The U.S. Crude Export Boom and the Midstream Opportunity
We favor Enterprise Products Partners, Tallgrass, and Magellan.
Expanding on our overarching theme of U.S. hydrocarbons (gas, oil, or natural gas liquids) increasingly being exported, we forecast U.S. crude-oil exports of 4 million barrels per day by 2020. This is sharply higher than consensus estimates of 1.4 million-2.25 million bpd. We believe the consensus forecasts underestimate the ability of U.S. oil supply, particularly from the Permian Basin, to increase over the next few years and lack a clear understanding of the ability for the United States to export oil in size, given that the country only essentially allowed exports to take place beginning in 2015. We think that only midstream companies that own all of the assets in the midstream value chain will be able to fully participate in the U.S. crude-oil export opportunity, limiting it to a few such as Enterprise Products Partners (EPD), Magellan Midstream Partners (MMP), and Tallgrass Energy (TGE).
Crude Exports Offer Underappreciated Opportunity
Crude-oil exports averaged about 2 million bpd in 2018 and will increase to 4 million bpd in 2020 and 5.7 million bpd by 2023. Meanwhile, consensus estimates are as low as 1.4 million bpd in 2020. A large part of the reason behind our optimism is our work on U.S. crude-oil supply, which we expect will increase to 14 million bpd by 2020 from 9.4 million bpd in 2017, led by the Permian Basin. As a result, there is a substantial opportunity for midstream entities to build new crude-oil export terminals and support their economic moats, with Enterprise Products Partners our favorite way to play this trend. Numerous midstream entities have announced projects under development to start up in the next few years.
PIRA Energy and IHS Markit estimate that U.S. crude exports will reach 2.25 million bpd and 1.4 million bpd in 2020, respectively. The U.S. Energy Information Administration forecasts exports of 1.8 million bpd by 2020. The International Energy Agency expects U.S. crude-oil export capacity of 4.9 million bpd by 2023. Our estimates of 4 million bpd in 2020 and 5.7 million bpd by 2023 are well ahead of consensus.
The U.S. government banned the export of most U.S. crude oil until December 2015, when restrictions were lifted. Since then, U.S. oil exports have increased quickly and were as high as 3 million bpd on a weekly basis in mid-2018. Through November 2018, the U.S. has exported an average of 1.9 million bpd per month. Asia is the most important export center, with over a third of U.S. barrels heading there in November 2018. Until August 2018, China was a major importer of U.S. barrels, having averaged 378,000 bpd of exports year to date. However, with the escalation of trade tensions between the U.S. and China, U.S. exports to China halted abruptly and fell to zero from August to November 2018. The lost exports to China were replaced by higher exports to other Asian countries, namely South Korea, Japan, and India. Media reports indicate that U.S. oil exports to China resumed in January 2019 in modest quantities.
We expect China’s oil demand will increase until 2030 and peak above 17 million bpd compared with just above 13 million bpd in 2018. In our opinion, consensus is missing the boat on China’s medium- and long-term oil demand, but in different directions. Over the next five years, we’re bullish. We think that by relying on China’s inflated official GDP growth figures, other forecasters are underestimating China’s income elasticity of oil demand. Looking from the bottom up, near-term improvements in China’s fuel-efficiency standards will be more modest than headline requirements suggest. Also, others have mistakenly identified the decline of China’s investment boom as the death knell for diesel growth. Still, in the long run, electrification of China’s vehicle fleet looms large for oil. We think adoption of electric vehicles is primed to take off in China, reaching 25% of sales by 2028 and catalyzing peak oil demand for China as early as 2030.
Building Our U.S. Oil Export Forecast
Our model for U.S. crude exports is built around four factors: projected U.S. crude-oil consumption (as measured by EIA product supplied data), crude-oil supply, crude-oil imports, and crude-oil exports.
