Herding Behavior in Cryptocurrency Market:Analyzing the Influence of Portfolio Composition and Bitcoin

Abstract

This thesis explores the phenomenon of herding behavior in the cryptocurrency market, where investors often mimic their peers instead of making independent decisions. The study focuses on the impact of herding during significant market fluctuations, particularly during the recent bull market in cryptocurrencies. It also examines how different market portfolio compositions, including the role of Bitcoin, influence this behavior. Despite the growing interest in cryptocur- rencies, there remains a notable gap in the literature regarding the complexities of herding be- havior within this volatile asset class. The study employs the Cross-Sectional Standard Deviation (CSSD) by Christie and Huang (1995) and Cross-Sectional Absolute Deviation (CSAD) by Chang et al. (2000) to quantify herding behav- ior across various cryptocurrencies, employing a quantitative approach. The sample period con- sists of two subperiods: a stable market from June 20, 2022, to October 15, 2023, and a signifi- cant bull market from October 16, 2023, to November 11, 2024. The hypotheses are grounded in psychological insights, which suggest that herding behavior is exacerbated during periods of high volatility. The results suggest that the herding behavior differed between the two subperiods. In the initial subperiod, “rational” investor behavior was observed in both the CSSD and CSAD methods, in- dicating a lack of herding. Conversely, the second subperiod observed a substantial negative coefficient in the CSSD model for the market cap-weighted portfolio that excluded Bitcoin. This coefficient suggested that herding behavior became apparent, particularly among altcoins, as market sentiment improved. The analysis demonstrated that the herding measurements were significantly influenced by the compositions of the market portfolios and Bitcoin’s impact. The discussion underscores the inconclusive nature of the findings, and the unique features of the cryptocurrency markets examined. It also addresses the limitations of the methodology and data, such as the absence of investor sentiment factors. This analysis supports the acknowledg- ment of the complexities of herding behavior and emphasizes the necessity of future research to further investigate these dynamics, particularly the influence of investor sentiment on herd- ing behavior and the evolving nature of investor actions in the cryptocurrency landscape.