By Allen Good
We think wide Gulf Coast differentials are sustainable through 2014 and could move much higher as an additional 1.4 million barrels per day of pipeline capacity begins delivering crude to the Gulf Coast. Much of the additional supply is likely to be light crude, for which there is little incremental demand. Considering this, combined with an inability to export excess supply, a build in inventory is likely to occur. Our analysis indicates that inventory levels could breach the November 2013 peak, when Light Louisiana Sweet/Brent widened to $14/bbl, by March and again later in the year. As a result, Gulf Coast light crude differentials could soar beyond the market's current expectation of $8/bbl, much like the West Texas Intermediate/Brent differential did previously. In this scenario, Marathon Petroleum (NYSE:MPC) and Valero (NYSE:VLO) are best-positioned to benefit. While trading near our fair value estimates, neither stock is currently pricing in a blowout in Gulf Coast differentials. Tesoro (NYSE:TSO) remains our preferred refiner, as it trades at the greatest discount to our fair value estimate and holds the greatest potential for earnings improvement under current market conditions.
Why Did Gulf Coast Differentials Widen in the Fourth Quarter?
Gulf Coast light and heavy crude differentials widened substantially in the fourth quarter as low capacity utilization and a heavy turnaround schedule created a build in inventory. Turnaround activity affected approximately 700 mb/d of capacity and drove utilization down to a low of 87% in October and 91% for the fourth quarter. While in line with historical averages for that time of the year, the abundance of domestic crude and already-reduced levels of imports resulted in a spike in inventories. As a result, LLS prices averaged almost $8/bbl less than Brent prices during the fourth quarter after trading at a premium for the first three quarters of the year. Further benefiting margins was the concurrent drawdown in product inventories, which resulted in strong product margins. As a result, Gulf Coast refiners processing LLS at times realized higher margins than Mid-Continent refiners processing WTI.
Late December brought higher utilizations as refineries resumed operations after turnarounds and companies sought to reduce their crude oil inventory for tax reasons. The favorable market conditions during the quarter led to a relatively strong round of earnings reports for Gulf Coast refiners such as Valero and Marathon Petroleum.
Why Differentials Could Widen Further in 2014
The more important question is what will happen to Gulf Coast differentials in 2014. We expect similar conditions in 2014, albeit with continued volatility, to those we saw in late 2013, including the potential for much wider light crude differentials. Our conclusion is based on increased pipeline capacity delivering crude to the Gulf Coast, new supply from pipeline additions being primarily light crude, the relatively low level of imports that can be displaced to accommodate new supply, and the inability to divert crude to other regions.
Increased Pipeline Capacity
In 2014, an additional 1.4 mmb/d of pipeline capacity will begin delivering crude to the Gulf Coast. The third phase of the Seaway reversal and the opening of TransCanada's (NYSE:TRP) Gulf Coast Connector pipeline collectively will add about 1.15 mmb/d. Other pipelines running from Texas will add another 290 mb/d. These additions come after 1.4 mmb/d was already added in 2012-13. The amount of pipeline capacity opening to the Gulf Coast suggests that inventories could soar as they did in the fourth quarter. Timing remains a question, however.
The flooding of the Gulf Coast region may not occur until later in 2014, as a lack of crude may result in pipelines running below capacity until the second half of the year. The Gulf Coast Connector officially started deliveries Jan. 21, but crude to fill the pipeline may prove elusive. With Cushing inventories no longer at record levels and the Keystone XL pipeline not in place, the Gulf Coast Connector may not have the necessary supply to run at its full 700 mb/d capacity. Crude availability should be less of an issue later in the year for the third phase of Seaway. It is scheduled to begin operating in mid-2014 and will coincide with the addition of the Flanagan South pipeline, which will bring an additional 585 mb/d of Canadian crude from Chicago to Cushing. Excess volumes could go toward filling the Gulf Coast Connector.
Pipelines Will Be Carrying Light Crude
The new pipeline capacity will largely be transporting light crude, by our estimates. The two largest pipeline additions - Gulf Coast Connector and Seaway - will probably carry a mix of light and heavy. It's more likely that the Gulf Coast Connector will carry a greater amount of light crude, given the absence of heavy volumes from Keystone XL. As such, it will have to rely more on excess Cushing inventory and Bakken supplies. The Seaway expansion will have greater access to Canadian heavy crude thanks to the Flanagan South pipeline. The rest of the pipeline capacity, from Texas, will primarily carry light crude.
