On July 6, 2013, 47 residents of Lac-Megantic, Quebec died horribly when a 74-car freight train carrying crude oil exploded in a fireball. It was the deadliest freight-train accident in Canadian history. Five bodies were never recovered and were assumed to have simply vaporized. DNA samples were required to identify others. Heat from the inferno was felt over a mile away.

In spite of this tragedy, moving crude by rail (CBR) is comparatively safe. Since 2013, safety standards have been tightened throughout North America. The International Association for Energy Economics (IAEE) found the incidence of spills with CBR to be less than for pipelines. However, this is a deceptive statistic. Once you adjust for the greater volumes of crude moved by pipeline versus CBR, as well as the greater distances covered, pipelines remain substantially safer. The report concludes that, “the risk associated with shipping crude oil is noticeably larger for rail deliveries than for pipeline deliveries.”

2,500 miles west, off the coast of Vancouver, live around 75 endangered killer whales. Their connection with Lac-Megantic is not obvious. Fortunately for the Orcas, vocal advocates have successfully made their continued survival more important than avoiding another freight train tragedy. By blocking the Trans Mountain Pipeline Expansion, they have ensured that more Albertan crude oil will reach its buyers by rail, given continued inadequate pipeline capacity.

The Trans Mountain Pipeline (TMX) was put into service in 1953, and has been in continuous operation ever since. It’s another example of the long life of installed energy infrastructure, which generally appreciates in value when properly maintained even while accounting rules allow for its depreciation. Kinder Morgan (KMI), which acquired the pipeline in 2005, had been frustrated in its efforts to more than double the capacity of TMX by adding a second pipeline alongside the first. The TMX Expansion’s approval became a provincial political football. Land-locked Alberta has few choices in exporting its crude oil.

British Columbia sought to prevent a new pipeline from Alberta reaching the Port of Vancouver. Oil that passes through on its way to export markets has little value to local residents, since British Columbia doesn’t need any more oil. Environmentalists worried about a spill, and regard Albertan oil-sands crude as exceptionally hostile to the environment (it’s true that an oil sands facility is not pretty). First Nations tribes claimed their water supplies were threatened, although of 133 indigenous groups consulted, 43 were in support. And the anticipated increase in tanker traffic at the port of Vancouver, from two to ten per week, risked displacing the dwindling community of killer whales.

It was this last point that prompted the Federal Court of Appeal to overturn the government’s approval of the expansion in late August. Although the absence of a meaningful dialogue with First Nations representatives was cited, failure to consider the increased tanker traffic, “was so critical that the Governor in Council could not functionally make the kind of assessment of the project’s environmental effects and the public interest that the (environmental assessment) legislation requires,” said the ruling, written by Justice Eleanor Dawson. Canadian law provides wildlife with considerable protections.

In a deft move, KMI had by now extricated themselves from the vagaries of Canadian energy policy. For months, they had complained about the impossibility of building a pipeline linking two provinces holding opposite views on its completion. Finally, in May they unburdened themselves of the whole sorry mess and sold TMX to the Canadian Federal government for C$4.5BN, viewed by many as a pretty full price. KMI shareholders dodged a bullet.

A few weeks after agreeing the sale of TMX, KMI estimated in a filing that completion of the expansion would cost C$1-1.9BN more than originally expected and take a year longer. On August 30^th, KMI shareholders formally approved the sale, hours after the Court of Appeal ruling. TMX was by now worth considerably less than the C$4.5BN paid — the bailout by Canadian taxpayers was complete.

Prospects for the TMX Expansion are uncertain, with a delay of two years or more seemingly inevitable. The Canadian government is considering an appeal, redoing its environmental review, and crafting legislation to force it through. Canada wants to complete the pipeline. Meanwhile, Alberta grumbles about contemplating separatism, and has in the past suggested that it might halt crude oil shipments to British Columbia altogether.

