Herding behavior empirical evidence in the Vietnamese stock market

Ho Chi Minh City, 2018

MINISTRY OF EDUCATION & TRAINING

HO CHI MINH CITY OPEN UNIVERSITY

-----------------------------------------------

PHAM THI THANH AN

HERDING BEHAVIOR:

EMPIRICAL EVIDENCE IN THE

VIETNAMESE STOCK MARKET

MASTER THESIS

FINANCE AND BANKING

Ho Chi Minh City, 2018

MINISTRY OF EDUCATION & TRAINING

HO CHI MINH CITY OPEN UNIVERSITY

-----------------------------------------------

PHAM THI THANH AN

HERDING BEHAVIOR:

EMPIRICAL EVIDENCE IN THE

VIETNAMESE STOCK MARKET

Major: Finance and Banking

Major code: 60 34 02 01

MASTER THESIS

FINANCE AND BANKING

Academic Supervisor:

Dr. Duong Minh Chau

i

DECLARATION

I, Pham Thi Thanh An, declare that the research work reported in this dissertation is

my own. It is submitted to fulfill the requirements for the Masters of Finance and

Banking degree at the Open University of Ho Chi Minh City.

This thesis has not, either in whole or in part, been submitted for a degree or diploma

to any other institution or university for a similar qualification except where otherwise

indicated and acknowledged.

There are no other research papers used in this dissertation that are not cited in

accordance with regulations.

This thesis has never been submitted to receive any degree at other universities or

training institutions.

Ho Chi Minh City, 2018

Pham Thi Thanh An

ii

ACKNOWLEDGEMENT

Firstly, I would like to express my sincere gratitude to my advisor Prof. Duong Minh

Chau for the continuous support of my Master thesis, for his patience, motivation,

and immense knowledge. His guidance helped me in the time of writing research

proposal and my thesis. I feel lucky thanks to having a devoted support and direction

of my advisor for my thesis.

Beside my advisor, I would like to thank to Ms. Ha, Ms.Uyen and all the employees

working in Office of Postgraduate Management at Open University Ho Chi Minh

City for helping me in procedure of my Master thesis.

My sincere thanks also goes to my classmates working and helping me in all the time

of doing my thesis. They are one of the motivations for me to complete my study.

Thanks for every night, every weekend in three years working with me. I appreciate

your enthusiasm.

Last but not the least, I would like to thanks my family: my parents and to my brother

for supporting me spiritually and support finance throughout my learning and writing

my Master thesis.

iii

ABSTRACT

The thesis was conducted to find out the existence of herding behavior and to identify

strong herding points and strong influential points on the Vietnamese stock market

from 2006 to 2016. Data collected includes non-financial companies on both Ho Chi

Minh Stock Exchange and Ha Noi Stock Exchange from 2006 to 2016. The data is

divided: the entire period from 2006 to 2016, individual years and 3 sub-periods

consisting of before crisis 2006-2007, during crisis 2008-2012 and after the crisis

2013-2016. The model used includes the Cross Sectional Absolute Deviation (CSAD)

model of Chang et al. (2000) to determine the presence of herding behavior and

Weighted Cross- Sectional Variance (WCSV) model of Xie et. Al (2015) to find

strong herding points and strong influential points on the Vietnamese stock market.

In addition, the thesis also compares two models: WCSV model based on Fama￾French three factors according to Arbitrage Pricing Theory (APT) and WCSV model

based on CAPM one factors through adjusted R2

. The purpose of the study is to

investigate the impact of herding behavior of individual investors as well as to

consider the suitability of the WCSV model when comparing strong herding spots to

those of prominent events that influence to the Vietnamese stock market in the period

from 2006 to 2016.

Regarding the CSAD regression model, the results show that the herding behavior

exists in the Vietnamese stock market on both HOSE and HNX in the overall period

from 2006 to 2016, individual years (except 2008, 2013, 2015 and 2016) and in three

sub-periods: before, during and after the crisis. A comparison of the two models

including WCSV model based on Fama-French three-factor model and WCSV model

based on CAPM one factor shows that the WCSV model according to APT has a

higher Adjusted R2

than the WCSV model based on CAPM one factor model.