We adjust our demand and supply forecasts to remove natural gas liquids and other liquids (refinery gains and biofuels, for example). Our oil demand forecasts use IEA estimates, while we use our own differentiated crude supply, import, and export forecasts. Broadly, we expect domestic U.S. crude-oil consumption excluding imports and natural gas liquids consumption to decline to 14 million bpd by 2023 from 14.6 million bpd in 2017. In contrast, crude-oil supply, led by the Permian Basin, should increase to 13.5 million bpd in 2023 from 9.4 million bpd in 2017. Imports should remain steady at 8 million bpd. Adding it all up, we expect U.S. crude-oil exports to increase to 5.7 million bpd by 2023 from 1.1 million bpd in 2017. Our export forecast is tempered to reflect likely infrastructure delays and the fact that existing export capacity is probably multipurpose (that is, it can handle refined products and oil), meaning that not all capacity will be 100% available for crude oil depending on economics at the time. The major change in U.S. supply over the next few years is mainly from the Permian Basin, where we expect production to reach 5.9 million bpd in 2023 from about 3.3 million bpd in 2018.
We think our U.S. production forecast is realistic, as we assume midcycle prices and rig counts take hold in 2021, which explains the slowdown in production growth beyond 2021. We believe the current light tight oil rig count is 750, at least 150 rigs higher than our long-term midcycle expectation of 500-600. Our production and export forecasts could shift depending on the actual number of rigs drilling in 2021 and beyond.
We expect crude-oil imports to remain flat at 8 million bpd over the next few years. At first glance, this forecast doesn’t seem to make sense, as readers might expect imports to be displaced by U.S. domestic crude-oil production. However, this is where the gravity specification of the crude oil being produced comes into play, as well as the fact that the U.S. refinery complex is configured largely to handle heavier grades of crude oil. U.S. shale plays are producing crude oil that is typically light crude with a gravity range of 35 degrees-plus (and in some areas of the Permian 40-50 degrees). However, U.S. refineries are generally configured to handle heavier crude (below 30 degrees), and the mix of crude imports is about two thirds heavy versus light, which has increased over the past few years.
A material portion of imports is heavy Canadian crude, which has increased to 3.7 million bpd in 2018 from 2.4 million bpd in 2012. Somewhat unsurprisingly, this has come at the expense of lighter crude, which has declined to 32% of imports in 2018 from 42% in 2013, but the absolute level of light crude imports has remained steady at 2.6 million bpd over the same period. This trend suggests that some refineries are in areas where it is uneconomic to build new pipelines to transport light U.S. crude to the refinery, meaning that even as exports increase, there will remain a reasonably steady demand for lighter imports. We also believe a good portion of these light oil imports is Saudi Arabian light crude, which is relatively sour and thus not a replacement for U.S. light crude. In addition, it is difficult and expensive to move light crude from the Gulf Coast to the East Coast, given the Jones Act, which requires the use of U.S. vessels. We assume that about 200,000 bpd of refinery capacity will be upgraded annually to take U.S. light crude, thus lowering demand for imports, but this will be offset by higher levels of Canadian heavy imports due to ongoing struggles from Venezuela and Mexico to export heavier crude, given economic unrest and production declines.
Our per-barrel oil price midcycle forecasts are $55 West Texas Intermediate and $60 Brent, with the $5 differential reflecting transportation costs. Despite a higher level of exports, we’re not making changes to our oil price midcycle forecast because our export forecast reflects changes in the global supply and demand for oil, not changes to the marginal cost of extracting the oil or the underlying geology. We expect the oil to be produced given existing economics available, and we’re highlighting where that oil should eventually flow.
The risks to our view include primary infrastructure constraints or delays on the export terminal side. If supply ramps up without a corresponding outlet to the global markets via an export terminal, it will be stranded in the U.S. and thus depress WTI prices and widen the Brent/WTI spread. We don’t expect much impact to spreads if demand either from the U.S. or international markets is higher than our expectations, due to the decades of drilling inventory left to be exploited in U.S. shale plays. We would simply expect to see more U.S. oil extracted quicker from the overall resource base, which has ample capacity.
There are a variety of tanker sizes that can transport crude oil, from the smaller Panamax (about 500,000 barrels of capacity) to the very large crude carriers, or VLCCs, which can carry 2 million barrels of oil. Tankers typically have one of three options for loading: direct loading at a marine terminal; partial reverse lightering, where a tanker is filled at an offshore trans-shipment area by a smaller tanker after obtaining the maximum possible amount at a terminal; or full reverse lightering, where a smaller tanker fills the larger tanker entirely offshore, since the larger tanker cannot source crude at the marine terminal because of size restrictions.
