Description

A bit more than three years ago, we wrote up Alliance, and our opening words were “Coal miners in the U.S. are unloved.”

To capture the mood today, take that statement and cube it. It seems as though there are more coal companies in bankruptcy than out of it.

Despite the mood, and because of the 27% distribution yield it offers, here is a second look at Alliance as a long idea. For full disclosure, we sold our Alliance stock in late 2013 in the 60s, but have now bought a small position in the low teens. There is not a rock-solid margin of safety here, but there is logic that could lead Alliance’s equity to be a double or more from today’s price.

For background on Alliance and the Illinois Basin where most of its operations are, see our write up from December 2012. Joe Craft has a great track record, generating ~20% unlevered returns on capital in a commodity business for more than a decade. How he deals with this tsunami of cheap gas and regulatory uncertainty is the question.

A first place to look is in Germany, which is like the Ghost of Christmas Future for the U.S. electricity market.

The German Electricity Market

Other than Denmark, which has paired offshore wind turbines and Norwegian hydro, no other developed country has achieved a penetration of wind- and solar-powered electricity as high as Germany’s.  In 2014, the Germans generated 16% of their electricity from wind and solar, compared to 5% in the U.S. On some windy and sunny days, their 74 GW of wind and solar capacity is enough to power their entire grid, which has an average demand of 59 GW.

There are three problems, however.

First, their wind and solar delivers only 13% of rated capacity on average. Sometimes the wind does not blow and the sun does not shine, especially at night. And so the Germans have to maintain a backup power system in order to keep the lights on.

Second, 13% capacity utilization means that the Germans have to overbuild their renewable system in order to reach their Energiewende goal of 80% green power by 2050. To get to 80% of demand with capacity that operates at 13% capacity on average, you need a capacity equal to 6x your average demand, plus massive amounts of storage. Compare that to the traditional grid, which has peak capacity of 2x average demand, and where supply is not intermittent.

These two issues drive the third problem, which is that German power prices are now very high – in 2014, 40 cents and 27 cents/kwh for retail and industrial users, respectively, compared to 13 cents and 7 cents in the U.S. If the average U.S. household had to pay 40 cents rather than 13 cents, its electric bill would increase $3,000/year. Whatever you think about Trump and Sanders voters today, this would not make them less angry.

Maybe the most fascinating thing about the Energiewende is what the Germans are using for backup fuel. They don’t have our cheap gas, and do not want to become even more dependent on Mr. Putin. And so almost half of their electricity is generated from coal, and coal-fired power was 1% higher in 2014 than in 2010, despite renewable power growing by 51% over the same period.

And Merkel’s decision to shut all of their nuclear plants by 2022 pretty much guarantees that coal’s role as a source of baseload and backup power will continue for decades, barring a breakthrough in storage that dramatically changes its availability and cost.

Here is a breakdown of how electricity was generated in the U.S. and Germany in 2014:

German coal-fired electricity is really cheap. The plants have been largely or fully-depreciated, and much of the coal is lignite, a cross between hard coal and peat. The cost per kWh of lignite-powered electricity is 2 cents. Without that cheap backup fuel, German power would be even more expensive, and their lights would not stay on.

How likely is that to change? Without some breakthrough in storage technology, how can renewable energy with intermittent availability and mid-teens capacity utilization power a developed country?

(It’s almost funny to contemplate the life of a utility manager today. His career choice was supposed to be safe and predictable – work at a monopoly that handles intermittent but predictable demand by overbuilding rate-based capacity that can be dispatched by the touch of a button. And Bob’s your uncle!)

The U.S. Electricity Market

With Germany as an example of what the future of renewable energy might bring, let’s take a look at the U.S. market.

In 2015, US coal production fell to 895 million tons, down from a peak of almost 1,200 million tons in 2008. Production so far this year suggests that 2016 production will fall below 750 million tons, and Wood Mackenzie is forecasting 829 million tons for 2017.

Here is a breakdown of production by basin, as well as the natural gas price and percent of US electricity generated by various sources. Note that the 2016 and 2017 gas prices are the June NYMEX futures price for each year, and that biofuel is ~2% of electricity generated.

To make sense of these numbers, here is a chart that summarizes the coal vs. gas dispatch decision for a utility. It assumes a 9,000 heat rate coal plant and a 7,000 heat rate combined cycle gas plant, and shows the Illinois Basin coal price that gives the utility the same fuel cost for each natural gas price:

At natural gas prices north of $4, CAPP producers with $70 operating costs could compete with gas. But the shale revolution brought us the Marcellus and Utica, which have added ~20 Bcfd of low cost gas production into a North American market producing about 90 Bcfd, driving gas prices down. This drove a shift from 2008 to 2014 of 7 percentage points of market share from coal to gas. A further 2 percentage points of coal’s market share was lost to renewables.

