Publication Type

PhD Dissertation

Version

publishedVersion

Publication Date

6-2020

Abstract

Chapter 1: Analyst report content and stock market anomalies A series of recent papers document that security analyst recommendations tend to contradict stock-mispricing signals. This seems at odds with the large prior literature on the investment value of analyst recommendations. What justifications do analysts make when they write reports on mispriced stocks? I use the latest techniques in machine learning and textual analysis to categorize the qualitative information in a large sample of analyst reports. I find that report content can be intuitively classified into five categories or topics: 1) Growth, 2) Earnings, 3) New developments, 4) Management transactions, and 5) Conviction. I then relate the frequency of each topic and the tone surrounding the topic to stock-anomaly mispricing signals. I find that although analysts are incorrectly optimistic about overvalued stocks in general, reports on new developments and management transactions have investment value after controlling for the predictive power of the mispricing signals. For undervalued stocks, while analysts are on average incorrectly pessimistic, reports on growth, new developments, and management transactions have investment value. Overall, this paper helps to understand how analysts provide value in their reports even when the report ratings appear to contradict well-known signals of mispricing.

Chapter 2: The information cycle and return seasonality (with Roger Loh) Heston and Sadka (2008) find that the monthly cross-sectional returns of stocks depend on their historical same calendar-month returns. We propose an information-cycle explanation for this seasonality anomaly—that firms’ seasonal release of information coincide with higher returns during months with such dissolution of information uncertainty, and lower returns during months with no information releases. Using earnings announcements and changes in implied volatility as proxies for scheduled information releases, we find that seasonal winners in information-release months and seasonal losers in non-information release months indeed drive the seasonality anomaly. Our evidence shows that scheduled firm-level information releases can give rise to the appearance of an anomalous seasonal pattern when stock returns are in fact responding to information uncertainty.

Chapter 3: Managerial and analyst horizons during conference calls It is alleged that public-firm managers face short-term pressures from investors. In this paper, I examine managers’ tendency to talk about the short versus the long term by analyzing the language in quarterly analyst conference calls. Using the word embedding model, I determine whether conference calls focus on the short or long term. I find that when firms fail to meet analyst expectations, both managers and analysts focus on the short term rather than the long term. However, in macro bad times, analysts question managers about the short term rather than the long term, while managers maintain the same long term-short term balance whether in good or bad macro conditions. Finally, I show that firms whose conference call participants focus more on the long term have negative initial market reactions, but stock prices recover in the subsequent months. subsequent months. The results are consistent with Wall Street exerting excessive short-term pressures on public firm managers.

Keywords

asset pricing, qualitative information, information cycle, return seasonality

Degree Awarded

PhD in Business (Finance)

Discipline

Finance | Finance and Financial Management

Supervisor(s)

LOH, Kiat Roger

Publisher

Singapore Management University

City or Country

Singapore

Copyright Owner and License

Author

Download

Share

COinS