From an economic perspective, it makes the most sense to load the largest possible tanker (the VLCCs) directly from a marine terminal given the time savings (demurrage) and cost savings (larger capacities spread fixed shipping costs over a larger amount of oil). Today, the vast majority of the VLCCs taking onboard U.S. crude undergo the multiday loading process with reverse lightering using an Aframax tanker (typical capacity of 750,000 barrels) because of draft limits; existing ports cannot support VLCCs’ depth rating of about 72 feet. Being able to skip reverse lightering can easily save $700,000-$1 million depending on tanker rates.
Taking Advantage of the Export Opportunity
As we expect oil exports to increase over the next few years, there’s an opportunity for midstream entities to provide new infrastructure and related assets. Currently, the major U.S. import/export asset is the Louisiana Offshore Oil Port, the only deep-water terminal in the Gulf of Mexico. Historically, LOOP just imported oil to the U.S., but it has recently begun to export oil and has about 300,000 bpd of capacity. This is inherently limited because of the logistical challenges of trying to import and export barrels without dedicated pipelines. LOOP has three buoys that can handle the VLCCs. The terminal transfers the oil to Clovelly, which has 71 million barrels of storage capacity. The storage terminals allow LOOP to store crude until it is needed to be blended for processing.
The opportunity is to load up U.S. crude at key ports (Corpus Christi and Houston Ship Channel) by docking VLCCs at marine terminals that have access to a deep enough channel to support VLCCs and building offshore terminals with the required infrastructure in deep enough water to support VLCCs. Given the potential delays in deepening the channels, companies are increasingly developing offshore terminals to meet the growing export opportunity. The new offshore terminal can then be connected to a large and well-established network of pipelines that can supply the needed crude.
The city of Corpus Christi is pursuing several projects to allow larger ships to enter the port by addressing height and depth limitations. A new Harbor Bridge will be complete by 2021 and provide 205 feet of draft clearance. Also, the channel depth will be increased to 54 feet from the current 47 feet, assuming receipt of funding from Congress; this will accommodate a fully loaded Suezmax tanker (about 1 million barrels of capacity) but not the more attractive VLCCs, which require 74 feet (and possibly up to 100 feet in the case of high winds or tides). This project is being pursued separately at the moment, primarily via funding from Carlyle Group. In 2017, the port of Corpus Christi exported an average of 306,000 bpd of crude oil.
LOOP offers several advantages. First, it is the only offshore terminal able to handle VLCCs. Second, its extensive storage allows it to blend different crude grades as needed by customers, as well as easily store the crude needed to load a VLCC. Third, it has connections to over 500,000 bpd of offshore production via the Mars and Thunder Horse pipelines that can be easily exported. Fourth, LOOP is connected to the currently lagging Capline pipeline, which can move 1.2 million bpd north to Patoka, Illinois. Once this pipeline is potentially reversed in a few years, it can export barrels from the Bakken and Canadian crude. Its major disadvantage is a lack of direct connections to crude delivered to Houston and Corpus Christi versus Louisiana. Its major Texas connection is via Shell Midstream Partners’ Zydeco pipeline, which can transport oil from the storage terminals in Houston, Port Arthur, and Lake Charles but does not have direct access to the Permian. We believe that Houston connectivity will be prized for new export terminals, given its access to cheap Permian crude.
We are in the beauty contest phase of the U.S. export infrastructure boom, in our view. To address the need for new exporting assets, we've seen number of recent announcements, several within the span of a few weeks. We’re currently tracking about 8.5 million bpd of U.S. crude-oil export terminals under consideration, which would take estimated total U.S. export capacity to 11.4 million bpd by 2023, well ahead of our expected 5.7 million bpd of actual exports during that period. Since not all of the projects have applied for permits (and one has obtained only some partial permits), which we expect could take 18-24 months, we don’t think all projects will receive the needed customer backing and contracts to go forward.
