The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

Information Asymmetry And The Role

Of Foreign Investors In Daily Transactions
During The Crisis; A Study Of Herding
In The Indonesian Stock Exchange
Ishak Ramli, Tarumanagara University, Indonesia
Sukrisno Agoes, Tarumanagara University, Indonesia
Ignatius Roni Setyawan, Tarumanagara University, Indonesia

ABSTRACT

The purpose of this study is to prove that there was herding behavior by domestic investors following that of foreign
investors in the Indonesian Capital Market (IDX) and that the herding was influenced by information asymmetry. It
began when global investors undertook international diversification to the IDX because the returns on their
portfolios were not on the efficient frontier during the crisis and because of the low correlation between Indonesia’s
economy and the American and European economies. Utilizing the IDX daily transaction data during the years
2009-2011, the herding behavior of domestic investors, which followed that of foreign investors, was tested by
Lakonishok models as was the influence of information asymmetry on the herding. It was found that the herding
behavior in the IDX occurred in buy, sell or entire herdings (buy and sell). There were 0.40 to 0.55 buy herdings
and 0.20 to 0.40 sell herdings during the crisis in 2008 and 2009. Buy herding then continued in 2010 onwards,
although with lower intensity (0.05 to 0.20); however, sell herding decreased dramatically, and there has been
almost no sell herding since then. Nevertheless, domestic investors did then sell in the opposite strategy, which was
to sell when foreign investors tended to buy. Subsequent findings demonstrated that herding occurred with the
influence of information asymmetry between domestic and foreign investors.

Keywords: Herding; Buy Herding; Sell Herding; Information Asymmetry; Feedback Trading

1. INTRODUCTION

T he global crisis began in 2008, and the liberalization of financial markets since the 1990s prompted
global investors to look for alternative investments in emerging countries that were less affected by
the crisis. Indonesia, as an emerging country, was an alternative investment destination country of the
international diversification. Indonesia is significant to global investors because of the advantage in the form of its
low correlation with the American and the European economies. At the time of the economic crisis in America and
Europe, the economies in emerging countries, particularly Indonesia, were not affected. This, then, caused Indonesia
and the Indonesian Capital Market (IDX) to gain an economic benefit from the crisis and the impact of financial
market openness. Financial liberalization, then, caused the IDX to benefit from the increased flow of funds from
global investors, which in turn raised the Jakarta Composite Index (JCI). The JCI increase, among the highest in the
world since 2006, would attract foreign investors to continue to diversify internationally in the capital market of
Indonesia (IDX).

Foreign investors were believed to have large capital funds, more knowledge, abilities, and mastery of information;
this then often turned out to be a reference for domestic investors in targeting their investment to the leading stocks
in the LQ 45 on the Stock Exchange for the past 5 years (Dvorak, 2005, and Aggarwal, et al., 2009). Then,
Panggabean (2006) and Setiyono (2012) found there was herding behavior in the IDX, in which domestic investors
mimicked the action of foreign investors’ trading strategy. Generally, domestic investors conducted herding based
on foreign investors trading strategy by following the transactions of purchases and sales of shares of the foreign
Copyright by author(s); CC-BY 269 The Clute Institute
The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

investors. The importance of herding then led many researchers to search for the determinants of herding,
particularly the herding of domestic investors following that of the foreign investors. The financial literature
highlights that the main determinant of herding is information asymmetry as a result of the impact of reputation
[Dvorak (2005)] and transaction costs [Chiang, et al. (2011)]. Domestic investors in the IDX have always been
considered inferior to foreign investors, so herding was bound to happen. We will highlight the transaction costs
argued by Chiang et al. (2011) because of the impact of reputation arguments.

In Chiang et al. (2011), herding will only occur when stocks are traded with high liquidity. Strong flows of trade
orders that are triggered by low transaction costs will increase the herding behavior of domestic investors.
Transaction costs, argued by Chiang et al. (2011), were confirmed by the results of previous studies, which found
herding after the next transaction and followed the pattern of previous transactions. The basis for the herding of
Chiang et al. (2011) is oriented more to lower transaction costs for domestic investors than to the reputation concern.
It is more reasonable than Dvorak (2005), in which the effects of the reputation supported herding, but it is the
transaction fee of Chiang et al. (2011) that is closer to the information asymmetry.

It is not yet conclusive whether the information asymmetry will be a determinant factor of herding. Al Shboul
(2012) found in Australia that the market capitalization and return growth was up to 51%, not only because of the
information asymmetry of the domestic investors but also because of their sophistication in the market as global
investors. However, there were some different conditions in Australia compared with Indonesia. Domestic investors
in Indonesia are more information asymmetric than Australians. This was why domestic investors in Indonesia herd
to the foreign investors, while Australians do not. Domestic investors in Australia can play the global investors’ role
to internationally diversify in the IDX. Despite the global and European crises, geographical proximity between
Indonesia and Australia would still be attractive for global investors to increase investing in the IDX as an
alternative.

Herding behavior of information asymmetry is one of the most powerful forms of aberration theory of EMH
(Efficient Market Hypothesis) of Professor Eugene Fama. The EMH assumes that investors will generally be
rational in expecting the price of the stock market to come to reflect all market information. However, in the real
world, rationality is not the case because many investors who get abnormal returns trigger other investors to do the
same. The actions of the investors are able to indicate their herding behavior in the capital market. Herding behavior
occurs when investors decide to emulate the decisions of others in the capital markets rather than follow their own
beliefs as well as information in their possession. Such behavior can be seen as rational for a number of reasons,
although it may not always lead to efficient market outcomes. Herding can be in the realm of rational utility
maximizing, for example, if the other participants in the market are expected to have better information or lack
information asymmetry [Hwang and Salmon (2004)]. Herding behavior is shown by the special characteristics of the
investor, and it is not only sophisticated investors can also conduct herding. Iihara et al. (2001) describes more
professional investors herding than amateur investors. Thus, the results of their studies support Chiang et al. (2011)
that herding and information asymmetry are interconnected (having causality).

Herding also depends on the size and the systematic risk of the company, but no herding was conducted by
professional investors who are less sensitive to these variables. In addition, herding conducted by both amateur and
professional investors is positively correlated with and significant to the volatility of stock market returns. Herding
conducted by amateur investors would cause market volatility. The increase in volatility caused by the information
asymmetry of the stock price dynamics is employed by amateur investors to not lose from professional investors
[Franke and Westerhoff (2011)]. Financial markets are known to be characterized by a number of facts relating to
matters with and or without preconditions in a time series order. It turns out that the price volatility in the capital
markets often occurred differently from the fundamental factors underlying them. Then, widely agreed upon in the
literature is the difficult to explain Efficient Market Hypotheses (EMH). The literature explains that the interaction
between actors of diverse and various investors (heterogeneity) in the capital market with limited rationality
supports the price volatility that occurs in the capital market. The investors have different information about the
growth level of performance and of dividend of the issuers in the future so that they have different expectations of
the value of the company. It is investors’ rationality utilizing information regarding the stock prices, dividends, and
historical performance in the future that maximizes expected utility.

Copyright by author(s); CC-BY 270 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

Wang (2000) and Aggarwal et al. (2009) found a size or a proxy of information asymmetry, that is, the difference in
bid and ask prices in the stock market that determines the highest and lowest prices of everyday transactions. The
existence of investors who have less or no information because they do not do the research would increase the risk
premium. Particularly when there are supply shocks in the turmoil situation, the asymmetry of information will
increase the risk premium, while supply shocks do not increase the risk premium in the situation when there is
information asymmetry. Information asymmetry among investors also may increase the volatility of stock prices and
returns with negative autocorrelation. The uninformed investors will behave as rationally as price hunters. They will
follow the herding of investors who have performed the research to gain better information about the company.

Hwang and Salmon (2004) stated that herding behavior occurs particularly when the market is relatively quieter than
in stressful times, and herding occurs because of shared values, particularly after an economic crisis. Herding
behavior was influenced by the financial crisis, the quite large stock trading volume, and the decisions of foreign
investors [see Al Shboul (2012) in his study of Australia]. They said that herding occurs in the stocks of large issuers
because it is the stocks of large issuers that are researched by mutual fund investors. Performing research requires a
considerable cost, which usually can only be paid by large listed companies. Generally, small fund investors do not
conduct research because it requires huge costs, so they generally do not have information or only have insufficient
information. Therefore, investors or portfolio managers who have information for conducting research will utilize
the information to select a portfolio of stocks that is expected to provide a sizeable return because it is below its
fundamental price. The portfolio manager will be able to maintain a high return because he has more information
and the stock portfolio is not disclosed. The SEC in the United States requires disclosure of stock mutual fund
portfolios every quarter (May 2004; previously every six months); this makes it easier for investors with small
funds, as they will select a mutual fund with big funds so they can follow the herding behavior.

In the Indonesian capital market (IDX), herding occurs by domestic investors to foreign investors, and the stock
price volatility is affected by the effect of the herding. [Setyawan and Ramli (2013)]. The implication is that the
more information the foreign investors and the fund managers have, the greater they control the capital market.
Excess stock price volatility then becomes the deciding factor for high returns. This triggers the hot money in
foreign funds related to the interest spread of the banking sector.

Thus, it is important to investigate herding behavior in the IDX because of its potential impact on the fluctuating
returns in the capital market, while according to Kremer and Nautz (2013), the intensity of the herding behavior
depends on the characteristics of stocks including returns and its volatility in the past. The instability of herding has
an impact on stock prices in the short term, mostly unplanned. It is done in part to avoid risk. Fernandez (2010)
suggests herding on individual stocks occurs when the market weakens, and there is an inverse relationship between
volatility and trading volumes, particularly in cases of the stocks with extreme returns. Most of the previous research
on herding focused only on the existence and extent of the herding behavior of investors, and to date, there is less
empirical evidence on the determinants of institutional herding behavior in emerging markets. The information
asymmetry as a determinant of herding is still not consistent, so it is still necessary to study evidence in the
Indonesian capital market (IDX). Understanding the determinants of institutional herding could provide better
solutions to reduce market volatility stemming from the irrational herding behavior.

In addition, previous studies that primarily examined herding behavior among institutional investors relied on
monthly or quarterly data, and because in the growing capitalization shares on the Stock Exchange there was often
short-term speculation (daily data) on particular individual stocks, it is necessary to perform better herding
observations utilizing short-term daily transaction data of each share. The study will be conducted employing each
stock’s daily transaction to see the herding behavior and whether information asymmetry becomes a determinant of
the herding in the period of the crisis (years 2009-2011). Determinants of herding in the literature are the liquidity of
the stock, stock price volatility and stock returns of the past, which are mostly derived from information asymmetry.
Measurements are conducted employing the herding measure of Lakonishok et al. (1992). The sizes are 0-1; 1
characterizes herding, and 0 otherwise. Information asymmetry is measured by the ratio of the difference between
the high and low prices of Aldea and Marin (2007), in which the greater is the ratio, the higher is the information
asymmetry.

Copyright by author(s); CC-BY 271 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

The study is intended to answer two research questions: 1) was there a herding behavior of domestic investors to
foreign investors during the crisis (2009 - 2011) utilizing daily stock transactions on the Indonesian Stock Exchange
(IDX) 2) Did the information asymmetry affect the intensity of herding behavior during the crisis (2009 -2011)
utilizing daily stock transactions on the Indonesian Stock Exchange (IDX).

2. LITERATURE REVIEW

2.1 Herding Concept

Herding is defined as the behavior of investors who follow the behavior of other investors during a certain period;
the investors put aside their own opinions of what to believe and follow the behavior of other investors [Davenow
and Welch (2004)]. An important implication of the formation of herding is that actors tend to rely on consensus
opinion and not on the past trading price of the underlying asset. As a result, according to Fernandez (2010), herding
could exacerbate asset return volatility and destabilize financial markets, particularly under the stress conditions
(turmoil).

From the theoretical point of view, the existence of herding can be rationalized in the context of the information
advantage enjoyed by some investors and information externalities, which can affect the capital structure, research
and development or merger decisions, panic in the industry [Devenow and Welch (2004)], the signal by investor
institutions, anxiety experienced by investors due to conflicting opinions (contradiction) and the incentive to hide
the inability by mimicking the decision making of more able managers. This is reasonable due to the pressure of
having the best-expected results from the shareholders.

The theoretical literature mostly conceptualized herding behavior as a behavior that is characterized by the actions
of individuals collectively buying and selling by following (trailing) a particular factor (performance) or style
(market portfolio), certain sectors, styles, or macroeconomic signals [Al Shboul (2012)]. Consequently, herding is
identified by utilizing the information contained at the movement of stock prices in cross-sections. Hwang and
Salmon (2004) argued herding would be stronger in the extreme market conditions compared to the stress capital
market situation due to potential of decreasing the cross-sectional standard deviation of return.

2.2 The Existence of Herding Hypotheses

Henker et al. (2006) examined whether herding in the broad market sector and industry occurs every day and
intraday in the Australian equity market. Employing 160 of the most actively traded stocks on the Australian Stock
Exchange for the period 2001-2002, they did not find any intraday sector herding behavior either in the entirety or in
industrial markets. However, herding that occurs intraday just occurs to a particular company stock. The difficulty in
distinguishing herding in normal market circumstances and the extreme market conditions led the researchers then to
distinguish the actual herding concept from the two following concepts: rational and irrational herding. The concept
suggests that investors adopt rational investment decisions of other investors to protect their own interests
[Bickchandani et al. (1992) and Davenow and Welsh (2004)]. Irrational herding occurs when investors blindly copy
other decisions despite having their own information [Nofsinger and Sias (1999) and Sias (2004)]. In addition to the
rational and irrational, according Davenow and Welsch (2004), there is still a semi-rational herding concept, where
an investor will be based on heuristics in decision-making of other investors, combined with their own information.
The motive of the heuristics is for the sake of long-term yield maximization.

