几种主要加密货币收益率波动之间因果关系的再分析  

A Reanalysis of Causality between Yield Fluctuations of Several Major Cryptocurrencies

作　　者：马瑶 胡玮 高成 

机构地区：[1]湖南师范大学数学与统计学院,湖南 长沙 [2]株洲市芦淞区统计局,湖南 株洲

出　　处：《统计学与应用》2021年第3期529-537,共9页Statistical and Application

摘　　要：以比特币为代表的加密货币是一种与传统货币完全不同的新型货币,因其高风险高收益而被大众所熟知。本文利用格兰杰因果关系定性检验和梁氏信息流因果关系定量分析对五种不同的加密货币之间的关联性进行研究。结果表明:格兰杰因果关系检验只能定性判断加密货币之间是否存在因果关系。而梁氏信息流因果关系定量分析从严格的物理意义上利用信息流方法不仅可以判断不同加密货币之间是否存在因果关系,而且还可以直接定量计算加密货币之间因果关系的大小,是投资者进行投资的重要参考信息。The cryptocurrency represented by Bitcoin is a new type of currency which is completely different from traditional currencies. It is well known by the public for its high risk and high yield. In this paper, Granger causality qualitative test and Liang’s information flow causality quantitative analysis are used to study the correlation between five different cryptocurrencies. The results show that the Granger causality test can only qualitatively determine whether there is a causality between different cryptocurrencies. The quantitative analysis of causality of Liang’s information flow can not only judge whether there is causality between different cryptocurrencies in a strict physical sense, but also directly calculate the magnitude of causality between cryptocurrencies quantitatively, which is an important reference information for investors to make investment.

关 键 词：格兰杰因果关系检验 梁氏因果关系分析 信息流 加密货币 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]