Herd behavior of investors in the stock market: An analysis of cross-country effects in the CEE

• Název práce v češtině: Stádní chování investorů na akciovém trhu: Analýza mezinárodních efektů v CEE
• Název v anglickém jazyce: Herd behavior of investors in the stock market: An analysis of cross-country effects in the CEE
• Klíčová slova: stádní chování, CEE, mezinárodní efekty, akciový trh
• Klíčová slova anglicky: herding, CEE, cross-country effects, stock market
• Akademický rok vypsání: 2017/2018
• Typ práce: diplomová práce
• Jazyk práce: angličtina
• Ústav: Institut ekonomických studií (23-IES)
• Vedoucí / školitel: PhDr. Jiří Kukačka, Ph.D.
• Řešitel: Mgr. Vojtěch Lerche - zadáno vedoucím/školitelem
• Datum přihlášení: 04.02.2018
• Datum zadání: 04.02.2018
• Datum a čas obhajoby: 19.06.2019 09:00
• Místo konání obhajoby: Opletalova - Opletalova 26, O105, Opletalova - místn. č. 105
• Datum odevzdání elektronické podoby: 10.05.2019
• Datum proběhlé obhajoby: 19.06.2019
• Oponenti: Mgr. Lukáš Vácha, Ph.D.

Seznam odborné literatury

Bekiros, S., Jlassi, M., Lucey, B., Naoui, K., & Uddin, G. S. (2017). Herding behavior, market sentiment and volatility: Will the bubble resume?. The North American journal of economics and finance, 42, 107-131.

Bikhchandani, S., Hirshleifer, D., & Welch, I. (1992). A theory of fads, fashion, custom, and cultural change as informational cascades. Journal of political Economy, 100(5), 992-1026.

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), 1651-1679.

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.

Chiang, T. C., Li, J., & Tan, L. (2010). Empirical investigation of herding behavior in Chinese stock markets: 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.

Demirer, R., Kutan, A. M., & Chen, C. D. (2010). Do investors herd in emerging stock markets?: Evidence from the Taiwanese market. Journal of Economic Behavior & Organization, 76(2), 283-295.

Devenow, A., & Welch, I. (1996). Rational herding in financial economics. European Economic Review, 40(3-5), 603-615.

Economou, F., Kostakis, A., & Philippas, N. (2011). Cross-country effects in herding behaviour: Evidence from four south European markets. Journal of International Financial Markets, Institutions and Money, 21(3), 443-460.

Gębka, B., & Wohar, M. E. (2013). International herding: Does it differ across sectors?. Journal of International Financial Markets, Institutions and Money, 23, 55-84.

Koenker, R., & Bassett Jr, G. (1978). Regression quantiles. Econometrica: journal of the Econometric Society, 33-50.

Mobarek, A., Mollah, S., & Keasey, K. (2014). A cross-country analysis of herd behavior in Europe. Journal of International Financial Markets, Institutions and Money, 32, 107-127.

Pochea, M. M., Filip, A. M., & Pece, A. M. (2017). Herding behavior in CEE stock markets under asymmetric conditions: a quantile regression analysis. Journal of Behavioral Finance, 18(4), 400-416.

Villatoro, F. (2009). The delegated portfolio management problem: Reputation and herding. Journal of Banking & Finance, 33(11), 2062-2069.

Předběžná náplň práce

Motivation:
Investing in the stock market may be a challenge when investors act like a crowd. Behavioral finance suggests that people follow the others especially during turbulent periods of fear, uncertainty, and panic. For the small investors who mimic the leading experts instead of inquiring costly analysis and for the financial intermediaries that do not care much about their reputation, following the market may be a fully rational action (Villatoro, 2009). On the other hand, completely ignoring own information and going against one’s beliefs blindly is considered as irrational (Devenow and Welch, 1996).