Also, incremental crude production growth this year in the Lower 48 and Canada will primarily be light oil, leaving little heavy oil to fill the new pipelines. In 2014, Canadian heavy oil production will increase only about 100 m/bd. In contrast, the Energy Information Administration expects U.S. crude production to increase 1 mmb/d in 2014, primarily from the Bakken, Permian and Eagle Ford, which all produce light crude.
Mid-Continent refiners don't appear to be overwhelmed with Canadian heavy crude either, removing a potential source of heavy crude to fill pipelines headed for the Gulf Coast. Through October 2013, PADD 2 and 4 imports of Canadian heavy crude totaled 1.3 mmb/d compared with heavy crude refining capacity of approximately 1.1 mb/d, suggesting little excess supply available for PADD 3. Assuming that the incremental 100 mb/d of Canadian heavy production and the 200 mb/d of excess heavy crude already flowing to PADD 2 move on the new 1.4 mmb/d of pipeline capacity, then the Gulf Coast could receive incremental light crude supply of 1.1 mb/d by the end of 2014.
Imports Suggest Little Ability to Absorb Additional Light Crude Supply
That much additional light crude is likely to result in a supply/demand imbalance along the Gulf Coast. Because refining capacity for light and medium crude is fixed at about 5mmb/d, there is no incremental demand for light crude currently. With domestic light crude already making its way to the Gulf Coast, refiners have actively been displacing higher-cost light crude imports. Light crude imports averaged 740 mb/d in 2013 through October compared with 1.3 mmb/d over the same period in 2012. October imports fell to 525 mb/d and just 150 m b/d for crude lighter than 35 degrees. Consequently, Gulf Coast refiners have very little ability to further reduce imports to accommodate additional domestic light crude supplies.
Furthermore, based on conversations with industry sources, Saudi Arabia has signaled a commitment to maintain its U.S. market share despite having to discount barrels to match lower U.S. prices. EIA data suggests this is the case, as Saudi Arabia supplied 75% of the light crude imports in October 2013 and its absolute volume was at the same level it was the year before, at about 400 mb/d. As a result, the ability of Gulf Coast refiners to reduce imports further may be negligible, adding to the potential for rising inventory.
We also estimate that refiners have little additional ability to process light crude optimally with current equipment, after striving to maximize light crude throughput to capture the wide WTI/Brent differential of the past three years. Some refiners are investing in equipment to process additional light crude, but the amounts are marginal compared with the amount of potential supply. Also, many of these investments will not be in service until 2015. Refiners could increase light crude processing beyond normally optimal levels, but they would need incredibly wide margins to do so in order to offset the lost efficiency. This would also lead to greater gasoline production, risking oversupply and depressing already weak gasoline margins.
Inability to Divert
Outside the Gulf Coast, domestic light crude can still displace imports on the East and West coasts. Refineries on both coasts continue to import light crude, with imports of 315 mb/d and 415 mb/d, respectively, in October. Transportation remains a hurdle, however. Shipment from the Gulf Coast to the East Coast relies on Jones Act vessels, which remain in short supply. Combining this with volatile spreads, movements have been sporadic. Also, many of the available vessels are better used moving Eagle Ford crude from Corpus Christi to the eastern Gulf. Shipments to Canada are possible and do not require a Jones Act vessel. However, volumes remain low.
Because of this, rail is the only real alternative to transport Bakken and Canadian crude to the East Coast (U.S. and Canada). Refiners and third parties have already been using rail, resulting in PADD 1 imports falling 500 mb/d over the past two years. The ongoing expansion of crude by rail receiving capacity in the region is likely to lead to eventual displacement of the remaining light crude imports, but that's unlikely until adequate export and unloading capacity is built beyond 2014.
The West Coast receives supplies of Bakken and Canadian heavy crude solely by rail. Currently there is about 200 mb/d of rail unloading capacity on the West Coast, which is expected to grow to 900 mb/d by 2016. However, that does little to route light crude away from the Gulf Coast in 2014. Also, when the new capacity opens, refiners will be just as interested in sourcing Canadian heavy crude as Bakken light crude. As a result, we see limited options for crude to move to either coast in substantially greater volumes in 2014, putting further pressure on the Gulf Coast.
