Printer Friendly

A Test of Herding in Performance of Indian Stock Exchange.

Introduction

The incidences of herding take place when people discount explicit information about firm and chase the market trend for making their investment decisions. The stock exchange is flooded with tons of information every day; however, investors do not show confidence in using these information's for making investment decisions. The common phenomenon is to copy the behavior of others investors, the 'imitative behavior' in investment is called herding. Banerjee (1992) studied existence of herd behaviour in case "everyone doing what everyone else is doing, even when their information suggests doing something different."

Herd behavior of investors is likely to be noticed in stress market conditions. Christie and Huang (1995) believe investors behave rationally, as explained in asset valuation theories, in their investment decision during normal market conditions. However, extreme situations force to produce intense emotions and people look to find comfort in following the crowd, as in the case of the financial market collapse in 1987, (Shiller, 1990) and in currency market (Frankel and Froot, 1986).

Caparrelli et al. (2004) studied the herd behaviour of investors by a very attractive illustration of two restaurants. One night, a customer decided to take his dinner in restaurant "A" on suggestion of an extremely reliable source. He observed full occupancy, with customers waiting, in another restaurant "B" whereas restaurant "A" hardly had any customer. So, he followed crowd and changed his choice to eat in restaurant A. He purely discarded his possessed information and went with crowd presuming that they had better knowledge. His conviction in the knowledge of mass may go wrong if some patrons had behaved in same manner and man who entered restaurant B initially may have wrong perception. Such fear of be a part of isolated consumers force to avoid restaurant "A" and become a part of the crowd by entering restaurant "B". Such similar behaviours may be observed by investors in stock markets. In the case of extreme disaster, 'herd behaviour' is reported as common term as a symbol of irrationality in market behaviour. Herding may direct towards mispricing of shares as capacity of decision making has been affected by subjective analysis of predictable systematic risk and return (Hwang and Salmon, 2004).

An effort has been made in this, to examine the evidence of herd behaviour in stock exchange of India, including period of recent financial crisis. The review of literatures explores asymmetry in herding behaviour with direction of market-wide return.

To provide a systematic approach to the study, the research paper has been divided into following sections; the review of past and published literature along with the argument on research have been discussed in section 2, Section 3 provides details of techniques need and methodology applied. The empirical evidences of herding in context of Indian share market have been covered in section 4 of the study; while Section 5 presents the final outcome of study, section 6 discusses the limitations and scope for future research.

Review of Literature

In last few years, empirical literature signifying the existence of herding in finance market has became popular in academic research. The section is dedicated to review the literature published on herding in different stock markets.

The work of Shiller and Pound (1986) has initiated the empirical study on herding behavior in stock market. Their work is a summary of a survey to show the influence of information by professionals on investment of major institutional investors in stock market. Banerjee (1992) developed an empirical model to show the impact of asymmetry of information and costly acquisition on investment pattern. Investors chase the market consensus by neglecting the intrinsic value of stock, which results to inefficient market condition. Lakonishok, et al. (1992) noted little evidence for presence of herding behavior in high capitalized stocks , while small stocks trading are being highly influenced by optimistic trading feedback among major corporate investors. Grinblatt et al. (1995) found proof for anti-herd behavior in investment pattern in mutual funds , whereas Wermers (1999) reported evidences of herding activity in investment pattern of some small mutual funds mainly in growth oriented schemes. Christie and Huang (1995) have supported efficient market theory for US stock market (NYSE) as herding behaviour has not been used as a factor for while calculation return. The extension of their work by Chang et al. (2000) studied herding behavior of investors in major stock exchanges of different countries, i.e., Hong Kong, USA, Taiwan, South Korea, and Japan. They reported absence of herding in stock exchanges of Hong Kong and USA, fractional presence of herding in Japanese market, and prominent presence in emerging markets of Taiwan and South Korea. Caporale et.al. (2008) tested behaviour in unfavorable market situations in ASE (the Athens Stock Exchange), and reported significant evidence of herd behaviour during the period of 1998-2007. Lodetti and Kallinterakis (2009) reported absence of herd behaviour in the Stock Exchange of Montenegro. Economou et al. (2011) and Singh and Lao (2011) supported that the possibility of presence of herding is more likely in stock exchanges of emerging economies. Many studies (Chang et al., 2000,McQueen et al., 1996;; Lee et al., 2013; Tan et al., 2008) have given evidence of directional asymmetry in price, i.e., return dispersion of a stock as a derivation of absolute market-wide return. The evidence of herding is stronger in downward movement condition in the financial market. Mustapha and Medhioub (2018) take into account the difference existing in market conditions and find evidence of herding behaviour during rising markets only for GCC(gulf Cooperation Market) markets. Economoua (2018) reported evidences of herding in stock exchanges of USA, UK and Germany from January 2004 to July 2014.

Some researches (Chiang et al., 2013; Caparrelli et al., 2004; Chang et al., 2000) have supported the evidence of strong herding during unfavorable market condition in different stock exchanges. The results of theses empirical works hold a lot of significance with regard to recent economic crisis as extreme volatility has been observed during event time. As financial markets in India offer a higher growth, studies pertaining to herding behaviour have gained lot of prominance. Prasad et al. (2012) study efficiency of Indian stock markets and note absence of herding. Lakshman et al. (2013) reported mild existence of herding in the stock exchange of India. Lao et al. (2011), Poshakwale et.al. (2014) reported herding in the stock exchanges on India during big market-wide movements.

There is non-uniformity among researchers with regard to herd behaviour, mainly in context of India Stock markets. Due to indefiniteness about the existence of herd behaviour in investment pattern in markets, the study re-examines one of the most debated topics "Existence of herding in the stock behaviour of Indian market". The major objective of this paper is to investigate existence of asymmetric relation between movements of stock and market-wide herding in a prominent Indian stock exchange, BSE. The paper examines the market-wide herding behaviour in extreme market condition in 2008 economic crisis.

Research Methodology and Description of Data

Methodology

In this section, the study explain an empirical methodology used to find the evidence of herding in stock markets investment pattern .The empirical models given by Christie and Huang (1995), extended by Chang, Cheng and Khorana (2000) suggested that herd behaviour can be evidenced by cross-sectional methods of assets valuation, and a cross-sectional dispersion on estimated average market returns, such movement indicates some kind of market consensus. The approach is evolved around the thought that CSAD (cross-sectional absolute deviation) value will be small throughout the herding duration. It depends on the conviction that herding cannot permit deviation in a stock returns from the average market return.

CSAD calculation is extended version of return dispersion model, used by Chang et al. (2000). Christie and Huang (1995) explained return dispersion as the best indicator for evidence of herding. Return dispersion indicates deviation in return of single stock from mean return in market. The measurement of cross-sectional standard deviation (CSSD) of return indicates the evidence of market herding. They also indicated the market movement between extreme and normal phases, and an investor would take investment decisions depending on the market average return during extreme movements while in normal condition, people invest rationally. As a effect, a stock returns will be in cluster in extreme market conditions, reducing CSSD in comparison to normal market condition. This is a disparity in balanced expectation of general asset valuation model, as dispersion grows in extreme market condition since each stock has dissimilar level of sensitivity to market mean performance.

Christie and Huang (1995) calculated value of CSSD for individual shares by equation (1). A small CSSD may indicate spurious and intentional herding.

[CSSD.sub.t] = [square root of ([[[summation].sup.n.sub.t=1] ([R.sub.I,t] - [R.sub.M,t]).sup.2]/N-1)] [R.sub.I,t] = Return of share I at time t; [R.sub.M,t] = Mean return of N returns for time t. N = Total no. of shares in sample. (1)

The calculated values of CSSD on returns over time were regressed to a constant and two dummy variables in order to recognize the extreme market condition. Dummy variables have been used with [D.sup.U] = 1 for returns lies in the acute right of return distribution, i.e. 1 percent to 5percent of upper tail (or else ZERO) and [D.sub.L] = 1 for market returns lies in acute left of the same data distribution, i.e. 1 percent to 5percent of lower tail (or else ZERO). Market-wide herding is reported by statistically significant negative values of coefficients in regression as per equation (2). Probable value of a coefficient denotes mean dispersion of returns without time period in selected sample correspondence to dummy variables of equation.

