HERDING BEHAVIOUR IN THE FINANCIAL MARKETS OF EASTERN AND CENTRAL EUROPEAN COUNTRIES: LEGACY OF SOCIALISM

HERDING BEHAVIOUR IN THE FINANCIAL MARKETS OF EASTERN AND CENTRAL EUROPEAN COUNTRIES: LEGACY OF SOCIALISM

Ivan Pidorenko

PhD student, Corvinus University of Budapest,

Hungary, Budapest

 

Human mind has a lot of flaws, inherited from our ancestors in the process of evolution. The evolution of homo sapiens’ brain had stopped around 40-50 thousand years ago, whereas society, environment, conditions of life and life expectancy have dramatically changed since that time. The combination of emotional and cognitive responses, generated by human brain not just had positive effect on the survival of human kind in severe conditions of savanna and jungles full of hazard, but also let them force out Neanderthals and therefore provided their survival as a specie. As far as these brain mechanisms could guarantee the success and prosperity of human kind, they have been intact by evolution for the following centuries. The main principle of human brain’s operation is that in average neocortex (80% of the brain, responsible for cognition) is subdued by limbic system (10% of the brain, ancient hormonal-instinctive area). This phenomenon has a biological explanations behind: 1) It provides decision making based on heuristics (quick, low-energy consuming solutions leading to suboptimal outcomes);  2) As far as it is an instinctive area, signals, generated by limbic system, and solutions, as a consequence, in contrast with cognition, directly lead to biological domination, what in modern society provides financial freedom, what in turn leads to providing to yourself necessary amount of food (own survival) and to reproduction (survival of the specie).

The negative side of this phenomenon, which is called «duality of consciousness» is that in realities of modern-world environment this kind of interaction between emotions and cognition can be destructive, violating individuals’ rationality and leading to numerous biases. The deficiencies of our mental processes are transferred to all spheres of life, and financial markets are not exception, what in turn causes inefficiencies like bubbles and crashes as a consequence, which fail to be explained neither by Neoclassical theory of decision-making in Finance, nor by Efficient Market Hypothesis (EMH), because they are centered around the concept of rationality, which is doubtful for the reasons stated above. However, recent studies in the field of behavioral finance, combining the knowledge from subdisciplines of Finance, as well as from Psychology, Sociology and Neuroscience shade light on the anomalies in financial markets, explain their nature, causes and consequences.

During conduction of our research, we will keep a focus on herding behavior of investors. It can be described as a situation, when market participants ignore their own beliefs and simply follow market expectations, imitating others for the reasons of high information costs, conforming and extreme risk-aversion. We will focus on financial markets of Eastern and Central Europe (Moscow Stock Exchange, Budapest Stock Exchange, Warsaw Stock Exchange and Prague Stock Exchange), because all of them are not mature enough and has just recently passed through transition. The latter by itself implies distortion of markets and rent-seeking behavior of investors. MOEX was reestablished in the period of transition, only in 1992 (originally as Moscow Interbank Currency Exchange and Russian Trading System), BSE a little bit earlier, in 1990, WSE in turn in 1991, and PSE in 1993 respectively. These conditions already imply an existence of anomalies. Even though for the last 30 years many regulations had been enacted: establishment of new institutions and elaboration of legislation in the sphere of investments, adapted to market conditions, specific economic policies, more plausible political and tax regimes, providing positive investment climate, associated with lower levels of risk for international investors, however financial markets of Eastern and Central European countries are still considered as young and emerging, and therefore more likely to be exposed to various anomalies.

Herding behavior was chosen as a bias for analysis, because socialism and power of ideology shaped the mentality of people in post-communist countries towards group way of thinking and other specific behavioral patterns, which could pave the way for this bias and to be reflected in financial markets as a consequence. Even though there are a lot of foreign institutional investors in these markets nowadays, the stock exchanges of Eastern and Central Europe keep the growing of domestic investor base as a key area of focus, which has been dramatically increasing for all of them for the last years. Therefore, the implication about the influence of socialism-shaped conforming mentality, transferred within a few generations and its reflection in the financial markets in the form of anomalies, specifically herding behavior, seems to be relevant. As it was demonstrated by Milgram’s experiment, Morck [12], people in average have a tendency to be loyal and to conform to authority: in more general terms historically at first it was a chief, then chief and religion, nowadays legislation and courts as well, whereas the chief was replaced by state authorities, and as for investors, it is the market. Therefore, in the situation of uncertainty they are expected to follow market consensus, or in other words, to exhibit herding behavior. The effect can be reinforced by the high information costs, tendency to risk-aversion, asymmetric market conditions, or like with accordance to our expectations about Eastern and Central European countries, immaturity of the markets and tendency to conformity as a legacy of socialism.

Various economic and political regimes can shape way of thinking and preferences of people, and the stronger the regime is, the more opportunities the authorities have to achieve it. Decision to establish pioneer movements demonstrates how bureaucracy under the socialist regimes made attempts to create new generation of «socialist» disciplined, conformist and loyal to the government men and women. They existed as mass organizations in all post-communist countries, including Soviet Union, Czechoslovakia, Poland and Hungary. The members were children from 6 to 14 years old, politically loyal and exceptionally zealous students, wearing red or light blue neckerchiefs as a symbol of belonging to organization. According to Droit [8], to wear a neckerchief meant to have personal act of commitment and love for socialism. A piece of cloth, originated by authorities in the face of Communist party, became a symbol of «socialist» identity for millions of talented representatives of the younger generation, who expected to give up their revolutionary ambitions in exchange for administrative ones, based on the rule of conformism in daily life and losing any personal identity. This way communist parties had shaped the society of loyal conformists in the face of pioneers and ex-pioneers, and used it as an instrument of ideology’s propaganda and source of reinforcing of its own power. It could predetermine the behavioral patterns of not only the contemporaries of pioneer movements, but even the new generations of people, following after the collapse of socialism.

Vlachoutsicos [7] pointed that in the culture under the socialism the group took a priority over the individual, and the common trait of society were: tendency to maintain stability, risk avoidance and extreme conservatism. But at the same time the contribution of the individuals was vital for survival of the group. Summing up, such factors as prevailing of group values towards individual values, the dominance of the group way of thinking towards personal thoughts and ideas, aspiration of maintaining stability, pointed by Vlachoutsicos [7] are associated with the forces, which drive investors to mimic each other in the financial markets.

Alesina and Fuchs-Schündeln (2005) compared behavioral aspects of GDR and FRG population after reunification and came to conclusion that citizens of the former GDR part, due to the influence of socialism had more preferences towards redistribution and government intervention. In addition, the effects for older generation were greater. They tended to be less self-initiative and more risk-averse. It goes in line with a fact that political regimes shape mentality, behavioral disposition and habits of the individuals with specific differences between people, belonging to different age categories.

Mamontov and Kozhevnikova [11] found that collectivism in Soviet times brought collegiality and sense of community to the population. The process was reinforced by mutual aid, labor cooperatives, and power of ideology implanted and spread by bureaucracy in the era of socialism. Soviet, and in general, post-communist mentality and group way of thinking, demonstrated by the authors, was transferred to the new generation as well, and still, more or less, may find its reflection in the main spheres of life in the countries with socialism trace: Russia, Czech Republic, Poland and Hungary are not exceptions.

Semenov [13] specified that mentality, besides the impact of political regime, can be defined by history, religion and cultural inheritance of the nation. He defined mentality as a historically developed group long-term unity of conscious and unconscious values, norms, installations in them cognitive, emotional and behavioral expression of certain levels of population. He classified types of modern Russian mentality in a following way: collectivistic-social, West capitalistic, orthodox, criminal-mafia-controlled and mosaic-conformist. The current existence of the first one and the last one makes us assume that it can turn into herding behavior, if the individuals, belonging to these two particular types of mentality are market participants. The latter and other arguments, stated above can be considered as determinants of herding behavior, or at least similar behavioral patterns among investors in the financial markets of Eastern and Central European countries. 

Cross-sectional standard deviation model (CSSD) is an extension of CAPM model, proposed by Christie and Huang [5]. It is applied for testing of herding behavior in financial markets. It is widely used method, which is based on finding the difference between stock returns and market returns. If market participants trust market expectations and follow them, returns of their portfolios will not vary significantly from the market returns, and the difference between variance of investor’s returns and the market returns will tend to be less than the mean, highlighting herding behavior of market participants. And on contrast, if returns of stocks mostly vary from market returns, it will show that the variance of returns will be greater than the mean and herding behavior is not in a presence.

CSSD model is represented in the following regression equation:

Equation 1:

                                                    (1) 

Where:

t – is a period of time;

 - is a dummy variable and equals to 1, if market generates returns, which belong to the upper tail in the time period t, and 0 otherwise;

 – on the opposite, represents dummy variable, which equals to 1, if market returns belong to the lower tail returns, and 0 otherwise in the time period t;

 and  - represent the coefficients of dummy variables   and  and if they are significantly negative, it indicates the herding bias of investors, and on contrast, if positive, it testifies its non-existence;

 - is a constant;

 - is a return of a particular company’s stock in the time period t;

 - is a return of the market;

N - represents the number of companies’ market portfolio consists of;

 - is random error term;

In fact, opinions of market participants about extreme returns can be different, and we should admit that returns may fluctuate rapidly. Moreover, according to Chang et al.  [3] herd behavior can happen for the return distribution, but become more pronounced with market stress. In addition, they point that CSSD model is too rigorous for testing the evidence of herding bias. Therefore, we will apply the CSAD model, which was proposed by Chiang and Zheng [4]

Cross-sectional absolute deviation model (CSAD) is represented by the following regression equation:

Equation 2:

                                                (2)

To make the detection of herding bias easier, we will calculate cross-sectional absolute deviation using the following formula:

Where:

t – is a period of time;

 - is the mean of market returns in cross-section in the t period of time;

 - represents the absolute values of returns of particular stocks;

 – represents market returns in the period t;

 - allows us to keep linearity of relationship between CSAD and returns of market portfolio;

 and   - are the coefficients of regression;

N - represents the number of stocks in market portfolio;

 - is a constant;  

 - is random error term;

CSAD model keeps the randomization of market participants’ behavior and adjusted to different conditions of the market by adding  variable. Moreover, as far as the evidence of herding behavior was mostly detected when markets are volatile, adding of  allows us to keep linearity of relationship between CSAD and returns of market portfolio.

Taking into consideration that stock markets during bull and bear phases can be asymmetric, we will conduct non-linearity test, as it was suggested by Caparrelli et al. [2] and Tan et al. [14] for bull and bear markets separately, as well as for market as a whole, as it was proposed by Chang et al. [3], Chiang and Zheng [4]. Comovement of market portfolio’s returns and cross-sectional absolute deviations of returns of the stocks in the composition at decreasing rates will testify herding behavior of investors. And on the opposite, increasing rates will signify inefficiency of the market, but without herding.

For bull markets the following model will be applied:

Equation 3:

                                       (3)

In turn, for bear markets:

Equation 4:

                       (4)

Finally, for testing of asymmetric behavior of stock markets in general we will use the following model:

Equation 5:

                        (5)

Where:

t – is a period of time;

 - is a cross-sectional absolute deviation;

 - is an absolute market return;

 - is a market return squared (for bull, bear and all market respectively);

 - is a dummy variable equal to 1 if  , and 0 otherwise;

 - is a dummy variable equal to 1 if  , and 0 otherwise;

,  ,  and  - are the coefficients of regression;

  - is random error term;

For the bull and bear market models negative and statistically significant coefficient   , and for the latter one negative and statistically significant coefficients and   respectively, will indicate herding behavior of market participants.

Moreover, we will explore patterns of herding by testing behavior of investors during high and low liquidity days. Variable, which captures liquidity, reflects the breadth and depth of the market (high trading volumes), and on the opposite, variable, which holds illiquidity, shows thinness and depth of the market (low trading volumes). Both variables altogether exhibit asymmetric behavior of the stock market.

For estimation the following regression model will be applied:

Equation 6:

                (6)

Where:

t – is a period of time;

 - is a cross-sectional absolute deviation;

 - is an absolute market return;

 - is a market return squared;

 - is a dummy variable equal to 1 if trading volume falls higher than 90^th percentile, and 0 otherwise;

 - is a dummy variable equal to 1 if trading volume falls lower than 10^th percentile and 0 otherwise;

,  ,  and  - are the coefficients of regression;

 - is random error term;

Negative and statistically significant coefficients  and  will demonstrate herding behavior of investors during low and high liquidity days.

Christie and Huang [5] and Tan et al. [14] suggest to use daily timeframe in the cross-section, because due to the higher extent of fluctuations on the longer timeframes, what increases variability and makes almost impossible to detect herding bias. The data will be extracted from Thomson Reuters Datastream database and the dataset will be composed of the shares of the companies from the biggest stock exchanges of Eastern and Central Europe: MOEX (Moscow Exchange), WSE (Warsaw Stock Exchange), PSE (Prague Stock Exchange) and BSE (Budapest Stock Exchange) starting from February 8, 2010 to December 10, 2021. For the calculation of individual returns and market returns the simple return method will be used.

The question this research will be answering is:

Is there any evidence of herding behavior in the stock markets of Eastern and Central European countries?

The hypothesis developed after a preliminary review of literature are the following:

Individuals follow expectations of the market (dispersion between market participants’ returns and return of the market is significantly higher than or equal to the mean of market portfolio’s return at 10% significance level)

Where  is a regression coefficient of the CSAD model’s regression equation.

Individuals follow expectations of the market (dispersion between market participants’ returns and return of the market is significantly higher than or equal to the mean of market portfolio’s return at 10% significance level) taking into consideration: differences for bull and bear market asymmetric patterns during high and low liquidity days

Where  and   are regression coefficients of the models, applied for testing of asymmetric patterns of investors’ behavior in different market phases.

Individuals follow expectations of the market (dispersion between market participants’ returns and return of the market is significantly higher than or equal to the mean of market portfolio’s return at 10% significance level) taking into consideration: differences during high and low liquidity days.

Where  and   are regression coefficients of the models, applied for testing of asymmetric patterns of investors’ behavior during high and low liquidity days respectively.

The cross-sectional absolute deviation model (CSAD) from the Equation 2, proposed by Chiang and Zheng [4] was applied for examination of herding behavior in the financial markets of Eastern and Central Europe from our sample. Negative and statistically significant  coefficients will indicate that   increasing with decreasing rate with a magnitude of market returns, and investors resort to herding behavior in particular market. The outputs with estimations of unconditional herding for both value-weighted and equally-weighted market portfolios are represented in the Table 1 and Table 2.

  Table 1.

Unconditional value-weighted herding estimations

          Table 2.

Unconditional equally-weighted herding estimations

 

The output of the Table 1 shows that for all the markets from our sample coefficients of  are positive and statistically significant at all alpha levels, what demonstrates that   in average increases with increasing of the magnitude of market movements. In addition,  coefficients for all the markets are positive and statistically significant as well, what indicates that in average  increases at increasing, but not a decreasing rate with the magnitude of market returns. It suggests that market participants in Russian, Hungarian, Polish and Czech stock exchanges in average prefer not to follow market consensus, but make different judgements about various stocks instead of, regardless what is the opinion of the market about current situation. Therefore, we fail to reject  hypothesis at all the major significance levels for value-weighted (benchmark) portfolios as a measurement of market returns.

Table 2 shows that as in the previous case, coefficients of  are positive and statistically significant at all alpha levels, and their values, besides Czech, are higher than for value-weighted portfolios. It testifies that magnitude of market returns has a higher impact on  in case of equally-weighted market portfolios. Most probably it can be explained by higher degree of market returns’ deviation of the stocks with lower weights in the benchmark, what can happen for example for the reason of less transparent informational environment and their lower liquidity.

In addition, all  coefficients, besides Hungary are positive and statistically significant, what indicates that in average  increases at increasing, but not a decreasing rate with the magnitude of market returns. The coefficient of  for Hungarian stock market is negative, but not statistically significant at 1%, 5% and 10% levels of alpha, because p-value is equal to 0,722, therefore there is not enough statistical evidence that in average investors tend to herd in Budapest Stock Exchange.