Through the WCSV model based on Fama-French three factors as the basic APT, the

results show that strong herding points and strong influential points appear in years,

periods of relevance to prominent events that strongly influence the Vietnamese stock

market. In general, strong herding points and strongly influential points appear

iv

energetically in the periods consisting of before crisis 2006-2007 and during crisis

2008-2012. However, in 2008, the CSAD model does not find the existence of

herding behavior, whereas the WCSV model yields strong herding spots at high

incidence. In addition, there is generally no difference between the results given in

the two models as well as there are not many differences between the selected

timelines. Thus, the above results show that there is a compatible combination of

CSAD and WCSV models in determining the existence of herding behavior,

eliminating a hypothetical factor of the WCSV model and determining the strong

herding points and strong influential points correspond to the time of prominent

events on the Vietnamese stock market.

v

TABLE OF CONTENTS

Page

Declaration .................................................................................................................. i

Acknowledgement...................................................................................................... ii

Abstract ..................................................................................................................... iii

Table of contents.........................................................................................................v

List of figures........................................................................................................... vii

List of tables............................................................................................................ viii

Abbreviations..............................................................................................................x

CHAPTER 1. INTRODUCTION ............................................................................1

1.1. Background......................................................................................................1

1.2. Research objectives..........................................................................................4

1.3. Research questions...........................................................................................4

1.4. Research Subject and Scope ............................................................................4

1.5. Contribution .....................................................................................................4

1.6. Structure...........................................................................................................5

CHAPTER 2. LITERATURE REVIEW AND HYPOTHESES ..........................7

2.1. Behavioral Finance ..........................................................................................7

2.1.1. Behavior Finance ..................................................................................7

2.2. Herding Behavior.............................................................................................8

2.2.1. Definitions of Herding Behavior...........................................................8

2.2.2. Types of Herding Behavior.................................................................10

2.3. Empirical Evidence........................................................................................12

2.4. Hypotheses.....................................................................................................23

CHAPTER 3. RESEARCH METHODS ..............................................................24

3.1. Data Description ............................................................................................24

3.2. Research methods..........................................................................................24

vi

3.3. Selection of Models.......................................................................................25

3.3.1. The Return Dispersion Model by Chang et al. (2000)........................25

3.3.2. The Return Dispersion Model by Xie et al. (2015) ............................26

3.3.3. Test for strong herding points and strongly influential points............29

CHAPTER 4. DATA ANALYSIS AND FINDINGS ...........................................32

4.1. Descriptive Statistics......................................................................................32

4.1.1. Descriptive Statistics...........................................................................32

4.2. Empirical Results...........................................................................................41

4.2.1. Results of testing the existence of herding behavior through CSAD

model.............................................................................................................41

4.2.2. Empirical results of testing strong herding points and strong

influential points............................................................................................48

CHAPTER 5. CONCLUSIONS AND RECOMMENDATION .........................58

5.1. Implications ...................................................................................................58

5.2. Limitations and Recommendations ...............................................................61

REFERENCES..........................................................................................................63

APPENDIX A ...........................................................................................................71

APPENDIX B ...........................................................................................................75

vii

LIST OF FIGURES

Page

Figure 4.1: Time series of the WCSV models during the period from 2006 to 2016

on HOSE ...................................................................................................................32

Figure 4.2: Time series of the WCSV models during the period from 2006 to 2016

on HNX .....................................................................................................................33

viii

LIST OF TABLES

Page

Table 3.1: A summarization of strong herding point testing ....................................29

Table 3.2 A detailed summarization of testing strong herding points in the thesis..30

Table 4.1: Summary statistics of dependent variables and independent variables of

WCSV model based on Fama-French Three Factor Model for 2006-2016 on HOSE.

...................................................................................................................................36

Table 4.2 : Summary statistics of dependent variables and independent variables of

WCSV model based on Fama-French Three Factor Model of Year 2006-2016 on

HNX. .........................................................................................................................37

Table 4.3: Description of variables in WCSV model in three sub-periods: before,

during and after crisis on HOSE ...............................................................................38

Table 4.4: Description of variables in WCSV model in three sub-periods: before,

during and after crisis on HNX .................................................................................39

Table 4.5: The correlation matrix of variables in WCSV model based on Fama￾French three factors on HOSE ..................................................................................40

Table 4.6: The correlation matrix of variables in WCSV model based on Fama -

French three factors on HNX ....................................................................................40

Table 4.7: Testing multicollinear of variables in WCSV model based on Fama￾French three factors over the period from 2006 to 2016...........................................41

Table 4.8: Testing multicollinear of variables in WCSV model based on Fama￾French three factors in three-sub periods: before, during and after crisis ................41

Table 4.9: OLS estimators of CSAD model on HOSE.............................................44

Table 4.10: OLS estimators of (a) WCSV model based on Fama –French three

factors as the basic APT (b) WCSV model based on CAPM one factor on HOSE .45

Table 4.11: OLS estimators of CSAD model on HNX.............................................46

Table 4.12: OLS estimators of (a) WCSV model based on Fama –French three

factors (b) WCSV model based on CAPM one factor on HNX ...............................47

Table 4.13: Summary of strong herding points and strong influential points with

single year data from 2006 to 2016 on HOSE. .........................................................53