Lakonishok et al. (1992) developed it specifically to detect the size of herding among pension fund managers. They
analyzed the correlation of trading patterns among a group of investors to buy or sell a particular asset in the same
time period, and they found that there was no convergence of trade between pension fund managers. Research
demonstrates that many who apply the size of Lakonishok et al. (1992) reported this, and there is strong evidence of
herding through the correlation between stock returns and trading volumes [Hiemstra and Jones (1994) and Wei et
al. (2009)]. However, to capture the differences in the behavior of traders, a group of researchers found weak
evidence of the existence of herding through the correlation between individual and aggregate stock returns and
trading volumes [See Alemanni and Onelas (2009)].

Copyright by author(s); CC-BY 272 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

Another study investigating the group herding behavior among foreign investors in emerging markets reported no
evidence of herding [Chen (2001), Koutmos and Saidi (2002) and Park and Sabourian (2011)]. Although these
studies have made enough contributions to the literature, there will still be discussions about different herding
measurements (as a correlation pattern among group of investors to buy and sell a particular asset) between
developed and developing countries. They have discussed differences in the magnitude of herding between
developed and developing countries, and many other studies still indicate there is herding in the market.

Chang et al. (2000) analyzed the markets in the US, Hong Kong, Japan, South Korea, and Taiwan, and they did not
find evidence of herding in the US and Hong Kong, but there was only some evidence for Japan. However, Chang et
al. (2000) reported significant herding evidence in emerging markets, such as South Korea and Taiwan. Choe et al.
(1998) studied the Korean currency crisis period in 1997 and found foreign investors were more likely to herd than
domestic investors. Empirical evidence for the case of IDX herding has been discovered by Gunawan et al. (2011),
Setiyono (2012) and Setyawan and Ramli (2013). Based on the previous description, an alternative hypothesis as
follows could be proposed:

Hypothesis 1. In the daily stock transactions in the Indonesia Stock Exchange (IDX), there was herding behavior by
domestic investors following that by foreign investors during the crisis (2009-2011).

2.3 Hypotheses of the Herding Determinants

Herding intensity is negatively related to the size of the market, which indicates that institutional herding in the
Taiwan stock market is primarily driven by information asymmetry [see Zhou and Lai (2009)]. The size of the
market in Taiwan will be perceived differently by investors. Chamley (2004) suggested that the problem of
information asymmetry in the capital market can explain the volatility of the stock price. Imperfectly informed
investors in the capital markets may cause the stock price to be more volatile than the capital market of perfectly
informed investors. There are two factors that contribute to the change in stock price: changing expectations of
future cash flow and a lack of information. Usually stocks that have less information are small in capitalization
compared to the large stakes. Small stocks are generally less researched by capital market researchers compared to
the large stakes. This is due to the relatively high cost of research. Conducting research for large stocks are expected
to provide greater profits, considering their trading stocks will be more liquid than those of small stocks.

The liquid or illiquid stock situation discussed in the literature mainly tested the impact on the stock returns [Lee and
Rui (2002) and Li and Wang (2010)]. Assets that react strongly to changes in the overall market liquidity crisis
make the illiquidity to cause investors’ expected returns to be higher, and investors tend to require systematic
liquidity premiums [Campbell, et al. (2009)]. Other research that focuses on the direct impact of illiquidity on stock
prices is Amihud and Mendelson (1986) and Amihud (2002). Amihud and Mendelson (1986) argued that the spread
(the difference between the stock prices) is a liquidity-based transaction cost for traders. When liquidity increases,
transaction costs (spread) decrease. Consequently, the share price will rise. However, illiquidity will reduce
investors’ interest due to the reduced attractiveness of the stock. When illiquidity occurs, corporate issuers will
perform various actions to raise the liquidity of their shares in more ways than discounting the price.

Amihud (2002) established the relationship between illiquidity and stock prices. He found illiquidity decreases the
share price on the basis of an increase in price volatility. During volatility, the asset liquidity premium increases. As
a result, the stock becomes more risky. Amihud (2002) used a general equilibrium model to connect the liquidity of
the stock market and its impact on asset prices. Amihud (2002) found that the trade in the stock market will be liquid
(increased in liquidity) when trading has a complete match between the seller and the buyer. Lou and Shu (2014)
found that in China's stock market, the illiquidity factor is derived from information asymmetry. Kremer and Nautz
(2013) found that the herding intensity depends on the characteristics of the stock, including stock liquidity, as well
as on historical stock returns and volatility of stock prices (unintentional herding). However, in contrast to the
intentional herding theory, herding is more inclined to the more liquid and larger shares. Herding intensity depends
on the volatility of the past within the asymmetric information situation. The volatility leads to increased sell
herding and decreased buy herding.

Copyright by author(s); CC-BY 273 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

When investors are uninformed about the real dividend growth rate, expectations of future cash flows are less
varied. This has the effect of reducing price volatility. However, it illustrates that there is uncertainty of cash flows
over these shares in the future. Investors then demand higher premiums to accommodate the information asymmetry,
and the stock price becomes more sensitive in trade deals. This would increase price volatility. The net changes in
the volatility of the price (highest and lowest price) depend on factors expectations or information asymmetry that
has a dominant influence on the stock price. When information asymmetry is considered more important or more
dominant, then the prices become more volatile as investors are uninformed.

Wang (2000) found that the information asymmetry among investors can lead to increased price volatility. In the
information asymmetry situation, the investors who have more information will have the advantage over the
uninformed investors. Therefore, the uninformed investors will face miscast (adverse selection) problems when they
respond to various types of information, so they will buy the stock at a price higher than the actual price
(overpriced). They demand an additional premium for the risk they trade off to the investors who have more
information. This situation makes it more elastic and will increase the price volatility. The existence of uninformed
investors may cause the risk premiums to become much higher than in the situation of symmetric information (type
of efficient markets). When many investors are less informed (uninformed investors), the price then is formed on the
basis of less fundamental information. This will result in the greater uncertainty of future cash flows. Therefore, in
stock investing, the uninformed investors will require higher premiums. Lou and Shu (2014) stated that as
uninformed investors increase, the premiums they will assign will increase. In situations where the uninformed
investors increase, the asymmetry of information will be greater, and the greater the difference between the highest
and lowest prices (spread). Therefore, investors are increasingly demanding higher premium as compensation.