Differential and Inventory Outlook
Forecasting near-term differentials can prove difficult as demonstrated by the volatile fourth quarter. However, we think inventory levels can offer insight as to the timing and magnitude of widening differentials. Our model of Gulf Coast inventory levels based on the aforementioned factors suggests that during the early and later part of the year, differentials could be their widest, with the likelihood of sustained wide differentials throughout 2014.
Our analysis indicates that inventory levels could breach November 2013 peaks, when LLS/Brent widened to $14/bbl, by March as a result of the ramp of the Gulf Coast Connector and seasonally low refinery utilization compounded by turnarounds at several facilities. Inventory levels should moderate into the summer as refinery utilization increases. However, by the end of 2014, inventory could swell as additional pipeline capacity begins operating in the second half of the year and refinery utilization falls. Our forecast is highly sensitive to import volumes, pipeline operating rates and refinery capacity utilization. However, we think the relative levels building in early spring and moderating through the summer before building at year-end are accurate.
We think this scenario is supportive of wide LLS/Brent differentials throughout the entire year, with narrowing in the summer, but with LLS remaining at a discount. Also, our forecast assumes light crude imports falling to near zero through the course of the year. If imports continue, then inventory may be higher than our forecast, pushing differentials wider.
What Refiner Is Best-Positioned to Benefit From Widening Gulf Coast Differentials?
Any refiner with Gulf Coast exposure should benefit from a blowout of the LLS/Brent differential. However, relative positioning in the Gulf Coast, light crude throughput and distillate yields are also important. Based on these factors, Marathon is the best-positioned, given its concentrated portfolio and lighter crude slate, followed by Valero, which has greater absolute Gulf Coast exposure, particularly to the western gulf, and higher distillate yields.
Gulf Coast exposure. The build of light crude inventory in the Gulf Coast should result in the widening not only of the LLS/Brent differential but also of the WTI/Brent differential. In this case, both Mid-Continent and Gulf Coast refineries should benefit. However, Mid-Continent product markets remain susceptible to oversupply and weak margins, as evidenced in November when Gulf Coast LLS margins exceeded Mid-Continent WTI margins. Gulf Coast refiners also run the risk of weak gasoline margins due to oversupply if light crude processing rises, but they have the ability to ship product inland or internationally. As a result, we favor the Gulf Coast. Marathon, Valero, and Phillips 66 (NYSE:PSX) all have relatively high levels of Gulf Coast exposure, but Marathon has a portfolio concentrated in the two regions set to benefit the most.
Western Gulf Coast exposure. We expect the entire Gulf Coast region to benefit from a widening LLS differential. However, West Gulf Coast refiners are likely to see additional benefit. The aforementioned pipeline additions will deliver crude to locations in Texas, creating the potential for a glut of crude to form in the western Gulf. The final expansion of Shell's (NYSE:RDS.A) (NYSE:RDS.B) Ho-Ho pipeline to 375 mb/d in early 2014 along with potential rail routes directly to refineries in Louisiana probably will prevent a glut in the west. However, even if crude flows freely to the east, refineries in Texas will still enjoy a transportation differential benefit of $2-$3/bbl depending on their exact location. Valero holds the superior position with nearly 75% of its Gulf Coast refining capacity in Texas.
Distillate production. Weak demand and high inventories weighed on gasoline margins last year. Distillate margins, on the other hand, remained strong, averaging nearly $10/bbl higher than gasoline margins. We expect 2014 margins to be similar. With refiners processing greater amounts of light crude and with demand for gasoline tepid, oversupply of gasoline remains a threat. Meanwhile, distillate domestic demand and exports are growing, alleviating pressure on U.S. supply. Gasoline exports remain relatively flat, despite U.S. refiners tapping new markets. Phillips 66 has superior distillate yields to Marathon and Valero, leaving it better positioned. However, we think the last two refiners' Gulf Coast exposure compensates for their lower distillate yields.
Light crude throughput. In the fourth quarter, both heavy and light Gulf Coast differentials widened relative to international benchmarks. While we expect a similar widening of light differentials in 2014, heavy differentials may not behave similarly, given continued reliance on imports. That said, refiners able to process heavy crude will still benefit from quality differentials and greater distillate yields. Wider light crude differentials, though, will benefit refiners processing greater amounts of light crude. Between the two Gulf Coast leveraged refiners, Marathon has the lighter crude slate, as indicated by relative coking capacity and lower complexity (11.5) compared with Valero (14.2) and Phillips 66 (12.1).
