Using daily data from August 9, 2015, to July 7, 2020, this study examines the effects of economic policy uncertainty (EPU) on the returns of four cryptocurrencies: Bitcoin, Ethereum, Litecoin, and Ripple. To this end, two new measures of EPU (Twitter-based economic uncertainty and Twitter-based market uncertainty) are considered. A Granger causality test using the recursive evolving window approach shows a significant causality between the Twitter-based EPU measures and the BTC/USD exchange rate from October 2016 to July 2017. Moreover, a significant causality was noted from the EPU measures to the ETH/USD exchange rate from June 2019 to February 2020 and from the EPU measures to the XRP/USD exchange rate from January 2020 to February 2020. The Twitter-based EPU measures primarily positively affect the returns of the related cryptocurrencies during these periods. These results are robust to different measures of Twitter-based EPU and different econometric techniques. Potential implications, including the COVID-19 era, are also discussed.

本研究使用 2015 年 8 月 9 日至 2020 年 7 月 7 日的每日数据，研究经济政策不确定性 (EPU) 对四种加密货币回报的影响：比特币、以太坊、莱特币和瑞波币。为此，考虑了 EPU 的两个新措施（基于 ​​Twitter 的经济不确定性和基于 Twitter 的市场不确定性）。使用递归演化窗口方法进行的格兰杰因果检验表明，2016 年 10 月至 2017 年 7 月，基于 Twitter 的 EPU 指标与 BTC/USD 汇率之间存在显着的因果关系。此外，从 EPU 指标到 ETH/ 2019 年 6 月至 2020 年 2 月的美元汇率以及从 EPU 措施到 2020 年 1 月至 2020 年 2 月的 XRP/美元汇率。在这些时期，基于 Twitter 的 EPU 措施主要对相关加密货币的回报产生积极影响。这些结果对于基于 Twitter 的 EPU 和不同的计量经济学技术的不同度量是稳健的。还讨论了潜在的影响，包括 COVID-19 时代。