De Long et al. (1990) suggested that the trade within the information asymmetry in the capital market could increase
price volatility due to a greater risk premium. Such high expectations of stock returns depend only on the risk of
stock future cash flows. Furthermore, according to De Long et al. (1990), although the trade within the increased
price volatility is without information asymmetry, this does not affect the risk premium because it will not change
the stock fundamental risk. However, if there is information asymmetry in the capital market, the trade will occur
with different information, and this will affect the quality of the price of the personal information of each investor of
the expected future cash flows.

Amihud (2002) demonstrated that there was a significant negative correlation sequence of long-term stock returns.
There was a successive negative correlation between the stock returns and the average of excess return (the
difference between actual returns and expected returns) that can be derived from a negative average of the
underlying variables. However, in the presence of information asymmetry, the uninformed investors then can only
learn from the state variables or from the realized returns (past). The future investors’ expected return will depend
on the average return in the past, and this will allow generating negative stock returns. Herding is due to the stock's
liquidity, volatility and past stock returns derived from information asymmetry [see opinions Testa (2012) and
Bootrz and Kremer (2013)].

Hypothesis 2: information asymmetry affects the intensity of herding.

3. RESEARCH METHODS

3.1 Research Data

Employing daily stock transaction of each share traded, the study was conducted during the crisis period of 2009-
2011. We used the Iq prime plus daily transactions data, such as transactions of each foreign buy and sell stocks, and
domestic buy and sell stocks with the high and low prices.

Copyright by author(s); CC-BY 274 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

3.2 Operational Variables

3.2.1 Herding Behavior

Setyawan and Ramli (2013) described herding behavior of investors that follows the decisions taken by other
investors. Operationally, the herding will be measured by the herding measure of Lakonishok et al. (1992),
Nofsinger (1996) and Neal et al. (2002). This herding size has a major component of the purchase and sale of shares
interaction between foreign investors and domestic investors. The code symbol size is according to herding research
by Lakonishok et al. (1992), Nofsinger (1996) and Neal et al. (2002), and is represented by Hi with the formulation:

Hi = 1/N ∑ ∑ ABS (Bijt/ (Bijt + Sijt) – pit) – AFijt (1)

For the effectiveness of the above Hi value, the data Bijt and Sijt are required to be present in the database trading.
Bijt and Sijt are buy and sell transactions between foreign investors and domestic investors. If there is an indication
of trading between domestic and foreign investors, the potential for domestic investor herding appears.

Component Notes
Hi herding measure from investor group i
Bijt number of buy trades from investors group i, in stock j; on day t
Sijt number of sell trades from investors group i, in stock j, on day t
Pit the proportion of trades by group i; across all stocks; on day t that are buys (i.e., the average of Bijt/(Bijt+Sijt)
over j)
AFijt adjustment factors, i.e., expected value of the absolute value of (Bijt/ (Bijt + Sijt) – pit)) with the assumption that
Bijt follows a binomial distribution with the probability pit of success.

Based on the formula of herding measure, if the Hi value is higher, it means there is no indication of herding
behavior. The more positive the Hi is, the bigger the buy herding behavior indication, while if it is more negative,
there is an indication of herding because of the behavior of the sell. When the value of Hi is close to zero, it means
there is no indication of herding behavior. For the case when the value of the IDX Pit and AFijt are difficult to
estimate, it is possible to be equated with zero. Purpose ABS is the absolute value when the results are compared
with the minus positive results.

Information Asymmetry

“Information asymmetry is present when one party to a transaction has more or better information than the other
party. (This is also called a state of asymmetric information). Most commonly, information asymmetries are studied
in the context of principal-agent problems”. (Aldea and Marin, 2007)

IAjt = [(HTijt – HTejt)/( HTijt + HTejt) 0,5] x 100% (2)

IAjt = Information Asymmetry of Transaction on Stock j on day t.

HTi jt = The Highest Price of Transaction on Stock j on day t.

HTe jt = The Lowest Price of Transaction on Stock j on day t.

3.3 Analysis Method

3.3.1 Herding Existence Model of Hypothesis Testing

To test the potential of herding in the Indonesian Stock Exchange (IDX), in addition to using the amount of herding
as size in B.1, we will use descriptive chart patterns. Using the dummy variable of the amount of herding, which is
measured by 1, and 0 otherwise, it was proven that the existence of herding is one point only (discrete). The pattern
in the chart shows the continuity of herding that occurs in the IDX. We consider the continuity of herding to be as

Copyright by author(s); CC-BY 275 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

important as the dynamics of the herding patterns, in terms of them being stationary or fluctuating. When the
movement indicates stationary herding, there is an indication of information asymmetry as a determinant of herding,
or else the fluctuated herding movement signals that there is a stronger effect of information asymmetry.

Frequency chart patterns and sizes were employed by Lakonishok et al. (1992) in a herding study. The indications of
herding will be proven (H1 will be accepted) when the pattern of the frequency and size of herding by Lakonishok et
al. (1992) are not too stationary, and they will be characterized by a certain trend. The total (H), buy (Hb) and sell
(Hs) herding were tested.

3.3.2 Herding Determinants Model of Hypothesis Testing

For this hypothesis testing, we will be using analysis of VAR or VEC. VAR analysis is used when herding
relationship patterns and information asymmetry are linear, while the VEC analysis is used when the relationship
between the two is non-linear. The terms of VAR or VEC analysis is the existence of a causal relationship between
herding and information asymmetry with a Granger Causality test.
Technically, a VAR or VEC analysis model of herding and information asymmetry can be described as follows:

Hit = β0 + β1Hi(t-1) +… + βnHi(t-n) + ψ1IAi(t-1) + …..+ ψn IAi(t-n) + ε it (3)

IAit = δ0 + δ1IAi(t-1) + ….δnIAi(t-n) + ω1Hi(t-1) + ….. ω nHi(t-n) + ε it (4)

Hypothesis 2 (H2) is accepted when β1, … βn, ψ1….. ψn, δ1, … δn and ω1, … ωn have a significant influence on H
and IA. [p-value ≤ 0.05].

4. RESULTS AND DISCUSSION

4.1 Identification of Herding During 2009-2011

Descriptive statistical analysis of the frequency herding behavior (Freq_herding), either buy or sell, demonstrates
that the average buy herding is larger than sell herding. Likewise, the median is greater for freq. herding to buy than
freq. herding to sell. This suggests that more domestic investors conduct herding when they buy shares. While there
is herding of domestic investors to the foreign investors in the context of the sale of shares, or freq. herding to sell,
domestic investors do not always conduct herding. Looking at the not supportive economic situation (not good),
then domestic investors tend to conduct a good buy or sell herding. This happened in 2009. However, when the
economy started to recover in the first quarter of 2010 to 2011, domestic investors tended to conduct buy herding
rather than sell herding. Maximum Freq. herding to buy is at 0.275168, while the result for the maximum freq.
herding to sell is 0.323077 (see table 4.1).

This suggests that when the current economic situation is volatile, the panicked domestic investors follow the
trading transactions undertaken by foreign investors, either sale or purchase of shares, even when freq. herding to
sell exceeds freq. herding to buy. This is supported by the smaller standard deviation of freq. herding to buy than the
standard deviation of freq. herding to sell. The domestic investors respond better to a volatile economic situation and
become more panicked by selling shares compared to buying stocks by following foreign investors who sell their
shares. Frequency of herding against foreign investors buying has a smaller deviation than the frequency deviation
of sell herding. Domestic investors are more convinced by the foreign investors in purchasing stocks, so they are
more stable following the purchase of shares bought by foreign investors. When selling stocks in the normal or not
volatile economic situation, domestic investors do not get hung up on the sale of the shares by foreign investors, but
rather on the stock return. Except in a turbulent economic situation, they tend to follow the behavior of foreign
investors selling shares.

Copyright by author(s); CC-BY 276 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

Table 4.1. Freq. herding to buy and Freq. herding to sell in 2009 – 2011
Component Freq. Herding To Buy Freq. Herding To Sell
Mean 0.141794 0.093766
Median 0.123529 0.019231
Maximum 0.275168 0.323077
Minimum 0.000000 0.000000
Std. Dev. 0.053798 0.103082
Skewness 0.255314 0.568319
Kurtosis 2.267630 1.535438
Jarque-Bera 14.77969 63.58243
Probability 0.000617 0.000000
Sum 63.09835 41.63214
Sum Sq. Dev. 1.285057 4.707241
Observations 445 444

Figure 4.1 provides a more detailed picture of the freq. herding to sell; in 2009, there was a fairly large freq. herding
to sell, even exceeding freq. herding to buy. This reinforces the view that in the current economic turmoil, domestic
investors tend to follow the trading behavior of foreign investors in buying and selling shares, particularly in selling
shares. While in a normal or stable economic situation, the domestic investors do not really follow sell herding
because they believe that in the sale of shares, they will follow the pattern of returns in accordance with their desired
return.

When the economic situation is in a normal state, freq. herding to buy is declining but still larger than freq. herding
to sell (Figure 5.2., 5.3.). After 2009, the economic situation in Indonesia tended to be stable and even increasing, so
freq. herding to buy continues although the intensity decreases. However, freq. herding to sell is down dramatically
and occurs only in small to zero frequency.

Figure 4.1. Freq. Herding to Sell in 2009 - 2011

   
.35

.30

.25

.20

.15

.10

.05

.00
10000 20000 30000 40000 50000 60000

The buy herding by domestic investors following foreign investors in emerging markets indicates that domestic
investors still lack confidence in the results of the analysis of their own work, and the domestic investors’ number of
transactions is much smaller than the foreign investors’ transactions conducted. Therefore, they prefer to follow the
purchase transaction of foreign investors. In addition, market analysis knowledge of domestic investors on the price
of a share is far less than that of the foreign investors. It makes it much safer for them to follow the purchase
transaction of foreign investors who are believed to have more expertise than domestic investors.
Copyright by author(s); CC-BY 277 The Clute Institute
The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

The data types of stocks with zero or no herding have a large enough frequency or greater per week compared with
the average data of the stocks that have a herding transactions by domestic investors, and it usually occurs in the
second or third tier stocks, which are less attractive to foreign investors. This suggests that the majority of domestic
investors tend to invest in stocks that are not traded by foreign investors, with a small total volume transaction. The
price of the stock is not influenced by foreign investors. Domestic investors traded the stocks with each other in
stocks that are outside of the seed stocks or outside the 45 blue chip stocks. The portfolio developed by the domestic
investors are more in stocks other than the LQ 45 index stocks, so the information asymmetry among domestic
investors was allegedly not large enough. Because the supply and demand of shares involve domestic investors’ lack
of knowledge or lack of information, the information asymmetry is not large enough, except in the case of insider
trading. Insider trading is alleged by the large enough volume of transactions that is above the average transactions
volume during the period (per week), although there is no information that could lead to an increase in the amount of
the transaction.

Figure 4.2. Freq. Herding to Buy in 2009 - 2011

Freq  Herding  Buy

.28

.24

.20

.16

.12

.08

.04

.00
10000 20000 30000 40000 50000 60000

Figure 4.3. Freq. Herding to Buy and Freq. Herding to Sell in 2009 - 2011

.35

.30

.25

.20

.15

.10

.05

.00
10000 20000 30000 40000 50000 60000

Freq  Herding  Buy

Freq  Herding  Sell

Copyright by author(s); CC-BY 278 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

4.2 Herding Identification During 2009-2011

During 2009 (the unit data of 100-300), the total herding (H) chart did not fluctuate too much at the beginning and
then declined in the second quarter following. It was between .40 and .50. The buy herding (Hb) and sell herding
(Hs) were similar and did not fluctuate too much either; it was between .20 and .30. A different case is in the second
quarter following; the buy herding dropped to .10 - .15, and the sell herding to .00 - .02. After the second quarter of
2009, there was still buy herding (Hb) with smaller frequency, but it was not with the sell herding. There was not
sell herding since. For the years 2010-2011 (the unit data 300-700), there was buy herding, but it was not with the
sell herding. This indicates that during the crisis (beginning of 2009), larger herding occurred in the IDX, and it still
continued afterwards but with a smaller frequency. The buy and sell herding occurred during the crisis, but there
was only buy herding afterwards. Domestic investors are more inclined to conduct buy herding.

Figure 5.4. Herding (H, Hb and HS) in 2009 – 2011

0.6

0.5

0.4

0.3

0.2

0.1

0.0
100 200 300 400 500 600 700

H HB HS

Table 5.2 Normality of Herding (H, Hb and Hs) in 2009 – 2011 with Proxy from Lakonishok, et. al. (1992)
Component H Hb Hs
Mean 0.185524 0.130625 0.054899
Median 0.138122 0.122222 0.017442
Std. Dev. 0.122408 0.047988 0.083068
Jarque-Bera 231.5994 15.71057 354.9736
Probability 0.000000 0.000388 0.000000

4.3 Identification Information of Asymmetry During 2009-2011 Utilizing The Size of Aldea and Marin (2007)

In Figure 5.5, the visible range of the value of information asymmetry is 0.005 to 0.015. It indicates that the
information asymmetry is not fluctuating too much. There was an indication that the domestic investors had
difficulty in obtaining additional information in deciding trading strategies. However, the rather stable value of
information asymmetry implicates high liquidity in stock exchange trading among domestic investors. The high
liquidity of the stock exchange will have a positive impact on the performance of the Jakarta Composite Index (JCI),
which is expected to lead to special attraction for foreign investors. The resulting implications include that the high
information asymmetry could improve herding behavior between foreign and domestic investors on the stock
exchange.

