Abstract:

The aim of this study is to determine the existence of herding behavior in cryptocurrency returns using the CSAD herding behavior measurement method developed by Chang et al. (2000). In the study, the eight largest cryptocurrencies in terms of market capitalization and the Daily closing data of the MVDA index covering the period of 06.12.2018-11.03.2022 are used to represent the cryptocurrency market. As a result of the investigation, according to both the general market and the up-down market conditions, the estimation findings of the ordinary least squares (OLS) provide evidence for the existence of anti-herding behavior (negative herding) in the cryptocurrency market. On the other hand, while the results of the GARCH model provide evidence supporting the OLS prediction result for the presence of anti-herding behavior in the down market situation, they provide statistically insignificant results in other situations. Obtaining evidence of anti-herding behavior in cryptocurrency markets shows that traders in the relevant market do not have a market-wide consensus, that is, they trade according to their own assessment.

Keywords: