CROSS-SECTIONAL HERDING BEHAVIOR IN DHAKA STOCK EXCHANGE : A CASE OF BEHAVIORAL FINANCE
DOI:
https://doi.org/10.58964/JBA45N104Keywords:
Dhaka Stock Exchange, Herding Behavior, Behavioral Finance, Equity Return Dispersion, Bangladesh, Cross Sectional Absolute Deviation, Cross Sectional Standard DeviationAbstract
This paper aims at exploring the herding behavior in Dhaka Stock Exchange (DSE) Ltd. Cross Sectional Standard Deviation (CSSD) and Cross Sectional Absolute Deviation(CSAD) approaches proposed by Christie and Huang (1995) and Chang et al. (2000) were undertaken to identify herding behavior or individual irrationality in investment decision making process in DSE. Exploring the models proposed by Christie and Huang (1995) and Chang et al. (2000), the author proposed the use of an alternative non-linear model which can better reflect the market volatility and uncertainty of Bangladesh stock market. Daily closing price data of DSE 30 has been used for the study period of 2013-2023. DSE 30 was used as the market portfolio measure as this index includes the most liquid, active and investable stocks. Daily closing price data of DSE 20 was used for 2007-2012 period as DSE 30 was launched in 2013. The daily closing price data of individual stocks of DSE 30 index were taken after adjusting for right offerings, cash dividends and stock dividends. The test results broadly found herding behavior in different cross sectional analysis and different market conditions in different time phases. Herding behaviors were evident during 2007-2023 period during phases of large price movements, pre-crash period from 2007-2010, stock market crash period from 2010-2011, pre-COVID period from 2012-2019, COVID period from 2020-2021 and post-COVID period from November 2021-June 2023. From the test results, it can also be concluded that herding behavior is more prominent in small and medium capitalized portfolio, high and moderate trading volume scenario and in bullish and bearish market whereas the herding behavior is not evident in large capitalized portfolio, low trading volume scenario. These tests gave better results with the non-linear models as market does not necessarily maintain a linear correlation with individual or portfolio return dispersion. This paper contributes to the behavioral finance literature by applying an alternative non-linear model to detect irrational behavior of investors of emerging and frontier market like Bangladesh.