Is there a structural change in the persistence of WTI–Brent oil price spreads in the post-2010 period? ☆ Wei Chen a , Zhuo Huang b, ⁎, Yanping Yi c a The Ohio State University, Department of Agricultural, Environmental, and Development Economics, United States b Peking University, National School of Development, PR China c Shanghai University of Finance and Economics (SUFE), School of Economics, PR China abstract article info Article history: Accepted 16 June 2015 Available online xxxx Keywords: Price spread Structural change CUSUM of squares-based test Unit root test In recent years, WTI oil has traded at a sizable discount against Brent oil, and this divergence has enlarged the price spreads. We investigate whether there has been a structural change in the persistence of WTI–Brent crude oil price spreads in recent years, i.e., a change from a stationary to a non-stationary time series. The CUSUM of the squares-based test of Leybourne et al. (2007b) is performed in which the breakpoint is not pre-specified, and the estimated breakpoint is found to have occurred in 2010. We conduct various unit root tests for the price spreads in two sub-samples, defined as before and after the estimated breakpoint, and find empirical evidence supporting our persistence change hypothesis. Two alternative persistence change tests are also performed to make our conclusion more robust. © 2015 Elsevier B.V. All rights reserved. 1. Introduction Crude oil price fluctuations substantially influence the global economy in many ways. First, high oil prices have a negative effect on economic growth, as documented in many empirical studies (Kilian, 2008; Kilian and Vigfusson, 2011). Inflation pressures induced by posi- tive oil price shocks are more likely to trigger tight monetary policies that curb economic growth. As a consequence, oil prices can be used to predict the economic growth rate (Narayan et al., 2014). Moreover, the oil price is a key signal in business activities and financial markets. Narayan and Sharma (2011) and Phan et al. (2015) find that firm returns across different sectors in the stock market respond significantly to oil price shocks. When the oil price reaches US$100 or more per barrel, the psychological barrier effect lowers firm returns (Narayan and Narayan, 2014). The oil price is also a significant determinant and predictor of firm return volatility (Narayan and Sharma, 2014; Sharma and Narayan, 2012). Investors can make substantial gains by using the oil price to forecast a firm's return volatility. In the international crude oil markets, West Texas Intermediate (WTI) and Brent Crude are the most influential benchmarks for light crude oil in North America and Europe, respectively. Their spot and futures prices are widely used as representative crude oil prices in the petroleum business, financial trading, and policy making. As Fattouh (2011) states, WTI is the basis for pricing oil imports into the U.S., the largest oil consumer in the world. Brent, produced in the North Sea, plays a fundamental role in the pricing of about 70% of the international oil trade. In the decades before 2010, the prices of WTI and Brent moved in tandem, and the WTI–Brent spreads (WTI minus Brent) were seldom wider than the $5 per barrel band. However, since early 2010, WTI has traded at a sizable discount against Brent, and this divergence has enlarged the price spreads, which reached a historical record of -29 dollars per barrel in late September 2011. These unusual and substantial spreads have profound economic implications and have received considerable attention in the media and from policy makers. First, as Brigida (2014) illustrates, enter- prises may rely on global oil prices to guide their petroleum-related business, and governments need representative oil prices to make ener- gy policies. Second, Janzen and Nye (2013) discuss the effect of the spread on oil-producing countries that export oil to the U.S. As the oil from these countries is mainly priced by WTI, the decreasing prices have negative macroeconomic effects on their business investments and fiscal budgets. In addition, Fattouh (2011) notes that the abnormal spreads reveal the deficiency of WTI as a useful international price benchmark, because it no longer reflects the world's supply–demand balance. For these reasons, WTI–Brent spreads are a hot topic in current international oil markets. A number of studies have ascribed the recent oil spreads to supply shocks in the WTI spot market. Fattouh (2011) points out that unlike Brent, which is waterborne and exportable, WTI is easier to dislocate Economic Modelling 50 (2015) 64–71 ☆ Zhuo Huang acknowledges financial support from the Youth Fund of the National Natural Science Foundation of China (71201001) and from the Ministry of Education of China, Humanities and Social Sciences Youth Fund (12YJC790073). Yanping Yi is also affil- iated with the Key Laboratory of Mathematical Economics (SUFE), Ministry of Education; supported by Pujiang Project of Science and Technology Commission of Shanghai Municipality (13PJC048) and Program for Changjiang Scholars and Innovative Research team in SUFE (IRT13077). ⁎ Corresponding author. Tel.: +86 10 62751424. E-mail address: zhuohuang@nsd.pku.edu.cn (Z. Huang). http://dx.doi.org/10.1016/j.econmod.2015.06.007 0264-9993/© 2015 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Economic Modelling journal homepage: www.elsevier.com/locate/ecmod