Canadian crude oil will continue to reach its buyers, although more of it will move by rail. The International Energy Agency expects CBR shipments to double over the next two years due to lack of pipeline capacity. Suncor, Canada’s largest oil and gas producer, won’t expand crude oil production until it sees progress on pipeline approvals. The persistent $25-35 per barrel discount of Western Canadian Sedimentary crude to the WTI benchmark is directly linked to limited transportation choices, and must be regarded as a huge success by Alberta’s neighbors in British Columbia. Since the U.S. is both its biggest energy customer and nowadays its biggest competitor, Canada’s position is unenviable.

American readers will be relieved that such extreme dysfunction doesn’t exist in the U.S. Recall though, Boston’s annual winter imports of liquefied natural gas from Trinidad and Tobago to keep the lights on.  Pipeline opposition isn’t limited to our northern neighbor (see An Expensive, Greenish Strategy).

The investment takeaway is that opposition to new pipelines increases the value of those already installed. When shippers resort to rail to move crude oil, their bargaining position on pipeline tariffs is weak. Oil and gas companies suffer lower revenues, consumers higher prices, and railway lines benefit if demand is sustained long enough to justify their capex. But for pipeline companies, installed pipelines with decades of useful life can represent a scarce resource. If the possibility of excess pipeline capacity concerned markets during the 2014-16 energy sector collapse, we are now headed solidly in the other direction.

We are long KMI.

Before jumping at an attractive yield, investors should pause to consider its consistency.

In recent years, traditional MLP investors were victims of one of the greatest betrayals in U.S. financial history. Older, wealthy Americans were drawn to companies that paid out most of their cashflow in distributions. For years, America’s energy came from roughly the same places in the same amounts, which meant little need for pipeline operators to re-invest in the business. The Shale Revolution changed all this – new sources of oil and gas required new infrastructure (see Will MLP Distribution Cuts Pay Off?).

MLPs decided to grow, redirecting cashflows from payouts to new projects. K-1 tolerant, income seeking investors were the quintessential long term buyer sought by every CEO. All they wanted was stable income. Although for years MLPs provided this, in 2014-15 many of them seized the opportunity to be growth businesses, which redirected cash away from investors. Alignment of interests was lost. MLPs in aggregate demonstrated that Distributable Cash Flow would be paid to investors only as long as they didn’t have any better uses for it. Payouts were cut, trust was shattered. Today’s sector remains 28% below its August 2014 high, and its recovery offers something for everyone (see Growth & Income? Try Pipelines).

Stable and growing dividends remain highly valued. Although the Alerian MLP ETF (AMLP) has cut its distribution by 30% (see It’s the Distributions, Stupid!), Alerian used to post a chart with 6% average ten year distribution growth.  Index components change, and Alerian was using the historic growth of existing index members regardless of how long they’d been in the index. This introduced a survivor bias, in that MLPs cutting distributions used to be dropped while the newly IPO’d ones they added were typically growing quickly. This confused many, because it failed to match actual investor experience. It seemed that everyone but Alerian knew MLP distributions were being cut. Facing growing criticism (see MLP Distributions Through the Looking Glass), Alerian revised their chart to better match reality.

However, energy infrastructure overall has provided far more distribution stability than shown by MLPs. The American Energy Independence Index consists of the biggest pipeline companies in America, which are mostly corporations although it includes a few MLPs as well. It yields 5.5% based on 2018 dividends, a payout that is up 10% on 2017. We expect dividend growth of almost 11% next year. While income seeking investors are naturally drawn to attractive yields, the reliability of the payout is critical.

The chart shows that Alerian MLP Index dividends fell far more than for energy infrastructure as a whole. Moreover, they still have a long way to recover back to the levels of 2014. Although we think MLP distribution cuts are mostly behind us, the group’s history is one that income seeking investors shouldn’t soon forget. By contrast, American Energy Independence Index dividends dipped but quickly recouped their losses. The members of this index have demonstrated far more reliability with their payouts.

The Alerian MLP Index yields 7.3%, although it’s only accessible via structurally flawed MLP-dedicated funds that pay corporate tax (see The Tax Drag on MLP Funds). Because tax expense for such funds flows out of their NAV, in reality, the yield is lower. The energy infrastructure sector’s 5.5% yield is accessible through conventional, RIC-compliant funds with no tax drag. Corporations have managed their cashflows, including payouts, far more reliably than have MLPs. Moreover, those lower payouts have meant more retained earnings to be reinvested back into their businesses.  This is what will drive the 10% annual dividend growth we expect 2017-19, a level MLPs failed to achieve even in the boom years leading up to the 2014-15 collapse.

