Two essays on business description content trend and the cross-section of stock returns

Update Item Information
Title Two essays on business description content trend and the cross-section of stock returns
Publication Type dissertation
School or College David Eccles School of Business
Department Entrepreneurship & Strategy
Author Zhang, Yan
Date 2017-08
Description This dissertation examines the business description firms include in their 10-Ks. Using the Latent Dirichlet Allocation topic extraction methodology, I identify the highest trending topic for each industry and each firm's loading on this topic. In the first chapter, after controlling for risk, I find that firms with higher loading on the highest trending topic are more likely to experience lower future returns. These findings are consistent with the notion that investors are willing to pay more for firms that use trendy language, even though the higher prices are not justified by their fundamental values. In the second chapter, I disentangle the relations between 10-K trend loading, earnings management, and analysts' forecast errors during the 10-K release. I find that 10-K trend loading and earnings management are conditionally uncorrelated, and that analysts can make forecast errors contrary to the level of 10-K trend loading, correcting the short-run market underreactions.
Type Text
Publisher University of Utah
Subject Latent Dirichlet Allocation; Stocks
Dissertation Name Doctor of Philosophy
Language eng
Rights Management (c) Yan Zhang
Format application/pdf
Format Medium application/pdf
ARK ark:/87278/s6p026s8
Setname ir_etd
ID 1302714
Reference URL https://collections.lib.utah.edu/ark:/87278/s6p026s8
Back to Search Results