Herding behavior in emerging equity markets during COVID-19 : An investigation of Russian, Taiwanese, and Vietnamese stock markets
Ojala, Axel (2022-12-22)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2022122273381
https://urn.fi/URN:NBN:fi-fe2022122273381
Tiivistelmä
This thesis investigates market-wide herding within Russian, Taiwanese, and Vietnamese stock markets during the COVID-19 pandemic. Moreover, the existence of asymmetric herding and industry-specific herding are also examined in more detail by utilizing OLS regressions. Due to the recency of the pandemic and the inconclusive evidence that has been published within the research field, there exists a clear need for further herding-related studies. Thus, an in-depth examination is conducted for three emerging markets that can be considered to be appealing areas for research. At the time of publication, this thesis is also one of the few academic studies to test how market-wide herding has emerged inside the Russian stock market.
The main methodology for this study is based on regression analysis where stock return dispersions are used to quantify the level of herding. Moreover, herding is measured by utilizing the cross-sectional absolute deviation (CSAD) approach which can be seen to ultimately stem from the studies conducted by Christie and Huang (1995) and Chang et al. (2000). Besides observing herding during the chosen sample period (01.01.2018-06.05.2022), regression tests are also conducted during shorter subperiods as the sample period is divided into three separate time periods: pre-COVID period, outbreak period, and post-COVID period.
The results of this thesis suggest that market-wide herding exists mainly inside Vietnamese stock markets. Surprisingly – and in contrast to numerous previous academic studies – no herding is detected within the Taiwanese stock markets. Inversely, the regression tests imply that Russian and Taiwanese markets have been more prone towards anti-herding behavior during the pandemic time. Based on the results of the empirical part, Russia and Vietnam seem to experience market-wide herding only during down-market days whereas no herding is observed during rising market days in any of the three markets. Finally, industry-specific herding is found to exist only within Vietnamese stock markets.
As the findings of this thesis are considered as a whole, it is justified to state that the observed results are inconclusive to a large extent. Due to the unique market characteristics of the chosen stock markets and their historical tendency for market anomalies, one could have expected more pronounced herding-results. In general, it is reasonable to argue that the current research methods within the research field include several limitations and thus can be seen as a partial reason for the inconclusive evidence that highlights herding-related research. Therefore, it is suggested that future research would concentrate more on the shortcomings of the current measures and steer focus towards the development of new herding-related methodologies.
The main methodology for this study is based on regression analysis where stock return dispersions are used to quantify the level of herding. Moreover, herding is measured by utilizing the cross-sectional absolute deviation (CSAD) approach which can be seen to ultimately stem from the studies conducted by Christie and Huang (1995) and Chang et al. (2000). Besides observing herding during the chosen sample period (01.01.2018-06.05.2022), regression tests are also conducted during shorter subperiods as the sample period is divided into three separate time periods: pre-COVID period, outbreak period, and post-COVID period.
The results of this thesis suggest that market-wide herding exists mainly inside Vietnamese stock markets. Surprisingly – and in contrast to numerous previous academic studies – no herding is detected within the Taiwanese stock markets. Inversely, the regression tests imply that Russian and Taiwanese markets have been more prone towards anti-herding behavior during the pandemic time. Based on the results of the empirical part, Russia and Vietnam seem to experience market-wide herding only during down-market days whereas no herding is observed during rising market days in any of the three markets. Finally, industry-specific herding is found to exist only within Vietnamese stock markets.
As the findings of this thesis are considered as a whole, it is justified to state that the observed results are inconclusive to a large extent. Due to the unique market characteristics of the chosen stock markets and their historical tendency for market anomalies, one could have expected more pronounced herding-results. In general, it is reasonable to argue that the current research methods within the research field include several limitations and thus can be seen as a partial reason for the inconclusive evidence that highlights herding-related research. Therefore, it is suggested that future research would concentrate more on the shortcomings of the current measures and steer focus towards the development of new herding-related methodologies.