Description

We re-examine the safe haven and hedge properties of two currency classes, distinguished by the regulation and centralization level. While the role of a long-term hedge and safe haven during crisis periods, such as COVID, has been explored in the literature for various assets, this paper fills a significant gap by utilizing intraday transaction data and addresses numerous issues related to integrated covariance estimation. Integrated covariance is estimated using realized covariance, and several estimators have been proposed over the past two decades, with their continuous improvements in balancing between bias and efficiency under conditions of microstructural noise, price jumps, discrete observations, asynchronicity, and irregularity of the data. To confront these challenges, the first methodological part of the paper concentrates on an extensive simulation study to determine a unique estimator that consistently performs best under varying market conditions. The second methodological part conducts an empirical study in re-examining the safe haven and hedge properties of three FX currencies (U.S. Dollar, Swiss Franc, and Japanese Yen) and one cryptocurrency (Bitcoin) against the U.S. equity market (S&P500) in the period from June 2013 to May 2022. Finally, all currencies are ranked by their safe haven and hedge performances when a unique estimator from the simulation study is employed, providing an evidence that Bitcoin, as a decentralized and unregulated currency, acts as a strong safe haven but a weak hedge. Contrary, highly regulated and centralized FX currencies are found to play a both roles with a dominance of Swiss Franc.