[CSSD.sub.t] = [alpha] + [[beta].sub.1] + [D.sup.U.sub.t] + [[beta].sub.2] [D.sup.U.sub.t] + [e.sub.t] (2)

Even if the cross-sectional standard deviation (CSSD) of stock returns is used as an instinctive measure to study herd behaviour of investors, it can significandy impact the presence of outliers. To reduce the effect of such outliers, Christie and Huang (1995) and Chang et al. (2000) projected exercise of the cross-sectional absolute deviation, (CSAD), as a improved gauge of dispersion as shown in equation (3) and performed regression similar to equation (2) in correspondence with equation (3) as shown in equation (4). Similarly statistically significant and negative values of [[beta].sub.1], and [[beta].sub.2] would indicate the presence of market-wide herding.

[CSAD.sub.t] = [[summation].sup.n.sub.t=1] [absolute value of ([R.sub.I,t] - [R.sub.M,t])]/N (3)

[CSAD.sub.t] = [alpha] + [[beta].sub.1] [D.sup.U.sub.t] + [[beta].sub.2] [D.sup.L.sub.t] + [[epsilon].sub.t] (4)

To measure herding on whole market return instead of only distribution on extremes market-wide returns as projected by Chang et al. (2000) and Christie and Huang (1995) is shown in equation (5):

[CSAD.sub.t] = [alpha] + [[gamma].sub.1] [absolute value of [R.sub.M,t]] + [[gamma].sub.2] [([R.sub.M,t]).sup.2] + [[epsilon].sub.t] (5)

The result of regression derives on relation between market return and CSAD, the regularly used asset valuation model assumes that returns dispersion has linear relation with market return .The positive value of 1 in equation-(5) shows absence of herding. However, if herd behaviour is there, dispersion would increase or decrease considerably in less amount of the market returns. Square values of market returns are included to confine this impact in the model, and subsistence of non-linearity by significant and negative value for 2 also act as evidence of herding.

Chang et al. (2000) examined for existence of asymmetric association between herding and directional movement of financial market. The above hypothesis has been tested using regression method as equations (6) and (7). A non-linear relation between CSAD and market-wide return indicates absence of herding. The non-linear relationship will be proved by statistically significant negative value of coefficient [[gamma].sub.2]

[mathematical not reproducible in ascii] (6)

[mathematical not reproducible in ascii] (7)

[absolute value of ([R.sup.UP.sub.M,t])] and [absolute value of ([R.sup.DOWN.sub.M,t])] indicate the absolute values of stock return during UP and DOWN of market respectively. This also makes a right comparison between and [[eta].sup.UP.sub.2] [[eta].sup.DOWN.sub.2] possible. The study examines evidence on herding on daily as well as monthly frequency from selected data set.

Data Collection

The study is conducted with the data obtained from the BSE Ltd, oldest stock exchange of India. As none of Indian stock exchanges has index for all listed shares, study is based on the broadest benchmark index, the S&P BSE All Cap Index. The Index is a broad index representing approx 95percent of the total market capitalization and comprises more than 910 shares that are listed on the BSE from various sectors mainly like finance, IT, FMCG, Transport Equipments, healthcare, Oil and gas, Capital Goods, Metals and Chemicals etc. Although this index is started in second quarter of 2015 but historical data has been made available from Sep 2005 by method of back calculation for constituent stocks. The study is based on data collected for period Sep-2005 to Mar-2016, incorporating the period of pre and post financial crisis 2008.

The impact of recent global financial crisis can be identify from January 21, 2008 by single day fall of 1408 points, and became severe in month of Sept-2008 on collapse of top investment bank Lehman Brothers and many more. So, the data sample has been divided into 2 broader categories of Pre i.e. Sept-2005 to Dec-2008 and Post i.e. Jan-2009 to Mar-2016. The return from all cap index has been treated against return from market (BSE SenSex) for the same period on daily as well as monthly basis to find the herding in Indian stock exchange during study period.

Empirical Evidence

The section of literature highlights empirical results in the context to identified research problem. Table I contains summary of descriptive statistics for return of market portfolio ([R.sub.m,t]) on daily and monthly data. Aggregate data for Sep-2005 to Mar-2016 indicate that returns of constituent stocks are extremely volatile in the limit of 26.06 percent. Predictably, this observed volatility in study time is due to consideration of recent financial crisis. Minimum daily return of -10.48 percent was recorded on October 24, 2008 and maximum daily return of 15.57 percent was recorded on May 18, 2009.

Negative return of 1166 days and 51 months has been observed during the study period of, as up to 5percent of tail either in case of upper limit or case of lower limit will assign value of 1 for dummy variables [D.sup.U] and [D.sub.L] respectively. Table-II reports the regression result for [CSAD.sub.t] as per model proposed in Eq-4.

Table II shows the result of regression of subgroups of sample, as daily data on daily return for whole sample period, monthly data on monthly return for whole sample period , Pre-crisis on daily return for period Sept-05 to Dec-08 , Post-crisis on daily return for period Jan-09 to Mar-16 and during crisis on daily return for period Jan-08 to Dec-08. All variable [alpha], [[beta].sub.1] and [[beta].sub.2] are statistically significant since, P-values are less than 5percent .There is no prominent evidence of herding during study period as the values of all coefficients are positive, except value of [[beta].sub.2] in case of monthly data. The negative significant value of [[beta].sub.2] proves herding in case of falling market on monthly investment pattern. Other values of coefficient show strong evidence of anti-herding for study period in daily as well as monthly investment pattern.

The second part of table II, represents the total market regression result as per equation 5 with the significant values of dummy variables. The positive and statistically significant values of [[gamma].sub.1] and [[gamma].sub.2] are evidences for absence of Herding in sub groups. Negative value of y, in daily return of whole sample is caused by weak evidence of herding in post-crisis period in Indian stock exchange, the negative value of [[gamma].sub.2] shows asymmetry non-linear relation of CSAD along with [R.sub.M,t] in study period. Furthermore, these resulted values are supporting consistency in investment pattern. Table-II show results of regression for directional market movements to check the asymmetric relation between Market return and CSAD.

The third part of table-II contains regression results for upward direction of market, [R.sub.M,t] > 0, and all values of [[gamma].sub.1.sup.UP] and [[gamma].sub.2.sup.UP] are significant and positive. This indicates anti-herding in Indian stock exchanges in daily investment pattern even in upward movement of market. Since [[gamma].sub.2.sup.UP] has negative value for monthly return pattern, it leaves a mild hope for evidence of directional herding in upward movement.

In similar way fourth part of table-II presents regression result for downward direction of market, [R.sub.M,t] < 0, and negative value of [[gamma].sub.1.sup.DOWN] indicates presence of herding in daily investment pattern of Indian market in downward movements, which leads to conclusion that returns of stocks may show a convergence in falling market condition , in some extent only. But other values for [[gamma].sub.1.sup.DOWN] and [[gamma].sub.2.sup.DOWN] do not support herding in large extent.

Conclusion

This literature has examined the existence of market-wide herding in the stock exchange of India during 2005-2016, a period before, during and after recent financial crisis of 2008. The study has been divided into subgroups and based on daily as well as monthly return data. The analysis of empirical results in section-4 does not completely disclose any strong proof of market-wide herding in during study period. However, some weak evidences of herding were reported during movements in market. Even the periods of extreme movements do not show any strong evidence for the same, this in turn, strengthens the notion of asymmetric nature of herding. In fact, Investment pattern shows evidences of rationality in investment decision based on information. This might be a major cause of mild effect of financial crisis 2008 in India. In other words, findings provide empirical evidence of a very positive outlook of investment in the Indian stock exchange mainly after global turmoil of 2008.

Limitations and Direction of Future

The research work has been conducted on constituents' shares of All cap Index i.e. 910 in numbers having 95 percent market capitalization, not on all listed shares in Indian stock exchanges. The other major limitation of the study is that, the paper only discusses the existent of herding not the cause. Further studies in this context must consider various sector and industry wise factors for better understanding of herding in investment pattern in Indian stock exchanges.

References

Ahsan, A., & Sarkar, A. (2013). Herding in Dhaka stock exchange. Journal of Applied Business and Economics, 14 (2), 11-19.

