Identifying the peak point of systemic risk in international crude oil importing trade. (1st June 2019)

• Record Type: Journal Article
• Authors: Du, Ruijin; Dong, Gaogao; Tian, Lixin; Wang, Yougui; Zhao, Longfeng; Zhang, Xin; Vilela, André L.M.; Stanley, H. Eugene
• Is Part Of: Energy. Volume 176(2019)
• Page Start: 281
• Page End: 291
• Publication Date: 2019-06-01
• Subjects: Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2019.03.127
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: The fluctuations of international crude oil markets have caused significant attention around the world and aroused strong interest in the forecasting of the systemic risk in crude oil trade. Based on the oil imported values data of 34 major oil-importing countries from January 2005 to June 2017, we calculate the cross-correlation functions of time lags and construct a sequence of time-evolving oil import correlation networks according to the similarities between countries. The probability distribution of time lag shows that the time lag effect is not sensitive to positive correlations, but obvious for negative correlations. There is a longer time-lag effect in the years when positive correlations are stronger. Further, we use a percolation analysis to quantify the structural change in the correlation network. The key result is that abrupt percolation transition is leading spikes in systemic risk with advance of 3–11 months suggesting that this event could function as an alarm. Therefore, percolation transition in the correlation network of oil-importing countries can be used as a means to estimate signals about future systemic risk. The methodology and results presented in this paper bring a fresh perspective to the study of systemic risk in crude oil importing trade, and they facilitate risk early-warning research in other energy systems that also have interactions among their elements. Highlights: A sequence of time-evolving oil import correlation networks is constructed. The time lag effect is obvious for negative correlations. Abrupt transition is leading spikes in systemic risk with advance of 3–11 months. Percolation analysis can be used to estimate signals about future systemic risk.