Beschreibung:
Fossil energy consumption, carbon emission and agriculture plants are three critical poles of global environmental protection. Thus, the prices of crude oil, carbon emission allowance and agriculture commodity futures should be closely connected for their inherent fundamental nexus and financialization characters. However, by now little is known about these interdependence (spillover) effects, especially at extreme market conditions across different time frequencies. In this paper, we aim to quantify not only the normal (mean quantile) static and dynamic spillover effects among them at both time and frequency domains, but the more important extreme spillovers spanning various time horizons by using a recently proposed new quantile-frequency spillover approach. The empirical results find that, the total spillover at extreme market situations among oil, carbon and agriculture commodity futures are much larger (over two times) than the one at normal conditions, and the total spillover increase significantly during three crisis periods from 2007 to the present. Second, in general soybean and corn are the most powerful information senders over other futures in time domain, while carbon emission allowance futures is an obvious spillover receiver at both normal and extreme market conditions across various time frequencies. Third, both the total spillover and the net ones are centered at short-term frequency (i.e., 1 to 4 weeks), suggesting that most hedgers/investors have more consistent investment behaviors in short term when trading crude oil, carbon and agriculture commodity futures. Finally, we find that although there is no clear difference in the total spillover between extreme bearish and bullish market environment, crude oil and agriculture commodity futures play significantly distinct roles in net spillover senders/receptors at different time horizons. These findings have valuable implications for policy makers, relevant producers/consumers, as well as futures investors