Banz, R. (1981). The relationship between return and market value of common stocks. Journal of Financial Economics, 9 (1), 3-18.

Banerjee, A. (1992). A simple model of herd behavior. The Quarterly Journal of Economics, 107 (3), 797-817.

Barberis, N., & Thaler, R. (2003). A survey of behavioral finance. Handbook of the Economics of Finance, 1, 1053-1128.

Bikhchandani, S., & Sharma, S. (2000). Herd behavior in financial markets, IMF Staff Papers, pp.279-310.

Caparrelli, F., D'Arcangelis, A., & Cassuto, A. (2004). Herding in the Italian stock market: A case of behavioral finance. The Journal of Behavioral Finance, 5 (4), 222-230.

Caporale, G., Economou, F., & Philippas, N. (2008). Herding behaviour in extreme market conditions: The case of the Athens Stock Exchange. Economics Bulletin, 7 (17), 1-13.

Chang, E., Cheng, J., & Khorana, A. (2000). An examination of herd behavior in equity markets: An international perspective. Journal of Banking and Finance, 24 (10), 1651-1679.

Chiang, T., Li, J., Tan, L., & Nelling, E. (2013.) Dynamic herding behavior in Pacific-Basin markets: Evidence and implications. Multinational Finance Journal, 17, 165-200.

Christie, W., & Huang, R. (1995). Following the pied piper: Do individual returns herd around the market? Financial Analysts Journal, 51 (4), .31-37.

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.

Economoua , F.,b, Hassapisc, C., & Philippasd, N. (2018). Investors' fear and herding in the stock market", Applied Economics, retrieved from : https://doi.org/10.1080/00036846.2018.1436145.

Grinblatt, M., Titman, S., & Wermers, R. (1995). Momentum investment strategies, portfolio performance, and herding: A study of mutual fund behavior. The American Economic Review, 85 (5), 1088-1105.

Koutmos, G., & Booth, G. (1995). Asymmetric volatility transmission in international stock Markets. Journal of International Money and Finance, 14 (6), 747-762.

Lakonishok, J., Shleifer, A., & Vishny, R. (1992). The impact of institutional trading on stock Prices. Journal of Financial Economics, 32 (1), 23-43.

Lakshman, M., Basu, S., & Vaidyanathan, R. (2013). Market-wide herding and the impact of institutional investors in the Indian capital market. Journal of Emerging Market Finance, 12 (2), 197-237.

Lao, P., & Singh, H. (2011). Herding behaviour in the Chinese and Indian stock markets. Journal of Asian Economics, 22 (6), 495-506.

Lee, C., Chen, M., & Hsieh, K. (2013). Industry herding and market states: evidence from Chinese stock markets, (Quantitative Finance, 13 (7), 1091-1113.

McQueen, G., Michael, P., & Steven, T. (1996). Delayed reaction to good news and the crossautocorrelation of portfolio returns. The Journal of Finance, 51 (3), 889-919.

Mustapha, C., & Medhioub, I. (2018). Herding behavior in Islamic GCC stock market: A daily analysis. International Journal of Islamic and Middle Eastern Finance and Management, retrieved from https://doi.org/10.1108/IMEFM-08-2017-0220.

Poshakwale, S., & Mandal, A. (2014). Investor behaviour and herding: evidence from the national stock exchange in India. Journal of Emerging Market Finance, 13 (2), 197-216.

Prosad, J., Kapoor, S., & Sengupta, J. (2012). An examination of herd behavior: An empirical evidence from Indian equity market. International Journal of Trade, Economics and Finance, 3 (2), 154-157.

Schwert, G. (1989). Why does stock market volatility change over time?', The Journal of Finance, 44 (5), 1115-1153.

Schwert, G. (1990). Stock volatility and the crash of'87. Review of Financial Studies, 3 (1), 77-102.

Shiller, R., & Pound, J. (1986). Survey evidence on diffusion of interest among institutional investors (No. 794), Cowles Foundation for Research in Economics, Yale University.

Tan, L., Chiang, T., Mason, J.,& Nelling, E. (2008). Herding behavior in Chinese stock markets: An examination of A and B shares. Pacific-Basin Finance Journal, 16 (1), 61-77.