We think three projects face hurdles: the Trafigura Texas Gulf Terminals, the Oiltanking joint venture, and JupiterMLP. We think the Trafigura project faces political opposition, mainly from the Port of Corpus Christi, which is pursuing its own export terminal efforts with Carlyle. It also has the added hindrance of depending on channel dredging to obtain optimal economics, which may not be complete until 2022. The Oiltanking joint venture, with a projected startup in 2022, seems too late and too large, at 2 million bpd of export capacity, to be needed. While we do expect the JupiterMLP pipeline to be built, as it has some customer commitments and serves all three major deep-water port hubs, we see less of a need for the incremental export capacity, given its weak Houston connectivity from its base in Brownsville, Texas. Assuming these three projects do not move forward, this reduces our expected export capacity to 7.9 million bpd in 2023, which ensures a reasonable level of utilization, given our export projection of 5.7 million bpd.
Financial disclosures on new terminals are scant, as only three (Trafigura, Enterprise Products Partners, and Texas Colt/Oiltanking) have applied for a permit from the U.S. Maritime Administration. Enterprise has estimated about 18-24 months for permit approvals before construction can begin, assuming a final investment decision is made. However, we do know that LOOP charges $0.55-$0.56 a barrel to export plus $0.02-$0.06 per barrel for storage, depending on the length of time needed.
Altogether, we estimate that the 5 million bpd of new export capacity being added over the next five years represents an investment of $3 billion-$4 billion. When we include the related pipelines and associated storage infrastructure, this could total $6 billion-$8 billion. If we assume a 60% EBITDA margin and a 6 times EBITDA multiple, this represents industry EBITDA of $350 million for the export terminals and a midpoint of $700 million in industry EBITDA when including the related pipeline and storage infrastructure. While the financial impact may be somewhat modest, as we expect Enterprise’s 2019 EBITDA to be about $7.7 billion, we think these projects are important because of the future expansion opportunities available at potentially high returns, but also the opportunity to land unrelated projects and investment opportunities because of the high-profile nature of these efforts.
Enterprise Products Partners, Magellan, and Tallgrass are all undervalued and great ways to play this industry shift, in our opinion.
Enterprise Products Partners Is a High-Quality Export Giant
We think Enterprise Products Partners is an important player in this expansion opportunity, as it is involved in one of the largest Houston-area expansions in crude-oil export capacity at 2 million bpd. It also is the largest crude-oil exporter in the market already, with substantial capacity across its existing asset base. We’re particularly impressed by the partnership’s position on the Houston Ship Channel in terms of storage and export opportunities. Enterprise controls 21 million barrels of storage with six deep-water dock ship and two barge docks, and it can accommodate Suezmax tankers, which are the largest tankers that can navigate the channel. The company also controls the Beaumont West, Freeport, and Texas City systems, which adds additional dock access. We think the dock access is important, given geographic constraints, and it allowed the company to export 34 million barrels of crude oil in 2016 following the lifting of the export ban in December 2015, as well as another 17 million barrels of processed condensate. Replicating Enterprise’s crude-oil asset portfolio will be very difficult, in our view. Beyond the actual economics of the terminal, we expect Enterprise will be able to extract further fees via its world-class marketing operations.
We view its marketing operations as a strong asset that lets Enterprise collect significant additional fees from its network versus being a pure toll-taker. With its marketing operations, Enterprise takes ownership of the hydrocarbons and seeks to exploit differentials based on time, location, or product arbitrage across the hundreds of connection points in Enterprise’s system. For oil exports, the marketing operations allow Enterprise to reliably source cheap crude oil and the right time and location to maximize its profits from wide differentials in a given international market. This type of asset is very difficult to replicate because of the complexity of Enterprise’s system, and the few producers that do undertake marketing activities focus on the relatively few basins they operate in versus the entirety of the U.S. oil and gas complex. It also lets Enterprise take full advantage of profitable opportunities in secondary markets across its pipelines for capacity not being used under firm contracts, versus ceding those fees to the shipper. Other examples include taking advantage of seasonal changes in demand for propane, upgrade opportunities for raw NGLs to be converted to higher-grade and more profitable olefins, and being able to use Enterprise’s network to move product to markets where differentials are the widest.
The marketing operations are embedded in the natural gas, oil, and NGL teams at Enterprise, providing insights to help them make investments across the portfolio. We believe the marketing operations provide an added benefit in terms of sourcing and developing relationships with producers to serve as committed shippers for future investments but also sources of internal demand (opportunities to Enterprise to take ownership of hydrocarbons, for example) to support incremental investments. Peers without this robust level of marketing operations will face higher hurdles in terms of being able to obtain commitments for large investments.
Our fair value estimate for wide-moat Enterprise Products Partners is $35.50 per unit, which implies a 2019 EBITDA multiple of 13.4 times, 2020 enterprise value/EBITDA multiple of 12.5 times, and 2019 distribution yield of 5.0%.
We expect the main driver for Enterprise will be its NGL segment, as the demand pull from the Gulf Coast and international markets lets Enterprise take advantage of the export opportunities and lucrative differentials via its comprehensive asset base. The diversity of Enterprise’s asset base, its ability to take advantage of just about any profitable opportunity that appears in U.S. midstream, and its largely fee-based earnings stream all support healthy single-digit growth prospects over the next few years. We expect EBITDA to increase about 7% annually over the next five years and distributions about 7% annually on average over the same time frame, though the growth is weighted more toward the tail end of our forecast period. In general, we anticipate that differentials will remain stable across Enterprise’s portfolio, though we model a decline in some NGL differentials to reflect our expectation that current spreads will narrow alongside a decline in Brent prices to our long-term price of $60 a barrel.
Tallgrass Energy Is Also Cheap
For Tallgrass, the Plaquemines Liquids Terminal represents a great way to leverage the Pony Express pipeline in a clever way, as it provides another reason for shippers to use the pipeline. It also should benefit to some extent from its marketing operations, though these efforts are much smaller than Enterprise Products Partners’. Pony Express sources from six supply sources and has five streams of oil capable of being batched; this diversity of supply and the fact that the pipeline is the cheapest transportation option for producers in these regions are considerable benefits. Despite being the cheapest, the pipeline offers a premium benefit: the ability to batch crude, which results in a higher-quality crude for refineries and is their preferred option versus a common stream (where various crudes are mixed to a specified standard), as the pipeline serves over 550,000 bpd of refinery demand. This type of batch pipeline is more expensive to operate, as it requires storage at both ends of the pipe to handle the different batches. We do have concerns about the location of the assets, as Pony Express obtains its oil from the DJ Basin, Bakken, Powder River, and Niobrara, which are not some of the lowest-cost basins, raising concerns about the long-term production outlook. But given that the pipeline is only a small portion of the combined basins’ overall production, even in a weak to declining production environment we would still expect Pony Express to obtain solid volumes.
We believe Tallgrass has a narrow economic moat as a result of efficient scale. Its two major pipeline assets are the Rockies Express and Pony Express. It also owns storage and gathering/processing assets that provide it with additional control over the hydrocarbons that feed into its pipelines. Of Tallgrass’ revenue, 97% is subject to firm fee contracts, with only 1% exposed to commodity prices. Tallgrass’ anticipated returns on invested capital (after consolidating 75% of REX) of 12% exceed our cost of capital, further supporting a narrow moat rating.
We’ve seen several positive developments for Pony Express recently, led by Tallgrass investments. The first is a five-year extension of Continental Resources’ contract to 2024 at blended rates of around $3.35 a barrel, by our estimates, about a 10% discount to recent rates versus the 30% decline we feared. The company also recently acquired about $50 million worth of storage terminals around Pony and completed a successful open season for an incremental 37,500 bpd, pushing volumes on the pipeline to record levels. Finally, Tallgrass is engaging in a planned 300,000 bpd expansion of Pony Express to support its Seahorse pipeline effort. Kinder Morgan has signed on to the initiative, and the pair have an agreement to combine assets to obtain incremental capacity.
Our Tallgrass fair value estimate is $28 per share, which includes 75% of REX’s debt, implies a 2019 EBITDA multiple of 12.3 times, 2020 EBITDA multiple of 9.6 times, and 2019 distribution yield of 7.7%. Broadly, we think the market has had a consistently negative outlook for Pony Express’ and REX’s future and fails to give the management team credit for landing new contracts for both pipelines to offset expiring ones and its skill at finding new development projects to actually increase volumes.