Copyright by author(s); CC-BY 279 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

Figure 5.5. Information Asymmetry in 2009 – 2011 with proxy from Aldea and Marin (2007)

0.025

0.020

0.015

0.010

0.005

0.000
100 200 300 400 500 600 700

IA

Figure 5.6. The visible evidence of normality for information asymmetry is shown in the histogram pattern above,
i.e., the value of the Jarque-Berra is very high, and the amount of information asymmetry exceeds the average value
of the standard deviation. The normality of information asymmetry will support stationary patterns when tested with
herding.

Figure 5.6. Test of Normality from Information Asymmetry in 2009 – 2011 with proxy from Aldea and Marin (2007)
200
Series:  IA
Sample  1  735
Observations  735
150
Mean          0.004332
Median    0.003626
Maximum    0.023791
100
Minimum    0.000000
Std.  D ev.      0.002652
Skewness      2.254088
50 Kurtosis      10.90092

Jarque-­Bera  2534.162
Probability  0.000000
0
0.000 0.005 0.010 0.015 0.020

4.4 Granger Causality and the VEC Model

4.4.1 Granger Causality Analysis on the 4th Lag

The Granger causality analysis is effective on the 4th lag. Reciprocal relationships were found between total herding
and buy herding, total herding and information asymmetry, and sell herding and information asymmetry. Reciprocal
relationships between total herding and buy herding, total herding and information asymmetry, and sell herding and
information asymmetry indicate that the three relationships can be estimated by the VAR or VEC model. The VAR
model could be employed if a reciprocal relationship is linear, whereas if the reciprocal relationship between these
three models is nonlinear, the VEC model would be more relevant.

Copyright by author(s); CC-BY 280 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

The cause of the non-linear nature of the relationship among herding and various determinants is the herding
measure, which is the ratio of the number of purchases and sales of stocks of foreign and domestic investors. The
herding amount obtained of 0-1, cannot be in the form of a straight diagonal line of data distribution when mapped
in P-Plot, which is usually used as normality check.

4.5 VEC Model

4.5.1 Total Herding and Buy Herding

The value co-integration equation of the buy herding significant t-test indicates the existence of a long-term
equilibrium relationship between total herding and buy herding. Most of the long-term period shares to buy
transactions on the stock exchange have been turned out by the foreign investors. The actions of the foreign
investors to buy shares in a period will be followed by the domestic investors.

The VEC regression model results found that there is significant influence of herding in the past compared with
herding now. This supports the theory of technical analysis stock: history repeats itself. In the transactions of
purchases and sales of shares on the stock exchange, foreign investors will look at historical performances. In
addition, there is also a significant effect of buy herding in the past on the current herding. The historical
performance of all transactions will be important information for foreign investors; this then is followed by the
domestic investors.

4.5.2 Total Herding and Information Asymmetry

In this section, it is found that there is a visible long-term relationship (co-integration) between total herding and
information asymmetry. The existence of co-integration between herding and information asymmetry indicates that
the action of purchases and sales of shares of foreign investors is followed by the domestic investors. This also
results in increasing price volatility as a result of a wide range of stock prices in the order book.

Related to the VEC model regression results, herding behavior is influenced by information asymmetry. However,
information asymmetry will not be influenced by herding behavior. The information asymmetry is influenced by the
information asymmetry of the two and three previous periods. This finding indicates that the domestic investors
followed the action of the foreign investors due to the lack of information. This is caused by the differences in
access to the latest information and most comprehensive data between foreign investors and domestic investors that
are still imposed in the IDX.

4.5.3 Sell Herding and Information Asymmetry

The VEC model analysis between sell herding and information asymmetry found that sell herding and information
asymmetry had a co-integration. The co-integration between the two is the act of the sale of shares of foreign
investors, followed by the domestic investors. The co-integration is more due to the information asymmetry. In the
long term, when the price volatility is increasing on the stock exchange, this will be encouraged by domestic
investors who conduct panic selling.

The information asymmetry shown affects the sell herding during 1, 2 and 3 periods previous. However, as an
endogenous variable, the sell herding as exogenous in VEC does not have a significant effect on information
asymmetry. The information asymmetry is only affected by the information asymmetry itself from the previous
period. These findings suggest that the information asymmetry the domestic investors suffer could potentially lead
to panic selling when they do not have enough confidence in and experience with a share.

Copyright by author(s); CC-BY 281 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

4.6 Results Discussion

Hypothesis 1: Testing (Existence of Herding)

The existence of herding in the study can be proved by the results of herding frequency movements during 2009-
2011. The frequency of buy herding and sell herding movement is shown in the results of different graphs. Sell
herding frequency movement was the dominant frequency in the period 2009-2010 and less dominant in the period
2010-2011. Different things occurred to produce a consistent buy herding movement that was dominant in the period
2009-2011. The dynamic movement of buy and sell herding confirms the existence of herding in the IDX and
automatically receives H1.

Employing the amount of the existence of herding of Lakonishok et al. (1992), it is found that there was also a
dynamic movement of buy herding and sell herding during the period 2009-2011. The dynamic buy herding
movement appears more dominant than sell herding. The magnitude of sell herding has a range of 0.05 to 0.10,
smaller than buy herding, which has the range of 0.20 to 0.30. The results also indicate that domestic investors are
more inclined to conduct buy herding than sell herding. Evidence of the existence of herding in the IDX supports the
studies by Neal et al. (2002), Bowe and Domuta (2004), Gunawan et al. (2011), Setiyono (2012) and Setyawan and
Ramli (2013).

Hypothesis 2: Testing (Determinants of Herding)

At the time of the analysis of the dynamic movement of information asymmetry with the graphical analysis, the
results demonstrated that a movement pattern is not as dynamic as herding. When we utilize the scatter plot analysis
(distribution of data) with a concern for herding and information asymmetry, a quadratic pattern is detected with the
characteristic of information asymmetry forming a parabolic curve with a maximum value of 0.20.

When regression was performed to test hypothesis 2, the VEC (Vector Error Correction) model was then
implemented, which preceded the Granger causality analysis; it was proved significant in the 4th lag. The VEC
analysis results prove the existence of herding influenced by information asymmetry, and more surprising, changes
in the IDX information asymmetry will lead to sell herding. These results support the study and confirm H2, Testa
(2012) and Brootz and Kremer (2013).

5. CONCLUSIONS AND SUGGESTIONS

5.1 Conclusions

There are two objectives in this study: 1) Confirming the existence of herding in the IDX and 2) Confirming whether
information asymmetry could be a determinant of herding. The results demonstrated that herding behavior still
occurs on the stock exchange, but of a different pattern from the previous studies by Gunawan et al. (2011),
Setiyono (2012) and Setyawan and Ramli (2013). We split herding into buy herding and sell herding. The purpose
of this herding solution is to see whether there is a donation effect of information asymmetry in accordance with
Testa (2012) and Brootz and Kremer (2013).

