- Refiners make money by cracking crude oil throughputs into valued-added products (i.e., yields). Crack spreads are cyclical and volatile.
- Refiners have adapted to margin volatility by engaging in derivatives contracts which off-set short and medium commodity price risks and by investing in assets which are able to process cost-advantaged crudes and optimize yields of higher value products.
- Vertically integrated refiners are further able insulate themselves from commodity risks and exert more pricing power.
- Ceteris parabus, long-term crack spreads will be upheld simply due to the fact that markets tend to value refining assets at or below their replacement costs (RCN).
- Compliance and regulatory measures are a more serious threat to the long-term viability of domestic refiners since they often elicit unintended economic consequences.
The Fundamentals of Refining
Refiners make money by “cracking” unrefined throughput (e.g., crude oil, condensates, natural gas) in produce value-added yields (i.e., products) like gasoline, diesel, and residual fuel oils. The price difference a throughput and its yields is known as the “crack spread”. Since clean energy regulations have made it economic to produce ethanol, many refiners have sought to enter the ethanol business. In ethanol production, the price difference between throughputs and yields is known as the “crush spread”. Since both ends of the refining business deal with commodities, the prices of which no single entity can exert control, profits are subject to sometimes wildly cyclical swings. Since volatile earnings can lead to volatile stock prices, investors are well-advised to inquire as to the differences between cyclical pressures, secular trends, and an erosion of an individual company’s competitiveness. Furthermore, since commodities cycles can take years or even decades to perfect, it is important to take a long-view on the fundamentals.
Crudes are classified and priced by density and sulfur content1. Heavier crude oils contain a higher proportion of long-chain hydrocarbons and, as a result, are more difficult to refine into lighter products such as gasoline. Sweetness is a measure of sulfur content; sweet crudes contain less than 0.7% sulfur. Sour crudes require complex capital equipment to produce value-added products which meet ever more stringent regulatory specifications. Since light and sweet crudes require the least amount processing, they deservedly trade at a premium to heavy and sour crudes. International pricing benchmarks typically use light-sweet crudes, such as Brent and West Texas Intermediate (WTI).
Figure (1): Crude Oil Benchmarks by Density and Sulfur Content
Source: Refining 101: Basics of Refining and Coking
Cracking is the process and breaking down large, unrefined hydrocarbon molecules into smaller, refined products like gasoline, diesel, fuel oils, and other value-added products. Crack spreads are usually cited as A:B:D or A:B:C:D, where A is the number of barrels of crude oil, B is the number of barrels of gasoline, C is the number of barrels of diesel or and D is the number of barrels of distillate (e.g., fuel oil, kerosene); ‘A’ always equals the sum of terms which proceed it. In other words, cracks represent a yield to throughput relationship.
Common crack spreads are 3:2:1 and 6:3:2:1. Although individual refiners differ with respect to the types of throughputs they can handle and the relatives proportions of petroleum products they can produce, common cracks roughly approximate the inherent yields of a moderately complex refinery which throughputs light crudes2. Refining companies often provide more specified indicators margins which focus on regional price structures. Because indicator margins lead financial statement data, it is possible to anticipate refiners’ profit margins through monitoring real-time commodity price data. Figures (2) and (3), below, exhibit the two most common cracks; this data is readily derived using data from the Energy Information Agency’s (EIA) webpage for commodity spot price data.
Figure (2): New York Harbor Crack Spread
Source: EIA Spot Price Data
Note to Figure (2): The New York Harbor crack assumes Europe Brent feedstock; New York Harbor Conventional Gasoline, New York Harbor Ultra-Low Sulfur No 2 Diesel, and New York Harbor No. 2 Heating Oil throughputs.
Figure (3): Gulf Coast Crack Spread
Source: EIA Spot Price Data
Note to Figure (3): The Gulf Coast Crack assumes 50% Cushing Hub Oklahoma WTI and 50% Europe Brent feedstocks; U.S. Gulf Coast Conventional Gasoline, U.S. Gulf Coast Ultra-Low Sulfur No 2 Diesel, and U.S. Gulf Coast Kerosene-Type Jet Fuel throughputs.
A brief visual inspection of the long-term crack spreads reveals that they are extremely volatile and cyclical. No one can reliability predict underlying commodities or their price differentials. For this reason, managements of refining companies typically assume 5 to 7 year commodities cycles when making capital budgeting decisions. With light-sweet oil prices currently hovering around the $45-50/bbl handle, just as recently as last June, BofA analysts had expressed confidence that Brent would continue to trade around $100/bbl3. Last September, EIA expressed with 95% confidence that NYMEX WTI would stay above $80/bbl through January 20154. Last October, Goldman Sachs analysts predicted that WTI would stay above $90 and Brent above $100/bbl for the following three months5. The list of failed predictions goes on ad nauseam. Rather than attempt to make investment decisions by predicting crack spreads, investors are better served by identifying refiners which are able to more readily adapt to varying market conditions.
Although refiners’ profit margins are extremely levered to commodity price differentials, refiners can hedge away a lot of the short and mid-term risk through engaging in derivatives contracts which offset adverse commodity price movements. Crack and crush spreads are very popular among traders who construct portfolios which mimic very specific throughputs and yields. Marginal refiners, especially those whose competitiveness rests on temporary pricing advantages, often fail to hedge. As a result, these marginal players will tend shake out if/when squeezed. In these regards, the refining industry is no different than any other.
Refiners have further adapted to margin volatility by investing in complex refining assets which are able to optimize yields on a given type of throughput. A refinery using less complex processes will be constrained in its ability to optimize its mix of crude supply and refined products. More complex refining assets will generally be able to produce more stable and predictable margins since they are able to more readily adjust throughputs to take advantage of variable cost structures by maximizing the yields on varying grade of crudes. Highly specialized refining assets, which combine several stages of thermal and catalytic processes, are able to produce significant quantities of high-value specialty petrochemicals.
The optimal complexity of a given refinery is a function of its optimal product mix (i.e., demand) and access to unrefined throughputs (i.e., supply). More complex refineries do not always result in higher returns on capital. For example, although highly sophisticated refining assets which are designed to process heavy crudes can also typically process cost-advantaged light crudes, this would be an inefficient use of capital. As a result of the recent North American shale oil glut (which tends to be light and sweet), many refiners and mid-stream companies have invested heavily in relatively simple topping and distillation units to take advantage of cost-advantaged crude.
Key measures of complexity include Equivalent Distillation Capacity (EDC) and Nelson’s Complexity Index (NCI). EDC, sometimes referred to as complexity-barrels, approximates the potential valued-added capacity of a given refinery. EDC uses factors from NCI to roughly quantify the construction costs (i.e., asset replacement values) of a refiner’s assets relative to the cost of a basic atmospheric distillation unit6 7 8. While the exact calculation of EDC can vary, in aggregate it correlates very strongly to the replacement cost (RCN) of a refiner’s core assets (i.e., PP&E). NCI, the ratio of EDC to basic atmospheric distillation capacity, ranks refineries by relative complexity. Simple refining assets, such as topping facilities which primarily rely on light crudes and an ultra-light form of crude called condensate, will have NCIs of between 1 and 5. Specialized petrochemical facilities can have NCIs above 15. According to data from the Oil & Gas Journal’s 2015 Worldwide Refining Survey and the EIA’s 2014 Refining Capacity Report, U.S.-based refining assets are among the world’s best (U.S.-based refineries had an NCI of 11.85; global refineries ex U.S. had an NCI of 6.22)i.
Construction of new and maintenance of existing refineries is incredibly capital intensive. Many refining companies historically and presently trade below their RCNs, which discourages new constructionii 9. In addition to requiring large upfront investments, refineries require significant amounts of recurrent capital spending, consume large amounts of precious metallic catalysts, and regularly shutdown to perform planned maintenance and upgrade cycles that the industry calls turnarounds. The frequency and timing of turnarounds in part defines how closely actual production capacity compares to full capacity.
Vertically integrated refiners further insulate themselves from commodity price risks. Gasoline and diesel fuels originate from intermingled pipelines; there is no indication that a branded station gets its fuel from a company refinery10. Even though gasoline is a fungible commodity, premium brands can charge more per gallon due to the perception of brand value which is driven mainly by patented detergents (i.e. Chevron with Techron®) which are blended at the rack. EPA regulations now mandate certain detergents, so now, more than ever, gasoline is just gasoline11. Even so, refiners which can sell fuel at retail vice wholesales prices exert better pricing power and will tend to produce more sustainable margins. Furthermore, refiners with integrated upstream and midstream operations are able to more reliably source cost-advantaged crude.
Refiners which have vertically integrated chemical manufacturing operations are further able to exert pricing power. Highly specialized refining equipment and chemical processes are able maximize yields on high-value olefins, oxygenates, aromatics, and various other types of petrochemicals which are used in plastics, food additives, textiles, cosmetics, and more (the chemistry of hydrocarbons is truly wondrous). Although yields of higher-valued petrochemicals is related to refining complexity (i.e., NCI), at some point down the value chain, petrochemical operations stop being classified as commodity oil and gas companies. Sometimes the differentiation between commodity and proprietary is simply a matter of perception and messaging. We all assume that gas prices react to crude oil prices, but who would ever expect the same for bottled water or makeup? All that said, refiners with complex assets and proprietary processes which allow for further reach down the petrochemical value chain will have more entrenched pricing power.
Figure (4): Sample Petrochemical Value Chain for Ethylenes
source: NovaChem, Joffre Expansion: What We Make
Over the short and medium terms, refining margins could come under pressure due to Brent’s unsustainable premium to WTI. Historically, Brent actually traded at a slight discount to WTI. However, the combination of the crude oil export ban enacted in 1975 and expanded domestic production has resulted in a dislocation between Brent and WTI beginning in 2010. Access to cost-advantaged domestic crude oil benchmarked to WTI has allowed previously marginal refiners with relatively simple assets to purchase easily refined light crude at domestic prices and sell refined products at international prices. This structural advantage is unlikely to persist. Lower oil prices discourage new drilling projects and output of shale-wells typically fall by 60-70% within the first year12. Regulators are also playing with the notion of rolling back the export ban. However it plays outs, as the market for North American crude rationalizes itself, the Brent-WTI premium is likely to further collapse. Refiners with complex assets and well-rationalized logistical infrastructure will adapt; some refiners, on the other hand, will struggle.
Figure (5): Brent to WTI Price Differential
Source: EIA Spot Price Data
Over the long-run, the replacement costs of refining assets support normalized (i.e., cyclically-adjusted) crack spreads. Even though refining assets are long-lived, traders highly regard short-term profits. Furthermore, since refiner’s short-term profit margins are extremely susceptible to perturbations in the crack spread, stock returns of refining companies tend to move in tandem with three-month crack spreads. However, because markets systemically undervalue refining assets, refining companies are incentivized to invest in other revenues streams and even use stock buybacks as a cost-advantaged way to invest in capacity. Limited new domestic capacity coming online means that existing refineries will retain whatever (little) pricing power they have and that refineries will be operating at heightened utilization rates for an extended period. In short, market valuations of refining assets below their RCNs actually supports refiner’s margins going forward. In the absence of another market shock, complex U.S.-based refiners with rational supply chains should continue to produce long-term per-barrel throughput margins of between $8 to $11; regional variations do exist, of course.
Refiners will continue to see cyclical and seasonal volatility in their margins. This is not new information. Personally, I find regulatory and compliance pressures more worrisome. For example, specifications changes from 2003 to 2006 have progressively lowered the maximum sulfur content in gasoline from 300 ppm in 2003 to 30 ppm in 200613. In order to remain compliant, refiners had to invest in more complex capital equipment which is also more expensive to operate and maintain. Marginal refiners who were unable to keep invest in more complex assets were forced to either quit or consolidateiii. The introduction of biofuel credits (primarily Renewable Identification Numbers (RINs), needed to comply with the U.S. federal Renewable Fuel Standard (RFS)), so called “cap-and-trade”, also weigh heavily on margins (they are typically included in the Cost of Goods Sold). Moreover, capital expenditures related to regulatory compliance are also rapidly rising.
While some statutory laws are needed to curb excesses, regulations which attempt to alter economic reality (e.g., cap-and-trade, artificially high demand from biofuel requirements, artificially cheap supply from corn subsidies, bio-fuel blender’s tax credits), whether they help or hurt corporate profits, create economic distortions which often result in unintended consequences that just shift the true economic costs elsewhere (which, in turn, somehow justifies new sets of regulations)14. Over the long-term, regulatory and compliance measures are likely to present greater headwinds than cyclical margin hiccups.
i. ^ NCIs of global (ex-U.S.) based refiners are derived from Oil & Journal’s 2015 Global Refining Survey. NCIs of U.S.-based refineries are derived from EIA’s 2014 Refinery Capacity Report.
ii. ^ The U.S.’s newest complex refinery in the U.S. was built in 1976 in Garyville, Louisiana. Although newer refineries have since been built, they have all been simple facilities (i.e., topping and/or vacuum distillation)15. New refinery construction also mirrors U.S. peak oil production in 197016 17.
iii. ^ According to Portfolio123’s data holdings, there were 86 publicly traded companies reporting under GICS Code 10102030 (Oil & Gas Refining & Marketing) in April 2006. One year later, there were only 36 companies reporting under the same GICS Code. Similarly, according to EIA’s report, “Number and Capacity of Petroleum Refineries”, there were 301 operational refineries in the U.S. in 1982; that number has dropped to 182 in 201418.
Footnotes [ + ]