Predicting Cryptocurrency Returns Based on Economic Policy Uncertainty: A Multicountry Analysis Using Linear and Quantile-Based Models

Abstract

In this paper, we examine whether the economic policy uncertainty (EPU) index can predict the cryptocurrency returns for countries with the highest number of Bitcoin nodes, which include U.S., Germany, France, Netherlands, Singapore, Canada, the UK, China, Russia, and Japan. To the extent cryptocurrencies are a speculative asset, we expect that an increase in EPU drives the prices of cryptocurrencies below fundamental values due to the flight to quality and that they subsequently correct. Therefore, we hypothesize that EPU is a positive predictor of cryptocurrency returns. Furthermore, we expect that the positive predictability of EPU is stronger in the long run than the short run since mispricing takes time to correct. Using ordinary least squares, multivariate augmented regression, and quantile regression, we find that EPU positively predicts cryptocurrency returns in the short run for subsequent 1-month returns. Moreover, we find stronger predictability of EPU for the subsequent returns over a longer horizon of 6- and 12-month than 1-month, which is again consistent with our hypothesis. Thus, cryptocurrencies might not act as a hedge or safe haven against other financial assets during uncertain times.