Decomposing Crude Price Differentials: Domestic Shipping Constraints or the Crude Oil Export Ban?

AuthorAgerton, Mark

    In 1975, United States President Gerald Ford signed the Energy Policy and Conservation Act, which prohibited the export of domestically produced crude oil and created the Strategic Petroleum Reserve. Decades later in the late 2000s and after many years of declining U.S. crude oil production, the combination of horizontal drilling and hydraulic fracturing techniques enabled companies to produce oil and gas from geological formations that had been, heretofore, uneconomic. These technological innovations sparked a production renaissance; by the end of 2014 U.S. production had reached levels not seen since the 1970s.

    As U.S. production rose, price differentials between domestic and international crudes grew to unprecedented levels. At one point, the most widely cited U.S. crude benchmark, West Texas Intermediate (WTI), was trading at more than a $25 discount to the international benchmark, Brent crude (see Figure 2). The large discount spurred a debate over its cause and whether the discount could be eliminated by removing the export ban.

    In December 2015, the export ban was lifted. During the preceding debate, those for and those against lifting the ban tended to associate it with the domestic crude discount. Those against lifting the ban argued that allowing crude exports would cause increases in domestic prices of refined products, such as gasoline, and would reduce the the security of the nation's energy supply. Proponents of lifting the ban refuted these concerns. First, they argued that exporting crude oil would not increase gasoline prices; if anything it would lead to a decrease in gasoline prices (Yergin et al., 2014; Ebinger and Greenley, 2014; Medlock, 2015). (1) Second, they argued that increasing domestic prices to parity with international ones would spur new investment and oil production, creating hundreds of thousands of domestic jobs (Yergin et al., 2014; Ebinger and Greenley, 2014). (2) Our interest is not in the merits of lifting the export ban; rather, we are interested in whether the export ban or an alternative explanation, scarce pipeline capacity, caused the domestic crude oil discounts.

    Yergin et al. (2014) and Ebinger and Greenley (2014) give the following economic argument for why the export ban lowered domestic crude oil prices. Since there is no ban on the import of crude oil--only its export--refineries in the U.S. are able to purchase both foreign and domestic crude. Because domestic and foreign crudes are substitutes in production, the two generally trade at similar prices. After decades of declining domestic crude oil production, domestic refineries had gradually reconfigured themselves to process available imported crudes. Oils from unconventional sources (termed light-tight oils, or LTOs) have a different chemical composition than foreign crudes, and domestic refineries were not optimized to handle the large quantities of LTOs that shale plays were producing. Because of the mismatch between refinery configurations and domestic crude characteristics, refiners could only process LTOs profitably if they could purchase them at a discount. Even though lighter crudes, like those from shale, were trading at a premium to heavier ones in the international market, the export ban eliminated foreign sources of demand: selling to domestic refineries at a substantial discount was the only option. The combination of the export ban and refineries' inability to process the new LTOs caused sustained price differentials. A necessary condition for the export ban to have acted as a binding constraint is that domestic refineries were unable to absorb new sources of domestic crude without significant additional cost.

    An alternative economic explanation, which we argue was more important, has to do with shipping constraints within the U.S.--not the prohibition on exporting the crude outside of the country. In addition to large discounts between domestic and international crudes, the shale boom coincided with large price discounts within the U.S. For a time, new volumes of crude overwhelmed existing pipeline capacity between locations like North Dakota's Bakken formation in the mid-continent and refineries located in the Gulf Coast region. The ensuing excess demand for pipeline capacity within the U.S. created a wedge between Brent and WTI prices. Facing insufficient pipeline capacity, oil producers in the mid-continent had two alternatives: store increasing amounts of crude oil in mid-continent inventories, or resort to more costly transportation modes like rail and barge. Several studies have associated internal shipping constraints with internal price differentials (Upton, 2015; Borenstein and Kellogg, 2014; Kaminski, 2014; Buyuksahin et al., 2013; Fattouh, 2007,2010, 2009; Kao and Wan, 2012). In fact, McRae (2015) argues that vertically integrated ConocoPhillips delayed pipeline expansions for the purpose of sustaining the price differential, thereby improving refinery profits. In contrast to the previous studies, we empirically evaluate the roles of these two possible physical constraints--internal and external--and consider the policy implications.

    The degree to which the WTI-Brent discount was due to a constraint on external trade with other countries (refinery constraints in conjunction with the ban) or internal trade between producing and refining regions (pipeline congestion) is an empirical question. If the constraint was internal, then the opportunity to arbitrage spatial differences in price would have led to new pipeline construction and the elimination of the discount whether or not the export ban was in place. However, if the discount was due to a mis-match of refining capacity with new U.S. crude supplies, then an earlier lifting of the export ban might have raised domestic wellhead prices for oil producers, increasing their profitability and mitigating the extent to which domestic refineries had to make investments or operational changes to handle this new source of crude.


    2.1 Refining and export restrictions

    Crude oil is an intermediate good, and there are two major sources of demand for domestic crude production: refining and export to the world market. Refineries transform crude oil inputs into petroleum product outputs. Global petroleum product prices track international crude oil prices closely because oil is the primary input in the production process. The 1975-2015 export ban implied that domestic refining was the only major source of demand for domestic crude oil.' Domestic refiners were thus the only firms who could freely arbitrage price differences between discounted domestic crude and undiscounted international petroleum product prices. Producers were unable to do so. (4)

    Crude oils are heterogeneous in their chemical compositions, and refineries are fine-tuned to a slate of particular crude oils. Historically, the U.S. has produced "light sweet" crude that has a relatively low density ("light") and relatively low sulfur content ("sweet"). U.S. refineries were originally built to process this domestically produced light sweet crude. Starting in the 1970s, domestic oil production fell and demand for refined products grew. Over time, refineries retooled and adjusted their diets to use a higher share of cheaper "heavy-sour" crude from overseas.

    Refineries had several options to adjust to increased domestic supplies of light sweet crude. As prices of products and particular crude oils change, refiners can, subject to constraints, modify the mix of inputs while maintaining an overall chemical composition. (5) For instance, if light crude is relatively inexpensive, a refinery might purchase more light crude and more heavy crude, causing the refinery to substitute away from a medium grade crude to take advantage of the relatively inexpensive light crude. This mixing gives the refinery flexibility to change its purchases to adapt to changes in relative input prices.

    Refiners can also make physical plant additions and alterations to allow for a different mix of crude to be processed, though these capital expenditures can be expensive. While refiners are unlikely to make significant changes to their equipment and operations in response to a transitory shock, they are able to make significant changes to accommodate structural changes in crude availability. (6)

    The top two panes of Figure 1 show that as domestic oil production increased, both U.S. production and exports of petroleum products increased dramatically. Simultaneously (as shown in the bottom pane), the average API gravity of Gulf Coast refiners' crude inputs (the inverse of crude oil density) increased sharply. This suggests that refiners were either changing their diets or that there was an increased utilization of simple refineries that were already configured to handle these light oils. It also suggests that weak domestic demand and low input prices allowed refiners to sell more petroleum products abroad.

    2.2 Producers and transportation constraints

    U.S. oil production, transport, and refining is reported regionally by Petroleum Administration Defense Districts (PADDs). Much of the oil production and refining demand takes place in PADDs 2, 3, and 4, which are the Midwest, Gulf Coast, and Rocky Mountains, respectively. One of the biggest sources of new shale oil, North Dakota, is in PADD 2. PADD 2 also contains Cushing, Oklahoma, where WTI is traded and priced. Much of the nation's refining lies in PADD 3 along the Gulf Coast.

    As shown in Figure 2, for over a decade, West Texas Intermediate (WTI) and Louisiana Light Sweet (LLS) traded in close proximity to each other (and Brent). However, beginning in the late 2000s, a large price gap emerged. With sufficient transportation infrastructure, a profit-maximizing producer or buyer of crude oil in the mid-continent would have seen an arbitrage opportunity, transport its oil to the Gulf Coast, and sell it...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT