The inescapable truth about uncertainty is that it is different from risk. Frank Knight, a leading economist of the Chicago school, first pointed out in 1921 that a poker player can figure out the odds of success when holding a pair of kings. But when dealing with macroeconomic factors or the future impact of disruptive forces, you cannot know the distribution of all potential outcomes. There are an awful lot of jokers in the pack.

Over the past ten years, the universe of independent power producers (IPPs) has slowly dwindled. The largest fuel switch in the electric industry's history—from coal to natural gas—changing consumer preferences, new technologies, weaker commodity prices, and greater distributed generation have put tremendous pressure on the traditional IPP model. Taken together, these disruptions have affected the dispatch profile of conventional IPP generation in a manner that is difficult to predict, and we believe that these uncertainties will continue for some time. Moreover, the increasing availability and use of renewable power sources, which have often performed well when conventional power runs short, and a more robust power system with limited fuel interruptions is perhaps permanently affecting electricity demand.

We think that IPPs have no choice but to evolve and adapt to the changing marketplace. This strategic shift is now playing out with surprising speed. Not only are IPPs deleveraging, but they're also aggressively transforming their portfolios and entire business strategies in ways that so far appear to be favorably affecting many companies' credit profiles. One major shift we are seeing is that IPPs are cutting debt. Another is that they are reinvigorating their business by seeking an integrated model for both wholesale and retail power sales.

Yet even as they pursue this strategy their stock prices have sometimes declined. This could be because of a downturn in natural gas prices, concern over proliferation of renewables (especially in Texas), or the potential impact on cash flow from adverse and pending legislation (notably in Illinois). Should stock prices not rebound, some market participants believe that IPPs could look to go private, as Calpine Corp. did in 2018.

With the future unclear for the IPPs, S&P Global Ratings is updating its credit views on the sector, with an emphasis on how the IPP business model has changed over the past two years. Although credit profiles are generally improving, we will need to see certain improvements before we contemplate higher ratings for these companies, and investors, who have recently requested our views on the industry, will want some answers as well.

A Problematic Business Model

Although the IPP model was adequate at the outset of competitive markets, it no longer creates long-term value.It is clear to us that IPPs that do nothing will have a much weaker investment proposition in two to three years. As a result, we have seen IPPs take a number of steps to improve their credit quality. Some of these include improving the financial risk profile of the company through deleveraging. A company has some control over this, as it is more a question of management's willingness to adjust the company's capital structure. But transformational changes, such as pursuing a new business model of integrated retail power and wholesale power (i.e. merchant power) sales could have significant execution risks that would affect our view of a company's overall business. These are the fundamentals that are now problematic for the industry:

• More renewable power is resulting in lower round-the-clock power prices.
• Lower commodity prices mean less opportunity for IPPs to earn energy margins from relative fuel differences. An efficient gas plant, for example, earns higher margins when gas prices are high.
• Increased energy efficiency and more distributed generation have resulted in persistently low and declining demand for electricity.
• Renewable portfolio standards (RPS) are contributing to declining load growth expectations. Behind-the-meter solar generation, for example, can be thought of as a net reduction in demand.
• Finally, weather patterns are unpredictable and beyond that, climate change may intensify, altering the long-term need in different regions for heating or cooling.

The Strategic Shift

We think there is no real positive catalyst for IPPs now, and that any upside will likely reflect seasonal spikes in natural gas, higher capacity prices, or the approval of energy price formation reforms. Few of these seem to us to be long-term fixes. Nevertheless, IPPs are responding to the changing market realities in a variety of ways—some of which will be more effective in the short term, some over longer periods. We see the following factors as characteristic of these strategic changes:

Leaner cost structures

Given the adverse market environment, IPPs are aggressively cutting costs and becoming more efficient. Some of these cost reductions are low hanging fruit. For instance, in a market that is oversupplied with generation, it is unlikely that many development efforts will be undertaken. Others can be more subtle, which while beneficial in the short term in that they boost short-term earnings or cash flow, may be detrimental in the long term because they may impair growth or require significant investment in subsequent repairs and improvements.

Deleveraging

We believe deleveraging will continue to take center stage through 2020, as IPPs use whatever excess cash flow they have to aggressively reduce debt. We also expect to see IPPs refinancing debt in order to push maturity further into the future. We see the current industry focus on lowering debt not only as a bondholder-friendly move, but also as a market economics-driven imperative. The only variable is the exigency of deleveraging. That will depend on just two key factors: The cash flow conversion (EBITDA to free operating cash flow) capability of the generation fleet and the expected life of the assets in the fleet. The IPP with higher environmental spending and shorter-lived assets would have to deleverage faster.

