“…The statistical models used in existing empirical literature to examine the safe-haven properties of assets include – but not limited to - quantile regression approach ( Baur and Lucey, 2010 ; Baur and McDermott, 2010 ; Bianchi et al, 2020 ; Śmiech and Papież, 2017 ), bivariate cross-quantilogram framework ( Manohar and Guntur, 2021 ; Shahzad et al, 2019 ; Uddin et al, 2019 ), time-varying Joe–Clayton copula method ( Nguyen and Liu, 2017 ; Tiwari et al, 2020 ), DCC GARCH approach ( Akhtaruzzaman et al, 2020 ; Akkoc and Civcir, 2019 ; Kinateder et al, 2021 ), rolling window approach ( Bouri et al, 2021 ; Bouri et al, 2020 ), and quantile cross-spectral coherency framework ( Le et al, 2021 ; Maghyereh and Abdoh, 2020 ; Naeem et al, 2020 ). However, these approaches can capture only one of the time or frequency dimensions of the data and fail to cover both the time and frequency characteristics of the co-moving time series simultaneously.…”