PREFACE A market is considered efficient when the stock price fully reflects the economic news and information. Thus, if no new information has been published, the changes in stock prices will be relatively small. In other words, there rarely exits a substantial increase or decrease. In the Vietnam, however, the market is quite different. We found many trading sessions in which the VN-Index significantly increased although there is no good information about the whole economy as well as the business situation of enterprises was announced. In addition, there are also many trading versions that VN-Index dropped up, even though none of negative information has been found. Such fluctuation and signs in stock market allow some financial analysis and investors to pay more attention to the concept “herding behavior” to explain investor psychology. Thus, what is herding behavior? Whether or not herding exits in Vietnamese stock market? If yes, how much it give impact to Vietnamese stock market? And is there any correlation between stock exchanges in Vietnam? In 1
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
PREFACE
A market is considered efficient when the stock price fully reflects the
economic news and information. Thus, if no new information has been published,
the changes in stock prices will be relatively small. In other words, there rarely
exits a substantial increase or decrease. In the Vietnam, however, the market is
quite different. We found many trading sessions in which the VN-Index
significantly increased although there is no good information about the whole
economy as well as the business situation of enterprises was announced. In
addition, there are also many trading versions that VN-Index dropped up, even
though none of negative information has been found. Such fluctuation and signs in
stock market allow some financial analysis and investors to pay more attention to
the concept “herding behavior” to explain investor psychology. Thus, what is
herding behavior? Whether or not herding exits in Vietnamese stock market? If
yes, how much it give impact to Vietnamese stock market? And is there any
correlation between stock exchanges in Vietnam? In this research, we conduct a
research to find down the answer for these controversial issues.
We named a topic “Herding Behaviour in Vietnam’s stock market” for such
purpose. This research consists 05 main parts including:
Chapter 1 : Introduction
Chapter 2 : Literature Review
Chapter 3 : Methodology
Chapter 4 : Data analysis and Findings
Chapter 5 : Recommendation
1
CHAPTER 1: INTRODUCTION
1. Background of Stock Market in Vietnam
1.1. Overview of stock market in Vietnam
With rapid growth and development in more than a decade since its birth, the stock market in Vietnam has become a potential channel of investment for financial institutions, credit funds and individual investors. More importantly, stock market has made a significant contribution to the industrialization and modernization of the country
In reality, driven by greed and fear and mislead by extremes of emotion and
the impulse of the crowd, investors passively form irrational expectation for
the companies’ future performance and the overall economy. As a
consequence, stock prices overestimate or underestimate their fundamental
values.
The behavior of an investor to imitate the actions of others or to follow the
movements of market, instead of following his own information and
strategy, is usually regarded as “herding”. Possibly herding is among the
most mentioned but least understood terms in the financial lexicon.
This paper examines whether herding behavior exists in HOSE and HNX
Exchange markets. By applying the methodology proposed by Chang,
Cheng, and Khorana (2000) to examine Vietnamese stock data, we provide
evidence showing that there is herding behavior in HOSE exchange market.
However, no supportive evidence for herding behavior is found in HNX
market. There is no concrete evidence illustrating the correlation between
investment decision between HOSE and HNX markets.
2
1.2. Purpose of the research
This paper provides a thorough investigation of herd behavior and then
comes up with the correct answer for such issue.
We confirm the results of previous studies regarding the existence of herding
and also propose a new measure of herding based on a run test. Once
herding has been shown to be significant in our data, we firmly believe that
the herding does exist in Vietnam stock market.
Finally, we indicate the connection between herd behavior and changes in
stock’s return in a stock market to illustrate that the herding behaviors are
consistent with the changes in surveyed stock’s return.
1.3. The significance of research
Behavioral Finance, a field of finance that proposes psychology-based
theories to explain stock market anomalies, has given learners a better
understanding about the determining factors that result in the particular
behavior and performance of institutions and individuals from around the
globe, enhancing the understand the psychology and the emotions underlying
the decisions behind creating the goal.
Behavioral finance study comes up with Herd Behavior - the tendency for
individuals to mimic the actions (rational or irrational) of a larger group.
Over the past decade, financial economists have become increasingly
passionate in herd behavior in stock markets.
Because a strong herd mentality can even affect financial decision-making
process, understanding herb behavior is utmost important in judging the
efficiency of the stock market, in particular, and the whole economy, in
general.
3
In this paper, we provide readers with evidences to examine the existence of
herd behavior in Vietnam stock market, to measure how much it impact the
investor’s decision making process and to document the correlation between
stock exchanges in Vietnam, if existing.
1.4. Research questions
This research focuses to provide the response for the following questions:
1. Whether or not herding behavior exists in HNX and HO exchange floors
since 2008 until now?
2. Does herding behavior exist in these trading floors when the market goes
down and goes up?
3. If yes, in which situation does herding behavior appear to be stronger, up or
down?
4. Is there any correlation between the investor in HNX and HO in term of
investing imitation and herding behavior?
1.5. The limitation
In this paper, we focus on herding behavior and research information in
Vietnamese stock market with listed stocks conducted in both HNX and
HOSE trading floors in the period from 2008 to 2010.
Thank to the research of Ms. Tran Thi Hai Ly about the herding behavior
that was published in Finance and development magazine on June, 2010, the
existing herding behavior in HOSE stock market was captured in the period
from January 1, 2002 to December 31, 2008. However, the research of Ms
Ly did not mention the situation in HNX stock Exchange since its operation
on March 8, 2005.
4
We, therefore, continue and consolidate this research by collecting,
analyzing and processing the figure and data in both two trading floors until
now. In the scope of this research, we plan to use data of all stocks in these
stock exchanges
CHAPTER 2: LITERATURE REVIEW
In fact, the existence of herd behavior among particular participants in markets has
been analyzed empirically in a number of studies. In this part, we will briefly look
at and examine some of the methods that have been employed.
Several measures have been developed to investigate herd behavior in financial
markets, including:
Lakonishok, Shleifer, and Vishny (1992) (LSV) based their criterion on the
trades conducted by a subset of market participants over a period of time.
Wermers (1995) proposed a portfolio-change measure (PCM) which is
designed to capture both the direction and intensity of trading by investors.
Christie and Huang (CH) (1995) investigates the magnitude of cross-sectional
dispersion (or volatility) of individual stock returns during large price changes.
Chang, Cheng and Khorana (2000) have recently suggested a variant of the
CH method, showing that under CAPM assumptions, rational asset pricing
models suggest that the equity return dispersion, measured by the cross-
sectional absolute deviation of returns, should be a linear function of market
returns.
Nofsinger and Sias analysis adopt a different approach to examine the
relative importance of herding by institutional and individual investors.
5
In the scope of this research, we focus on two main measures and its application in
both Vietnam and foreign context. These two measures chosen is LSV method and
CH one thank to their significant roles and applications.
1.1. LVS measure of herding
Lakonishok, Schleifer, and Vishny (1992) (hereafter called LSV) define
herding as “the average tendency of a group of fund managers to buy and
sell particular stocks simultaneously relative to what would be expected if
managers traded independently” (Bikhchandani and Sharma 2001).
The LSV measure is based on trades conducted by a subset of market
participants over a period of time. This subset usually consists of a
homogenous group of fund managers whose behavior is of interest
(Bikhchandani and Sharma 2001).
In LSV’s paper, they denote B(i,t) [S(i,t] as the number of investors in this
subset who buy [sell] stock I in quarter t and H(i,t) as the measure of herding
in the stock I for quarter t. The measure of herding used by LSV is defined
as follows: H(i,t) = p(i,t)- p(t) – AF (i,t)
According to Bikhchandani and Sharma (2001), the LSV (1992) measure of
herding behavior is deficient in two aspects:
Firstly, this measure only uses the number of investors on the two sides
of the market (extreme market conditions), without taking the amount
of stock they buy/sell into account, to assess the extent of herding in a
particular stock.
Secondly, it is impossible to identify inter-temporal trading patterns
using the LSV measure. To specify, the LSV measure could be used to
test whether herding in a particular stock persists over time, that is to
6
evaluate whether E [H(I,t) H (I, t-k)]= E [H(I,t)], but it cannot inform us if it
is the same fund that continue to herd.
In case of Vietnam stock market, a number of market participants are hardly
to be measured correctly since an individual investor can illegally open more
than one account in a security company. About 85% investors in stock
market are individual investors while the quantity of stock they trade in the
market, according to the second drawback of LSV, cannot be measured.
Until now, the application of LSV measure in Vietnam stock market to find
out herding behavior is not employed yet.
