Crumbling Boardwalk Empire Doesn't Shake MLP Foundation

 Boardwalk Pipeline Partners' sharp sell-off last month prompted many questions regarding the stability and risk profile of master limited partnerships and midstream firms more broadly. In addition to cutting our fair value estimate for Boardwalk nearly in half, we cut its moat rating from wide to none and raised its uncertainty rating from low to high. Do other MLPs deserve the same re-evaluation?

We contend that Boardwalk is an isolated case, the product of fundamental changes in gas flows caused by the emergence of Marcellus shale gas, a constrained balance sheet, and the inability of growth projects to replace cash flows lost to declining pipeline utilization and lower rates on contract renewals. While virtually all gas pipelines in the eastern United States must contend with the impact of Marcellus gas, Boardwalk's balance sheet and project slate are company-specific problems that do not impugn peer companies. Midstream energy is in the midst of interesting times, and while shifting fundamentals challenge Boardwalk, these same changes create opportunities for the majority of pipeline operators we cover. Despite our re-rating of Boardwalk, our moat and uncertainty ratings for the remainder of our midstream coverage remain intact.

How the Marcellus Shale Turned the Pipeline Map on Its Head
Historically, natural gas flowed from producing regions in Texas, Louisiana, and the Gulf of Mexico to major markets in the Midwest and Northeast. The major legacy gas pipelines--Columbia Gulf ( NiSource (NI)), Texas Eastern ( Spectra Energy Partners ), Texas Gas (Boardwalk), Tennessee Gas ( Kinder Morgan Energy Partners ) and Transco ( Williams Partners )--maintained fairly steady levels of utilization from year to year, with weather and seasonality being the primary determinants of throughput.

Enter the Marcellus. Located in Pennsylvania and West Virginia, this massive shale gas play has grown from 2% of U.S. dry gas production to nearly 20%, with production increasing by 10 billion cubic feet per day since 2010.

In 2013 the Marcellus produced an average of 10.2 Bcf/d of gas, rising to just shy of total demand in the New England and Middle Atlantic states. Over the past three years, Marcellus gas has displaced roughly 8 Bcf/d of gas imports to the region. According to EIA data, roughly half of the displaced gas originated in the Southeast--that is, these are volumes that would have otherwise been transported along the major legacy systems mentioned above. Furthermore, Marcellus production will continue to increase, topping 16.5 Bcf/d by the end of 2015 (averaging 15.7 Bcf/d for the year), and backing out regional imports in all but the coldest months.

Outlook for Legacy Pipelines
Rising Marcellus production will have two countervailing impacts on legacy long-haul gas pipelines. First, there clearly is much less need to move gas from the Gulf to the Northeast, and even less of a pricing incentive. We can see this reflected in gas price basis differentials, where Northeast gas hub pricing has shifted from a premium to Henry Hub pricing, to a steep discount. It now costs Northeastern end users more (considerably more, counting transportation fees) to source gas from the Gulf versus the Marcellus, and there's little reason to expect this dynamic to shift back. On balance, we expect to see continued pressure on pipeline volumes for carriers formerly supplying Northeast markets. As existing contracts roll off, we expect to see reduced volumes, as well as lower revenue per unit as shippers cut rates to buy throughput.

On the other hand, the gas surplus building in the Northeast needs to find a home. Legacy pipelines that can accommodate Marcellus gas and provide access to new markets are well-positioned to see increased volumes and command premium rates. Of the Marcellus debottlenecking projects we believe will move forward, the majority build off of the big pipes from the Gulf. Williams' Liedy Southeast Expansion project is a good example of how we expect these pipelines to adapt to the Marcellus. The project loops an existing lateral off of Transco, bringing 0.5 Bcf/d of additional Marcellus gas onto its system to access markets on the Atlantic seaboard.

Williams has also proposed its Atlantic Sunrise project, which would provide 0.5-1.0 Bcf/d of capacity to flow Marcellus gas all the way to Station 85 in Alabama. Spectra's Sabal Trail pipeline, expected in service in late 2017, will then allow Marcellus gas to flow south to Miami. Eventually we see the need for Marcellus gas to head to markets along the Gulf of Mexico, and we expect additional project announcements in the coming quarters. In addition, while the Marcellus has been more of a challenge than an opportunity for Boardwalk thus far, its Ohio-to-Louisiana project, which will move 600 mmcf/d south along Texas Gas by the end of 2016, is a sign of things to come.