If many investors act in the same way over certain time, the overall market is affected systematically. Bikhchandani et al. (1992) explain that such correlation among a large group of people could lead to the whole market's false decision making. As consequence, investors need to buy more assets to ensure the same level of portfolio diversification and the pricing of assets is incorrect and does not correspond to the fundamentals.

Demirer, Kutan, and Chen, (2010) conducted research on the Taiwanese market and recommend the selection of emerging economies. In support, Pochea, Filip and Pece (2017) who focused on the Central and Eastern Europe (CEE) name the causes: “…weak reporting requirements, poorer accounting standards, lax enforcement of regulations, and costly information acquisition which end in lack of transparency and propensity to herd.“(2017, p. 401)

Apart from considering one separate market, Economou, Kostakis, and Philippas (2011) focused on Portugal, Italy, Greece, and Spain (so-called “PIGS”) and the inter-countries herding effect during the crises of 2007-2008. The Greek cross-sectional dispersion of returns is connected to the dispersions of the remaining “PIGS”. This implies that portfolio diversification in these countries is less effective and the region is more unstable during market stress conditions.

One possible way is to examine cross-country effects within the region, i.e. whether the Polish investors herd around the Czech market, for example. However, the thesis aims at exploring the effect of foreign developed markets in explaining domestic herding of CEE countries, following the paper of Chiang and Zheng (2010). They indicate that return dispersions in the U.S. play a significant role in explaining the non-US market’s herding behavior in Asian and Latin-American markets. The similar relationship could hold for the CEE as strong dependencies vis-a-vis the advanced western markets can be expected.

Hypotheses:
Hypothesis #1: There is no herding effect in the stock market. (evaluated for each CEE country)
Hypothesis #2: There is no herding effect in the local stock market conditional on the dispersions of returns in different quantiles of data.
Hypothesis #3: There is no cross-country herding effect between the advanced markets and an individual CEE markets.
Hypothesis #4: There is no cross-country herding effect between the advanced markets and an individual CEE markets conditional on the dispersions of returns in different quantiles of data.

Methodology:
In the first step, daily prices of firms will be obtained from the Thomson Reuters Eikon Database. The CEE region (following the OECD classification) consists of 12 countries: Albania, Bulgaria, Croatia, the Czech Republic, Hungary, Poland, Romania, the Slovak Republic, Slovenia, and the three Baltic States: Estonia, Latvia, and Lithuania. For cross-country regressions in Europe, the relevant benchmark country is Germany (result of Mobarek, Mollah and Keasey, 2014).

To assert the herding effect empirically, two main approaches of Christie and Huang (1995) and of Chang et al. (2000) are widely used in literature. Christie and Huang propose the cross-sectional standard deviation of returns (CSSD) approach. Lower values of dispersion indicate that individuals do herd, i.e. follow the market consensus and the individual returns are literally less dispersed than the information distributed among the agents. Formally, the model follows:
〖CSSD〗_t= α+ β^L I_t^L+ β^U I_t^U+ ε_t ,where 〖CSSD〗_t=√((∑_(i=1)^n▒(R_(i,t)-R_(m,t) )^2 )/(N-1)). R_(i,t) is the return of firm i at time t, R_(m,t) is the cross-sectional average of returns of n firms in the aggregate market portfolio. I_t^L and I_t^U are dummy variables for the extreme lower tail and upper tail of market returns (say 10%). Rational asset pricing model predicts that when the market return increases, so should proportionally the dispersion of individual returns. The predicted positive and linear relationship is rejected if there are significantly negative coefficients β^L and β^U. It is suggested that during extreme periods, investors tend to panic and follow the market consensus and thus the dispersion of returns diminishes, indicating herd behavior.

The second approach proposed by Chang et al. (2000) focusses on the relationship between the cross-sectional absolute deviation (CSAD) and the market portfolio return. Formally, the general model has the following form:
〖CSAD〗_t= γ_0+ γ_1 R_(m,t)+ γ_2 R_(m,t)^2+ ε_(t ,) where 〖CSAD〗_(t )= 1/N ∑_(i=1)^N▒|R_(i,t)-R_(m,t) | . Note that the CSAD is not a measure of herding itself as in the case of the CSSD, but the relationship between these two variables is essential. As Chang et al. point out, the CSAD approach should be preferred because it allows for nonlinearity and the estimation is more robust to outliers. Herding is indicated by the CSAD either being a negative function of the market return or an increasing function at a decreasing rate. Significantly negative coefficient γ_2 would imply herding.

These straightforward models can be further extended. Economoua, Kostakisb, and Philippasa (2011) adjust the model for the asymmetric effect of positive and negative returns:
〖CSAD〗_t= γ_0+ γ_1 (1-D) R_(m,t)+γ_2 DR_(m,t)+γ_3 (1-D) R_(m,t)^2+ 〖γ_4 DR_(m,t)^2+ ε〗_(t ), where the dummy variable D is equal 1 in the case market return is negative at time t, and 0 otherwise. Behavioral finance suggests that people panic during severe times whereas for the periods of positive market returns, optimistic investors are more likely to decide alone and less likely to follow the market consensus.

After the time series models for each individual country are estimated (hypothesis 1), the focus of this thesis will shift to the examination of the cross-country effect of herding (hypothesis 3). In the study of Chiang and Zheng (2010), the CSAD of the U.S. market is included in the specification:
〖CSAD〗_t= γ_0+ γ_1 R_(m,t)+ γ_2 R_(m,t)^2+γ_3 〖CSAD〗_(US,t)+γ_4 R_(US,m,t)^2 〖+ ε〗_(t ). The domestic market is herding together with the U.S. market if the coefficient γ_4 is negative and significant, i.e. the local investors herd around the U.S. market.

To improve the econometric modelling, several studies of herding benefit from the quantile regression method (Bekiros, Stelios, et al., 2017). Another paper of Chiang, Li, and Tan (2010) concludes that within the least squares framework, the herding was not present in the Chinese stock market. But investors indeed herd conditional on the CSAD in the lower quantile region. In the hypotheses 2 and 4, this pattern will be explored and the conclusions from the hypotheses 1 and 3 can be thus restated. The QR framework is proposed by Koenker and Bassett (1978).
Expected Contribution:
I will contribute to the existing literature in three ways. First, the most recent data will be utilized to evaluate the intra country effect in the CEE countries after the financial crisis of 2008. Pochea, Filip, and Pece (2017) investigated herding behavior on the dataset from 2003 to 2013 in the Central and Eastern Europe. The time span includes the crisis with high volatility when investors should herd more extensively than usual. I will focus on the period after the crisis when the market has relatively calmed and has probably undergone certain structural changes, as the authors claim.

The main contribution consists of the cross-country effect analysis between the advanced markets and the CEE countries. The paper of Chiang and Zheng (2010) examined herding behavior in the global market. Nevertheless, the CEE countries were not included because the timespan started far before the CEE area’s transformation in 1990’s. The study of Mobarek, Mollah, and Keaseynot (2014) did not cover CEE countries either. Then, the results can indicate some countries herd more extensively while others do not. Portfolio diversification and policy implications can be formulated.

Third, in the thesis, I extend the Chiang and Zheng’s methodology by applying the quantile regression approach which is rather a new econometric method in the focus on cross-country herding effect (see Gebka and Wohar (2013)).

Outline:
Introduction: I will explain the rational and irrational herding and the connections to the efficient market hypothesis. Also, there is a need to explain the economic value of portfolio diversification and describe how the financial market is affected by herding.
Studies on herding: I will briefly contrast and compare the relevant studies on herding that focus either on the CEE area or on the cross-country effects.
Data: I will explain the data, time span, frequencies and provide relevant statics.
Methods: I will explain the base models of CSSD and CSAD and the quantile regression approach. I will mention necessary estimation issues, the Newey-West estimators and further.
Results: I will discuss my regression outputs and do a comparison of CEE countries.
Concluding remarks: I will summarize my findings and their implications for policy and future research.