When analyzing the dynamic movement of buy herding and sell herding frequency utilizing the amount of herding
of Lakonishok et al. (1992), the second type of herding presents the dynamic movement patterns that prove the
existence of herding continuously during the period 2009-2011 that remained since the crisis. These results extend
the existence of discrete herding occurring in earlier studies.

The second objective of this study was also successfully demonstrated by the effect of information asymmetry, in
particular, on sell herding. This result is unique to the case of the stock exchange because it means the phenomenon
of panic selling that occurs is caused by herding and information asymmetry in accordance with the opinions of
Testa (2012). Domestic investors who are exposed to information asymmetry tend not to disagree with foreign
investors’ trading position but instead always support them.

Copyright by author(s); CC-BY 282 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

5.2 Suggestions

This research still needs to be improved in terms of the measurement of buy herding and sell herding because it
employs only the negative indication of the proxy of herding by Lakonishok et al. (1992), which leads to sell
herding, and a positive indication for buy herding. Another important aspect to improve is to reconstruct the proxy
of information asymmetry of Aldea and Marin (2007) from daily data into the intraday level.

AUTHOR INFORMATION

Ishak Ramli was born in 1952; Formerly Finance and Administration Manager for thirteen years, Finance Director
for thirteen years and Human Resources Director for eleven years; has been teaching Corporate Finance, and
Management Accounting for 26 years, and Capital Market & Investment for six years. E-mail: ishakr@[Link]

Sukrisno Agoes was born in 1947; Public Accountant since 1985; Professor in Auditing; has been teaching
Auditing and Financial Accounting for 36 years. E-mail: sukrisno.a@[Link]

Ignatius Roni Setyawan was born in 1973; has been teaching Corporate Finance and Investment for fifteen years.
Email: ignronis@[Link]

REFERENCES

Alemanni, B. and J.R.H. Onelas (2009), Herding Behavior by Equity Foreign Investors on Emerging Markets, Working Paper
[Link],. 1 – 28.
Al-Shboul, M. (2012), Asymmetric Effect and the Herd Behavior in the Australian Equity Market, International Journal of
Business and Management, Vol. 7. No.7, 121-140
Aldea, A. and D. Marin (2007), Asymmetric Information- Adverse Selection Problem, Economics and Applied Informatics,
Years III, 1-8.
Aggarwal, S., S. Faircloth, C. Liu and S.G. Rhee (2009), Why Do Foreign Investors Underperform Domestic Investors in
Trading Activities? Evidence from Indonesia, Journal of Financial Market Vol. 12,32-53
Amihud, Y. and H. Mendelson (1986), Asset Pricing and the Bid-Ask Spread, Journal of Financial Economics 17, 223-249.
Amihud, Y. (2002), Illiquidy and Stock Return: Cross-Section and Time-Series Effects, Journal of Financial Market 5, 31-56.
Bickchanadani, S., D. Hirshleifer and I. Welsch (1992), A Theory of Fads, Fashion, Custom and Cultural Change in
Informational Cascades, Journal of Political Economy,100, 992-1026.
Bootrz, C. and S. Kremer (2013), The Impact of Information Risk and Market Stress on Institutional Herding, Working Paper
from Freie Universitat Berlin, Germany, 1-28.
Bowe, M. and D. Domuta (2004), Investor Herding During Financial Crisis: A Clinical Study of the Jakarta Stock Exchange,
Pasific-Basin Finance Journal Vol.12, 387 – 418
Chamley, C. P. (2004), Rational Herds: Economic Models of Social Learning., Cambridge University Press.
Chandra, M. (2012), Pengukuran dan Analisis Perilaku Herding pada Saham IPO Bursa Efek Indonesia tahun 2007-2011, Tesis
MM UI, 1-139.
Chang, E C, J. [Link] and A. Khorana (2000), An Examination of Herd Behavior in Equity Market: an International
Perspectives, Journal of Banking and Finance, 24, 1651-1679.
Chiang, T.C., J. Li, L. Tan and E. Nelling (2011), Dynamic Herding Behavior in Pasific–Basin Markets: Evidence and
Implications, Working Paper from Drexel University, 1-38.
Choe, H., B.C. Kho and R.M. Stulz (1998), Do Foreign Investors Destabilize Stock Market? The Korean Experience in 1997,
Working Paper from [Link], 1- 40
Davenow, A. and I. Welch (2004), Rational Herding in Financial Economics, European Economic Review Vol. 40. No.2, 603 -
615
De Long, J.B, A. Shleifer, L/H. Summers and R. J. Waldmann (1990), Noise Trader Risk in Financial Market, The Journal of
Politucal Economy, Vol.98, No. 4, 703-738.
Dvorak, T. (2005), Do Domestic Investor Have an Information Advantage? Evidence from Indonesia, Journal of Finance 60 (2),
817-839
Fernandez, V. (2010), Asymmetrical Individual Return Herding, Academia Revista Latino Americana de Administracion, Vol.
45, 20-39.
Franks, R. and F. Westerhoff (2011), Estimation of a Structural Stochastic Volatility Model of Asset Pricing, Journal of
Computational Economics Vol. 38, No.1, 53-83.
Gunawan, H. Wijayanto, N.A. Achsani dan L.O. A Rahman (2011), Pendeteksian Perilaku Herding pada Pasar Saham Indonesia
dan Asia Pasifik, Forum Statistika & Komputasi, Vol. 16 No.2, 16-23.
Copyright by author(s); CC-BY 283 The Clute Institute
The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