Bilateral contracting

We see the future of the IPP sector as drawn along a spectrum of risk between lower-risk contracted solar and wind assets, and higher-risk legacy base-load coal and nuclear assets. What remains to be seen is where each company aligns itself. There is also a bid-ask chasm between what buyers of long-term power are willing to pay, and what sellers of power are willing to potentially leave on the table in order to contract long-term. This effectively curtails equity upside in exchange for some degree of certainty.

IPPs are also focusing on increasing the contracted proportion of their aggregate cash flow. Many companies are pivoting to a contracted strategy. Companies have started disclosing contractual details routinely in their guidance each quarter. While companies are seeking to extend power contracts, there is a distinct focus on contract erosion risk (i.e., the ability to recontract at similar terms) or alternatively, counterparty reneging risk. S&P Global Ratings focuses on these issues when we talk with management teams.

Retail power

Up until now, we have seen IPPs and diversified power companies expanding retail offerings in an effort to complement their volatile wholesale revenue. However, in the context of disruptive renewables and better storage facilities, the retail power business becomes the next battleground for survival. With the grid expected to become much smarter over the next two decades, getting inside a customer's premises could mean opening up an entire avenue of growth through sales of everything from thermostats and smart meters at one end, to photovoltaic (PV) solar systems and eventually battery solutions. In this way, retail business provides a great defense against the rise of renewable power sources.

We note that while no company requires its retail supply business to source its hedges from the company's wholesale business, they are sometimes motivated to do so. We expect a natural evolution toward wholesale hedges greater than 50%, reducing the need for collateral support for the retail operations. Lenders now also appear to be more comfortable with an 'asset-backed' retail power strategy.

We now view an asset-backed retail business as core to a wholesale generation platform, which typically insulates market risk for two years (the average length of a retail contract). Merchant generators have increasingly mitigated the impact of declining wholesale prices by expanding their retail business.

From a credit perspective, the profitability of the retail and wholesale businesses moves up and down in counter cycles. With high power prices, capital charges are also high and cut into gross margins. Yet customers are less inclined to lock in prices at these levels. As a result, at elevated prices, we expect fixed-price sales to fall, reducing total capital requirements and increasing average margins on existing retail volumes. At low power prices, capital charges decline. Although customer migration ensues, gross margins for retail volumes rise due to increasing headroom between locked-in retail prices and wholesale prices. Thus, although the generation business' profitability declines when prices are low, the retail business' profitability improves, and vice versa.

Merging or going private

Counter-intuitively, despite current free cash flow generation and high cash flow conversion, the pubic equity markets have not appreciated (in stock price valuations) the strategic shift undertaken by IPPs. We think this is largely because potential equity investors are unsure about how this phase in the evolution of the power markets will shake out. They are specifically questioning who will be, or should be, the eventual owners of power generation assets, or what the expected life of conventional generation assets would be.

We have seen 'inefficient equity market valuations' driving mergers, acquisitions, and IPPs going private. In 2018, Calpine was taken private in a $5.5 billion deal. Contrary to expectations, the purchase by Energy Capital Partners (ECP) has turned out credit favorable for Calpine, as the new owner has followed through on the deleveraging plan that the company had announced. But we think that game is still, to use a baseball metaphor, in the bottom of the fifth. We continue to engage with ECP and Calpine's management about ECP's eventual exit strategy. However, the deleveraging at peer power companies somewhat changes the exit point that ECP may have initially targeted at the time of the acquisition.

Can IPPs Achieve Investment Grade?

This is the most often asked question about IPPs, specifically as it relates to Vistra Energy Corp (Vistra) and NRG Energy Inc. (NRG). While we'll limit our response to these two companies, we will also discuss how the credit quality of other IPPs has evolved further on.

Our business risk and financial risk assessments of both NRG and Vistra are fair and significant, respectively. Our issuer credit ratings on these companies are currently 'BB/Stable/--', which is still a couple of notches from investment grade. Compared to the financial measures that the companies project, our financial ratios—as reflected in adjusted debt to EBITDA—are about 0.35x-0.5x weaker because of debt-like imputations (we impute debt for asset retirement obligations, capitalized operating leases, and unfunded pensions and OPEB's) as well as our lower cash flows expectations, which are based on our assumptions of the forward power curves. So relative to their targets of 2.75x-3.0x adjusted debt to EBITDA, our financial ratios trend closer to 3.15x-3.25x. Similarly, our adjusted FFO to debt ratio expectations are about 2%-3% lower. Both companies, however, have shown the willingness and ability to reduce leverage. We have been explicit that if they continue to execute on their plan, we could raise their ratings to 'BB+'.

But the questions remains: Can they achieve an investment-grade rating? Over the past two years, even as IPP's began cutting debt, none had publicly and clearly stated a desire for investment-grade ratings. However, with volatile commodity prices, we think an investment-grade rating brings significant benefits, and slowly, the IPP's have indicated their desire to achieve that status.

