More to cryptos than bitcoin: A GARCH modelling of heterogeneous cryptocurrencies

【Author】 Fung, Kennard; Jeong, Jiin; Pereira, Javier

【Source】FINANCE RESEARCH LETTERS

【影响因子】9.848

【Abstract】This paper explores the risk and return characteristics of a large and diverse cross-section of 254 cryptocurrencies that differ in traded volume and main usage. First, we find long memory, volatility clustering, heavy tails, and negative leverage effects to be common features of cryptocurrencies' return behavior. Second, GARCH models accounting for these features provide the best goodness-of-fit properties. About 80% of cryptocurrencies are well described by Student's t (stud) GARCH specifications with the TGARCH-stud chosen for about 20% of the sample. We then compare out-of-sample 1%-Value-at-Risk (VaR) forecasts under 48 specifications using standard backtesting procedures. Heavy-tailed VaR specifications outperform all normally distributed alternatives. Throughout the analysis, differences emerge when results are broken down by traded volume and usage categories. Overall, our findings have important implications for investors, policymakers, and regulators for the understanding and measuring of market risk in the cryptocurrency market.

【Keywords】Cryptocurrency; Volatility; GARCH; Value-at-Risk

【发表时间】2022 JUN

【收录时间】2022-07-10

【文献类型】实证性文章

【主题类别】

区块链治理-市场治理-市场分析