Matthias Burghardt. Retail Investor Sentiment and Behavior

Transcription

1 Matthias Burghardt Retail Investor Sentiment and Behavior

2 GABLER RESEARCH

3 Matthias Burghardt Retail Investor Sentiment and Behavior An Empirical Analysis RESEARCH

4 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at Dissertation Karlsruhe Institute of Technology, st Edition 2011 All rights reserved Gabler Verlag Springer Fachmedien Wiesbaden GmbH 2011 Editorial Office: Stefanie Brich Anita Wilke Gabler Verlag is a brand of Springer Fachmedien. Springer Fachmedien is part of Springer Science+Business Media. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the copyright holder. Registered and/or industrial names, trade names, trade descriptions etc. cited in this publication are part of the law for trade-mark protection and may not be used free in any form or by any means even if this is not specifically marked. Cover design: KünkelLopka Medienentwicklung, Heidelberg Printed on acid-free paper Printed in Germany ISBN

5 Acknowledgements I am indebted to many people for their support throughout the preparation of this thesis. In particular, I am grateful to my advisor Prof. Dr. Christof Weinhardt who gave me his support and advice while granting me the freedom to explore different ways to combine academic rigor and practical relevance in interdisciplinary research. In addition, I would like to thank Prof. Dr. Marliese Uhrig-Homburg for co-advising this thesis. My thanks also go to Prof. Dr. Ute Werner and Prof. Dr. Ingrid Ott for serving on the board of examiners. I would like to thank my colleagues from the research group Information and Market Engineering at the Institute of Information Systems and Management (IISM) for many fruitful discussions and their valuable comments. I am grateful to Dr. Henner Gimpel who was always ready to discuss problems with me and to share his ideas how to solve them. I also thank Dr. Stefan Seifert for his advice on the direction of this work. Many thanks also go to Dr. Ryan Riordan for being my co-author on many papers and for providing me with encouragement and support. Ryan Riordan and Martin Wagener deserve special thanks for proofreading major parts of the thesis and for providing me with constructive comments. Parts of this research have been done while I enjoyed the hospitality of London Business School. I am grateful to Prof. Bruce Weber who made this possible. Finally, I would like to express my gratitude to Oliver Hans, Dr. Christoph Mura, and Christoph Lammersdorf for their constant support while I was a research fellow at Boerse Stuttgart. This work would not have been possible without their commitment and dedication to this project. Matthias Burghardt

6 Table of Contents Acknowledgements... v Abbreviations... xi List of Figures... xiii List of Tables... xv Introduction Motivation Research Outline Overview and Structure Related Publications... 8 Related Theoretical and Empirical Work Introduction to Behavioral Finance Are Financial Markets Efficient? Challenges to Efficient Markets Emergence of Behavioral Finance Theoretical Work Empirical Work Noise Traders Investor Sentiment Individual Investors Correlated Trading Discussion Behavioral Finance Correlated Trading Return Correlation Market Efficiency Conclusion Investor Sentiment Construction Classification of Sentiment Measures Related Work Advantages and Disadvantages... 37

7 viii Table of Contents Sentiment Measures in Research The Closed-End Funds Discount Meta-Measures Sentiment Measures in Practice Survey-Based Measures Market-Data-Based Measures Meta-Measures Summary Statistics Evaluation of Sentiment Measures Direct Sentiment Measures Indirect Sentiment Measures Direct vs. Indirect Sentiment Measures Sentiment Measures vs. Market Returns Review of Results Conclusion Construction of the Euwax Sentiment Index Introduction Securitized Derivatives European Warrant Exchange Key Facts Data Set Basic Index Calculation Sentiment Analysis Number- vs. Volume-Based Measures Product Types Order Types Order Volume Groups Submitted Orders Leverage Comparison with Other Sentiment Measures Indirect Sentiment Measures Direct Sentiment Measures Review of Results Conclusion Retail Investor Herding Introduction Related Work Definitions of Herding Empirical Findings Discussion... 98

8 Table of Contents ix Evidence of Market-Wide Herding Data Herding Measure Construction Results Review of Results Market-Wide Herding on a Broker Level Data Results Review of Results Stock-Level Herding Data Herding Measure Results Review of Results Conclusion The Predictive Power of Retail Investor Sentiment Related Work Data and Methodology Data Set Methodology Results Pre- and Post-Portfolio-Formation Returns Control Variables Portfolio Holding Returns Robustness Checks Conclusion Conclusion and Future Work Conclusion Summary of Contributions Future Work Appendix A Appendix to Chapter Bibliography

9 Abbreviations AAII American Association of Individual Investors ADV/DEC Advances/Declines Ratio ADV-DEC Advances minus Declines ASX Australian Stock Exchange CAUD Consolidated Equity Audit Trail Data CBOE Chicago Board Options Exchange CCI Consumer Confidence Index CEFA Closed-End Funds Association CEFD Closed-End Funds Discount CEFD Closed-End Funds Discount col. column DAX Deutscher Aktienindex DJIA Dow Jones Industrial Average e.g. for example EMH Efficient Markets Hypothesis et al. et alii EUWAX European Warrant Exchange G-Mind German Market Indicator i.e. that is ICE Index of Consumer Expectations ICS Index of Consumer Sentiment II Investors Intelligence IPO Initial Public Offering ISE International Securities Exchange ISSM Institute for the Study of Security Markets LSV Lakonishok, Shleifer, and Vishny (1992) MSH Morgan Stanley High-Tech Index NASDAQ National Association of Securities Dealers Automated Quotations NAV net asset value NYSE New York Stock Exchange p/c put/call ratio REX Rentenindex S&P Standard & Poor s SMH Signed Market Level Herding TAQ Trade and Quote database TRIN Traders Index

10 xii Abbreviations U.S. UBS UMH VAR VDAX VIX VWD ZEW United States Union Bank of Switzerland Unsigned Market Level Herding Vector autoregression Volatility DAX Volatility Index volume-weighted discount index Zentrum für Europäische Wirtschaftsforschung

11 List of Figures Figure 1.1: Thesis Structure... 7 Figure 3.1: CEFD History of Tri-Continental Corp Figure 3.2: Baker and Wurgler (2006) Composite Sentiment Figure 3.3: Sentix Private Investors Sentiment Figure 3.4: Cognitrend Bull Bear Index Figure 3.5: AAII Sentiment Index Figure 3.6: Michigan Consumer Sentiment Figure 3.7: Consumer Confidence Index Figure 3.8: UBS/Gallup Index of Investor Optimism Figure 3.9: G-Mind Figure 3.10: CBOE Put/Call Ratios Figure 3.11: VDAX-New Figure 3.12: ISEE Sentiment Index Figure 3.13: Investors Intelligence Bearish Sentiment Index Figure 4.1: Euwax Sentiment Index ( ) Figure 5.1: Pairwise Broker Correlation Figure 6.1: Portfolio Holding Returns Figure 6.2: Portfolio Holding Returns by Market Capitalization Figure 6.3: Portfolio Holding Returns for Small Caps

12 List of Tables Table 2.1: Overview of Empirical Work on Trading Imbalances Table 2.2: Evidence of Return Correlation Found in Empirical Work Table 3.1: Summary Statistics of Sentiment Measures Table 3.2: Correlation of Weekly and Monthly Sentiment Measures Table 3.3: Correlation of Daily Indirect Sentiment Measures Table 3.4: Correlation between Indirect and Direct Weekly Sentiment Measures Table 3.5: Correlation between Indirect and Direct Monthly Sentiment Measures Table 3.6: Correlation of Sentiment Measures and Market Returns Table 4.1: Summary Statistics Table 4.2: Multivariate Regression Results for Order Volume Table 4.3: Multivariate Regression Results for Order Number Table 4.4: Leverage Table 4.5: Daily Sentiment Correlation Table 4.6: Weekly Sentiment Correlation Table 4.7: Monthly Sentiment Correlation Table 5.1: Signed and Unsigned Market Herding Measures Table 5.2: Signed and Unsigned Market Herding Measures by Market Period Table 5.3: Daily Herding Measures on a Broker Level Table 5.4: Daily Herding Measures by Broker Type Table 5.5: Correlation between Broker Types Table 5.6: Summary Statistics by Geographical Origin of Company Table 5.7: Summary Statistics on Market Capitalization Table 5.8: Modified LSV Herding Measure by Product and Order Type Table 5.9: Modified LSV Herding Measure by Firm Size Table 5.10: Comparison of LSV Measures across Studies Table 6.1: Return Statistics by Product Type Table 6.2: Return Statistics by Order and Product Type Table 6.3: Return Statistics for Submitted Orders Table 6.4: Controlling for Market Returns Table 6.5: Controlling for Momentum Table 6.6: Controlling for Market Capitalization Table 6.7: Portfolio Holding Return Statistics Table 6.8: Annual Portfolio Holding Returns Table A.1: Summary Statistics on Executed Order Volume Table A.2: Summary Statistics on Submitted Orders