Multiscale Tail Risk Interdependence between Precious Metals

This paper investigates extreme risk interdependencies between four precious metal markets in different periods and in different time-horizons. Several wavelet approaches are used for this task - coherence, correlation and cross-correlation. Wavelet coherence shows strong extreme risk connection in the longer time-horizons, particularly between gold and other markets and between platinum and palladium. Wavelet correlations further strengthen wavelet coherence results, but also show that high correlation is present even in the short time-horizons for the gold-silver and platinum-palladium pairs. Wavelet cross-correlations reveal that gold and silver lead platinum and palladium in short term, whereas this situation reverses in the longer time-horizons. This indicates that investors in the bigger markets closely monitor extreme risk developments in the smaller markets in longer time-horizons and take them as a forecast what might happen in the future. On the other hand, bigger markets react faster to global shocks due to higher trading volumes, which is the reason why they lead smaller markets in short term.