In the meantime, however, there's a stark contrast between gas pipelines that can bring Marcellus volumes onto their systems, and those that cannot. Spectra, Williams, and Kinder Morgan have been somewhere between opportunistic and aggressive in their pursuit of Marcellus projects, in part by virtue of the physical locations of their lines (which all run through the heart of the Marcellus) and further enabled by strong financial backing. Boardwalk's Texas Gas, however, as well as pipelines such as ANR ( TransCanada (TRP)), Trunkline ( Energy Transfer Partners ), and Columbia Gulf (NiSource), extend only into the Midwest, serving utility and power customers along their routes but not physically crossing the Marcellus. These pipes have been slower to adapt to Marcellus growth, but they're feeling the pain of reduced volumes.

Competitive Responses
These pipelines have three competitive responses to declining volumes.

Backhauling. The simplest to implement, backhauling involves delivering gas to destinations in the opposite direction of pipeline gas flow, usually accomplished by swaps rather than by moving physical gas.

Reversals. By retooling pipeline compressors, and delivery and receipt points, a pipeline can reverse the physical flow of gas on the line. Requires greater capital-expenditure investment.

Conversions. Natural gas pipelines can convert all or part of capacity to move natural gas liquids, crude oil, or refined products instead of gas. Often pursued along with a reversal. Requires greater investment than a reversal.

​

Columbia Gulf completed a backhaul arrangement in 2013, allowing for 0.5 Bcf/d of gas to move from north to south, while Energy Transfer filed with the Federal Energy Regulatory Commission last year to convert Trunkline Gas from natural gas to crude oil service and reverse flows to the Gulf. Boardwalk is seeking to convert and reverse a portion of Texas Gas to NGL service, as part of the Bluegrass pipeline project (a JV with Williams), and announced the Ohio-to-Louisiana project, reversing another portion of Texas Gas to physically move 0.6 Bcf/d to the Gulf by mid-2016. Several of these projects will go forward (the jury is still out on Bluegrass), and we would expect more to come.

Why Boardwalk Stumbled
If Boardwalk is victim to the same changing industry fundamentals, and is responding to these changes in a similar fashion, why was it forced to cut its distribution, and are other MLPs in danger of distribution cuts? Our answer here is no, for a couple of reasons.

First, Texas Gas drives an estimated 40% of Boardwalk's cash flows, a significantly higher concentration than any other MLP. Similar revenue decreases on segments of Trunkline, Tennessee Gas, or ANR just won't have the same impact on Energy Transfer, Kinder Morgan, or TransCanada, respectively. In addition, while Texas Eastern and Transco, respectively, represent roughly one fourth of Spectra Energy Partners' and Williams Partners' EBITDA, volume growth from new Marcellus projects here will offset declines elsewhere on their systems.

Second, Boardwalk lacked the dry powder to build its way out. High leverage and tight distribution coverage limited Boardwalk's capital spending options over the last several years, despite an accommodative general partner. Leverage has hovered close to its bank debt covenant of less than 5 times debt/EBITDA, thanks to a large cost overrun on older growth projects, ongoing capital spending for new projects, and funding distributions with debt, due to cash-flow shortfalls. Boardwalk's distribution coverage has been spotty during the past two years, sometimes covering and at times falling just short of full coverage. Thus, with management's expectation for a 30% decrease in 2014 cash flows, the partnership would have required some kind of equity infusion to maintain covenants, growth projects, and its distribution.

The Marcellus and Midstream Moats
While Boardwalk's distribution collapse surprised us and led us to cut the partnership's moat rating from wide to none, we do not believe that our remaining midstream coverage is threatened by rising Marcellus gas production. Rather, we see meeting the infrastructure challenges presented by the Marcellus' impact on gas flows as one of the great investment opportunities for midstream firms. We believe the dislocations in gas markets caused by rising Marcellus production will provide firms with a solid slate of attractive projects, many of which will actually serve to enhance excess returns by driving higher system utilizations, providing access to new markets, and supported by long-term, fee-based contracts.