1.2. Modification of the LSV measure of herding
Wermers (1995) develops a new measure of herding that captures both the
direction and intensity of trading by investors. This new measure which is
called a portfolio-change measure (PCM) of correlated trading, overcomes
the first drawback listed above of LSV measure. Intuitively, “herding is
measured by the extent to which portfolio weights assigned to the various
stocks by different money managers move in the same direction” according
to Wermers (1995).
The PCM measure has three main drawbacks which have been summarized
in Bikhchandani and Sharma (2001) paper:
First of all, according to PCM measure, the buy or sell decision by the
amount traded should be weighted, but doing this introduces another
bias since larger fund managers tend to get a higher weight.
Second, Wermer’s statistic which looks at changes in fractional weights
of stocks in portfolios may yield spurious herding as weights of stocks
that increase (decrease) in price tend to go up, even without any buying
(selling). Taking the average of beginning and end-quarter prices to
7
determine portfolio weights may correct for it as Wermers claims. It,
however, depends on exactly how it is done.
Finally, the justification of using net asset values as weights in
constructing the PCM measure is not clear (Bikhchandani and Sharma
2001).
2. The CH measure of herding
Christie and Huang (1995) (hereafter called CH) investigates the magnitude
of cross-sectional dispersion (or volatility) of individual stock returns during
large price changes. If the dispersion is small during the large price changes
then they suggest that there is evidence of herding.
Christie and Huang (1995) propose that the market impact of herding can be
measured by considering the dispersion or the cross-sectional standard
deviation (CSSD) of returns. In CH paper, they mention that traditional
asset-pricing theory predicts as a results of varying stock sensitivities to
market returns, the dispersion of return increases with the aggregate market
return
The rationale behind the use of this dispersion measure is that if the herding
occurs in the whole market, returns on individual stocks would be more than
usually clustered around the market return as investors suppress their private
opinion in favor of the market consensus (Henkers et al 2003).
Since dispersion measures the average proximity of individual returns to the
mean, when all market returns move in perfect unison with the market,
dispersion is zero. When individual returns differ from the market return, the
level of dispersion increases.
By using daily and monthly returns on U.S equities, CH finds a higher level
of dispersion around the market return during large price movements,
evidence against herding (Henkers et al 2003).
8
As Richards (1999) points out that the CH test looks for particular form of
herding and only in the asset-specific component of returns. It does not
allow for other forms of herding that may show up in the common
component of returns. Therefore, although CH test can be regarded as a very
accurate estimation of particular form of herding, the absence of evidence
against this form of herding should not be construed as showing that other
types of herding do not exist (Henkers et al 2003).
2.2. CCK- the modification of CH measure of herding
Chang, Cheng and Khorana (2000) (hereafter called CCK) propose a
modification to the model presented by CH. This model uses the cross-
sectional absolute standard deviation (hereafter CSAD) of returns as a
measure of dispersion to detect the existence of herding.
Their model suggests that if market participants herd around indicators, a
non-linear relationship will result between the absolute standard deviation of
returns and the average market return during periods of large price
movements.
CCK develop a more sensitive means of detecting herding by including an
additional regression parameter to capture a potential non-linear relationship
between security return dispersions and the market return. This alleviates the
limitation inherent in the Christie and Huang approach, which require a
greater magnitude of non-linearity in the return dispersion and mean return
relationship to identify herding (Henkers et al 2003).
Application of CCK in foreign countries
CCK uses monthly data of individual returns to analyze and find out that
under CAPM assumption, rational asset pricing models suggest that the
9
equity returns dispersion, measured by the cross-sectional absolute deviation
of returns, should be a liner function of market returns
They find a significant non-linear relationship between equity return
dispersion and the underlying market price movement of the South Korean
and Taiwanese markets, providing evidence of herding within these
emerging markets. However, they do not find evidence to support the
presence of herding in the developed markets of the U.S, Hong Kong and
Japan (Henkers et al 2003).
Application of CCK in Vietnam
Research about Herbing behavior in Vietnam stock market
- The research of M.A Tran Thi Hai Ly about the herding behavior was
published in Finance and development magazine on June, 2010. Ms Ly use
CCK measure to research the herding behavior in Vietnam and successfully
conclude the existing herding behavior in HoChiMinh stock exchange in the
period from January 1, 2002 to December 31, 2008.
- M.A Tran Thi Hai Ly used the point that if market participants herd around
indicators, a non-linear relationship will result between the absolute standard
deviation of returns and the average market return during periods of large
price movements.