As the second chart shows, when payouts are cut less in the short run, they grow faster over the long run. Energy infrastructure is a growth business that offers attractive yields. Investors who favor companies with a reliable history of dividends are likely to fare better.

We are short AMLP.

Jim Carville’s admonition during Bill Clinton’s 1992 run for President was, “It’s the economy, stupid!”

In its August 2018 edition, The Utility Forecaster warns readers to approach MLPs “with caution.” Too risky for income investors is their conclusion. MLP buyers have been badly abused, and Chief Investment Strategist Robert Rapier reminds readers of the many “simplifications” that triggered unwelcome tax bills, as well as the multiple distribution cuts. Without doubt, MLP prices have followed distributions.

In one important respect though, Rapier adopts a simplistic yet incorrect explanation. “…during a long downturn in oil and gas prices, contracts expire and MLPs had to renew agreements under less favorable terms. Many MLPs found themselves doing what was once unthinkable – they had to cut distributions.”

It’s conventional wisdom that the 2014-15 oil collapse hurt pipeline company operating earnings, which caused payout cuts. But the numbers don’t support this narrative. The Kinder Morgan (KMI) chart below shows their Distributable Cash Flow (DCF) per share alongside a significantly more volatile stock. KMI’s DCF per share is little changed from 2015 to 2016, but the share price fell by half. They cut their payout twice: once when combining Kinder Morgan Partners (KMP) with KMI (“simplification”, in which KMP investors received KMI stock with a lower payout as well as a tax bill); and again later when KMI cut its dividend. KMI was learning that MLP investors want income over the promise of growth.

To fund their growth projects, KMP was paying out most of its DCF and then seeking to recoup some of it through secondary offerings.  In effect, investors were being asked to reinvest a portion of their distributions back into the company. Many holders found this unattractive, since they spend the income. So KMP’s yield rose, which made issuing equity too expensive. KMI concluded MLP investors no longer suited their purpose, and left to become a corporation. Today, KMI yields 5.1%, with a payout more than 2X covered by its $4.7BN DCF.

MLP buyers can be focused on distributions to the exclusion of anything else. Two recent examples highlight:

In late July, American Midstream Partners (AMID) slashed its distribution by 75% so as to, “…significantly reduce leverage, provide capital for strategic growth opportunities, and create long-term value.” Although these all sound like desirable objectives for a total return investor, AMID’s stock fell 43%.

Meanwhile, Hi-Crush Partners (HCLP) tripled its distribution and now yields 23%. The higher payout is unlikely to persist, but if it’s sustained for a year the company will convert to a corporation. Whether or not this is good for investors, its stock rose 28% on the day of the announcement.

MLP investors want their income.

The chart from Bank of America is even more striking. On the MLPs they cover, they show steadily growing EBITDA with improving leverage, alongside a declining Alerian MLP Index. Falling MLP distributions clearly drove index performance more than improving financials.

It’s as if MLP investors look at payouts and little else.

Probably the simplest measure of MLP payouts is to look at dividends on the Alerian MLP ETF (AMLP), which are 30% lower than in 2015.

There were nonetheless some companies whose operating performance sagged. Plains All American (PAGP/PAA) relied in part on its Supply and Logistics division to support its distribution. When arbitrage opportunities dried up in 2016-17, almost $800MM in EBITDA evaporated, leading to a second distribution cut. Today, increased Permian volumes are boosting cashflows once more. On last week’s earnings call they forecast 2019 EBITDA growth of 14-15%.

But episodes such as PAGP were the exception – operating results for the most part held up.

As memories of the 2014-15 bear market recede, we believe the conventional explanation for it will shift. MLP prices didn’t collapse because of weak operating performance. They fell because DCF was redirected to pay down debt and finance new projects, all to achieve growth (see Will MLP Distributions Pay Off?). Income seeking MLP investors don’t want their income redirected in this way. Hence, persistently weak MLP prices which have led the shift to a corporate structure for those companies wishing to access a far larger pool of buyers.

BofA Merrill is forecasting 2016-20 distribution growth for 27 of the 32 midstream infrastructure names they cover. JPMorgan forecasts 6-10% growth 2017-19. We expect our American Energy Independence Index to grow its dividends by 9% this year and 11% in 2019.

These forecasts are supported by growing pipeline demand. There are bottlenecks in moving crude oil and natural gas out of the Permian Basin in west Texas, and in getting natural gas out of the Marcellus in Pennsylvania. New pipelines to transport Canadian heavy oil from Alberta continue to face political challenges. The Shale Revolution is driving volumes higher.

With pipeline demand and dividends both growing, the sector is poised to continue its rally.

We are invested in KMI and PAGP. We are short AMLP.

Includes corrected text and a revised chart compared to an earlier version, with respect to Kinder Morgan

The Alerian MLP ETF (AMLP), with its tax-burdened structure (see AMLP’s Tax Bondage), is by design unable to come close to the return on its benchmark, the Alerian MLP Infrastructure Index (AMZI). Because AMLP doesn’t qualify for treatment as a Regulated Investment Company (RIC), it’s subject to corporate income taxes just like any other corporation. The fund’s tax liability has acted as a substantial drag on returns; since inception in 2010, it has returned just 1.99% (through June 30th), less than half its benchmark of 4.39%.

When AMLP’s portfolio rises, its NAV lags by the amount of the additional tax liability. When its holdings fall, AMLP reduces its taxes owed which cushions the blow somewhat. In fact, the tax drag acts to dampen its volatility. Tax-impaired returns ought not to be that attractive, although occasionally an investor notes that he likes the resulting lower volatility. While MLPs have been choppy in recent years, embracing this type of tax-encumbered, reduced price movement seems wrongheaded; after all, AMLP holders presumably expect a positive return, and the tax structure lowers it.

AMLP is simply the biggest of around $50BN in flawed funds. The chart below is from an Oppenheimer prospectus which helpfully explains how taxes hurt returns on their funds.

In an additional wrinkle to the pay-off profile, taxes only accrue when AMLP has unrealized gains. In such cases it holds a Deferred Tax Liability (DTL). When it has unrealized losses, as currently, it doesn’t accrue a tax loss carryforward, or Deferred Tax Asset (DTA). At such times the tax-induced low volatility disappears, and AMLP moves up and down with the market.

We are approaching an inflection point. AMLP’s unrealized losses are shrinking as the sector grinds higher. Its current DTA (which is offset with a valuation allowance, thus negating it) will flip to a DTL with the next 4-5% rise in its portfolio. At that point, further gains will trigger taxes owed. Its modestly lower volatility will return. With an approximate tax rate of 23%, AMLP will provide its holders with only 77% of the upside or all of the downside.

Like the approach of a surfer’s perfect wave, this situation offers the AMLP short-seller a rare, asymmetric opportunity. Just as an AMLP holder is accepting diminished upside for full downside risk, the AMLP short will enjoy the complete benefit on the downside with constrained losses on the upside. Combining a short AMLP with a diversified long position in energy infrastructure names or a properly structured, RIC-compliant, non-tax-burdened fund creates an interesting trading opportunity.

Such a long/short portfolio should be profitable in a rising market and roughly breakeven in a down market. There are financing costs, but AMLP has $10BN outstanding and so is not that hard to borrow.

Because AMLP is 100% invested in MLPs, a diversified long portfolio that includes energy infrastructure corporations further allows the investor to bet on a continued shrinking of MLPs as a subsector of energy infrastructure. As the biggest MLPs convert to corporations where they can access a far bigger set of buyers, this leaves fewer names with a lower median market cap (see Are MLPs Going Away?). MLP fund investors face the additional risk that one of the larger funds elects to drop the tax burden by selling 75% of its MLPs, becoming RIC-compliant. This could depress MLPs and hurt valuations of all MLP-dedicated funds.