Henker, J., T. Henker and A. Matsios (2006), Do Investors Herd Intraday in Australian Equities, International Journal of
Managerial Finance Vol. 2, No.3, 196-219.
Hiemstra, C. and J.D. Jones (1994), Testing for Linear and Non Linear Granger Causality in the Stock Price-Volume Relation,
Journal of Finance Vol. 49, 1639 – 1664
Hwang, S. and M. Salmon (2004), Market Stress and Herding, Journal of Empirical Finance Vol. 11, 585 – 616
Iihara, Y., H.K. Kato and T. Tokunaga (2001), Investor’s Herding in the Tokyo Stock Exchange, International Review of
Finance Vol. 2 No. 1, 71-98.
Koutmos, G. and R. Saidi (2002), Positive Feedback Trading in Emerging Capital Markets, Applied Financial Economics Vol.
11, 291 – 297
Kremer, S. and D. Nautz (2013), Short-term Herding of Institutional Traders: New Evidence from German Stock Market,
European Financial Management, Vol. 19., No.4, 730-746.
Lakonishok, J., A. Shleifer and R.W. Vishny (1992), The Impact of Institutional Trading on Stock Prices, Journal of Financial
Economics 32, 23 - 43.
Lou, X. and T. Shu (2014), Price Impact or Trading Volume: Why is the Amihud (2002) Illiquidity Measure Priced? Working
Paper from Lehigh University, 1-51.
Nofsinger, J.R. and R.W. Sias (1999), Herding and Feedback Trading by Insitutional and Individual Investors, Journal of
Finance Vol. 45 No. 6, .2263 – 2295
Neal, B.C., S.L. Jones, [Link] and R. Neal (2002), Herding and Feedback Trading by Foreign Investors: The Case of
Indonesian during the Asian Financial Crisis, Working Paper from European Financial Management Association, 1 –
20.
Park, A.A. and H. Sabourian (2011), Herding and Contrarian Behavior in Financial Market, Econometrica, 79, 973-1026.
Setiyono (2012), Perilaku Pengikutan (Herding Behavior) dan Efisiensi Pasar Saham, Ringkasan Disertasi FE-UGM, 1-44.
Setyawan, I.R. dan [Link] (2013), Pengujian Fenomena Herding Behavior di Bursa Efek Indonesia Terkait Peran dan Kontribusi
Investor Asing Selama Periode Observasi 2006-2011, Laporan Penelitian Hibah LPPI UNTAR dengan no. SPK: 906-
SPK-LPPI/Untar/IX/2012, 1-69.
Sias, R.W. (2004), Institutional Herding, Review of Financial Studies Vol. 17, 165 – 206
Testa, A. (2012), Path-Dependent Behavior with Asymmetric Information about Trader’s Type, Discussion Paper, University of
Bonn, 1-36.
Wang, J. (2000), Foreign Trading and Market Volatility in Indonesia, Working Paper from UNSW, pp. 1-25
Wei, L., S.G. Rhee and S.S. Wang (2009), Differences in Herding: Individual versus Institutional Investors in China, Working
Paper [Link], 1 - 28.
Zhou, R.T. and R.N. Lai (2009), Herding and Information Based Trading, Journal of Empirical Finance 16 (3), 388-393.

Copyright by author(s); CC-BY 284 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

APPENDIX

Pairwise Granger Causality Tests

Null Hypothesis: Obs F-Statistic Probability
HB does not Granger Cause H 731 11.2304 7.6E-09
H does not Granger Cause HB 13.7490 8.2E-11
HS does not Granger Cause H 731 11.2304 7.6E-09
H does not Granger Cause HS 0.59546 0.66603
IA does not Granger Cause H 731 3.29053 0.01098
H does not Granger Cause IA 2.55494 0.03778
HS does not Granger Cause HB 731 13.7490 8.2E-11
HB does not Granger Cause HS 0.59546 0.66603
IA does not Granger Cause HB 731 1.80201 0.12657
HB does not Granger Cause IA 2.79296 0.02545
IA does not Granger Cause HS 731 3.07651 0.01579
HS does not Granger Cause IA 2.33977 0.05376

VEC Model of Total herding and Buy herding

Error Correction: D(H) D(HB)
0.230447 0.222268
CointEq1 (0.03747) (0.02569)
(6.14954) (8.65229)
-0.538892 -0.059217
D(H(-1)) (0.08545) (0.05858)
(-6.30619) (-1.01086)
-0.381836 -0.070385
D(H(-2)) (0.09120) (0.06252)
(-4.18680) (-1.12581)
-0.364917 -0.124379
D(H(-3)) (0.08955) (0.06139)
(-4.07487) (-2.02604)
-0.316925 -0.164306
D(H(-4)) (0.08055) (0.05522)
(-3.93469) (-2.97572)
0.341833 -0.180090
D(HB(-1)) (0.15335) (0.10512)
(2.22911) (-1.71313)
0.233172 -0.098826
D(HB(-2)) (0.15317) (0.10500)
(1.52235) (-0.94122)
0.323090 0.088022
D(HB(-3)) (0.14150) (0.09700)
(2.28337) (0.90746)
0.359472 0.194395
D(HB(-4)) (0.11739) (0.08047)
(3.06215) (2.41563)
-0.000682 -0.000228
C (0.00185) (0.00127)
(-0.36775) (-0.17958)
R-squared 0.289442 0.375356
Adj. R-squared 0.280560 0.367547
Akaike Information Criteria -8.599291
Schwarz Criteria -8.460870

Copyright by author(s); CC-BY 285 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

VEC Model of Total Herding and Information Asymmetry

Errzor Correction: D(H) D(IA)
0.002486 0.002061
CointEq1 (0.00576) (0.00025)
(0.43192) (8.32205)
-0.551717 -0.000224
D(H(-1)) (0.03732) (0.00161)
(-14.7819) (-0.13953)
-0.400188 0.000530
D(H(-2)) (0.04127) (0.00178)
(-9.69702) (0.29846)
-0.262418 0.000935
D(H(-3)) (0.04101) (0.00176)
(-6.39863) (0.52997)
-0.148139 0.001094
D(H(-4)) (0.03679) (0.00158)
(-4.02651) (0.69103)
-1.040831 -0.244567
D(IA(-1)) (1.11656) (0.04805)
(-0.93218) (-5.08985)
-2.608741 -0.075645
D(IA(-2)) (1.07675) (0.04634)
(-2.42279) (-1.63250)
-2.198762 -0.035876
D(IA(-3)) (1.01300) (0.04359)
(-2.17054) (-0.82295)
-0.303504 -0.034578
D(IA(-4)) (0.86213) (0.03710)
(-0.35204) (-0.93199)
-0.000730 -3.46E-06
C (0.00190) (8.2E-05)
(-0.38415) (-0.04225)
R-squared 0.252449 0.287775

Copyright by author(s); CC-BY 286 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

VEC Model of Sell Herding and Information Asymmetry

Error Correction: D(HS) D(IA)
-0.000569 0.002939
CointEq1 (0.00393) (0.00035)
(-0.14470) (8.38374)
-0.471031 -0.003662
D(HS(-1)) (0.03711) (0.00331)
(-12.6941) (-1.10703)
-0.312762 -0.002827
D(HS(-2)) (0.03990) (0.00356)
(-7.83862) (-0.79472)
-0.249787 0.000617
D(HS(-3)) (0.03976) (0.00354)
(-6.28167) (0.17396)
-0.145180 0.000117
D(HS(-4)) (0.03683) (0.00328)
(-3.94182) (0.03569)
-0.958616 -0.241414
D(IA(-1)) (0.53784) (0.04795)
(-1.78236) (-5.03497)
-1.366803 -0.074879
D(IA(-2)) (0.51937) (0.04630)
(-2.63166) (-1.61722)
-1.021588 -0.038732
D(IA(-3)) (0.48941) (0.04363)
(-2.08739) (-0.88772)
-0.043326 -0.038223
D(IA(-4)) (0.41661) (0.03714)
(-0.10400) (-1.02914)
-0.000469 -5.68E-06
C (0.00092) (8.2E-05)
(-0.51031) (-0.06941)
R-squared 0.204056 0.286646

Copyright by author(s); CC-BY 287 The Clute Institute

The Journal of Applied Business Research – January/February 2016 Volume 32, Number 1

NOTES

Copyright by author(s); CC-BY 288 The Clute Institute