As in the case of value-weighted market portfolio, we did not find enough statistical evidence of unconditional herding behavior in Russian, Hungarian, Polish and Czech stock markets, therefore we fail to reject  hypothesis for equally-weighted portfolios as a measurement of market returns.

For examination of herding behavior during asymmetric market conditions an extension of cross-sectional absolute deviation (CSAD) model from the Equation 5 will be applied. Negative and statistically significant  and  will indicate a tendency of investors to herd on the days of positive and negative returns, and therefore we will be able to estimate an impact of up and down markets on investors’ behavior and an extent of markets’ asymmetricity. The outputs with estimations of herding in rising and declining markets for both value-weighted and equally-weighted market portfolios are represented in the Table 3 and Table 4.

          Table 3.

Value-weighted herding estimations in rising and declining markets

 

 

According to Table 3, if a benchmark return used as a measurement of market return, all the coefficients of  and  for all the markets are positive and statistically significant (for Hungary at 5% level of significance, and for all the other countries at 1% level of significance). Therefore, we fail to reject  hypothesis. Regardless market returns are positive or negative, their cross-sectional absolute deviation increasing at increasing rate depending on their magnitude, and there is no asymmetric herding behavior in presence. In addition, for all the markets  coefficients are significantly higher than  coefficients, what demonstrates that  increases at a higher rate on the rising markets, than on the declining ones. It shows that even herding behavior for value-weighted portfolios is not in presence in the markets of Eastern and Central Europe from our sample, but the markets exhibit asymmetricity, depending on whether market returns are positive or negative. 

Table 4.

Equally-weighted herding estimations in rising and declining markets

 

Table 4 shows that all the coefficients of  are positive and statistically significant (for Russia, Hungary and Czech at 1% level of significance, and for Poland at 10% level of significance), so there is no statistical evidence of herding around market consensus in these markets on the days of positive market returns. Moreover, all the  coefficients for equally-weighted portfolios are significantly different from  coefficients for the value-weighted portfolios: (5,407 against 7,012 in Russian, 7,080 against 3,110 in Hungary, 1,861 against 3,443 in Poland and 18,20 against 4,248 in Czech). In case of Russia and Poland cross-sectional absolute deviation increasing at higher increasing rate with a magnitude of market returns, if the portfolio is value-weighted, and in case of Hungary and Czech if portfolio is equally-weighted.

As for the days of negative returns,   coefficients for Russian, Polish and Czech markets are positive and statistically significant at 1% level, and moreover they are significantly higher than for the value-weighted portfolios. It indicates a higher deviation of smaller individual stocks with increasing of magnitude of negative market returns, in comparison with their peers. However, as for the Hungarian market,  coefficient is negative (-3,182) and statistically significant at 5% level of alpha (p-value 0,02), what testifies that  in average increasing at a decreasing rate on the days of negative market returns. Therefore, it is the only one case of herding behavior during asymmetric market conditions we detected in the financial markets of Eastern and Central Europe from the sample. Herding behavior in Hungarian market goes in line with arguments of Chang et al. [3] and Yao et al. [15], who stated that investors tend to act in more homogeneous fashion during down market conditions. Christie and Huang [5] as well as Chiang and Zheng [4], argued that effects of herding have the most sustainable impact when markets are in down phases with respect to volatility and trading volume, due to behavioral patterns, human beings are prone to, and market participants tend to herd to higher extent during periods of panic, market losses and extreme trading volumes  Moreover, for equally-weighted portfolio as a measurement of market return, Hungarian market is asymmetric: investors in average exhibit herding behavior on the days of negative market returns, but make independent decisions about stocks on the days of positive market returns, what is reflected by the Figure 1.

 

Figure 1. Asymmetric behavior of investors in the Budapest Stock Exchange

         

Figure 1 shows that yellow dots decay with increasing of the magnitude of negative market returns, what basically graphically reflects that  in average increases at a decreasing rate during declining market conditions and testifies herding behavior of investors. Whereas, blue dots, on the opposite, have a tendency to go up at an increasing rate after positive market returns cross a specific threshold, what testifies that investors in average make their decisions independently from the market on the days of market rising.

In conclusion, we fail to reject  hypothesis for Russian, Polish and Czech markets, which states that investors in the financial markets of Eastern and Central Europe exhibit herding behavior during asymmetric market conditions, when equally-weighted portfolio’s returns are taken as a measurement of market returns, because we did not find statistical evidence for that. However, we cannot neither accept, nor reject  hypothesis for the Hungarian market, because investors in the Budapest Stock Exchange act differently on rising and declining markets: make independent decision in the former case, and herd around market consensus in the latter one. 

For examination of herding behavior on the days with high and low trading volumes an extension of cross-sectional absolute deviation (CSAD) model from the Equation 6 will be applied. Negative and statistically significant  and  coefficients will indicate a tendency of market participants to mimic each other on the days with extreme volumes traded (volume values lower than 10^th percentile for low liquidity days and volume values higher than 90^th percentile for high liquidity days).  The outputs with estimations of herding conditional on market liquidity for both value-weighted and equally-weighted market portfolios are represented in the Table 5 and Table 6.

Table 5.

Value-weighted herding estimations conditional on market liquidity

 

According to Table 5, the coefficient of   for the Moscow Stock Exchange is positive, but not statistically significant at the main levels of alpha (p-value is equal to 0,426), whereas for Budapest Stock Exchange it is negative, but not statistically significant as well (p-value is equal to 0,583). Finally, for the Warsaw Stock Exchange it is negative (-5,235) and statistically significant at 5% level of alpha (p-value is equal to 0,013), therefore we have enough statistical evidence to say that investors in Polish market in average likely to exhibit herding behavior during low-liquidity days, if the benchmark (WIG20 indice) is taken as a measurement of market returns. Most probably it is an inefficient form of herding, because during low-liquidity days both institutional and individual investors make less trades, and price movements are nearly no more than a noise.

As for the days of high market volumes, results of Table 5 testify that investors in the Moscow Stock Exchange do not tend to mimic each other on the days with high volumes traded, because  coefficient is positive and statistically significant at all the levels of alpha. As for the Hungarian and Polish markets, we do not have enough statistical evidence to say whether investors exhibit herding behavior or follow their own opinions about stocks, regardless what the market returns are.

Table 6.

Equally-weighted herding estimations conditional on market liquidity

 

Table 6 shows that if equally-weighted portfolio returns are taken as a measurement of market returns, in Moscow Stock Exchange and Warsaw Stock Exchange investors have a tendency to herd around market consensus on the low-liquidity days. The coefficients of  are negative and statistically significant at 1% level of alpha. For the Russian market it is equal to (-8,866) with a p-value equal to 0, and (-7,274) for the Polish market with a p-value equal to 0,002 respectively. As in the case of WIG20 indice with a benchmark as a measurement of market returns, investors in Russian and Polish markets may exhibit inefficient form of herding on the days with low trading volume. Moreover, in the Moscow Stock Exchange the degree of herding is higher on the low-liquidity days (-8,866 against -8,062).  In addition, the degree of herding on the low-volume trading days in the Warsaw Stock Exchange is higher, when equally-weighted portfolio is taken as a measurement of market returns (-7,274), than in case of WIG20 (-5,235). It testifies that  may increase with a magnitude of market returns at more decreasing rate due to the higher weight of smaller stocks in the equally-weighted portfolio. Coefficient  for Hungarian market is positive, relatively high (14,31) and statistically significant, what indicates that investors are more likely to form and follow opinions about stocks independently from the market on the low-liquidity days.  

As for the high-liquidity days,   coefficients for the equally-weighted portfolios in Russian and Hungarian markets are negative and statistically significant at all main levels of alpha (-8,062) and (-4,965) respectively with p-values equal to 0. It demonstrates that market participants in these markets exhibit herding behavior on the high-volume trading days, what goes in line with arguments and empirical evidences of Tan et al. [14], Yao et al. [15], Lam et al. [10], Gavriilidis et al. [9], who stated that days with high trading volumes attractive mostly for institutional investors, what leads to copying of their behavior by private investors, who possess less information in comparison with more sophisticated colleagues. Coefficient  is positive and not statistically significant, and therefore there is no evidence that investors in Warsaw Stock Exchange herd around market consensus on the high-liquidity days.

Summarizing output results from Table 5 and Table 6, investors in Warsaw Stock Exchange tend to herd during low-liquidity days, if WIG20 indice’s returns are taken as a measurement of market returns. As for the equally-weighted portfolios, investors in Russian market herd around market consensus both on days of extremely high and extremely low volumes traded, investors in the Hungarian market more likely to herd during high-liquidity days, whereas in Polish market in contrast, during low liquidity days. We have enough statistical evidence to reject  hypothesis and to state that investors in the financial markets of Eastern and Central European countries have a tendency to exhibit herding behavior, conditional on market liquidity.

Four stock markets of post-socialist Central and Eastern European countries (Russia, Hungary, Poland and Czech Republic) were empirically examined on the presence of herding behavior of investors. Hypothesis 1 was elaborated for the testing of unconditional herding, whereas Hypothesis 2 and Hypothesis 3 for the testing of herding, dependently on asymmetric market conditions: rising and declining markets as well as extremely high and extremely low-liquidity days. Both the models, based on value-weighted portfolios (keeping market returns equal to daily returns of MOEX, BUX, WIG20 and PX stock indices) and equally-weighted portfolios (holding markets returns equal to daily average returns of stocks in the composition of portfolio) were tested in this research.

The results do not provide evidence of herding behavior in general, however herding behavior of investors under asymmetric market conditions was detected. Market participants in Budapest Stock Exchange (equally-weighted portfolio) tend to herd around market consensus on the days with negative market returns, what goes in line with implications of Christie and Huang [5] and Chiang and Zheng [4], who argued that investors in the periods of panic and declining of the market have a tendency to follow market expectations in order to avoid of bigger losses. Moreover, on the days with low trading volumes, investors in the Warsaw Stock Exchange (value-weighted portfolio) prefer to mimic each other.  During low-liquidity days institutional investors trade less, and therefore private investors prefer to stay away from the market as well, following their more sophisticated peers, because price movements in such conditions are nearly a noise. In addition, if an equally-weighted portfolios are taken as measurements of market returns, investors in Moscow Stock Exchange exhibit herding behavior during both low and high-liquidity days, in turn investors in Warsaw Stock Exchange only during low-liquidity days, and in Budapest Stock Exchange only during high-liquidity days respectively.  In the case of Hungarian market, it goes in line with an argument of Gavriilidis et al. [9] and De Long et al. [6] who stated that on the days of extremely high trading volume in general private investors prefer to follow their more sophisticated peers, institutional investors, who possess more information, for the reason of having fear to lose an opportunity to buy in a good time and to lag behind the market.

The pattern of investors’ behavior in the financial markets of Eastern and Central European countries is quite similar: they do not herd around market consensus in general, however they tend to exhibit herding behavior with respect to asymmetric market conditions, and specifically on the days with extreme volumes traded. The only one difference is that in some markets market participant prefer to mimic each other on the high liquidity days (Budapest Stock Exchange), in some of them on the low-liquidity days (Warsaw Stock Exchange), and in others on both high and low-liquidity days (Moscow Stock Exchange). Another thing, what slightly differentiates the markets in terms of herding is a tendency of investors in the Hungarian market to follow market expectations on the days of negative market returns.

Even though, stock markets of post-socialist countries from the sample have been adapting to the new conditions and has been significantly developing for the last 30 years, they are not by chance still considered as emerging markets. The author finds enough evidence to state that market anomaly in the face of herding behavior, dependently on asymmetric market conditions still exists in the region and may reinforce inefficiencies of the market during periods of shocks and market crashes as well as during period of over optimism and market bubbles. Therefore, relevant state authorities in perspective should enact proper regulations in order to alleviate the negative effects of investors’ herding behavior. 

In this paper we did not differentiate between domestic and international investors, as well as between private and institutional investors because of unavailability of data, providing such classifications. According to intuition of the author, exactly domestic investors in the stock exchanges from the sample would more prone to exhibit herding behavior, due to the both more general reason : tendency of people to be loyal to authority and to conform, hardwired in the process of evolution to the human brain (in case of stock market, the direct authority is the market itself, and indirect authorities are institutional investors, because markets are mostly driven by them) , and more specific reason: group way of thinking, shaped by socialism in the past, transferred to society through generations, leading to extreme conservatism, maintaining of stability and risk-aversion,  what could be reflected in the financial market as a consequence. Therefore, if this kind of data would be possible to obtain, this research can form a solid foundation for the future development of studies, comparing the cross-country extent of herding of various groups of investors in the financial markets of Eastern and Central European countries, what in turn would make possible to state authorities and responsible institutions to enact more relevant regulations for stock markets of the region.

 

References:

1. Alesina, A. - Fuchs-Schündeln N. (2005): Good Bye Lenin (Or Not?): The Effect of Communism on People’s Preferences. NBER Working Paper No. 11700, pp. 1-12
2. Caparrelli, F.- D’Arcangelis, A.M.- Cassuto, A. (2004): Herding in the Italian Stock Market: A Case of Behavioral Finance. The Journal of Behavioral Finance, Vol. 5, No. 4, pp. 222-230
3. Chang, E.C.- Cheng, J.W.- Khorana, A. (2000): An Examination of Herd Behaviour in Equity Markets: An International Perspective. Journal of Banking & Finance, Vol. 24, No. 10, pp. 1651–1679.
4. Chiang, E.C.- Zheng, D. (2010): An Empirical Herd Behavior in Global Stock Markets. Journal of Banking & Finance, Vol. 34, No. 8, pp. 1911-1921
5. Christie, W. G. - Huang, R. D. (1995): Following the Pied Piper: Do individual returns herd around the market? Financial Analysts Journal, Vol. 51, No. 4, pp. 31–37
6. De Long, B. - Shleifer, A. - Summers, L. - Waldman, R. (1990): Positive Feedback Investment Strategies and Destabilizing Rational Speculation. Journal of Finance, Vol. 45, No. 2, pp. 379-395
7. 22. Dr. Vlachoutsicos, C. (1997): Russian Collectivism: An Invisible Fist in the Transformation Process of Russia. Retrieved: February 10, 2020
8. Droit, E. (2012): From Identity to Appearances: Clothing and Political Identity in the GDR. The Case of the Young Pioneers’ Neckerchief. Revue d’histoire moderne et contemproraine, Vol. 59, No. 2, pp. 108-143
9. Gavriilidis, K.- Kallinterakis, V.- Ferreira, M.P.L. (2013): Institutional Industry Herding: Intentional or Spurious? Journal of International Financial Markets, Institutions & Money, Vol. 26, No. 1, pp. 192-214
10. Lam, K.S.- Qiao, Z. (2015): Herding and Fundamental Factors: The Hong Kong Experience. Pacific-Basin Finance Journal, Vol. 32, issue C, pp. 160-188
11. Mamontov, V.- Kozhevnikova T. (2014): Collectivism and Individualism in Modern Russia. Asia Social Science; Vol. 10, No. 23, pp. 210-214
12. Morck, R. (2004): Behavioral Finance in Corporate Governance – Independent Directors, Non-Executive Chairs, and the Importance of the Devil’s Advocate. Working paper 10644, National Bureau of Economic Research, Massachusetts
13. Semenov, V. (2004) About Methodology and Methods of Modern Psychology. Vestnik of Saint-Petersburg University, Vol 6, No. 2, pp. 61-63
14. Tan, L.- Chiang, T.C.- Mason, J.R.- Nelling, E. (2008): Herding behaviour in Chinese stock markets: An examination of A and B shares. Pacific-Basin Finance Journal, Vol. 16, issue 1-2, pp. 61–77
15. Yao, J.- Ma, C.- He, W.P. (2014): Investor herding behavior of Chinese stock market. International Review of Economics and Finance, Vol. 29, issue C, pp.  12-29