Herding in the crypto market: a diagnosis of heavy distribution tails

Abstract

PurposeWith the unprecedented growth of digitalization across the globe, a new asset class, that is cryptocurrency, has emerged to attract investors of all stripe. The novelty of this newly emerged asset class has led researchers to gauge anomalous trade patterns and behavioural fallacies in the crypto market. Therefore, the present study aims to examine the herd behaviour in a newly evolved cryptocurrency market during normal, skewed, Bitcoin bubble and COVID-19 phases. It, then, investigates the significance of Bitcoin in driving herding bias in the market. Finally, the study gauges herding contagion between the crypto market and stock markets.Design/methodology/approachThe study employs daily closing prices of cryptocurrencies and relevant stocks of S&P 500 (USA), S&P BSE Sensex (Index) and MERVAL (Argentina) indices for a period spanning from June 2015 to May 2020. Quantile regression specifications of Chang et al.’s (2000) absolute deviation method have been used to locate herding bias. Dummy regression models have also been deployed to examine herd activity during skewed, crises and COVID-19 phases.FindingsThe descriptive statistics reveal that the relevant distributions are leptokurtic, justifying the selection of quantile regression to diagnose tails for herding bias. The empirical results provide robust evidence of crypto herd activity during normal, bullish and high volatility periods. Next, the authors find that the assumptions of traditional financial doctrines hold during the Bitcoin bubble. Further, the study reveals that the recent outbreak of COVID-19 subjects the crypto market to herding activity at quantile (t) = 0.60. Finally, no contagion is observed between cryptocurrency and stock market herding.Practical implicationsDrawing on the empirical findings, it is believed that in this age of digitalization and technological escalation, this new asset class can offer diversification benefits to the investors. Also, the crypto market seems quite immune to behavioural idiosyncrasies during turbulence. This may relieve regulators of the possible instability this market may pose to the entire financial system.Originality/valueThe present study appears to be the first attempt to diagnose leptokurtic tails of relevant distribution for crypto herding in the wake of two remarkable events: the crypto asset bubble (2016–2017) and the outbreak of coronavirus (early 2020).

Behavioural finance and cryptocurrencies

“…Mandaci and Cagli (2022), employing a novel Granger causality methodology with a Fourier approximation, find significant herding behaviour during the COVID-19 outbreak, whilst Lee et al. (2021) and Shrotryia and Kalra (2021), find that herding effect is present, only during down markets in periods of high volatility and up markets in periods of low volatility, and during normal, bullish and high volatility periods accordingly. Finally, in their study, Rubbaniy et al.…”

“…On the other hand, the analysis of Susana et al (2020) revealed that herding was present amongst all cryptocurrencies of the sample in normal conditions, but not during market upswing or downswing. Mandaci and Cagli (2022), employing a novel Granger causality methodology with a Fourier approximation, find significant herding behaviour during the COVID-19 outbreak, whilst Lee et al (2021) and Shrotryia and Kalra (2021), find that herding effect is present, only during down markets in periods of high volatility and up markets in periods of low volatility, and during normal, bullish and high volatility periods accordingly. Finally, in their study, Rubbaniy et al (2020), utilise over 100 cryptocurrencies in order to investigate the phenomenon.…”

“…They also suggest Bitcoin as a "transfer currency," and show experimentally that herding measures focused on such a transfer Main findings Coskun et al (2020) The findings for the entire sample point to anti-herding behaviour. Additionally, during the "up and down" market phases, there was no notable asymmetric behaviour Kurt G€ um€ us ¸et al (2019) The results of the study indicate no evidence of herding behaviour in the cryptocurrency market in both CSSD and CSAD approaches Panel C: Herding during the COVID-19 pandemic Lee et al (2021) Find that herding effect is present, only during down markets in periods of high volatility and up markets in periods of low volatility Mandaci and Cagli (2022) Findings reveal significant herding behaviour during the COVID-19 outbreak Mnif et al (2020) Results show that all, except for Bitcoin, become more efficient in the post-COVID-19 period; that is, less herding behaviour exists after the pandemic Rubbaniy et al (2020) Their empirical results reveal the presence if herding behaviour during extreme market conditions, with significant herding being found during COVID-19 post-lockdowns periods Shrotryia and Kalra (2021) Herding is present during normal, bullish and high volatility periods Susana et al (2020) The analysis revealed that herding was present amongst all cryptocurrencies of the sample in normal conditions but not during market upswing or downswing Yarovaya et al (2021) Findings show that although the COVID-19 pandemic increased the volatility in cryptocurrency markets, it does not seem to cause stronger herding in these markets Table 1. RBF 14,4 currency give a more exact depiction of dispersion in investors' opinions on the cryptocurrency market.…”

Investor behavior and cryptocurrency market bubbles during the COVID-19 pandemic

“…In the same way, Montasser et al (2022) have examined the efficiency of 18 cryptocurrency markets and showed that the COVID-19 pandemic significantly affects their behavior. Several studies have explained this fact through behavioral biases (Shrotryia and Kalra, 2021;Mnif and Jarboui, 2021;Shrotryia and Kalra, 2021). Most studies consider that cryptocurrency prices are fundamentally bubbles, as they have neither an intrinsic value nor provide dividends (Geuder et al, 2019).…”

Do birds of a feather flock together? Evidence from time-varying herding behaviour of bitcoin and foreign exchange majors during Covid-19

“…CSADs have been a popular measure of herding intensity within the cryptocurrency markets. Rubbaniy et al, 2021 , Shrotryia and Kalra, 2021 employ it on daily datasets of 101 and 83 cryptocurrencies respectively, from 2015 to 2020, and reveal the presence of herding during bullish and high-volatility periods. Using CSADs and GARCH on six cryptocurrencies from January 2015 to March 2019, Ballis & Drakos (2020) conclude that herding is present during both up- and down-market days.…”