3535385
ISBN: 978-1-267-85665-4
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest