Testing the Safe-Haven Properties of Green Bonds, Gold, and Bitcoin for Traditional Bonds: A Wavelet Quantile Correlation Approach

Abstract

In this paper, the safe-haven attributes of green bonds, gold, and bitcoin are compared to those of traditional bonds under various time periods and quantiles by using the WQC methodology. The results indicate that green bonds exhibited a stable safe-haven function at longer time horizons during the full sample period, whereas other assets did not have safe-haven features. During the COVID-19 pandemic, bitcoin exhibited safe-haven attributes at all time horizons, whereas gold demonstrated these characteristics over the short and medium terms. In the sample period during the Russia–Ukraine war, green bonds had strong safe-haven properties at shorter and middle time horizons, whereas bitcoin had these properties at longer time spans. In this paper, a multivariate network framework that includes green bonds, gold, and bitcoin is constructed, and the theoretical foundations that influence the safe-haven attributes of assets are detailed. In addition, this study clearly presents the safe-haven effects of assets under various sample periods, time horizons, and quantiles, thereby bridging the gap of existing studies that ignore time frequency. Thus, this paper provides advice for investors, regulators, and policy-makers, such as choosing portfolios on the basis of asset characteristics, monitoring asset disclosure, and encouraging the trading of safe-haven assets.