To be sure, an investment-grade rating has always bestowed some clear advantages; a lower cost of capital, lower collateral postings with counterparties, a better ability to transact with counterparties—especially industrial customers that prefer investment grade counterparties—and reduced liquidity needs. But now we see additional advantages. The higher rating is also about better access to capital and alignment with public equity investors who have been looking for simpler and less risky capital structures. Anecdotally, we get these questions more from equity investors, rather than fixed income investors, showing that investment-grade ratings appear to have relevance for shareholder value maximization (by maximizing the stakeholder value proposition).

IPPs, however, are not there yet. They still have to execute fully on, and sustain, their transformation strategy of deleveraging and reducing exposure to wholesale power markets. We present our corporate ratings matrix (see table 1) to underscore this point. This matrix combines an issuer's financial risk profile with its business risk profile to present an anchor.

Table 1

What the IPP's are controlling right now is their financial risk profile by targeting an adjusted net debt to EBITDA between 2.5x-2.75x. That helps them shift left on the scale. Moving left on the financial risk scale gets them to an anchor rating of bb+. But at a level of 2.75x-3.0x, their financial ratios are still higher than most refineries. (Phillips 66's debt to EBITDA is about 1.9x and Valero Energy Corp. is even lower at about 1.2x).

So to achieve investment grade, the IPPs also have move up on the business risk scale, and that takes time to do because rating agencies look at an issuer's history and financial performance under varying economic conditions. For a business model that is evolving, we typically assess volatility in cash flow (standard error of regression etc.) over a period of five years before we revise our business risk assessments—especially for companies undergoing a major transformation. That means we'll start considering investment-grade possibilities in the latter half of 2020 and early 2021 depending on how well these companies have executed their business plans and whether they have established a stronger record of managing retail and wholesale margins through the cycle.

Companies do not necessarily have to achieve a business risk profile of satisfactory to be investment grade. For example, if we believe that a company's free cash flow and/or cash flow conversion rate is sustainable and superior to similarly rated companies, we could apply a modifier to reflect that in our comparable rating adjustments.

As a separate but related point, the stock price performance of IPPs has languished in the past six months. Despite the cost cutting and deleveraging efforts, stock prices have retreated, resulting in speculation of potential take-private transactions. More chatter arose when Vistra's management stated in its second quarter earnings call that this could be a potential future option too.

From a strategic perspective, we are not really concerned whether a company is publically or privately owned and operated. In fact, we think Vistra's management alluded to it more as a response to a question on the call, in that it has a fiduciary responsibility to review all options. However, we still need to assess what a take-private transaction could mean. We would reasonably expect a financial sponsor that is writing a big check to contemplate some form of leveraging of the company, especially if the new owner expects sustained free cash flow in the short to medium term.

On balance, while the credit momentum for IPPs is positive, we will consider a take-private transaction as negative for credit quality given our belief that any financial sponsor would likely employ debt in its acquisition financing. On the other hand, purchase of a company by a strategic buyer—such as some recent acquisitions of retailers by big oil companies in Europe—could be credit neutral-to-favorable depending on the acquirer's strategic and financing plans.

Business Risk Is Slowly Improving

Adverse market conditions for the IPPs began in 2016 and in general they haven't improved. In the Pennsylvania New Jersey Maryland Interconnection (the PJM market), the demand forecast continues to be revised lower, while the market remains very well supplied as new combined cycle gas turbines (CCGTs) come on line to offset retired facilities. As a result, market heat rates have been under pressure. Implied heat rates in PJM have contracted by 10% over the forward curve. Moreover, capacity auction parameters point to lower pricing in the 2022 auction (date not yet set). Moreover, both capacity and energy price reforms are stalled at the Federal Energy Regulatory Commission (FERC) and PJM stakeholder process. In response, PJM and other regional transmission organizations or independent system operators (RTOs and ISOs) are increasingly intervening to prop up nuclear plants, with government intervention undercutting confidence in markets.

Moreover, we are seeing lower forward power prices in virtually all other independent power markets, including the Midcontinent Independent System Operator Inc. (MISO), the New York System Independent Operator (NYSIO), and ISO New England (ISO-NE). Even ERCOT witnessed weakness in its forward markets after the release of the May CDR report that projected higher than expected renewable supply. At a high level, we think these declines reflect some combination of lower natural gas prices and a mild start to summer 2019 that weighed on prompt prices, which then cascaded out onto the forward curve. (ERCOT's recent mid-August heat wave bucked the trend and lifted forward summer prices through 2021). In addition, prices fell because fewer generating assets than the markets expected were retired. All of this has led to forward curves in steep backwardation, exacerbated by a lack of liquidity in out years. PJM prices in 2020 and 2021 declined sharply, with Northern Illinois (NI) Hub around the clock prices falling nearly $3/MWh or roughly 11% for 2020 relative to 2019, and about $2.50/MWH in 2021. Similarly, PJM West Hub prices fell more than $4/MWh and roughly 13% to 14% in 2020 and 2021, respectively.

So for IPP business profiles to stay neutral or strengthen, we probably need to see some evidence that wholesale power market and capacity market conditions will stabilize. Then again, our focus on the wholesale market may be missing the forest for the trees. Frankly, over the past two years retail power has contributed to lower volatility and higher cash flow conversion than we anticipated and there could indeed be an element of adverse selection on our part in weighing in the wholesale market exposure disproportionately in the business risk assessments for IPPs.

These companies have asserted that the combined wholesale and retail model—the integrated model—has lowered the standard deviation to a quarter of the standard deviation for a standalone retail or wholesale business. During periods of high volatility, generation availability provides a backstop to load portfolio. On the other hand, during periods of low volatility, the IPPs are able to capture higher margins through a lower cost to serve their customers while optimizing the value of their fleet's dispatch.

We recognize that the integrated model has built up some credibility over the past two years because retailing power appears to be providing a hedge for wholesale power operations when they are regionally matched, reducing the financial impact of lower forward power curves. Volatility has indeed been lower (15% trough to crest) and the companies have had stronger cash flow conversion rates than refineries. However, it is difficult to believe that a capital-lite model providing a consumer non-discretionary service, such as electricity, will go uncontested. We think that companies with the integrated model have not been tested, either in the competitive landscape where the fight for market share could get bloodier, or in the form of extreme (or very mild) weather where the efficiency and efficacy of the integration is tested. For instance, in a recession we expect both wholesale power prices and retail power margins to decline, especially if weather doesn't cooperate.

Over the next 18 months, favorable credit momentum would be contingent on the consistency of execution. This includes our continuing assessment of the sector, particularly our view of the success or failure of an IPP's retail power segment and the predictability of its cash flow.

Vistra: A Case Study In Change

Vistra and NRG's business strategies started diverging about two years ago. While NRG shed a substantial portion of its generation and pivoted emphatically towards a retail dominated model, Vistra kept substantial generation capability. In fact, after its bankruptcy, Vistra had initially emphasized a model that would match load with generation, and management had evinced no interest in a business model that would be more exposed to generation. However, that strategy shifted as it became apparent that Dynegy Inc. was available for sale. The most likely reason Vistra ultimately acquired the company, in our opinion, was to add scale and realize cost synergies while diversifying its business away from ERCOT.

Chart 1

While there are indeed substantial cost synergies, the acquisition resulted in a portfolio that is now heavily invested in wholesale generation (see chart 1). Prior to the recently announced closures, the company had significant long exposure to generation with about 190 TWh of expected production, almost 40% of which is from coal-fired assets (see chart 2).

Chart 2

We continue to see risks in merchant generation. Solar and wind constitute about 80%-90% of new builds in almost all markets. The significant increase in solar, in particular, is a precursor to potentially backwardated merchant generation margins. The continued entry of renewables, which generate when the resource is available, is likely to affect market heat rates and keep generation prices depressed. Even as we believe ERCOT's backwardated forward power price curves will likely lift, given increasing load and less new supply, and that energy reforms in PJM are likely, we see Vistra's large exposure to wholesale generation as a significant risk.

The company, however, has invested heavily to build a substantial retail business that now serves as a hedge to its merchant generation business. In table 2 we've broken out Vistra's retail business by market. We can see the scale and diversity of the retail business and how much it mitigates the more volatile wholesale generation business—what IPPs refer to as the integrated portfolio.

Table 2

Chart 3

Could Vistra Improve Its Credit Profile?

We assess risks in Vistra's portfolio by breaking out the components of its cash flows (see table 3). We have done so for Vistra's expected 2019-2020 gross margins and EBITDA. The aggregate percentages may not add to 100% as these are ranges and represent a trend over the two years.

Table 3

Before the merger with Dynegy, Vistra saw itself as a load-to-generation matching company and not as an IPP. In fact, management saw limited value in gaining scale and scope if it resulted in incremental wholesale power exposure. The combination now has only about 85 TWh of retail load and a substantial larger wholesale generation capability at about 200 TWh. This translates into only a 25% proportion of retail business by gross margin (see chart 4) compared to NRG's 50%.

Chart 4

What also jumps out for Vistra is that the company has almost 55%-60% of its gross margins exposed to energy margins compared to only about 35% for NRG. In a market environment that is experiencing the continual onslaught of distributed generation and proliferating renewables, we think this is an exposure that may need to be mitigated with offsetting hedges.

The following factors could influence Vistra's business risk:

• Vistra is net short retail in ERCOT and PJM, but still benefits if we assume retail counter-cyclicality on the portion of their wholesale fleet in each market that is matched with retail. Before the Ambit acquisition, the company had a load to generation match of 53% in ERCOT (see chart 3). It has since improved to about 64%. We think that to mitigate this risk the company will either need to grow its retail business or will have to reduce its merchant exposure in ERCOT. While we expect growth in ERCOT retail, given that Vistra produces over a third of its ERCOT generation from coal-fired assets (33 TWH), we see the potential for incremental plant closures (over the 4.2 GW already announced) as a possibility.
• Despite the fact that Vistra has about 14% of its generation capacity in MISO, it produced only about 7% of its wholesale gross margins (3.5% of wholesale EBITDA) in the region. Before the recent announcement of coal-fired plant closures, the company had a relatively low 48% load-to-generation match in PJM/MISO, which has since improved to 55% with the announcement. In order to balance the wholesale–retail integration and improve its business risk profile, we thought the company would need to retire some Midwest units. Even after the recent retirements announcement related to Illinois multi-pollutant rules that reduce coal-fired generation by 10 TWH, almost 20 TWH of generation in MISO is still from coal-fired units. We think the company will have to close more of its power plants rather than build its retail business in order to mitigate exposure to wholesale generation.
• While the net attrition rate in Vistra's retail business is still low (we estimate 0.3%-0.4%), we think the company has to move out of ERCOT and grow retail nationally to increase its net residential customer count. In contrast, NRG is seeing overall portfolio growth largely because it is in markets where competition is still limited, with the incumbent utility serving about 65% of residential load. However, Vistra's ability to maintain margins when it expands into the eastern markets remains to be seen.
• There are growing tensions between pure IPPs and diversified companies such as Exelon Corp. and Public Service Electric & Gas Co. (PSE&G). Illinois, New York and New Jersey have already implemented zero emission credits (ZEC) standards that support specific nuclear plants with incremental revenue. There are also state lawmaking efforts at various stages of development in Pennsylvania and Connecticut that may lead to similar programs, and in Ohio this might also include coal plants. The success of these initiatives represents opportunity for Exelon Generation LLC and PSE&G to the detriment of the pure IPPs.
• Vistra's cash flow could decline in 2023-2024 if the Illinois legislation is enacted to procure clean capacity from utilities in PJM. Senate Bill 2132 and House Bill 2861 in Illinois promote carbon free generation over fossil generation and favors ExGen's nuclear fleet to the detriment of Vistra's.

The potential upside

In the medium term, we do see some factors that could prop up Vistra's cash flows.

• Despite backwardation in the forward power curves, the fundamental market tightness in ERCOT can drive the forward outlook higher. ERCOT continues to show demand growth even as supply has stalled. We think the summer reserve margin is probably closer to 9% and tighter than what the capacity, demand and reserves (CDR) report suggests. The key here will be the growth in renewable generation, especially solar. Over the span of a week in mid-August, ERCOT experienced a heat wave and record demand, including a 1.5 hour pricing at the $9,000/MWH cap (and several hours at elevated prices) that did drive forward prices higher.
• PJM's wholesale market reforms—such as the operating reserve demand curve (ORDC) revisions and inflexible baseload units setting marginal pricing—are still likely, albeit on a somewhat slower schedule.
• Vistra self-insures against outages in ERCOT by keeping about 1.2 GW of generation merchant and exposed to lower spot liquidation prices.
• A coal to solar act contemplated in Illinois could provide an incentive to redevelop coal plant sites into utility-scale solar and energy storage projects.

NRG: A Case Study In Integration

It is NRG's long-term strategy that is more interesting. NRG is the only company that has decisively pivoted in favor of its retail platform by integrating it with its generation assets. Power generation is a capital-intensive business. At a time when much of the new capital is going into renewables, NRG has decided to exit this space by selling all of its equity ownership stake in NRG Yield, now Clearway Energy Inc., and choosing instead to focus on its large retail power business (65 TWh). NRG has essentially decided to focus on retail power as its defense against the disruptive forces in the wholesale market.

NRG, however, still has a stake in some renewable assets, which it will likely monetize over time. The company has also started to sign some medium-term solar power purchase agreements (PPAs) to cover a portion of its retail needs. We are somewhat surprised at this because it was only six years ago that the company set up the first YieldCo as a capital replenishment vehicle for its long-term contracted conventional and renewable assets. We feel that the sale of NRG Yield may have had some agency issues between management and activist shareholders –Elliott Management and Bluescape Energy—who likely felt that renewables were trading at attractive prices and wanted to monetize them given their short-term investment horizon in NRG.

Nevertheless, NRG has been lightening its generation fleet. Compared to Vistra, NRG now has a much smaller wholesale generation footprint. This is because it has divested a substantial portion of its erstwhile generation portfolio. The disposition of the GenOn assets was related to its bankruptcy (Genon's assets were held through an excluded project subsidiary), and the sale of NRG Yield to Global Infrastructure Partners Inc. partners (GIP) pertained to the aforementioned strategic shift away from renewables. In an effort to focus more on its integrated retail and generation platform, the company has also shed legacy assets in areas with no retail presence, most notably about 3.5 GW of its South Central portfolio that included large assets like the Big Cajun and Cottonwood power plants. We present NRG's wholesale business in chart 5.

Chart 5

NRG's economic generation, i.e. the generation that can economically back its retail load, is now much smaller than Vistra's, at about 62 TWh. Despite the smaller portfolio, we note that over 50% of NRG's generation is still produced by coal-fired assets (see chart 6). NRG's generation, in our opinion, is also somewhat higher priced on a variable and fixed operating and maintenance (O&M) cost comparison because its assets are older.

Chart 6

The fact that NRG's strategy has been more focused on integrating its generation assets with its retail portfolio over the past two years is evident from the break-out of its margins (See chart 7). Retail now provides half of NRG's margins, which is about twice the contribution provided by retail margins in Vistra's overall gross margins.

Chart 7

Moreover, NRG has invested substantially in the retail power business. In table 4 we've broken out its retail business by markets:

Table 4

NRG has about 3.5 million residential and small business customers in 20 states and provinces. This business (mass retail) operates through a variety of brands and delivers most of the company's retail EBITDA. NRG's business solutions segment (LCI) has a sizable load that, along with the mass retail segment, is a key asset within NRG's integrated platform. Both mass retail and business solutions have a presence beyond Texas, and deliver products and services in addition to basic electricity.

NRG appears to have a mostly matched book in ERCOT and PJM, reducing its exposure to power curve volatility. However, even as NRG has a fully matched book in ERCOT, which should mitigate its exposure to commodity curve fluctuations, we feel there could be instances when it could get caught short should load turn out higher than its expectations. The company runs its eastern portfolio at low capacity factors and mostly benefits from capacity revenues rather than energy revenues.

Could NRG Improve Its Credit Profile?

In table 5, we assess risks in NRG's portfolio by breaking out the components of its cash flows. We have done so for NRG's expected 2019-2020 gross margins and EBITDA.

Table 5

Due to a combination of the expansion of its retail power business, backwardated wholesale margins, and our expectations of additional plant closures, we expect NRG's retail business to account for about 75% of the company's aggregate EBITDA by 2023, up from 50% now. For an EBITDA shift so dramatic, we are taking time to ensure the sustainability of the company's improving business risk. The continuing success of the transformation strategy is key to our assessment of the company's business risk profile.

The following factors will likely influence our ongoing assessment of NRG's business risk:

• Energy margins currently account for about 75%-78% of NRG's wholesale generation margins. NRG, however, still produces about 42% of its wholesale energy margins from coal-fired assets (32 TWH). We are interested in the performance of the company's Midwest fleet where we expect weakening margins. We also see the possibility of plant closures in the region.
• Compared to about 60 TWH of expected load in ERCOT, NRG has only about 40 TWH of economic generation (i.e. generation that routinely supports retail load). To mitigate the risks of getting caught short, the company also owns generation with the ability to flex up (typically with its out-of-market peaking gas-fired units) or by buying supply on the open market. As a result, we think that NRG's wholesale generation will have to pivot to match the growth in retail.
• We think NRG could be more exposed to extreme weather than Vistra. As such, the ERCOT portfolio will likely require additional capacity that is complementary to the retail load shape. To achieve this, we think that the company will have to consider a combination of options including: renewable and conventional medium-term PPAs; buying shape or block energy; developing distributed behind the meter generation for its C&I load; tolling existing units; and evaluating its generation development
• NRG has to capture the margin enhancement it expects through its transformational plan.
• NRG has cleaned up its balance sheet substantially since it sold NRG Yield, relinquished control over GenOn, and successfully executed its deleveraging plan. However, compared to Vistra and Calpine it still has many projects and subsidiaries that are not consolidated, including Petra Nova, Ivanpah, Agua Caliente, and its international assets. For greater transparency to its financials, we think NRG needs to continue simplifying its balance sheet.

The potential upside

Most of our market related comments that are relevant to Vistra's upside apply to NRG as well. NRG believes that given the changing power market, retail power has become the growth platform, with the wholesale generation business pivoting only to provide support to match retail shape. NRG believes that by shedding excess generation in areas with no retail presence it has integrated, rebalanced, and right-sized its portfolio and, moreover, that other companies will be forced to follow suit. The company has delivered a 45% increase in EBITDA from 2014 to 2018, typically with high cash flow conversion from EBITDA to free cash flow. It has also increased recurring customer count by over 1.1 million between 2013 and 2018.