This kind of asymmetry has value. Option premiums reflect the substantial upside/downside difference for the holder. Long-dated bonds (most dramatically, zero-coupon bonds) also offer a modest amount of asymmetry due to convexity, which causes investors to drive their yields somewhat lower than they’d otherwise be. In theory, AMLP should trade at a discount to NAV because of its tax-driven asymmetry, reducing the benefits of shorting and providing some additional return to holders to compensate for its flawed structure. Because the AMLP short is receiving the benefit of the tax drag, it’s one of the few ways you can really get Uncle Sam to work for you.

The benefits of shorting AMLP highlight the diminished return prospects of the holders. There are several other similarly tax-impaired funds, including mutual fund products from Oppenheimer SteelPath, Cushing Mainstay and Goldman Sachs, and a couple of small ETFs. While none of these are short candidates, they all suffer from the same structural flaws as AMLP. Their holders are unlikely to enjoy returns close to the indices, even while they incur unimpeded downside risk.

MLP-dedicated funds are unique in owing taxes at the fund level. Fund investors seeking energy infrastructure exposure should select a RIC-compliant, non-tax paying fund with less than 25% in MLPs.

We are short AMLP.

We have three funds that seek to profit from this environment:

Could the Shale Revolution be driving oil prices higher? It seems counter-intuitive – the U.S. is on course to be the world’s biggest oil producer by next year. And it was the additional shale supply that led to the 2014-15 oil collapse.

Yet, a growing chorus of industry voices is warning of an impending supply squeeze. This includes the Saudi Energy Minister, Khalid A. Al-Falih, who recently expressed concern about shortages of spare crude oil capacity. Last week David Demshur, CEO of oilfield services company Core Laboratories (CLB) forecast oil at $100 a barrel. He cited declining output in many key conventional plays globally, even while U.S. output is growing strongly. The International Energy Agency added their warning that, “…the world’s spare capacity cushion…might be stretched to the limit.”

Part of the problem relates to supply disruptions. Venezuela’s output continues to plummet because of chronic underinvestment. Renewed sanctions on Iran are impeding their exports sooner than expected. Saudi Arabia has promised to increase output to offset the loss of Iranian crude, but many question their ability to sustain output much above 11 Million Barrels a Day (MMB/D).

Meanwhile, global oil demand is expected to grow at around 1.5-1.7 MMB/D over the next year. Depletion of existing fields, estimated at 3-4 MMB/D annually, is worsening according to some observers. So the world needs at least 4-6 MMB/D of new supply to balance current consumption of around 98 MMB/D. Concern is growing of a shortfall.

Although blaming the Shale Revolution for a looming supply shortage sounds implausible, Shale’s lower risk profile is drawing capital investment away from conventional projects. You can see this in Exxon Mobil’s (XOM) five year plan to commit $50BN to North American oil and gas production. You can also see it in the dramatic decline in the size of large projects. As we’ve noted before (see The Short Cycle Advantage of Shale) conventional oil and gas projects take far longer to return their capital invested than shale projects.

The energy business has always been cyclical for this reason, because of the supply side’s slow response function. Shale’s ability to recalibrate output much more responsively to price can smooth out the cycle, if there’s enough short-cycle supply available. But there isn’t. North America is the major source of such projects and virtually the only source of shale activity. Rising U.S. output is nonetheless insufficient to provide adequate supply.

Conventional projects have always had to consider macro factors, such as global GDP growth, commodity prices and production cost inflation before committing capital. Add to those the likely path of government policies aimed at curbing fossil-fuel related global warming, and the unknown pace of technological improvement with electric vehicles. Today it’s probably as hard as it’s ever been to confidently allocate capital to a conventional oil or gas project with a 10+ year payback horizon.

Moreover, short-cycle projects like Shale lurk in the background, capable of wrecking the market with oversupply, yet able to protect themselves by quickly curtailing production.

This uncertainty is limiting the commitment of capital to conventional projects. If the warnings are prescient, oil prices will rise to a level that induces more investment back in to conventional projects. The market will self-correct. But it’s looking increasingly likely that higher prices will be required in the meantime.