But what of the rest of our midstream MLPs--could something like Boardwalk's distribution cut happen to another wide-moat stock? We think this is unlikely. Of the 20 midstream MLPs we cover, we rate 13 of these wide moat, five narrow moat, and only two no moat. Our wide-moat franchises on balance earn higher returns and post stronger distribution growth and coverage, with lower levels of financial leverage, than our narrow- or no-moat MLPs.

Reviewing MLP Uncertainty
We had rated Boardwalk as low uncertainty, based on its contract structure, with roughly 80% of revenue derived from capacity reservation charges, and contract tenures of about five years. A long-standing mantra in the pipeline space is that actual throughput is less relevant than contracted capacity; in retrospect it's clear we missed some of the signs that contract renewals were under pressure, driving steep rate reductions to win capacity commitments. Had we assigned a greater probability to significantly lower rates on renewing contracts, our fair value estimate would have been lower, and we would have demanded a greater margin of safety, designated by a higher uncertainty rating. However, none of this would have led us to conclude that Boardwalk would choose to cut its distribution aggressively and in effect attempt to take all its lumps at once. Our prior forecast called for meager distribution growth, tight coverage, and continued high financial leverage, with slow improvements toward the end of our forecast as growth projects began to offset weakness on legacy assets. Given Boardwalk's distribution cut, its continued exposure to declining volumes and lower contract renewals, and its limited financial flexibility going forward, we have moved its uncertainty rating to high.

In our MLP coverage, we rate only three other stocks as low uncertainty-- Magellan Midstream Partners ,  Holly Energy Partners , and Spectra Energy Partners. We continue to believe that these three partnerships exhibit stable, predictable cash flows secured by longer-term contracts, visible growth prospects, and healthy balance sheets to support both growth and distributions.

Thanks to long-term contracts and annual inflation tariff adjustments, we believe that Magellan and Holly can rely on steady, highly visible, fee-based revenues. Contracts for their interstate pipelines and storage terminals are long-term, ranging from three to 20 years, often on a “take or pay” basis, so customers pay to reserve capacity, regardless of actual utilization. Importantly, these firms transport and store refined products and crude oil, not natural gas, and are thereby insulated from the gas market changes we've been discussing. As long as end-user consumption of refined products remains relatively stable over time, Magellan and Holly will see stable volumes on existing assets. Shipping rates are often regulated by FERC, based on changes to the fixed goods Producer Price Index. These contractual annual price adjustments provide inflation protection and allow pipeline operators to increase cash flows even with flat volumes. As toll-road operators, they do not take title to the products that it transports and stores, so commodity exposure is very limited. We estimate that 80% of Magellan's and 100% of Holly's cash flow is fee-based.

Spectra Energy Partners is, in fact, a natural gas pipeline MLP similar to Boardwalk. Ninety-five percent of its revenues are fee-based and supported by long-term contracts. However, unlike Boardwalk, Spectra's core pipelines, Texas Eastern and Algonquin, experienced 98% revenue renewal on expiring contracts in 2013. With more than $4 billion in new projects to adapt its core gas pipelines to Marcellus flows currently underway, we see solid volume and cash-flow growth going forward, supportive of steady distribution growth while maintaining better than 1.1 times distribution coverage. These projects are expected to add $550 million in EBITDA by 2017, versus our $1.4 billion EBITDA estimate for the gas pipeline business in 2014; clearly, these are material additions to cash flow. Finally, we note that roughly 20% of Spectra's cash flow comes from long-haul crude oil and NGL pipelines, which enjoy long-term contracts and similar dynamics to Magellan and Holly liquids lines.

Given a first-class set of assets that generate highly stable cash flows, strong growth prospects, a solid balance sheet, and stable distribution coverage and growth, we are very comfortable rating Spectra Energy Partners as a wide-moat stock with low uncertainty. Moreover, the stock is currently trading at a 16% discount to our $58 per unit fair value estimate, making Spectra Energy Partners one of our top picks in the MLP space.