Detection of bubbles in WTI, brent, and Dubai oil prices: A novel double recursive algorithm

• Published in: Resources policy Vol. 70; p. 101956
• Main Authors: Ajmi, Ahdi Noomen, Hammoudeh, Shawkat, Mokni, Khaled
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.03.2021; Elsevier Science Ltd
• Subjects: Algorithms; Brent oil price; Bubbles; Crude oil; Dubai oil price; Economic activity; Economic conditions; Investors; Markets; Petroleum; Policy making; Price bubbles; Prices; Pricing; WTI oil price; Dubai United Arab Emirates; United Arab Emirates; C10; Price bubbles; C12; C58; Brent oil price; Dubai oil price; Q40; WTI oil price
• ISSN: 0301-4207; 1873-7641
• DOI: 10.1016/j.resourpol.2020.101956

Since the oil crisis of 1973, the evolution of oil prices has been subject to a complicated dynamic that is marked by considerable surges and bursts. This pattern is an expression that oil prices deviate from their fundamentals, thereby forming the so-called speculative bubbles. However, this pattern requires a sophisticated econometric tool to capture its episodes. In this paper, we use the novel double recursive algorithm of Phillips and Shi (2018) to detect and analyze possible occurrences of speculative bubbles in oil prices in three key regional markets, including the European Union (Brent), Asia (Dubai), and the United States (WTI). Using the available monthly data ranging from January 1982 to October 2020, the results suggest two episodes of common bubbles to all oil prices, occurring in July 1986 and March–July 2008. The Dubai oil price is the most affected by the bubbles. Further, political events, oil supply shocks, and global economic activity are the main factors contributing to this bubble behavior. We also provide policy implications of the findings, which should help policymakers make suitable decisions related to monitoring oil price shifts and their possible causes. Besides, by explaining the bubbles’ episodes, investors can be cognizant of any factors that lead to such oil market failures, which should help in avoiding them. •We use the novel double recursive algorithm to detect bubbles in oil prices.•We find two common oil prices bubbles (July 1986 and March–July 2008).•Dubai oil price is the most affected by bubbles.•Geopolitical events and global economic activity are the main drivers of this behavior.