In 2015, gas prices fell below $3, and now many non-CAPP producers were made uncompetitive, including higher-cost NAPP and ILB players, and PRB operators at the mercy of railroads to get their tons to utilities far away from Wyoming. Coal gave up another 6% of share to gas.

Now, let’s add the winter we just had. It was the warmest recorded in 130 years of record-keeping by NOAA, 4.6 degrees Fahrenheit warmer than average, and 6 degrees warmer than last winter.

During the three months of December-February, an average of 10 Bcfd less gas was burned to heat homes than normal, or ~900 Bcf. As a result, gas storage in the US is currently 2,500 Bcf, up 1,000 Bcf over the normal amount at this time, and the natural gas price at Henry Hub fell to $1.64/mcf in March, its lowest since 1999. At prices at and below $2, demand for even low-cost ILB coal fell.

Natural gas operators are terrified about what happens this summer, when normal gas injection in preparation for winter may overload the storage system. Where will the gas go? How low will the spot price go before other operators shut in production? Will the bank cut my credit line even more?

How are the natural gas producers doing? Cabot is perhaps the lowest cost gas operator in the country, having had the good fortune to find itself with acreage in Susquehanna County in the Northeastern Marcellus. This allows it to add gas with the drill bit at 40 cents/mcf. With cash operating costs of $1.15/mcf, that means that it can create molecules of gas for $1.55/mcf. So $1.64 Henry Hub should be thin gruel but OK, right?

Unfortunately Cabot is suffering from the ~80 cent basis differential in the Marcellus. The sudden appearance of a Middle East-scale gas field in Pennsylvania has caused a local glut of gas. The price that Cabot realizes in Pennsylvania is 80 cents less than Henry Hub on the Gulf Coast, which means it has been netting 85-90 cents for its $1.55 gas. Not easy to make up on volume.

Interestingly, to finance a new long-haul pipeline to the Gulf Coast, a market large enough to turn the Marcellus and Utica gas into end-products or LNG exports, requires operators to commit to firm transportation contracts for 10-20 years at 80-100 cents/mcf. So the true cost of Cabot’s gas appears to be more like $2.35 than $1.55.

And Cabot is the lowest-cost producer in the lowest-cost gas field. Dozens of gas operators are now in or close to bankruptcy, and the natural gas rig count has fallen to 92 rigs from 151 a year ago, and ~400 in 2013. Baker Hughes tells us that the U.S. rig count is now the lowest since 1948.

Cabot itself has announced a drilling budget for 2016 that is 58% lower than its 2015 spending.

All of this is a long way of saying that natural gas prices are not sustainable here. With lowest-cost Cabot probably requiring $2.75 gas to earn its costs and generate a return on capital, it seems likely that $3 gas will return. And $3 gas means $54 ILB coal.

What Does This Mean For Alliance?

Here are revenues/ton and operating expenses/ton for most of the major ILB players (the main one missing is Murray, which is privately held, but which now controls Foresight’s MLP). Note that by operating costs, we mean cash operating costs plus an estimate of maintenance cap ex (generally $4-5/ton, depending on the operator) and an allocation of corporate G&A.

Several trends stand out from this table.

First, Alliance has been able to maintain a premium in pricing relative to the others, partly reflecting cheaper transportation to customers’ plants, and partly reflecting a reputation for reliable delivery. This premium seems likely to persist.

Second, Peabody’s production has declined as the stress from their terribly-timed, all-cash acquisition of Australian metallurgical coal has led them to run their ILB operations for cash, neglecting basic maintenance.

Third, Foresight is the lowest-cost operator in the basin due to its long walls, but it has lower-quality coal geared for export markets. Backing out contracts to sell in Europe that will run-off unrenewed given low API2 prices, its revenue per ton is likely to fall. It is also in financial distress as a result of its transaction with Murray Coal, which has caused its debtholders to put their bonds to the company, resulting in a stand-off that has bankruptcy as a possibility. Foresight’s Deer Run mine has also been on fire intermittently for more than a year, and the problems there seem tough to overcome, putting 5.5 million tons of annual production in question. Given all of this, production at Foresight seems likely to decline.

Finally, three transactions in the last 18 months have reduced the number of operators competing in the ILB:  Murray’s taking control of Foresight, Alliance’s acquisition of White Oak’s long wall operation, and Hallador’s acquisition of Vectren. Eight players pricing coal have been reduced to five, and further consolidation seems likely.

It is possible for pricing to get worse from here, but it is also possible that a return to 120-130 million tons of demand for ILB coal would lead to firmer pricing.

But with demand below 110 million tons, investors are focusing on the status of contracts in the ILB for 2016 and 2017. This is the reason for AHGP’s 27% distribution yield.

Coal is sold by operators to utilities under long term contracts. Typically these define volumes for 3-5 years, and have set prices 100% for the coming year, roughly 2/3 for the year two, and 1/3 for year three. This structure allowed utilities to have stability in their fuel prices, especially compared to natural gas, which in the past could sky-rocket above $10/mcf if a hurricane hit the Gulf in the wrong place, or a cold snap froze the Northeast in winter time.

But the glut in natural gas has led utilities to be much less willing to commit to price coal for 2016 and 2017. On its Q4 earnings call, Alliance management announced that they were planning to produce 6.5 million tons less of ILB coal in 2016 than in 2015. This reflects their desire to not chase the price down with their clients, who are happy to be burning gas below $2. Even not chasing price down, Alliance is forecasting a 4-6% price reduction in 2016.

Also, Alliance has secured pricing commitments for just 19 million tons in 2017, or about 55% of 2016’s sales across the company (including non-ILB tons).  At this point in prior years, Alliance has typically had 90% priced for the following year.

Looking at the forward curve for gas, this makes sense. For June 2017, NYMEX gas is $2.68, or an ILB coal price equivalent of $48. Having sold coal for years in the 50s, Joe Craft is not eager to sell in the 40s. The utilities, looking at the glut of gas, are not eager to transact in the 50s. So there is a stand-off.

How does the stand-off impact Alliance unit holders? Let’s look at the impact on distributions of various tons sold and prices charged. Note that operating profit means EBITDA less maintenance cap ex, and that Alliance sells tons in other basins as well, driving “other operating profit.”

At a unit price for AHGP of $14, even the low-end distribution of $1.17/unit is an 8% yield. Obviously cutting the distribution would be a bad thing, but at $14 we are getting paid for some pretty dire outcomes.

And if gas stabilizes at $3/mcf, and ILB pricing firms up at $50 or higher, with an increase in volumes back to 130-140 million tons for the basin as a whole, a distribution at AHGP of $5/unit is unlikely to keep the AHGP price at $14/unit.

What Could Go Wrong From Here?

One possibility is a price put on carbon, either in the form of a tax or a cap-and-trade system. Burning coal releases roughly twice as much CO2 as burning natural gas, and so this will impact the decision by utilities to switch between the fuels. Working through the math, every $10/ton price put on CO2 will push the coal price down by $10/ton relative to gas. So our table from earlier would look like this:

Carbon prices implemented in the UK and the EU are at £18/ton and €6/ton, respectively. The EU price has been higher, but poor design of their cap-and-trade system has kept it in the single digits. In the UK, the carbon tax is pretty controversial. There is a lot of support for green power, but a growing concern about energy poverty due to high prices, and it looks as though Britain is about to lose its last steel plant largely due to power prices among the highest in Europe.

But let’s say a $30/ton carbon price was implemented in the U.S. and stayed in place. At $3/mcf gas at the wellhead, this would mean $24/ton ILB coal at the mine, which would mean no ILB coal mined. This would also mean little coal mined elsewhere, as the ILB is the lowest-cost basin per delivered BTU of energy (when including the cost of rail transportation) for much of the U.S.

So we would have to replace 800 million tons of coal with natural gas. That would require 40 Bcfd of incremental gas production, or 2x the Marcellus and Utica.

Would $3 at Henry Hub stimulate an incremental 40 Bcfd of supply on top of current North American production of 90 Bcfd, when we can see that $2.68 Henry Hub has driven a 58% drilling budget cut by the lowest-cost producer, and the lowest rig count since 1948?  It seems more likely that the wellhead price would increase, and some new balance would apply, in which low-cost coal from the ILB has a place to compete.

Another thing that could go wrong is that 1,000 Bcf of additional gas in storage is a lot of gas, and even a hot summer may not burn it off quickly. We may see gas prices at $1 or lower this August as we test the theoretical limits of North American storage up around 4,700 Bcf.

With gas prices that low, Alliance’s 2017 priced coal volumes might remain around 19-20 million, which may mean a distribution cut. Though we can argue that a cut on AHGP’s distribution to $1.50/unit is still a >10% yield at $14, the emotional response by MLP unit holders to distribution cuts elsewhere suggests that there is further downside. It may also be difficult to regain the confidence of MLP investors (who are a fairly stressed group of people at present) when and if coal prices and volumes recover in a better gas market.

Finally, the shale revolution may continue to be deflationary for gas prices, while the expense of getting coal out of the ground stays fixed, or increases due to further regulations. We can read about monster 30 or 40 MMcfd wells in the Marcellus or Utica, and there is some evidence that well productivity in other areas is improving. Maybe gas will stay in the $2s forever. But the behavior of producers in general, cutting rig count to a multi-decadal low, seems like a leading indicator of a different outcome.

None of this is certain, but $3 gas seems likely to return, with implications for ILB volume and price. It is difficult to justify an investment in Alliance based on a clear margin of safety, but it is an interesting option on rationality returning to the U.S. natural gas market.

Catalyst

An end to fear and loathing in the U.S. gas market