Please use this identifier to cite or link to this item:
Title: Investor Sentiment and Asset Pricing : Empirical Evidence from an Enhanced Investor Sentiment Index
Authors: Ung, Sze Nie
Issue Date: 2020
Publisher: Newcastle University
Abstract: This thesis covers three interconnected topics that investigate the impact of investor sentiment on stock returns. Given that investor sentiment is the central theme of this thesis, an accurate measure of investor sentiment is of great importance, and it is this theme which the thesis starts by exploring. With a new investor sentiment index which is superior to others currently available, the question of whether sentiment or fundamental factors play a more important role in driving stock returns is then explored. Finally, the thesis explores in greater depth channels through which investor sentiment drives stock returns as well as the pricing of rational and irrational risk factors. The first substantive chapter proposes an enhanced investor sentiment index, uniquely accounting for time-varying components in its construction. The poor time-series forecasting power of the often-used Baker and Wurgler (2006) investor sentiment index has long been a puzzle, and this study demonstrates that it is largely due to its implicit assumption that contributions of its individual index components to the aggregate sentiment index are timeinvariant. By capturing time-varying contributions of those components, the enhanced investor sentiment index not only demonstrates the basic property of a good sentiment measure (i.e. sentiment today predicts negatively the future aggregate stock returns), but also represents a superior measure of investor sentiment as compared to other sentiment indexes given that it is the only investor sentiment measure that has its sustained predictive power across different forecast horizons. Cross-sectionally, the new index also predicts significantly the time series of cross-sectional stock returns for portfolios sorted based on firm size, bookto-market ratio and momentum. The relative importance of investor sentiment to stock market fluctuations is explored in the second substantive chapter. Whilst most studies can be split into two distinct branches of the forecasting literature – forecasting power of investor sentiment versus fundamental return predictors – this chapter performs a battery of forecasting tests in evaluating the forecasting power of the enhanced investor sentiment index against a host of widely applied economic predictors in order to determine the main driver of stock market fluctuations. The results show that investor sentiment exerts a stronger influence on stock market movements, manifested by the superior forecasting power of the new index relative to the economic predictors, in both the statistical and economic sense. The third, and final, substantive chapter examines the channels through which investor sentiment affects stock market returns, i.e. the cash flow or discount rate channel, in light of the predictive ability of investor sentiment on stock market returns. This chapter constructs a four-beta model that separates the cash flow beta and the discount rate beta of Campbell and Vuolteenaho (2004) into rational and irrational components. The results show that the irrational beta in the cash flow channel receives a relatively greater weight than that in the discount rate channel, implying that the predictive power of investor sentiment is going through the cash flow channel. The findings also do not support the assumptions made in Campbell, Polk and Vuolteenaho (2010) that cash flow (discount rate) is mainly fundamental (sentiment) driven. Comparing the asset pricing performance of the four-beta model against alternative asset pricing models reveals that the four-beta model has a better model fit with a lower pricing error. The documented negative (positive) risk premia of irrational (rational) betas implies that investors are willing to pay a price (require a risk premium) for stocks that are sensitive to the irrational risk factors (rational risk factors).
Description: Ph. D. Thesis.
Appears in Collections:Newcastle University Business School

Files in This Item:
File Description SizeFormat 
Ung Sze Nie Hardbound E-Copy.pdfThesis4.55 MBAdobe PDFView/Open
dspacelicence.pdfLicence43.82 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.