- With these results, the author suggest that strong herd behavior exists in
Vietnamese market, and that herd behavior tend to be stronger in cases
where the market is flourished and boomed than that in downturn side.
Study about Psychology in securities investment in HoChiMinh City
This research is to understand the impact of psychological factors in the
securities investment in Vietnam and based on this analysis, the authors propose
10
solutions to improve professional relationships with investors in the securities
company in HCM City.
Subjects studied: 100 investors in the securities trading floor
Scope of study: In Ho Chi Minh city
Experimental method of investigation comprises two parts:
* Qualitative research: the research team attempted to find information and
opinions from the expert consultation including the consultant and the dream world
of securities certificates unhappy investors, organizing the focus groups to retrieve
information.
* Quantitative research: the team conducted a survey with 100
questionnaires to 100 investors in the securities trading floor.
This research concluded that communication is increasing its importance
because it is an effective method influencing social and emotional psychology and
that when financial markets grows up, investors and financial analysis gain focus
on the impact of information on the investor’s psychology.
The measure developed by Chang, Cheng and Khorana (2000) neither
consider the time-varying properties of beta in the CAPM nor herding
towards other factors which might be important in the interpretation of asset
returns. (Henkers et al 2003).
CHAPTER 3: METHODOLOGY
1. Source of information
11
Data used in this analysis (hereafter called DATA) are freely imported from
website: http://youthdragoncapital.com, the official website of the Youth
Dragon Capital Investment Fund.
DATA consists of prices and trading volumes of all stocks listed on the
Vietnam’s stock market (including HNX- the Hanoi Stock Exchange and
HSX- Hochiminh Stock Exchange) from their first trading day up to April 8,
2011; and the daily results of indices.
We do not use information on UPCOM trading floor due to its new
establishment and ineffective operation.
Including in the data is information about open price; close price, the highest
price and lowest price of stock each day (see the data table for more
information). However, for the purpose of calculating the daily return of
stocks, we only use the close price of each trading day.
Downloaded data are in the form of text.file, we import them into EXCEL
and do most of the data processing on EXCEL. Besides this powerful tool,
we also take advantage of two famous statistic and econometric softwares
which are MEGASTAT and EVIEW in our analysis.
Selected data includes information of 350 stocks in HNX and 265 stocks in
HSX (after excluding non-qualified stocks, using the sampling method
below).
However, in some stocks, due to the statistician’s carelessness, he/she just
either left the blank cells or filled in that with a random number, seriously
affecting the return calculation. Due to our group’s experiment, there are
about 131 stocks in HNX and 58 stocks in HSX had that error.
There are two ways to deal with the problem without affecting the final
The return of the two markets are closely related, we suspect that probably,
there is chance that investors in HNX stock market made their investment
decisions based on HSX investors’ decision. Hence, we go a step further to
examine this. If the evidence appears to conform to this argument, we say
that investors on HNX herd around HSX Index and vice versa.
To test for this hypothesis, we add an additional factor of cross-market
return squared into the model as follows:
1.
2.
The equation 1 is used to test whether HSX investors herd on HNX market;
whereas, the equation 2 is used to test whether HNX investors herd on HSX
market.
If HSX investors herd around HNX market or vice versa, we will expect that
both and in the two model were have negative signs, and is
statistically significant.
3. Testing & Results
Importing data from CROSS DATA file into EVIEW, and run the regression
equation, we got the following results
Dependent Variable: CSAD_HSX
Method: Least Squares
Date: 05/11/11 Time: 12:02
Sample: 1 1012
Included observations: 1012
Variable Coefficient Std. Error t-Statistic Prob.
25
C 0.018843 0.000691 27.26338 0.0000
ABS(RHSX) 0.362765 0.085204 4.257628 0.0000
RHSX^2 -3.474573 2.131059 -1.630444 0.1033
RHNX^2 -0.241265 0.345227 -0.698858 0.4848
R-squared 0.063709 Prob(F-statistic) 0.000000
The first regression model:
The overall model is statistically significant at 1% because F-statistic has
probability of nearly 0.
H0: β3 ≥ 0, no herding behavior of HSX on HNX
H1: β3 < 0, herding behavior of HSX on HNX
The key thing we need to look at is the sign of β3 and its significance. Here,
β3=-0.2412, negative. However, t-statistic is -0.6988 > t0.01= -2.32. Just we do not reject H0. So, there is no evidence of herding behavior of HSX on HNX market.