Date of Award

2013

Degree Type

Thesis

Degree Name

Master of Science in Industrial Engineering (MSIE)

Department

Industrial Engineering

First Advisor

Yuehwern Yih

Committee Chair

Yuehwern Yih

Committee Member 1

Mehmet Deniz Yavuz

Committee Member 2

Mark Lehto

Abstract

Classical finance theories neglect the impact of investor sentiment on stock returns. These theories assume that investors are rational and make decisions in a way that maximizes their wealth. However, a vast amount of research shows that investors' decisions are affected by their psychological biases and feelings. These findings suggest that investor sentiment may have an impact on stock returns. This hypothesis is the main motivation of this study. First, this study examines whether there is correlation among investor sentiment indicators, and whether sentiment indicators have an impact on stock returns in the US and other countries. Second, this study investigates whether a global sentiment exists in developed and emerging countries. Additionally, it examines the relationship between investor sentiment and anomalies. Finally, this study investigates a method that helps investors use sentiment information during trading process.

The results of this study suggest that there is correlation among sentiment indicators in the US. In addition to this, several US investment indicators have a significant relationship with the S&P 500 index. Similar findings are found in Japan, Germany, China and Turkey. Moreover, this study finds that local (country) sentiment indicators are significantly correlated. It seems there is a global sentiment which impacts many countries. This global sentiment is stronger in the years between 2008 and 2012 than in the years between 1985 and 1990 due to increased economic ties among countries. Additionally, countries' stock market indices are significantly correlated. Furthermore, this study suggests that size, book-to-market and momentum anomalies can be explained by investor sentiment. Finally, the last chapter of this study proposes a sentiment rating system for individual stocks. In this system, stocks are assigned to different rating groups based on their sensitivity to sentiment changes. For example, a stock with very limited susceptibility to sentiment changes has AAA rating. An AAA rating means that a particular stock is not affected by sentiment driven mispricing and unexpected macroeconomic news. Therefore, the rating information can be used by individual investors to understand stock' behavior under sentiment changes. In addition, it is found that stock groups, which have negative correlation with sentiment changes, may have differences in terms of risk and size.

Download

Share

COinS