What LLS/Brent Differential Is Priced Into Refiners' Shares?
The unexpected and quick widening of Gulf Coast differentials late last year quickly translated into share price gains as the market awoke to the possibility of a potential Gulf Coast differential blowout in 2014. However, our estimates suggest that a near-term blowout in differentials is not priced in, despite an approximate 30% appreciation of refiners' shares from October lows.
Most refiners, with the exception of Tesoro, are close to or above our fair value estimates, which are primarily derived from discounted cash flow models incorporating our long-term outlook for differentials. As a result, those shares above our fair value estimates are implicitly pricing in long-term differentials above $4/bbl LLS/Brent and $10/bbl WTI/Brent.
However, using DCF valuation for refiners must be weighed against current market expectations. Refiners continue to trade on a short-term outlook of six months to a year. Using a shorter-term lens, refiners' valuations do not look terribly expensive and are certainly not reflecting a differential blowout.
Most of our 2014 EBITDA estimates, which incorporate an LLS/Brent differential of $8/bbl and WTI/Brent differential of $12/bbl based on forward curves, are in line with consensus. Using our estimates, most refiners look fairly valued on a relative basis with almost every refiner stock near its historical median EV/EBITDA multiple. At the same time, it implies further share appreciation is possible if differentials widen beyond or even remain at current levels.
Neither our estimates nor consensus estimates include the impact of a potential widening or blowout in light crude differentials along the Gulf Coast. Share prices do not currently reflect a blowout either. In a case where the LLS differential moves to $18/bbl and the WTI differential to $22/bbl, Marathon and Valero would be the biggest beneficiaries, as the factors we discussed above would translate to sharply increased earnings. Conversely, they would also be the biggest losers if differentials narrowed materially.
Why We Like Tesoro
Marathon and Valero will be the primary beneficiaries in the event of a blowout in Gulf Coast and Mid-Continent differentials, and we see that as a likely event in 2014. However, a blowout in differentials would probably be required for the shares to move much higher. Also, we view such differentials as unsustainable over the long term. Any share price appreciation for Marathon and Valero from current levels would primarily be a result of short-term market forces and likely to reverse as differentials tighten back over time.
Tesoro will benefit less from such a blowout, but it remains our favorite refiner for the following reasons.
Valuation. Tesoro is trading at the greatest discount to our fair value estimate, which is based on midcycle conditions implying that future share price appreciation is possible without improvement in market conditions. Additionally, our valuations are more comprehensive than a forward multiples approach. They include the value of improvements in competitive position as well as ownership (and increases in value) of its limited partnership and general partner interest in Tesoro Logistics Partners.
Feedstock cost advantage. While capitalizing on domestic crude differentials at its Utah and North Dakota refinery, Tesoro's West Coast refineries (85% of capacity) have largely yet to benefit, with the exception of the Washington refinery, currently receiving 50 mb/d of Bakken crude by rail. Tesoro expects to increase West Coast rail shipments of advantaged crude by 325 mb/d by 2015. Supplanting Bakken crude for Alaska North Slope improves yields, resulting in a $3-$5/bbl margin increase. In contrast, other refiners have already seen much of the benefit of crude differentials in their earnings.
Improving California competiveness. Management executed a shrewd acquisition with the purchase of BP's (NYSE:BP) Carson facility, but it left Tesoro even more exposed (60% of capacity) to a historically tough California market. However, Tesoro can integrate the facility into its existing network while improving its competitive position by increasing distillate yields, reducing emissions and lowering operating costs. Combined with the aforementioned increase in cost-advantaged feedstock, Tesoro expects to increase annual EBITDA by almost $500 million by 2017.
Increasing free cash flow and shareholder returns. We expect free cash flow to steadily increase over the next five years as improvement initiatives increase profitability and capital spending peaks in 2015. With its balance sheet stronger after the dropdown of Carson assets to TLLP, increased shareholder returns are possible. Tesoro currently has about $500 million remaining on a $1 billion share-repurchase authorization and pays a dividend of $1 per share, yielding 2%. We expect both to grow in the coming years. We think Tesoro could spend another $1.5 billion on share repurchases through 2017. Combining this with improving profitability, we estimate that earnings per share would grow to $8 in 2017 from $3 in 2013. Increasing the dividend about 25% per year would result in a modest 30% payout ratio in 2017 while still leaving Tesoro with a healthy balance sheet.