We also expect Tallgrass to benefit from acquisition opportunities. It has been able to largely offset expiring contracts at REX and Pony Express with new assets via organic development and deals and new contracts. We see continued logical growth opportunities, particularly with the Rockies Express, with more opportunities around Pony Express being developed. Tallgrass will benefit from ongoing production growth in the Appalachian region to the extent that it provides flows to move toward Midwest demand centers, though we expect the region to be well supplied with takeaway capacity, meaning differentials will remain low. As a result, we project 2023 EBITDA at $1.4 billion compared with $860 million in 2018.
Magellan Is Once Again Pursuing Smart Investments
We think it is a good idea for Magellan Midstream Partners to pursue crude-oil export opportunities, as we think it leverages its Houston-area distribution infrastructure as well as its existing export-related assets. We don’t expect Magellan to benefit quite as much as Tallgrass and Enterprise Products Partners because it does not have any marketing operations, which would be a source of incremental fees. However, we think Magellan is undervalued because investors fail to appreciate the stability of its refined product network and are overestimating the immediate near-term impact of electric and autonomous vehicles. While we acknowledge that there are long-term concerns about the impact of EVs and AVs on gasoline demand, in particular, Magellan has time to reposition the network to ensure its stability.
We consider Magellan’s storage business to be better than average because of its position in Galena Park on the Houston Ship Channel, but we still see it as having no moat, given the lack of barriers to entry. Terminals are used to take advantage of changes in seasonal demand by marketers or traders, tankage constraints, or the specialized handling needs of the product. Rates are unregulated, and about 80% of capacity tends to be under contract at any given time for about three years, where the customer pays for capacity regardless of actual usage. Magellan’s Galena Park terminal has about 13 million barrels of storage and access to multiple pipelines as well as docks that can accommodate ship traffic. The facility is particularly important because of the growing demand for exports of refined products and NGLs, and access to five docks is an important competitive advantage, given geographic constraints. We have mixed feelings about the 50/50 joint venture with Valero Energy to develop the Pasadena marine storage terminal, which includes 5 million barrels of storage and two ship docks. We think the joint venture was needed for Magellan to possibly acquire access to Valero refineries, but we think this is an attractive asset that could have been fully funded by Magellan and thus retain full economics.
We do award Magellan a wide economic moat based on its refined product network. Its refined product pipelines are an extremely attractive business that has served as a reliable cash flow generator for Magellan to expand elsewhere. The infrastructure is fundamentally better positioned than most if not all natural gas and oil pipelines for a few reasons. First, as the 9,700 miles of pipes across 15 states connects refineries to end markets such as gas stations via 53 Magellan terminals, demand is highly stable, as it depends on consumer demand in terms of miles driven and fuel efficiency. Similarly, supply and thus flows are less volatile as they are aggregated by a refinery versus multiple gathering and processing units linked to levels of drilling activity. Second, with Magellan serving the midcontinent, which has benefited the most in recent years from an increase in light crude supply, it serves some of the most profitable refineries in the U.S., meaning it can easily boost prices 2%-3% annually in line with inflation with little pushback, as the costs are passed through to the end consumer. Third, we’d characterize the incremental fee opportunities from refined product storage terminals and butane blending as low risk and attractive, given the lack of volatility in Magellan’s end markets. Finally, with the completely stagnant demand outlook, there are virtually no new pipes being constructed, meaning Magellan faces no new competitive threats from that angle.
Our fair value estimate for Magellan is $72 per unit, which implies a 2019 EBITDA multiple of 14 times, 2020 EBITDA multiple of 13 times, and 2019 distribution yield of 5.8%. We expect Magellan to continue to benefit from a stable environment for its refined products footprint, which will serve as the cash cow for its investments in crude-oil pipelines and marine terminals. We expect revenue growth to be about 4% on average over the next five years, while earnings per unit should increase about 2%-3% on average over the same time frame. We do not see a need for the partnership to raise equity capital in the near term. We anticipate operating margin to top $1.8 billion in 2023 from $1.7 billion in 2018. We also expect that around 5%-8% distributable cash flow growth for 2018-20 is reasonable.
Stephen Ellis does not own (actual or beneficial) shares in any of the securities mentioned above. Find out about Morningstar’s editorial policies.