Crude Oil Price Fluctuation Analysis Under Considering Emergency and Network Search Data

Abstract

Generally, international oil prices fluctuate for a long time, and become more dramatic. Global emergencies are important factors that affect the stability of the oil market. Abnormal fluctuations of oil prices have an essential impact on world economic trends, and these price fluctuations destabilize investment income, bringing risk to producers and consumers alike. The event study method is widely used to understand the phenomenon of oil price fluctuations caused by emergencies. However, this method cannot determine the exact effect of an emergency in terms of the levels of oil prices. This is mainly because the time window can only be determined subjectively after the incident, and the actual influence of the emergency on the oil price cannot be clarified. Therefore, this paper takes network search data into the analysis framework to improve traditional case analysis. According to the latest report from CNNIC, as of December 2016, China had 731 million Internet users-equal to the total population of Europe-the country's Internet penetration rate had reached 53.2%, and there were more than 600 million users of search engines. Search engines record the user's search behavior, which reflects the attention paid by the user to emergencies, thereby offering both a new perspective and a new data base for market economy behavior analysis.

Dynamic Characteristics of Oil Attributes and Their Market Effects

“…Since 2003, the demand for crude oil in developing countries, mainly from BRICs countries of China, India, Brazil, and Russia, has grown rapidly. Among them, China has gradually become the main area of the world for crude oil consumption, which rose from 7.21% in 2003 to 13.88% in 2018 of the total world consumption. With the acceleration of financial innovation, the international crude oil market has experienced the evolution of the spot market, forward market, and future market, which established its financial attribute.…”

“…For example, Wu et al [12] predicted international crude oil market prices by using Google trend and Web text data based on a traditional web model. In addition, Dai et al [13] used network public opinion data on emergencies. Other studies mainly predicted international crude oil market prices from the perspective of sequence decomposition, which includes empirical mode decomposition (EMD), wavelet analysis, spectral decomposition, and variational mode decomposition (VMD) [14][15][16][17][18][19].…”