The aim of this study is to examine the existence of herding behaviour in the cryptocurrency market under uncertainty by employing cross-sectional absolute deviation (CSAD) of returns, ordinary least squares (OLS), generalized autoregressive conditional heteroscedasticity (GARCH) methods and Time-Varying Markov-Switching (TV-MS) model for both overall sample and sub-periods which was determined based on the results of Quandt-Andrews and Bai-Perron breakpoint tests. We utilized the daily data of the 14 leading cryptocurrencies in terms of closing price, market cap and transaction volume. We also used dummy variables to analyze whether or not an asymmetric behaviour occurred during the "up and down" market periods. Our results for the overall sample refer to an anti-herding behaviour in each model. However, the results of the TV-MS model for the 3rd sub-period (2/28/2017-1/16/2018) imply the existence of a herding behaviour in the low volatility regime, an anti-herding behaviour occurred during the high volatility regime and the effect of uncertainty was significant on the anti-herding behaviour. Finally, our results suggest that there was no significant asymmetric behaviour during the "up and down" market periods.
|Journal||Research in International Business and Finance|
|Publication status||Accepted/In press - 18 Jun 2020|