Thaler, R.H. (1987). Anomalies: the January effect. The Journal of Economic Perspectives, 1(1), 197-201.

Wermers, R. (1999). Mutual fund herding and the impact on stock prices. The Journal of Finance, 54 (2), 581-622.

Santosh Kumar

Assistant Professor, School of Management, IMS Unison University, Uttrakhand.
Table--I
Descriptive Statistics on Return on Market

                              Daily Data   Mothly Data

                     Market         CSAD        Market    CSAD

Observation            2612         2612           127     127
No of '-ve' return     1166           NA            51      NA
No of '-ve' return     1446           NA            76      NA
Minimum             -10.483        0.800       -26.910   3.016
Maximum              15.573        9.977        32.730   20.967
Mean                  0.053        2.661         1.156   6.180
Median                0.134        1.701         1.088   3.667
Standard Deviation    1.487        2.147         1.089   4.669

Source: Compiled by Researcher

Table--II
Regression Result for CSADt (as per Eq-4, 5, 6 and 7)
at 99% Significance Level (Sep-2005-Mar-2016)

                        Coefficients      P-value

                        [CSAD.sub.t], as per Eq-4

                        Daily Data

[alpha]                    2.4426031      0
[[beta].sub.1]             1.7534041      3.77E-11
[[beta].sub.2]             2.3841924      2.11E-15

Pre crisis (daily data)

[alpha]                    2.0013166      1.15E-86
[[beta].sub.1]             2.9449119      9.53E-10
[[beta].sub.2]             2.4747473      2.41E-07

Daily Data

[alpha]                    2.3827152      O.OOE+00
[[beta].sub.1]             -0.0444289     0.133180697
[[beta].sub.2]             0.0755519      3.14E-34

Daily Data

[alpha]                    1.790026       1.76E-67
[[beta].sub.1]             0.0848849      0.12024611
[[beta].sub.2]             0.1622721      1.87E-20

[CSAD.sub.t] as per Eq-6

Daily Data

[alpha]                    2.259845       1.63E-108
[[gamma].sup.UP]           0.217534       0.019
[[gamma].sup.UP.sub.2]     0.035593       0.0039

[CSAD.sub.t] as per Eq-7

                           Coefficients   P-value

[alpha]                    2.206631       6.60E-78
[[gamma].sup.down.sub.1]   -0.12212       0.303028
[[gamma].sup.down.sub.2]   0.085757       3.33E-05

                           Coefficients         P-value

                           Monthly Data

[alpha]                     4.838959768        4.80E-17
[[beta].sub.1]              1.464839946       0.600400759
[[beta].sub.2]             -2.307396311       0.473321176

                           During Crisis(daily data)

[alpha]                     2.860465707        2.75E-39
[[beta].sub.1]              2.00158135        0.000110828
[[beta].sub.2]              2.254250743        5.88E-06

                           Monthly Data

[alpha]                     4.853503917        1.30E-15
[[beta].sub.1]              0.073470114        0.2676601
[[beta].sub.2]              -0.00193058        6.23E-01

                           During Crisis(daily data)

[alpha]                     2.701466318        2.64E-38
[[beta].sub.1]              0.001146205       0.984001409
[[beta].sub.2]              0.088415462        1.21E-12

                           Monthly Data

[alpha]                      3.179559          0.007812
[[gamma].sup.UP]             0.424368          0.110863
[[gamma].sup.UP.sub.2]       -0.01279          0.212482

                           Coefficients         P-value

[alpha]                      5.672086          0.000168
[[gamma].sup.down.sub.1]     0.171305           0.64234
[[gamma].sup.down.sub.2]     0.000483          0.976796

Source: Compiled by Researcher
COPYRIGHT 2019 Foundation for Organisational Research & Education
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kumar, Santosh
Publication:Abhigyan
Geographic Code:9INDI
Date:Apr 1, 2019
Words:4213
Previous Article:Exploring Motivational Factors and Brand Equity as Drivers of Luxury Consumer Buying Behavior: An Empirical Study.
Next Article:Risk Perception of Mutual Fund Investors.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters