If Charles Dickens were to write a classic on U.S. power generation fuels (why wouldn't he?), we think he would likely call it 'A tale of Two fuels' and it would start something like this: It is the best of times—for natural gas-fired generation; it is the worst of times—for coal-fired generation. This version could even have its own Bast(ille): the Best Available Sustainable Technology. Renewable proliferation and the inevitable advance of regulations would have stormed coal, with shale powering the revolution that overthrows king coal's reign. And at the end, a modern day Madame Lafarge—perhaps a hard-core environmentalist?—would sit knitting beside the guillotine while a polluting, inefficient, uncontrolled coal generation unit wonders if it is going to a far, far better rest than it has ever known.

We expect a dramatic increase in coal-fired retirements over the next five years because of the advent of cheap and plentiful natural gas, and a push from sustainability conscious investors who are now far less interested in backing coal plants. Because of fuel-switching economics, uncertainty about how long coal-fired plants can remain in operation, and a decided shift in investor sentiment against owning them, we are significantly lowering the asset valuations we use for coal-fired generation in our post-default recovery simulations and assessments.

We have written about the impending decline in coal-fired generation since 2011. At that time, we felt that advancing environmental regulations would significantly impair the economics for coal-fired power generation. In fact—and perversely—it turned out during that period that complying with regulations that were not final, while appropriate from an environmental perspective, was a directional bet from an economic perspective. Plant operators that deferred environmental upgrades during those years generally benefited from postponed regulations or other changes in the market. For instance, a number of coal-fired generators deferred spending on scrubbers to reduce sulfur dioxide emissions because of uncertain carbon regulations. They perceived that if the carbon rules turned out to be more restrictive than anticipated, they would be investing in coal-fired assets that may turn uneconomical.

We expected merchant companies to retire units earlier as they were at greater risk from advancing regulations. Unlike regulated utilities, unregulated generators do not have cost-recovery mechanisms and must rely on market pricing to recover environmental capital spending. Yet, the earliest wave had more coal-fired units retiring in the regulated space. We have also seen increasing, and accelerating, retirements at both regulated utilities and merchant generators since. This is because the dynamics changed by 2014. It is now mostly economic rather than environmental considerations driving retirements.

More Gas

The shale gas boom has had a disruptive role for the electric generation business. Large shale gas discoveries and resurgent natural gas production have resulted from new drilling techniques, such as horizontal drilling and multistage fracturing. The marginal costs of production have declined as drilling rig efficiencies continue to improve and the disproportionate impact of sharply lower natural gas prices is now weighing significantly on power prices. This is because in most markets natural gas represents the marginal fuel—the fuel that sets market prices for power generation. Since the beginning of 2017, U.S. natural gas production has increased 25% to 95 billion cubic feet (bcf) per day from about 75 bcf/day, or just over 25% since the beginning of 2017(see chart 1).

Chart 1

While gas demand during peak winter days has exceeded 135 bcf/day, we note that the area under the natural gas production line and demand has been increasing both from the number of months production has stayed above demand (excess supply affects the reference gas price and forward curve), as well as the increasing intra-month difference between the supply and demand (causes increasing basis differentials during off shoulder months).

Not only has natural gas production increased dramatically over the past two years, we expect that it will stay this way through 2030 because of prolific shale plays. Table 1 shows that barring a fracking ban from a change in energy policy, incremental gas production from the shales will keep the domestic natural gas market supplied so well that we do not expect forward prices to show much volatility above $3/MCF (temporary dislocations in underserved markets is likely).

Table 1

The Economics Of Switching Fuels

Coal is losing out because, in short, natural gas costs are very low relative to coal costs. Moreover, greater energy efficiency resulting from building codes and appliance standards has slowed natural gas demand in industrial use and exacerbated coal's loss of market share for power production.

To provide insight into the economic incentives of coal-to-gas switching, we modeled natural gas and coal price equivalence to determine when switching from one fuel to the other becomes economical. Although our calculations generalize the coal-gas price equivalency, efficient coal units that burn cheaper Power River Basin (PRB) coal will typically have a lower equivalent gas price than inefficient units that burn more expensive Central Appalachian (CAPP) coal—leading to the switching competition we've seen between mid-merit combined cycle gas turbines (CCGTs) and efficient coal-fired units using Appalachian coal.

We note that coal-to-gas switching is more intense in the middle of the supply stack, and is usually along a sliding scale—first from inefficient coal units to efficient CCGT units—and as gas prices decline relative to coal prices, from efficient coal units to mid-merit CCGT units (see table 2). The last line in the table provides the natural gas price at which switching starts from coal-fired to gas fired generation. We note that because CAPP coal is more expensive, we are seeing switching between mid-merit CCGTs from efficient coal-fired units.

Table 2

Not surprisingly then, as natural gas prices have declined, fuel switching has accelerated over the past three years. Chart 2 chart shows the percentage of coal- and natural gas-fired generation as a proportion of overall power generation in the U.S.

Chart 2

Coal Plants Are Closing

We already know that the consumption of natural gas for power consumption in 2018 was about 13% more than in 2012, even as the weighted average price of natural gas was not materially different between those two years. We believe this increase is not just because of favorable economics, but because, for the following reasons, more coal-fired plants are being retired.

• Investors are increasingly avoiding coal-fired exposure because of sustainability goals, as well as environmental, social, and governing factors.
• Increasingly stringent environmental mandates, such as recent ones in Maryland and Illinois, have hastened the retirement of coal-fired assets..
• Operating costs are rising because of environmental compliance issues.
• There is more power available from renewable energy sources.
• Lower than expected demand that has affected coal-fired generation disproportionately as the marginal fuel in many regions.

The retirements have accelerated. Not only have retirements increased (tables 3,4), retirement of many units that were slated for closure several years later have been accelerated, resulting in significantly more retirements in 2014-2018, that we expected back in 2015.

Table 3

Table 4

As we see, while announced retirements between 2014 and 2018 in March 2014 were about 26 GW, actual retirements during this period were nearly double that at about 50 GW, with the large majority of the retirements taking place in the PJM Interconnection, MISO, and SERC. We highlight about 27 GW of retirements through 2023. Table 5 presents a list of some of the largest that have been announced.

Table 5

We expect this trend to accelerate as aging power plants become increasingly uneconomic and renewables grow more competitive. In fact, we expect an uptick in announcements for 2024-2025 to 25 GW, as renewables continue to proliferate and battery mandates begin to encroach. (S&P Platts, an affiliate company, expects batteries deployment to increase to 10 GW by 2023 from 1 GW currently). Moreover, we continue to see upside risk to these retirement levels, noting that nearly 100 GW of coal capacity potentially appear at risk. In addition to the units that are actual closures, there are about 7 GWs that intend to convert to natural gas. It is unclear to us at this point how many 'distressed' units will contemplate conversion to natural gas. We see plants that are at greater risk of retirements as having the following characteristics:

• They run at low capacities of less than 50% (see table 6).
• They are relatively small at less than 500 MW, and lack economies of scale to save costs. (Many units in the MISO system are like this.)
• They use CAPP coal, though some relatively inefficient ones survive on PRB coal.
• Their coal capacity that is located in the Marcellus/Utica shale, atop rich gas reserves. We estimate this at about 40 GW, and predominantly in plants supplying the PJM system.

Table 6 below presents a typical portfolio—this one presents confirmed and potential coal-fired retirements in the Tennessee Valley Authority (TVA) portfolio. As can be seen, the capacity factors are low for the units we flag for potential retirements.

Table 6

Among the regulated credits we cover, not surprisingly, American Electric Power Co. Inc., Southern Co., and Dominion Energy Inc. remain the most exposed to coal capacity at risk of retiring, while Vistra Energy Corp, NRG Energy Inc., and Talen Energy Corp. are the most exposed merchant players.

Falling Distressed Valuations

When we look at recent coal-fired transactions two facts jump out: First, over the past three years, strategic players (public companies and portfolio operators) were only on the selling side, while financial sponsors were the only active buyers. It is apparent that investor, lender, and shareholder sentiment has changed rapidly and dramatically about owning coal-fired generation. Second, we find that recent transactions have gone silent on pricing and sale multiples. That usually happens when the bid/ask spread starts widening.

Our discounted value assessment corroborates this conclusion. We believe that even though some efficient, mine-mouth coal-fired generation located in load pockets could potentially garner $250/KW-$300/KW based on cash flows, a buyer would likely bid only in the $100/KW range. This is because of uncertainty about asset life, not only because of regulatory factors, but because changing public sentiment and increasingly vocal anti-coal organizations are influencing, or forcing, closure and conversion decisions.

Table 7

We will assign no value to units that are below 500 MW and also operating at low capacity factors, which we generally define as less than 50%. Our revised valuations will assign $0-50/KW to coal-fired units below 500 MW that have higher capacity factors. A larger unit with heat rates of about 10,000 btu/kwh is assigned values in the range of $50/KW-75/KW. However, efficient units could have higher values.

An example here is illustrative. Affiliates of ArcLight Capital Partners LLC and Blackstone Group Inc. bought a portfolio of three gas fired assets and the James M. Garvin Power Plant—among the more efficient coal-fired units—from AEP in 2016. We assumed a going concern valuation of $500/KW for the CCGT's and $200/KW for the peaker, which means that payment for Gavin turned out to be around $400/KW. But this was in 2016. We think eroding economics and fewer buyers mean that now even a unit like Gavin would not sell for much above $150/KW-200/KW. Similarly, despite the $380/KW sale price for the Keystone and Conemaugh interest in 2014, we think a purchase offer for these older units will be at levels of only about $100/KW.

Because of a limited universe of buyers, we think that in a default scenario, efficient coal-fired units could have higher values if operated by the lender-owners. Our discounted cash flow analysis (rate @ 12% based on a cost of debt @ 9%-10%, cost of equity @ 15%-17% and a 50/50 debt/equity mix) resulted in a value of $200/KW-$250/KW. In our distressed simulations, we have typically not assumed asset lives beyond 2030. However, we think that coal fired units with long-term contracts that have change-in-law pass-throughs could have higher valuations.

We expect these revisions to have a limited impact. Our valuation for project financed coal-fired portfolios are already low, but could change still. Table 8 provides a snapshot of how weakly these portfolios have been performing from a market pricing perspective. However, this performance would not affect recovery ratings on issue-specific project finance ratings. While our recovery assessments for companies with meaningful exposure to coal, like Vistra Energy and NRG Energy, are lower, their unsecured debt ratings are not affected because their recovery ratings were capped at '3' and are still within that band.

Table 8

The problem for coal-fired generation is that as sustainability goals have become increasingly important to investors coal-fired generation has lost its attractiveness for long-term investors. Simply stated, market power is with the buyer because there are very few left. Latin sounds more poetic even when it rings a dirge: Res tantum valet quantum vendi potest. Or, a thing is worth as much as someone would pay for it.

Related Research

• Merchant Power Update: Something's Gotta Give, March 21, 2017
• Where Will Coal End Up In The U.S. Power Industry's Future? Sept. 29, 2014
• Industry Economic And Ratings Outlook: The U.S. Merchant Power Industry Is On A Long And Winding Road To 2015, August 22, 2013
• Taking It To The MAT(S)--The U.S. Merchant Power Sector Wrestles With Emissions Rules, Sep. 10. 2012
• Abundant Natural Gas And Looming Regulations Have U.S. Unregulated Coal-Fired Power Generation On The Ropes, Jan. 31, 2011