INTRODUCTION
Background
The evolution of global stock markets has led to more comprehensive regulations, highlighting the limitations of the Efficient Market Hypothesis (EMH), which posits that stock prices reflect all available information and assumes rational investor behavior (Fruiger, 2016) However, numerous economic phenomena cannot be adequately explained by EMH or traditional finance theories, prompting the emergence of behavioral finance This field examines how investor sentiment influences decision-making and challenges the notion of rationality in financial choices, particularly in capital markets (Fromlet, 2001) Additionally, Vieira and Pereira (2015) emphasize the importance of exploring theoretical frameworks that account for herding behavior among investors.
Herding is a fundamental human instinct that significantly influences decision-making, particularly in the context of investing This behavior provides insights into investor actions that traditional Efficient Market Hypothesis (EMH) frameworks fail to explain (Chang et al., 2000) Consequently, herding behavior can elucidate various anomalies observed in stock markets According to Chiang and Zheng (2010), herding can reveal the interactions among investors, highlighting its importance in understanding market dynamics.
Herding behavior has been identified as a key factor contributing to excess volatility in stock prices, leading to significant deviations from their fundamental values This phenomenon has been extensively studied for over two decades, establishing it as a well-recognized topic within the realm of behavioral finance.
Herding behavior in finance, as explored by early theorists like Banerjee (1992), Bikhchandani et al (1992), and Welch (1992), reveals that when numerous investors exhibit similar behaviors, others tend to disregard their own information and follow suit (Xie et al., 2015) Despite being a long-studied phenomenon, herding remains a compelling subject due to its ties to complex psychological factors Various researchers, including Christie and Huang (1995), Chang et al (2000), and Hwang and Salmon (2004), have proposed different theoretical methods to quantify herding behavior, leading to potentially conflicting results Christie and Huang (1995) introduced a measure based on the cross-sectional standard deviation (CSSD) of equity returns, while Chang et al (2000) expanded this approach using the cross-sectional absolute deviation (CSAD) to detect herding Hwang and Salmon (2004) further enhanced the measurement by incorporating multiple factors influencing security returns, including size and value (Khan et al., 2011).
Herding behavior is evident in both developed and emerging markets, with the Vietnamese stock market showcasing various documented anomalies Research by Faber et al (2006) highlights stock return anomalies through limit-hits and herding patterns Additionally, Nguyen (2012) identifies a momentum effect within the Vietnamese market Seasonal trends are explored by Friday and Hoang (2015), while Luu et al (2016) further investigate this phenomenon on the Vietnamese stock exchange Truong (2006) reveals a day-of-the-week effect, and Tran and Pham (2011) indicate that individual investors' decisions in this market are generally influenced by these anomalies.
Behavioral finance offers an alternative explanation for investor emotions that classical financial theories cannot account for, particularly in the context of the Vietnamese stock market Research by Dang (2016) highlights several characteristics of Vietnamese investors, including their tendency to mimic foreign investors' trading decisions, hold stocks with significant unrealized losses, and make forecasts based on global market trends Additionally, these investors often engage in speculative trading without discerning between good and bad stocks, relying on opaque information Herding behavior has been a focal point in studies since the market's inception, especially during periods of excessive volatility While initial research, such as Faber et al (2006), identified herding behavior qualitatively, there remains a lack of quantitative evidence Recent studies have employed models like CSSD, CSAD, and HS to analyze herding behavior, yet these have primarily confirmed its existence without detailing the frequency or timing of herding events To address this gap, this study introduces the Weighted Cross-Sectional Variance (WCSV) model, developed by Xie et al (2015) and based on Arbitrage Pricing Theory (APT), which enhances the accuracy and robustness of herding analysis This innovative approach aims to identify specific time points of significant herding activity within the Vietnamese stock market.
The research by Xie et al (2015) demonstrates a significant herding pattern, which is further analyzed using their WCSV model This thesis effectively highlights the influence of herding behavior over the years, providing valuable insights into its impact.
Hence, “Herding behavior: Empirical evidence in the Vietnamese stock market” could be seen as a title for the study.
Research objectives
This study aims to identify herding behavior in the Vietnamese stock market by employing the CSAD model developed by Chang et al (2000) Furthermore, it utilizes the WCSV model introduced by Xie et al (2015) to pinpoint instances of significant herding The analysis focuses on the Vietnamese stock market practices from 2006 to 2016.
Research questions
The study seeks to answer the following three questions:
1 Does herding behavior exist in the Vietnamese stock market?
2 Which years during the period from 2006-2016 do herding show the strong point and strong influential point?
Research Subject and Scope
This study utilizes the WCSV model to analyze herding behavior in the Vietnamese stock markets, specifically focusing on the Ho Chi Minh Stock Exchange (HOSE) and the Ha Noi Stock Exchange (HNX) The research encompasses data from all firms listed on both exchanges throughout the entire study period.
2006 to 2016 (integral data), individual year and in three sub-periods including: before crisis 2006-2007, in crisis 2008-2012 and after crisis 2013-2016.
Contribution
This thesis analyzes the structure of the Vietnamese stock market from 2006 to 2016 by examining the herding behavior of individual investors It employs two key models: the CSAD model, which identifies the presence of herding behavior, and the WCSV model based on the Fama-French three-factor framework, which highlights significant herding and influential points.
The dissertation employs the WCSV model, grounded in the Fama-French three-factor framework, to analyze the presence and effects of pronounced herding behavior in the Vietnamese stock market.
Analyzing the herding behavior of investors provides valuable insights into the Vietnamese stock market from 2006 to 2016, enabling investors to make informed investment decisions.
Structure
The structure of the thesis as follows:
Chapter 1 introduces the herding behavior, the study of many authors in many countries in the world, in Vietnam, through which to say the reason for choosing the study of herding behavior of the thesis Chapter 1 covers the choice of topic, research objectives, research questions, research scope, and research contributions.
LITERATURE REVIEW AND HYPOTHESES
Behavioral Finance
Behavioral finance is an emerging approach to financial markets that addresses the limitations of traditional financial theories by incorporating psychological factors into financial analysis and investment decisions This modern theory posits that some market participants act irrationally, leading to financial phenomena that classical theories cannot adequately explain By examining these anomalies, behavioral finance offers a more comprehensive understanding of market dynamics Various definitions of behavioral finance highlight its significance in the field.
Belsky and Gilovich (1999) define behavioral finance as "behavioral economics," which merges psychology and economics to elucidate the reasons behind individuals' seemingly irrational or illogical financial decisions regarding spending, investing, saving, and borrowing money.
Barber and Odean (1999) suggest that behavioral finance enhances traditional financial economics by integrating observable and systematic human deviations from rationality into conventional financial market models.
Shefrin (2000) defines behavioral finance as “a rapidly growing area that deals with the influence of psychology on the behavior of financial practitioners”
Sewell (2007) suggests that behavioral finance is “the study of the influence of psychology on the behavior of financial practitioners and the subsequent effect on
8 markets” This means that this aspect investigates what happens when investors decide trading investment based on intuitions and emotions (Subash, 2012)
Classical financial theory overlooks the decision-making processes of individuals, while behavioral finance emphasizes the psychological factors influencing both individual and institutional investors' trading decisions (Bodie, Kane, and Marcus, 2014; Reilly and Brown, 2012).
Behavioral finance examines the psychological factors that influence financial analysis and investment decisions, encompassing various categories such as information processing errors and irrational behavioral biases A significant aspect of this field is the herding behavior exhibited by individual investors, which is a focal point of this thesis Consequently, herding behavior is emphasized as a critical issue within the scope of behavioral finance.
Herding Behavior
Henker et al (2006) highlight that human imitation of others' actions is a widespread phenomenon observed across various social contexts and life stages This imitative behavior is also linked to the actions of individuals in economic scenarios The literature on behavioral economics presents multiple definitions of herding behavior, as discussed by Banerjee.
Herding, as defined by Bikhchandani et al (1992), refers to the phenomenon where individuals follow the actions of others despite having different private information This behavior reflects investors' intent to mimic the decisions of their peers Hirshleifer et al (1994) highlight that herding occurs when investors rely on the same information sources and interpret data similarly, leading to uniform financial choices Additionally, Christie and Huang (1995) note that herding is characterized by individuals who set aside their personal beliefs and make investment decisions based solely on the market's collective actions, even when they personally disagree.
Herding behavior, as defined by Nofsinger and Sias (1999), occurs when a group of investors trades in the same direction over a period, often influenced by the actions of others This phenomenon arises when individuals make similar decisions by mimicking the choices of their peers, leading to synchronized trading patterns in the market.
Herding behavior in investing is characterized by individuals mimicking the collective actions of the market rather than relying on their own beliefs or private information According to Chang et al (2000), investors often align their opinions with overall market trends, while Hwang and Salmon (2004) define herding as the tendency to replicate the decisions of others Patterson and Sharma (2006) further elaborate that herding occurs when groups of investors trade in the same securities simultaneously, often disregarding their private insights.
Herding behavior in financial markets occurs when investors mimic the trading decisions of others, often those perceived to be better informed, rather than relying on their own information and beliefs (Blasco & Ferreuela, 2008; Huang et al., 2015) This tendency to follow the crowd can lead to collective decision-making that drives stock prices away from their fundamental values (Chiang & Zheng, 2010; Chang et al., 2010) For less sophisticated investors, herding can be a rational strategy, as it is often more cost-effective to monitor successful investors than to utilize their own knowledge (Khan et al., 2011) Additionally, herding is viewed as an investment strategy where individuals align with market consensus or imitate the actions of financial experts (Chen, 2013; Yao et al., 2013) Ultimately, herding behavior reflects the inclination of investors to replicate the actions of their peers, disregarding their own insights (Filip et al., 2015; Vieira & Pereira, 2015).
In this study, herding behavior could be built on similar definitions of Christie and Huang (1995), Chang et al (2000), Houda and Mohamed (2013) which is defined
Individual investors often engage in the practice of mimicking the actions of others, leading them to invest in markets even when they disagree with the predictions or dismiss their own beliefs and private information.
Herding behavior has been extensively studied across various economic activities, including investment recommendations (Scharfstein and Stein, 1990), the price dynamics of Initial Public Offerings (Welch, 1992), and market fads, which refer to large groups of investors trading similarly (Wermer, 1999) Additionally, research has explored herding in relation to customs (Bikhchandani et al., 1992), earnings forecasts (Trueman, 1994), and corporate conservatism (Zwiebel et al., 1995).
Numerous theories explain the phenomenon of herding behavior among investors, which can be categorized into two main types: information-driven and behavior-driven factors Research by Bikhchandani and Shama (2000), Kremer (2010), and Hsieh highlights the underlying causes of this collective behavior in financial markets.
Herding behavior in financial markets influences market efficiency in various ways (Hsieh, 2013) Information-driven herding occurs when investors face similar decision-making challenges and share correlated private information, often stemming from similar educational and professional backgrounds (Hirshleifer et al., 1994; Falkenstein, 1996) This type of herding is linked to momentum trading strategies and reflects fundamental market conditions, thereby impacting stock prices consistently In contrast, behavior-driven herding arises when investors mimic the actions of others, potentially leading to market destabilization (Hsieh, 2013) The informational cascade model suggests that observing the trading behavior of others can prompt investors, lacking a common connection, to change their decisions, creating a cascade effect (Banerjee, 1992; Bikhchandani et al., 1992) Additionally, concerns about reputation can further drive behavior-driven herding among professional investors.
11 disregard their private information and trade according to the crowd because they are subject to the reputation risk of trading in a contrarian manner (Hsieh, 2013)
Bikhchandani and Sharma (2000) distinguish between intentional herding behavior and false herding behavior in investment decisions Intentional herding occurs when investors deliberately choose to follow the actions of other market participants In contrast, false herding arises when a group of investors, facing similar challenges in formulating an investment strategy, ends up making comparable trading choices without a conscious intention to mimic others.
Herding behavior in investing can be categorized into rational and irrational herding (Devenow and Welch, 1996; Lao and Singh, 2011) Rational herding occurs when investors mimic others, utilizing this behavior to gather valuable information that aids in predicting stock price movements (Filip et al., 2015) This indicates a complex interplay between rationality and emotion in investor decision-making (Vieria and Pereira, 2015), where rational herding facilitates an efficient reallocation of assets based on shared fundamental news (Hwang and Salmon, 2004) In contrast, irrational herding arises from instinctive tendencies, leading investors to imitate others without critical analysis, which can result in market inefficiencies and widespread erroneous decision-making (Bikhchandani et al., 1992).
Since 2010, market participants have shown enthusiasm for herding behavior, as it creates profitable trading opportunities by leveraging aggregate information over private insights when prices deviate from their fundamental values This phenomenon has also garnered the interest of academic researchers, as herding's influence on stock price movements can significantly impact risk and return, ultimately aiding in the development of asset-pricing models.
Irrational herding, along with behavior-driven and intentional herding, plays a significant role in investment decisions Specifically, irrational herding occurs when investors mimic the choices of others without thoroughly analyzing the intrinsic value of stocks This behavior can result in poor investment decisions and contribute to market inefficiencies.
Empirical Evidence
These summaries below including many previous researches studying herding behavior of investors in stock market from 1995 to 2017
Christie and Huang (1995) developed a method to measure herding behavior in the stock market, focusing on individual investors and assessing whether equity returns display herding tendencies Their study posits two hypotheses: first, that herding occurs during periods of market stress, regardless of average price movements; and second, that herding reactions are asymmetrical under intense market conditions Analyzing daily and monthly data from NYSE/AMEX firms between July 1962 and December 1988, the researchers found that return dispersions increased notably during volatile periods, suggesting that herding behavior is absent in the US stock market when individual returns mirror portfolio returns Subsequently, Chang et al (2000) examined herding behavior in international stock markets by introducing the cross-sectional absolute deviation (CSAD) model, which replaces the CSSD model from Christie and Huang Their analysis of daily stock price data from firms in the US, Hong Kong, Japan, South Korea, and Taiwan revealed that return dispersion, as measured by CSAD, increases more in rising markets than in declining ones, aligning with findings from McQueen et al (1996).
Recent research indicates that return dispersions in the US, Hong Kong, and Japan tend to increase rather than decrease during periods of extreme price movements, suggesting the absence of herding behavior in these markets This finding aligns with the conclusions drawn by Christie and Huang (1995) regarding the US market In contrast, herding behavior is observed in the emerging markets of South Korea and Taiwan during both upward and downward price movements, providing significant evidence that herding can exist in emerging markets.
Demirer and Kutan (2006) detect the existence of herding behavior in Chinese stock market from 1/1999 to 12/2002 using both individual firm and sector-level data
The study analyzed firms across various industries, calculating portfolio returns using an equally weighted approach and gathering index returns from the Shanghai and Shenzhen stock exchanges By employing the models of Christie and Huang (1995), Chang et al (2000), and Gleason et al (2004), the researchers developed a cross-sectional standard deviation to assess return dispersion, indicating that herding behavior tends to emerge during significant market movements and aligns with market consensus Furthermore, the research compared return dispersion during market upturns and downturns, revealing that dispersion is lower in downturns Contrary to previous literature, the study found no evidence of herding behavior in Chinese stock markets, suggesting that market participants make rational investment decisions, which could be beneficial for Chinese policymakers.
Tan et al (2008) utilize the CSAD metric to assess return dispersion, distinguishing their approach from CCK's model by specifically describing a single factor of the CAPM model Their methodology integrates the techniques of Christie and Huang (1995) alongside those of Gleason et al.
This study, conducted in 2004, explores herding behavior in the Chinese A-share and B-share markets, utilizing data from all firms listed on the Shanghai and Shenzhen stock exchanges from July 1994 onwards, without the need for beta estimation.
In December 2003, researchers examined the variation of herding behavior in different market conditions, focusing on asymmetric responses influenced by market returns, trading volume, and volatility Their findings revealed that herding occurs in both A-share and B-share markets on the Shanghai and Shenzhen exchanges when analyzed with daily data, contradicting earlier research by Demirer and Kutan (2006) The study showed that herding behavior is more pronounced in daily data compared to weekly or monthly data Notably, herding is more significant during market upturns with high trading volume and volatility, particularly among investors in Shanghai's A-share market, who exhibit stronger asymmetric reactions In contrast, B-share investors do not display these asymmetries The authors suggest that these differences stem from the distinct characteristics of A and B markets, where domestic individual investors dominate A-shares, while institutional investors prevail in B-shares.
Goodfellow et al (2009) investigate herding behavior among individual investors in the Polish stock market from July 1996 to November 2000, analyzing two distinct trading platforms to differentiate between individual and institutional investors The study aims to identify herding tendencies during both rising and declining market conditions Utilizing the means cross-sectional absolute deviation method, along with the models of Christie and Huang (1995) and Chang et al (2000), the findings indicate that individual investors exhibit herding behavior primarily during declining markets, while no such behavior is observed among institutional investors Furthermore, the research highlights a divergence in trading patterns, suggesting that individual investors are more susceptible to sentiment-driven investment decisions compared to their institutional counterparts during downturns.
Chiang and Zheng (2010) investigate herding behavior in global stock markets by applying a modified version of the CCK – CSAD model This adaptation aims to analyze asymmetric investor behavior across varying market conditions The study utilizes data from industry and market price indices across advanced, Latin American, and Asian markets, covering the period from May.
This study, covering the period from 1988 to April 2009, investigates herding behavior in various national markets during both up and down periods, incorporating a dummy variable for analysis The findings reveal that herding exists in most national markets, excluding the US and Latin America, which contrasts with previous research by Chang et al (2000) and Demirer and Kutan (2006), yet aligns with Tan et al (2008) for similar markets Notably, herding behavior is more pronounced in upward markets compared to downward markets in several Asian countries, particularly in China, Japan, and Hong Kong Additionally, the study highlights that asymmetric herding responses are most evident in five Asian markets, while advanced markets show little evidence of asymmetry, with the exception of Japan and Hong Kong Overall, the research underscores the distinct patterns of herding behavior across different markets and conditions.
US and Latin American markets during crisis periods
Chiang et al (2010) investigate herding behavior among investors in the Chinese stock markets, specifically focusing on A-share and B-market Utilizing the CSAD model, they employ two different estimation methods: Ordinary Least Squares (OLS) regression and quantile regression, to assess their efficiency The study analyzes data from all firms listed on the Shanghai and Shenzhen stock exchanges from January 1996 to April 2007 A key contribution of this research is the introduction of quantile regression analysis, which is found to be more effective than OLS regression Additionally, the paper explores herding behavior in both rising and falling markets using both estimation techniques, revealing that herding behavior is present across the overall market.
The study reveals distinct herding behaviors in A-share and B-share markets, with A-share markets exhibiting herding in both rising and falling conditions, while B-share markets only show herding during downturns Notably, quantile regression analysis indicates that B-share investors demonstrate herding behavior during rising market conditions, specifically within the 10% to 50% quantiles The findings suggest that the varied results stem from the inability to capture asymmetric reactions of market returns and the oversight of distributional information in the quantile regression analysis model.
Lao and Singh (2011) explore herding behavior in the Chinese and Indian stock markets, analyzing data from the top 300 firms in the Shanghai A-share market and the Bombay Stock Exchange from July 1999 to June 2009, focusing on stock prices and trading volume Utilizing the CSAD model, similar to that of Tan et al (2008), the study aims to identify herding in these emerging markets, particularly during extreme market conditions, while examining the asymmetric effects of market returns and trading volume The findings reveal a significant presence of herding in both markets, with stronger effects observed during upward and downward trends Notably, herding intensity differs between the two markets; it is more pronounced in the Chinese market during downturns with high trading volume, whereas in India, herding occurs during falling markets without a direct link to trading volume Additionally, the global financial crisis has intensified herding behavior in the Chinese stock market, providing further evidence of herding in emerging markets.
In their study, Houda and Mohamed (2013) investigate herding behavior in stock markets across Africa, Asia, Europe, and America from January 2006 to February 2009, utilizing the WSCI index as a market portfolio proxy They employ the CSSD model by Christie and Huang (1995) and the CSAD model by Chang et al (2000) to demonstrate that herding behavior varies between upward and downward markets The authors introduce two key assumptions: that securities returns will decline in an upturn and increase in a downturn Their OLS regression results indicate no herding behavior towards the WSCI Index over the entire period; however, when analyzing upturns and downturns separately, they identify asymmetric reactions, revealing that herding is more pronounced in upward markets The study proposes a novel method for testing herding and calculating excess volatility by integrating the CSAD model with E-GARCH Ultimately, it concludes that herding significantly influences market returns and increases volatility, highlighting distinct behaviors in different market conditions.
Yao et al (2014) investigate herding behavior in the Chinese A and B stock markets using data from all firms listed on the Shanghai and Shenzhen stock exchanges between January 1999 and December 2008 The study categorizes stocks by industry, market capitalization, and growth versus value characteristics, while also analyzing herding behavior during both rising and declining market conditions Utilizing CSSD to measure herding and employing OLS regression for estimation, the findings indicate that herding behavior is more pronounced in B-share markets, whereas no evidence of herding is found in A-share markets, aligning with conclusions drawn by Tan et al.
Herding behavior is prevalent at the industry level and is more pronounced among larger stocks compared to smaller ones Additionally, it is stronger for growth stocks than for value stocks Notably, research indicates that herding is more significant during declining market conditions than in rising markets, affecting both A-shares and Shanghai B-shares.
18 markets This conclusion supports the study in the case it needs some identifications to compare
RESEARCH METHODS
Data Description
This thesis utilizes data from daily stock prices, market capitalization, and book-to-market ratios for non-financial firms listed on HOSE and HNX from 2006 to 2016, encompassing a total of 2,738 observations The risk-free return is determined using the daily-compounded value of the 1-year fixed bank interest rate sourced from the State Bank of Vietnam Stock returns are calculated using the formula 𝑅 𝑡 = ln (𝑃 𝑡 ) − ln(𝑃 𝑡−1 ).
Research methods
This study utilizes the WCSV model to identify significant herding behavior and influential herding points in the Vietnamese stock market However, given that the WCSV model presumes the constant presence of herding, it is essential to first establish the existence of herding behavior in this market To achieve this, the research employs the CSAD model developed by Chang et al (2000) to test for herding behavior in the Vietnamese stock market.
A comparison of the WCSV model based on the Fama-French three-factor approach and the WCSV model utilizing a one-factor CAPM reveals that the former is more effective for identifying significant herding behavior and influential factors in the Vietnamese stock market.
The thesis categorizes the collected data into three distinct segments: the entire decade from 2006 to 2016, individual years from 2006 to 2013, and three specific sub-periods, which include the pre-crisis phase from 2006 to 2007 and the crisis period spanning 2008 to 2016.
The Vietnamese stock market underwent significant changes during three key periods: pre-crisis, during the crisis, and post-crisis, spanning from 2012 to 2016 The global financial crisis, which began in late 2007 and early 2008, initially impacted developed countries before extending its effects to developing nations, including Vietnam This timeline highlights the substantial influence of the global crisis on the Vietnamese financial landscape.
The Vietnamese stock market is currently experiencing a delayed crisis, influenced by a combination of global economic challenges and significant local events This situation can be divided into three distinct sub-periods: the pre-crisis phase from 2006 to 2007, the crisis phase from 2008 to 2012, and the post-crisis phase from 2013 to 2016 This classification aims to analyze and compare the differences across these time periods using the WCSV model.
DATA ANALYSIS AND FINDINGS
Descriptive Statistics
The data illustrates the trend of the dependent variable WCSV from January 2006 to December 2016 across both HOSE and HNX, encompassing a total of 2,738 observations This sample includes 498 observations prior to the financial crisis, 1,244 during the crisis, and 996 after the crisis for each stock market.
Figure 4.1: Time series of the WCSV models during the period from 2006 to 2016 on HOSE
As we can see from figure 4.1, WCSV model shows that there are some anomalies usually happening in some years such as 2006, 2007, 2009, and 2012
Figure 4.2: Time series of the WCSV models during the period from 2006 to 2016 on HNX
In figure 4.2, on HNX, WCSV has higher volatilities appearing in 2006, 2007,
Between 2006 and 2016, the average excess return in the Vietnamese stock market exhibited significant fluctuations, declining sharply in 2007 and 2008, rebounding in 2009, and then dropping to negative levels in 2010 and 2011 The market showed a reversal with increases in 2012, 2013, and 2014, followed by a decrease in 2015, and another rise in 2016 Overall, the average excess return remained negative throughout this period, indicating a lack of investor interest in the stock market, as many opted for Treasury Bills, which offer lower but more stable returns The years 2007 to 2011 were particularly marked by higher risk and lower profitability compared to Treasury Bills.
Since 2012, the Vietnamese stock market has drawn investors due to its stable average excess returns The minimum and maximum values of these returns have shown consistent distribution with minimal fluctuations Notably, the average SMB on HOSE remained negative for a significant portion of this period.
2008 to 2012, in 2015 and 2016 This suggests that in the years, the average return of large capitalizations companies was higher than the average return of companies with
From 2006 to 2016, the values of Max and Min for small capitalizations (SMB) exhibited similar trends, moving in the same direction throughout the entire period In contrast, the values of High Minus Low (HML) on the HOSE displayed divergent patterns The Min value showed a decreasing trend from 2006 to 2014, with only a rise observed in 2015 and 2016, while the Max value generally increased, experiencing a slight decline only in 2015.
From 2006 to 2016, the average HML values on HOSE consistently showed negative figures, highlighting that value stocks underperformed compared to growth stocks During the same period, the minimum value of WCSV on HOSE exhibited an upward trend, whereas both the maximum and average WCSV values experienced continuous fluctuations year over year.
Between 2008 and 2012, the Vietnamese stock market on HNX experienced negative mean excess returns, indicating a lack of investor interest, as many shifted to safer investment options like treasury bills However, from 2013 to 2016, the mean excess return showed fluctuations, initially rising from 2012 to 2013, then declining until 2015, before increasing again in 2016, suggesting a gradual attraction of investors, albeit at a low intensity Throughout this period, the average return on small capitalization stocks remained lower than that of large capitalization stocks, with negative SMB values recorded in several years, reflecting a shift in investor focus towards larger, less risky stocks Additionally, the mean value of HNX indicated that value stocks underperformed compared to growth stocks, although the overall trend from 2006 to 2016 showed an increase in the Min and Max values of HNX.
35 increases from 2006 to 2011 only falls in 2008, between 2012 and 2016, the mean value of HNX decreases in 2012, 2014 and 2015
Table 4.3 presents descriptive statistics for the independent variable WCSV and the dependent variables (Rm-Rf), SMB HOSE, and HML HOSE across three distinct sub-periods: prior to the crisis (2006-2007), during the crisis (2008-2012), and following the crisis (2013 onwards).
Between 2008 and 2012, the average excess return in the stock market was negative, indicating that investing in treasury bills yielded better returns than stocks Additionally, small capitalization stocks underperformed compared to large capitalization stocks during this period, as reflected by the negative mean value of SMB on HOSE Furthermore, the negative mean value of HML across all three periods suggests that value stocks consistently provided lower returns than growth stocks Lastly, the mean value of WCSV on HOSE showed a declining trend during and after the crisis, highlighting a decrease in performance over time.
Table 4.4 presents descriptive statistics for the independent variable WCSV and the dependent variables (Rm-Rf), SMB HNX, and HML HNX on HNX across three distinct sub-periods: before the crisis (2006-2007), during the crisis (2008-2012), and after the crisis (2013-2016) Notably, the mean value of the excess market return on HNX is negative for the crisis period of 2008-2012.
2012, indicating that the average return on investment in the stock market is lower than that of treasury bills The mean value of SMB is negative for the period 2008-
In 2012, data revealed that small capitalization stocks yield lower average returns compared to large capitalization stocks Additionally, the negative mean value of HML across all three periods indicates that value stocks underperform relative to growth stocks Furthermore, the mean value of WCSV has shown a decline since 2006.
2007 to 2008-2012 and then increases again in 2012-2016
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
Minimum -0.05512 -0.04616 -0.05770 -0.04624 -0.05253 -0.06655 -0.05392 -0.04089 -0.06052 -0.04998 -0.03634 -0.09643 Maximum 0.05184 0.04447 0.05027 0.04635 0.04307 0.03317 0.03630 0.03343 0.02984 0.02835 0.02882 0.08288 Mean 0.00363 -0.00005 -0.00529 0.00121 -0.00091 -0.00203 0.00019 0.00088 0.00000 -0.00034 0.00058 -0.00013 Variance 0.00040 0.00032 0.00052 0.00046 0.00019 0.00018 0.00016 0.00012 0.00014 0.00008 0.00008 0.00033 Standard deviation 0.02001 0.01788 0.02277 0.02150 0.01383 0.01324 0.01265 0.01095 0.01194 0.00897 0.00901 0.01814 Skewness -0.25906 0.21326 0.10759 -0.12878 -0.30170 -0.52291 -0.92016 -0.44220 -1.12920 -0.53505 -0.32035 -0.17804 Kurtosis 0.34978 0.04387 -0.43671 -0.49576 1.12717 2.29606 2.73360 1.76989 3.51109 3.83030 1.09964 1.84982
Minimum -0.74059 -1.49610 -2.30391 -2.60344 -3.09902 -3.29004 -3.13574 -2.73251 -4.30978 -3.53636 -1.97451 -6.01319 Maximum 0.51358 1.25115 2.03193 2.27518 2.88268 2.14916 3.30917 2.07625 2.38414 1.91889 2.25843 4.49532 Mean 0.02455 0.00162 -0.27110 0.07592 -0.12882 -0.30386 -0.00182 0.04427 0.05485 -0.03303 -0.02897 -0.05906 Variance 0.03974 0.21069 0.95374 1.11946 1.21202 0.86150 1.06190 0.50529 0.95386 0.33174 0.30184 1.11607 Standard deviation 0.19935 0.45901 0.97659 1.05805 1.10092 0.92817 1.03049 0.71084 0.97666 0.57596 0.54940 1.05644 Skewness -0.42705 -0.04761 0.21576 -0.29406 -0.14600 0.13297 -0.20717 -0.54254 -1.16280 -1.02900 -0.17894 -0.25470 Kurtosis 0.86875 0.37594 -0.43297 -0.26297 0.38624 0.39161 0.65585 1.68546 2.68747 6.62641 1.50968 2.67566
Minimum -0.26550 -0.55303 -0.63115 -0.95889 -1.23213 -1.29752 -1.10886 -1.24023 -1.27450 -0.65505 -0.97516 -2.11916 Maximum 0.33028 0.70855 0.88706 1.08977 1.08935 1.11350 1.00511 1.33992 1.25199 0.55789 1.18630 1.94509 Mean -0.00878 -0.01646 -0.06365 -0.01956 -0.05650 -0.16191 -0.05165 -0.04106 -0.03582 -0.04255 -0.10268 -0.10370 Variance 0.00747 0.03808 0.06481 0.09494 0.14212 0.17822 0.18533 0.13947 0.18755 0.04160 0.09096 0.15839 Standard deviation 0.08645 0.19515 0.25458 0.30813 0.37699 0.42217 0.43051 0.37345 0.43307 0.20395 0.30159 0.39798 Skewness -0.14884 0.44471 0.47635 0.28830 0.04622 0.36082 0.02369 0.39202 -0.21109 -0.05841 0.05537 -0.11686 Kurtosis 1.29696 0.82384 0.54081 0.63691 0.36737 0.37515 -0.48729 1.44613 0.39882 0.16668 0.83187 2.20248
Minimum 0.00000 0.00008 0.00000 0.00001 0.00007 0.00014 0.00013 0.00011 0.00013 0.00013 0.00017 0.00000Maximum 0.01831 0.05748 0.01307 0.03813 0.01587 0.02081 0.02143 0.00647 0.00992 0.00574 0.00738 0.02565Mean 0.00044 0.00116 0.00063 0.00091 0.00061 0.00068 0.00050 0.00036 0.00043 0.00041 0.00050 0.00092Variance 0.00000 0.00002 0.00000 0.00001 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000Standard deviation 0.00122 0.00430 0.00104 0.00265 0.00128 0.00148 0.00138 0.00044 0.00072 0.00058 0.00054 0.00103Skewness 12.90249 10.45661 8.37312 11.75713 8.66245 11.05952 14.25942 11.27384 10.50598 6.77767 9.06488 12.43252Kurtosis 187.98482 125.98670 89.36054 158.36394 90.38994 141.25532 214.95840 153.47140 129.41521 52.34132 109.14353 245.56242
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
Minimum -0.07534 -0.06590 -0.09643 -0.05819 -0.06415 -0.04266 -0.04749 -0.04581 -0.06669 -0.04801 -0.02715 -0.09643 Maximum 0.04685 0.08288 0.06279 0.05852 0.06236 0.03510 0.03521 0.02590 0.03574 0.03452 0.03524 0.08288 Mean 0.00192 0.00068 -0.00412 0.00212 -0.00152 -0.00286 -0.00005 0.00089 0.00110 0.00004 0.00030 -0.00013 Variance 0.00041 0.00047 0.00084 0.00065 0.00039 0.00017 0.00020 0.00010 0.00021 0.00008 0.00006 0.00033 Standard deviation 0.02023 0.02162 0.02897 0.02552 0.01979 0.01314 0.01424 0.01004 0.01463 0.00880 0.00804 0.01814 Skewness -0.32081 0.40428 -0.09947 -0.19962 0.02568 0.31651 -0.19491 -0.63083 -0.93833 -0.68021 -0.06166 -0.17804 Kurtosis 0.86656 1.61442 -0.38106 -0.39501 0.55183 0.54008 0.39236 1.95115 2.42745 4.32875 1.90812 1.84982
Minimum -0.39977 -2.28428 -3.40389 -3.91461 -6.01319 -4.21693 -3.71603 -2.71414 -5.10920 -3.24437 -2.04225 -6.01319 Maximum 0.33108 2.54076 3.41809 3.43981 4.44131 3.25526 4.49532 2.03926 2.59853 1.58308 1.77959 4.49532 Mean 0.00646 0.05082 -0.28175 0.08592 -0.14871 -0.39255 -0.01970 0.04814 0.06947 -0.02955 -0.04327 -0.05906 Variance 0.00883 0.45410 1.71413 2.03109 3.04367 1.57823 1.49543 0.42892 0.86220 0.25478 0.21485 1.11607 Standard deviation 0.09398 0.67387 1.30925 1.42516 1.74461 1.25628 1.22288 0.65492 0.92855 0.50475 0.46352 1.05644 Skewness -0.00303 0.06255 0.08366 -0.20088 -0.10229 0.25675 0.09567 -0.24372 -1.19300 -1.39349 -0.41580 -0.25470 Kurtosis 2.33849 1.31956 0.08139 0.02893 0.17947 0.87502 1.11273 2.25567 4.60324 7.56310 2.18331 2.67566
Minimum -0.32860 -0.84863 -1.04250 -0.98686 -1.46482 -2.04042 -1.32861 -1.56638 -2.11916 -0.80541 -1.25924 -2.11916 Maximum 0.35277 0.69905 1.06139 0.91113 1.00596 1.34490 1.61783 0.99521 1.94509 0.60978 0.90944 1.94509 Mean -0.00518 -0.04950 -0.06380 -0.07749 -0.14904 -0.24089 -0.10557 -0.11823 -0.11040 -0.05767 -0.16239 -0.10370 Variance 0.00394 0.06078 0.11706 0.11762 0.21223 0.29413 0.29171 0.14114 0.28977 0.07038 0.10897 0.15839 Standard deviation 0.06276 0.24655 0.34213 0.34295 0.46069 0.54234 0.54010 0.37569 0.53830 0.26530 0.33011 0.39798 Skewness -0.19488 0.26397 0.09256 0.15172 -0.25185 0.17042 0.19999 -0.04609 0.16871 -0.20873 -0.21925 -0.11686 Kurtosis 7.43734 0.86053 0.29214 -0.14017 0.31746 0.66002 0.15242 0.85136 2.51305 -0.30797 0.47352 2.20248
Minimum 0.00000 0.00014 0.00000 0.00007 0.00030 0.00041 0.00025 0.00039 0.00048 0.00038 0.00049 0.00000Maximum 0.01296 0.02565 0.00547 0.00507 0.00684 0.01304 0.00295 0.01042 0.00501 0.00563 0.00660 0.02565Mean 0.00064 0.00137 0.00059 0.00084 0.00094 0.00096 0.00077 0.00104 0.00100 0.00095 0.00103 0.00092Variance 0.00000 0.00001 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000Standard deviation 0.00096 0.00245 0.00090 0.00049 0.00078 0.00090 0.00034 0.00094 0.00054 0.00047 0.00056 0.00103Skewness 9.13659 7.82085 2.54886 3.11797 4.82614 10.18843 2.04258 7.04640 3.74707 4.89619 5.33833 12.43252Kurtosis 112.44220 70.51251 7.80244 20.94899 28.90918 130.40445 8.53878 61.73445 22.14839 41.80331 42.97900 245.56242
Table 4.3: Description of variables in WCSV model in three sub-periods: before, during and after crisis on HOSE
Table 4.4: Description of variables in WCSV model in three sub-periods: before, during and after crisis on HNX
The whole period from 2006 to 2016
Tables 4.5 and 4.6 present the correlation coefficients between the independent and dependent variables in the WCSV model, utilizing the Fama-French three-factor approach for both HOSE and HNX, calculated using the Pearson correlation coefficient The results indicate that, on HOSE, the correlation coefficients between WCSV and most independent variables are significant, with the exception of SMB Conversely, all variables on HNX exhibit significant coefficients Notably, WCSV shows a positive correlation with the squares of SMB, HML, and the excess market return, as derived from the WCSV formula based on the Fama-French three-factor model, which serves as the foundation for the Arbitrage Pricing Theory (APT).
Table 4.5: The correlation matrix of variables in WCSV model based on Fama-
French three factors on HOSE
(on HOSE) SMB 2 HML 2 (Rm –Rf) 2 WCSV
Table 4.6: The correlation matrix of variables in WCSV model based on Fama -
French three factors on HNX
(on HNX) SMB 2 HML 2 (Rm –Rf) 2 WCSV
Table 4.7: Testing multicollinear of variables in WCSV model based on Fama-
French three factors over the period from 2006 to 2016
Variable VIF HOSE VIF HNX
Source: Author’s calculation 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
Variable VIF HOSE VIF HNX
Tables 4.7 and 4.8 indicate the absence of multicollinearity among the variables for the entire period from 2006 to 2016, as well as in three specific periods, with a Variance Inflation Factor (VIF) of less than 5 This confirms the appropriateness of the regression model utilizing these dependent and independent variables.
CONCLUSIONS AND RECOMMENDATION
Implications
This thesis investigates herding behavior in the Vietnamese stock market using the CSAD model, while also evaluating the WCSV model based on Fama-French three factors The study focuses on strong herding and influential points during significant events affecting individual investors Data was collected from non-financial companies listed on HOSE and HNX between 2006 and 2016, comprising 2,738 observations The analysis reveals that most variables are significant and predominantly align with the WCSV model based on Fama-French three factors.
The comparison of Adjusted R² values among the CSAD model, the WCSV model based on Fama-French three factors, and the WCSV model based on CAPM one factor reveals that the WCSV model utilizing three factors achieves the highest Adjusted R² among the three selected models.
The CSAD model presented in the thesis effectively identifies herding behavior in the Vietnamese stock market from 2006 to 2016, with notable exceptions in 2008, 2013, 2015, and 2016 for HNX This analysis covers three distinct periods: the pre-crisis phase (2006-2007), the crisis phase (2008-2012), and the post-crisis phase (2013-2016) The findings reveal a strong correlation between herding behavior and significant influential points, validating the applicability of the WCSV model, which is based on the Fama-French three-factor framework, to the Vietnamese stock market.
The thesis analyzes data collected from 2006 to 2016, examining yearly trends and three distinct periods: before, during, and after the crisis The findings indicate minimal differences across the selected timeframes, suggesting a consistent analytical process throughout the study.
From 2006 to 2012, the herding behavior of individual investors in the Vietnamese stock market was prominently observed, particularly during significant market events The boom period from 2006 to 2007 saw a surge in capital inflow and individual investor participation, often leading to a disregard for fundamental and technical analysis This behavior contributed to a collective mindset among investors However, the global financial crisis from 2008 to 2012 negatively impacted the market, resulting in a decline in investor numbers and a drastic fall in stock prices, which further fueled the herd mentality In contrast, the post-crisis period from 2013 to 2016 exhibited a lack of clear herding behavior, with minimal strong herding points identified in 2013, 2015, and 2016 according to the CSAD model.
2015 and 2016 - the latest two years
In 2013, the CSAD model indicated a positive γ2, suggesting the absence of herding behavior, contrasting sharply with the strong herding and influential points consistently observed in 2008 across all selected periods Notably, during the three phases—before, during, and after the crisis—2013 exhibited significant herding points on the HNX, differing from other years Furthermore, the WCSV model, based on the Fama-French three factors, revealed inconsistencies in certain years within the Vietnamese stock market.
The analysis of the period from 2006 to 2016, including specific sub-periods before, during, and after the crisis, indicates that the WCSV model based on Fama-French three factors is most effective in identifying strong herding and influential points in the Vietnamese stock market While there are variations in the manifestation of these points across different periods, the overall trend remains consistent In contrast, the CSAD model used by Xie et al (2015) for the Chinese stock market did not detect herding behavior in individual years but noted it over the entire period from 2006 to 2013 The WCSV model aligns well with the dynamics of the Chinese market, suggesting its applicability in identifying significant herding points Additionally, the differences in coefficients within the WCSV regression model highlight the unique characteristics of the Vietnamese market compared to the findings of Xie et al (2015).
The analysis from 2015 revealed positive coefficients for independent variables such as excess market return, SMB, and HML in relation to the WCSV dependent variable In the context of the Fama-French three-factor model applied to the Vietnamese stock market, it was found that excess market return and HML showed a positive correlation with WCSV, with the exception of 2006 on the HOSE Conversely, SMB exhibited a negative correlation with WCSV, except during the years 2006, 2008, and 2009 on the HNX.
This thesis modifies the WCSV model by Xie et al (2015) by removing one assumption regarding the constant presence of herding behavior It integrates the CSAD model by Chang et al (2000) with the WCSV model to analyze herding behavior and identify significant herding and influential points in the Vietnamese stock market from 2006 to 2016 The primary aim is to explore the impact of investor herding behavior and its evolution in Vietnam's stock market over the years.
Limitations and Recommendations
The thesis has notable limitations, primarily due to its reliance on data exclusively from non-financial companies, excluding securities firms, financial institutions, and banks Consequently, significant events in the banking and finance sectors over the years may not be adequately represented in the findings of the selected model.
The WCSV model proposed by Xie et al (2015) is a relatively new framework that has yet to be extensively utilized across various stock markets globally, leaving its applicability in different countries largely unexplored This thesis aims to implement this innovative model in Vietnam, moving beyond traditional models used by previous researchers, to provide stronger evidence of herding behavior and identify significant herding instances within the Vietnamese stock market.
The WCSV model has notable drawbacks, primarily the lack of evidence supporting herding behavior, which relies heavily on assumptions Its main objective is to identify significant herding and influential points relevant to the author's thesis Additionally, the model predominantly utilizes historical data to analyze market morphology and the psychological effects observed in previous years, but it has not demonstrated reliability in predicting future market trends.
Due to time constraints and concerns regarding data accuracy, the model has not yet been recognized as a standard, primarily because it relies on secondary data collected from websites like cophieu68.vn and cafef.vn, raising questions about its reliability.
Despite the existing flaws in the dissertation and its applied model, this research serves as a foundation for future studies The utilization of the WCSV model to identify herding behavior and to refine assumptions or compare with alternative models is crucial for determining the most suitable approach for the Vietnamese stock market, highlighting a key area for subsequent investigation.
Banerjee, A.V (1992), “A Simple Model of Herd Behavior”, The Quarterly Journal of Economics, 107(3), 797-817
Barber, B.M., Odean, T (1999), “The Courage of Misguided Convictions”, Financial
Barberis, N., Thaler, R (2003), “Chapter 18 A survey of behavioral finance”,
Handbook of the Economics of Finance, 1, Part B, 1053-1128
Belsky, G., Gilovich, T (1999), Why Smart People Make Big Money Mistakes and
How to Correct Them: Lessons from the new science of behavioral economics,
Belsley, D.A., Kuh, E., Welch, R.E (1980), Regression diagnostics: Identifying influential data and sources of collinearity, John Wiley & Sons, United States
Bikhchandani, S., Sharma, S (2000), Herd Behavior in Financial Markets: A Review, IMF
Bikhchandani, S., Hirchleifer, D., Welch, I (1992), “A theory of fads, fashion, custom and cultural change as informational cascades”, Journal of Political Economy, 100, 992-1026
Blaso, N., Ferreuela, S (2008), “Testing Intentional Herding in Familiar Stocks: An Experiment in an International Context”, Journal of Behavioral Finance, 9(2),
Blasco, N., Corredor, P., Ferreurela, S (2012), “Does herding affect volatility? Implications for the Spanish stock market”, Quantitative Finance, 12, 311-327 Bodie, Z., Kane, A., Marcus, A.J (2014), Investment, United States, Mc Graw Hill Education
Chang, E.C., Cheng, J.W., Khorana, A (2000), “An Examination of Herd Behavior in Equity Markets: An International Perspective”, Journal of Banking & Finance, 24(10), 1651-1671
Chen, T (2013), “Do Investors Herd in Global Stock Markets?”, The Journal of Behavioral Finance, 14(3), 230-239
Chiang, T.C., Li, J., Tan, L (2010), “Empirical investigation of herding behavior in Chinese stock market: Evidence from quantile regression analysis”, Global Finance Journal, 21(1), 111-124
Chiang, T.C., Zheng, D (2010), “An empirical analysis of herd behavior in global stock markets”, Journal of Banking & Finance, 34(8), 1911-1921
Christie, W.G., Huang, R.D (1995), “Following the Pied Piper: Do Individual Returns Herd around the Market?”, Financial Analysts Journal, 51(4), 31-37
Demirer, R., Kutan, A.M (2006), “Does herding behavior exist in Chinese stock markets?”, Journal of International Financial Markets Institutions and Money, 16(2), 123-142
Devenow, A., Welch, I (1996), “Rational herding in financial markets”, European
Faber, A., Nam, N.V., Hoang, V.Q (2006), Policy impacts on Viet Nam stock market:
A case of anomalies and disequilibria 2000-2006, FPT University
Falkenstein, E (1996), “Preferences for stock characteristics as revealed by mutual fund portfolio holdings”, Journal of Finance, 51, 111-135
Fama, E., French, K (1992), “The Cross- Section of Expected Stock Returns”,
Filip, A., Pochea, M., Pece, A (2015), “The herding behavior of investors in the CEE stocks markets”, Procedia Economics and Finance, 32, 307-315
Friday, H.S., Hoang N (2015), “Seasonality in the Vietnam stock index”, The International Journal of Business and Finance Research, 9(1), 103-112
Froot, K., Scharfstein, D, Stein, J (1992), “Herd on the street: Informational inefficiencies in a market with short-term speculation”, Journal of Finance, 47, 1461-1484
Fromlet, H (2001), “Behavioral Finance – Theory and Practical Application”,
Frugier, A (2016), “Returns, volatility and investor sentiment: Evidence from European stock markets”, Research in International Business and Finance, 38, 45-55
Gleason, K.C., Mathur, I., Peterson, M.A (2004), “Analysis of Intraday Herding Behavior among the Sector ETFs”, Journal of Empirical Finance, 11(5), 681-
Goodfellow, C., Bohl, M.T., Gebka, B (2009), “Together We Invest: Individual investors’ trading behavior in Poland”, International Review of Financial Analysis, 18(4), 212-221
Graham, J.R (1999), “Herding among Investment Newsletters: Theory and Evidence”, The Journal of Finance, 54(1), 237-268
Henker, J., Henker, T., Mitsios, A (2006), “Do investor herd intraday in the Australian equities market?”, International Journal of Managerial Finance,
Hirshleifer, D., Subrahmanayam, A., Titman, S (1994), “Security analysis and trading patterns when some investors receive information before others”,
Houda, B.M., Mohamed, F (2013), “Herding During Market Upturns and Downturns: International Evidence’, The IUP Journal of Applied Finance,
Hsieh, S.F (2013), “Individual and Institutional herding and the impact on stock returns: Evidence from Taiwan stock market”, International Review of Financial Analysis, 29, 175-188
Huang, T.C., Lin, B.H., Yang, T.H (2015), “Herd behavior and idiosyncratic volatility”, Journal of Business Research, 68(4), 763-770
Hwang, S., Salmon, M (2004), “Market Stress and Herding”, Journal of Empirical
Hwang, S., Salmon, M (2004), A New Measure of Herding and Empirical Evidence, Financial Econometrics Research Centre
Khan, H., Hassairi, S.A., Laurent, J (2011), “Herd Behavior and Market Stress: The Case of Four European Countries”, International Business Research, 4(3), 53-
Kremer, S (2010), Herding of institutional trader, Humboldt-University of Berlin
Lao,P., Singh, H (2011), “Herding behavior in the Chinese and Indian Stock Markets”, Journal of Asian Economics, 22(6), 495-506
Litimi, H., Bensaida, A., Bouraoui, O (2016), “Herding and excessive risk in the American stock market” A sectoral analysis”, Research in International Business and Finance, 36, 6-21
McQueen, G., Pinegar, M.A., Thorley, S (1996), “Delayed reaction to good news and the cross autocorrelation of portfolio returns, Journal of Finance, 51, 889-
Nofsinger, J.R., Sias, R.W (1999), “Herding and Feedback Trading by Institutional and Individual Investors”, Journal of Finance, 54(6), 2263-2295
Patterson, D.M., Sharma, V (2006), Do Traders Follow Each Other at the NYSE,
Reilly, F.K., Brown, K.C (2012), Investment Analysis & Portfolio Management,
South-Western Cengade Learning, United States
Schafstein, D.S, Stein, J.C (1990), “Herd Behavior and investment”, American Economic Review, 80, 465-479
Sewell, M., (2007), Behavioral Finance, University of Cambridge
Shefrin, H (2000), Beyond greed and fear; understanding behavioral finance and the psychology of investing, Harvard Business School Press
Shleifer, A., Summers, L H (1990), “The noise trader approach to finance”, Journal of Economic Perspectives, 4, 19-33
Subash, R (2012), Role of Behavioral Finance in Portfolio Investment Decisions: Evidence from India, Charles University in Prague
Tan,L., Chiang, T.C., Mason, J.R., Nelling E (2008), “Herding behavior in Chinese stock markets: An examination of A and B shares”, Pacific –Basin Finance Journal, 16 (1), 61-77
Trueman, B (1994), “Analyst Forecasts and herding behaviour”, Review of Financial
Vieira, E.F.S, Pereira, M.S.V (2015), “Herding Behaviour and Sentiment: Evidence in a small European market”, Revista de contabilidad, 18, 78-86
Welch, I (2000), “Herding among security analysts”, Journal of Financial Economics, 58, 581-622
Welch, I (1992), “Sequential sales, learning, and cascades”, Journal of Finance, 47, 695-732
Wermers, R (1999), “Mutual Fund herding and the impact on stock prices”, Journal of Finance, 54, 581-622
Xie,T., Xu, Y., Zhang, X (2015), “A new method of measuring herding in stock market and its empirical results in Chinese A-share market”, International Review of Economics and Finance, 37, 324-339
Yao, J., Ma, C., Pe, W.H (2014), “Investor herding behaviour of Chinese stock market”, International Review of Economics and Finance, 29, 12-29
Zwiebel, J (1995), “Corporate conservatism and relative compensation”, Journal of
Nguyen Trung Chinh (2008), Hiện tượng tâm lý bầy đàn trên thị trường chứng khoán
Việt Nam – Lý thuyết và thực nghiệm
Luu Tien Chung, Pham Hung Cuong, Pham Long (2016), “Seasonality Effect on the Vietnamese Stock Exchange”, International Journal of Financial Research,
Dang Van Dan (2016), “Lý thuyết tài chính hành vi trên thị trường chứng khoán Việt Nam và những vấn đề đặt ra”, Tài chính, 3(2), 15-18
Le Thanh Ha, Nguyen Duc Hung, Ngo Quoc Thai, Tran Nhat Tuan (2011), “Nhận diện và lý giải hành vi bầy đàn trên thị trường chứng khoán Việt Nam”,
Financial Review, National Economics University
Nguyen Thu Hang (2012), “Momentum effect in the Vietnamese stock market”,
Nguyen Duc Hien (2012), Hành vi của nhà đầu tư trên thị trường chứng khoán Việt
Nguyen Duc Hien, Tran Thi Lan Huong (2013), “Hành vi bầy đàn trên thị trường chứng khoán Việt Nam – Khuyến nghị và chính sách”, Nghiên cứu Kinh Tế, 11, 12-19
Truong Duc Loc (2006), Equalization and stock-market development, University of Groningen
Tran Thi Hai Ly (2010), “Hành vi bầy đàn trên thị trường chứng khoán Việt Nam, nguyên nhân và một số giải pháp”, Phát triển & Hội nhập, 5, 18-25
Tran Chung Thuy (2010), Nhận dạng và đo lường hiệu ứng đám đông trên TTCK Việt Nam và phân tích các nhân tố ảnh hưởng, National Economics University
Bui Nguyen Thanh Truc (2012), Nghiên cứu tâm lý bầy đàn trên thị trường chứng khoán Việt Nam
Tran Thi Thanh Tu và Pham Viet Hung (2011) đã nghiên cứu ứng dụng lý thuyết tài chính hành vi để kiểm định hành vi đầu tư trên thị trường chứng khoán Việt Nam Nghiên cứu này giúp hiểu rõ hơn về các yếu tố tâm lý ảnh hưởng đến quyết định đầu tư của nhà đầu tư trong bối cảnh thị trường Việt Nam.
Doan Anh Tuan, Hoang Mai Phuong (2017), “Nghiên cứu “Hành vi bầy đàn” trên thị trường chứng khoán Việt Nam”, Tạp chí Khoa học Đại học Đà Lạt, (1), 96-108
Nguyen Cao Ve (2010), Đo lường hành vi bầy đàn trên thị trường chứng khoán Việt
Vo Xuan Vinh, Phan Dang Bao Anh (2017), “Further evidence on the herd behavior in Vietnam stock market”, Journal of Behavioral and Experimental Finance,
Table 1: Summary of strong influential points appear in weak herding sense over
Year from 2006 to 2016 on both HOSE and HNX
Table 2: Summary of strong influential points appear in weak herding sense over Year from 2006 to 2016 on both HOSE and HNX without assumption
Data HOSE no assumption HNX no assumption
Table 3 presents the OLS estimators for three models: (a) the CSAD model, (b) the WCSV model utilizing the Fama-French three factors, and (c) the WCSV model based on the CAPM one factor, all applied to the HOSE without assumptions The table details the coefficients and Adjusted R² values from the OLS regression analysis conducted on yearly and comprehensive data from 2006 to 2016.
Standard errors are shown in brackets.
HOSE WCSV model (based on Fama - French Three Factor Model no assumption)
Table 4 presents the Ordinary Least Squares (OLS) estimators for three models: (a) the CSAD model, (b) the WCSV model utilizing the Fama-French three factors, and (c) the WCSV model based on the CAPM one factor, applied to the Hanoi Stock Exchange (HNX) without any assumptions This table includes the coefficients and Adjusted R² values from the OLS regression analysis conducted on annual data from 2006 to 2016 Standard errors are provided in brackets for reference.
HNX WCSV model (based on Fama - French Three Factor Model no assumption)
The WCSV model, grounded in the Fama and French three-factor APT framework, provides a comprehensive analysis of herding behavior from 2006 to 2016 This study categorizes the frequency of herding points into strong, weak, and inadequate classifications, along with identifying influential points that significantly impact market trends.
Strong Weaker Inadequate Strongly influential
Table 2 presents a comprehensive analysis of herding behavior, categorizing it into strong herding points, weaker herding points, inadequate points, and strong influential points This examination is conducted using the WCSV model, which is grounded in the Fama and French three-factor model, utilizing single-year data from 2006 to 2016.
Strong Weaker Inadequate Strongly influential
Table 3 provides a comprehensive analysis of herding behavior, categorizing points into strong herding, weaker herding, inadequate, and strong influential points This assessment utilizes the WCSV model, grounded in the Fama and French three-factor framework, and spans three distinct sub-periods: the pre-crisis phase (2006-2007), the crisis period (2008-2012), and the post-crisis era (2013-2016) within the broader timeframe of 2006 to 2016.
Strong Weaker Inadequate Strongly influential