While we have noted that retail earnings primarily flow straight to cash, we are still seeking validation that these earnings are reliable and that NRG's business model has evolved to a retail/wholesale load match platform that is much less meaningfully commodity cyclical.

ExGen: Advocating Successfully For Nuclear Energy

While ExGen is operationally a nuclear company, as the owner of the largest price-taking merchant nuclear portfolio, economically, it is a natural gas company. Given its large exposure to wholesale power markets, it also has the most to lose from disruptive forces.

Yet we assess ExGen's business risk profile as satisfactory because the company's business strategy now includes petitioning for supporting 'compensation structures' for its fleet's zero-emissions attributes. As a consequence of commodity price decline and more renewable power sources, an increasing number of ExGen's nuclear plants have reported negative cash flow. Moreover, the May 2018 PJM capacity auction for the 2021-2022 planning year resulted in the largest volume of nuclear capacity ever not selected in the auction, including all of ExGen's Dresden unit, and portions of its Byron and Braidwood assets.

Even as cash flows weakened, however, the company has successfully argued for administrative relief for its beleaguered units, citing environmental, fuel diversity, and reliability concerns. Regulators are also now more sensitive to the nuclear industry's distress and have approved ZEC revenues for ExGen's nuclear plants in Illinois, New York, and most recently in New Jersey (Clinton, Quad Cities, Fitzpatrick, Ginna, Nine-Mile Point and Salem). By our estimates, the Illinois, New Jersey, and New York ZECs boost ExGen's revenues by about $630 million. We see these revenues as quasi-regulated and consider them as less risky in our assessment of ExGen's business risk profile.

Chart 8 presents the impact of a decline (and increase) in commodity prices by two standard deviations (what we have shown in the charts as the P(5) and P(95) cases, respectively). We note that the company's gross margin floor expectations—the P(5) stress—had fallen significantly in 2018 from 2017 (not shown here). The substantial uplift in margins that the company reported for expected 2018 margins in its fourth quarter 2016 earnings call represent the impact of the approved ZEC revenues.

Chart 8

Nevertheless, three of ExGen's nuclear plants in Illinois will face economic hurdles if current power market conditions continue. In order to protect the value of its portfolio, ExGen, along with other companies, is working with PJM to get the FERC to carry out baseload, scarcity pricing and other capacity market reforms. But short of that, the 1,845-MW Dresden, 2,347-MW Byron, and 2,386-MW Braidwood nuclear plants in Illinois are showing increased signs of economic distress which could lead to early retirement.

The Illinois legislation has recently proposed a framework to procure clean capacity for Illinois utilities in PJM. House Bill 2861 (The Clean Energy Progress Act) aims at addressing potential changes related to PJM capacity, as well as revenue and support for existing clean energy resources, including zero emissions nuclear generators. This legislative agenda might facilitate ExGen securing capacity payments for its entire 10GW (ComEd zone) nuclear fleet directly from the state of Illinois (with the Illinois Power Authority as the procurement agent) by taking the ComEd zone out of the PJM capacity market under a zonal fixed resource requirement (FRR) construct, while the ComEd zone would continue to participate in the PJM energy market. Even as ExGen's nukes are having a tougher time addressing market conditions, ExGen could see a significant value re-rate if it achieves its legislative agenda in Illinois this fall. The prospect of an approval of this Illinois legislation would help the company's business.

ExGen generates a significant portion of its earnings from its retail power operations, which helps mitigate wholesale business volatility. Through retail and wholesale channels, ExGen provided nearly 212 TWh of load in 2018 (about 140-150 TWh retail and 60-70 TWh wholesale), or nearly 5% of total U.S. power demand. ExGen's Constellation brand is currently a market leader in commercial and industrial (C&I)load served, and about fourth nationally in residential load served. ExGen has expanded its retail business and improved its load-to-generation mix to the point where the total retail and wholesale load served has grown by about 8% in the past three years. While ExGen still has a net long generation position, because of its C&I focus, it sells much more volume in the retail space at about 165 TWh-170 TWh compared to Vistra's retail load of about 85 TWh-90 TWh or NRG's 70 TWh-75 TWh. However, ExGen's retail margins are lower than those companies' due to its focus on C&I rather than residential sales.

Chart 9

In most locations, ExGen has adequate intermediate and peaking capacity for managing load-shaping risks. However, we believe the company will still need to buy and sell generation in the market to manage its portfolio needs, which exposes it to considerable commodity risk. Moreover, ExGen has a significant open position in the Midwest to merchant markets, though a somewhat tighter position in the Mid-Atlantic and New England, where it risks finding itself short when loads and power prices are high. In particular, ExGen's load-to-generation matching in the Mid-West is just 36%, a region where it faces the maximum pressure from wind generation. The situation is better in the Mid-Atlantic, however, where matching is at 71%.

Calpine Corp.: A Case Study In Bilateral Contracts, But Still A Wholesale Business

In comparison with NRG and Vistra, Calpine has the highest share of wholesale merchant power in its total business. However, relative to those peers, Calpine's generation has a reasonable geographic spread. The company produces almost 100 TWh of generation from its units (see chart 10).

Chart 10

This generation is relatively efficient and of recent vintage as evident from the fact that the company's economic generation is meaningfully higher than NRG's for an incrementally higher level of installed GWs. But for 725 MW of geothermal capacity at the Geysers, the fleet is fired on natural gas as seen in the following charts.

Chart 11

We note that Calpine's cash flows are reasonably well split between its three markets, with the West now contributing the most. For now, Calpine's business strategy revolves around its relatively long-term contracts. Nearly 55% of Calpine's gross margins are locked in through 2023 with hedges, capacity revenues, and contracts. Almost 95% of Calpine's free cash flow from its California operations through 2022 are from its geysers and contracted gas-fired assets.

In situations such as these, we focus on re-contracting risk; specifically the risk of lower prices at renewal time. We see re-contracting risks in California for Calpine's two major assets: Los Esteros and Russell City (both 2023 expiration) . The Otay Mesa contracts expires in October 2019, but Calpine is contracting for resource adequacy payments with San Diego Gas & Electric Co. through 2024. However, debt fully amortizes over the PPA periods, and the absence of significant amortization after the contract period should limit cash flow backwardation. The rest of Calpine's portfolio assets—Delta, Gilroy, Metcalf, Pastoria, among them—are already at market prices. We see some risk at the Geysers because some renewable energy certificates (RECs) are contracted at about $20/MWh but currently trade lower. REC prices vary by contract, although in other contracts they are embedded in a bundled rate.

Stronger markets have been a big boost for margins, EBITDA, and free cash flow at Calpine. The power markets in Texas and Californian have fundamentally improved, in large part due to retirements. In ERCOT, significant retirements have ensured that about 6 GW of generation remains 'at-risk', and new supply has been delayed, even as demand steadily increases. In California, gas peakers previously slated for retirement have been re-contracted under resource adequacy (RA) agreements at prices that have doubled (to $5.00-$7.00 per KW-mo. from $2.00-$2.50 per kw-mo.) due to increasingly tight dispatchable supply.

Table 6

The California Public Utilities Commission (CPUC) recently noted that gas-fired generation will be the largest contributor of capacity to the California system through 2030. In its resource optimization analysis, the CPUC included approximately 20 GW of existing gas-fired generation through 2030 to meet its system capacity needs. But we don't ascribe much value to Calpine's gas fleet after 2023 given the rapidly changing landscape in California. As a result, we believe that the company will have to deleverage quickly to adapt to potentially lower cash flows and also grow its retail operations.

Compared to NRG (50%) and Vistra (25%), Calpine's retail power operations account for only about 10% of its aggregate gross margin and EBITDA (see chart 12).

Chart 12

We think that retail operations are important to shaping the company's long-term strategy--from the pure play gas-fired generating company that it is today, to the integrated power merchant that it aims to become. Integrating retail presents a management challenge given the exposure to wholesale markets, but is a move the company likely has to make, in our view. Calpine entered into a retail venture with Champion Energy Corp in 2016 and bought the commercial and industrial (C&I) operations from Noble Energy Inc. in 2017 and housed under Calpine Energy Solutions(CES. The company expects to run the three customer channels (direct large C&I, indirect small C&I, and residential mass markets) through two platforms (Champion and CES). We estimate an average uplift in margins of about $5.00 per MWh and expect almost $350 million in margins from this segment.

While we believe acquiring retail operations was a strategically sound decision given reduced liquidity in power markets, the ability of Calpine's commercial team to match its wholesale generation with retail load will determine its success (see chart 13).

Table 7

Chart 13

Apart from the much lower retail component, the other significant difference between Calpine and Vistra/NRG is its financial leverage. While the company has reduced its leverage significantly, it is higher on debt to EBITDA relative to both NRG and Vistra by about 2x. Stated differently, Calpine has about 40% higher debt compared to NRG but a similar EBITDA, or has the same level of debt as Vistra but 50% lower EBITDA. While we are unsure of the endgame ECP is pursuing for Calpine, we have assumed that it will hold the company for five years before exiting through an initial public offering. However, we think the decision by Calpine's peers to reduce their leverage substantially has shifted the goal post for Calpine too In order to undertake an IPO, Energy Capital Partners may have to increase its holding plans and reduce Calpine's leverage over and above the $2.7 billion of debt it has committed to cut by the end of 2019.

Investors have asked us what impact a potential leveraging (and/or sale) of the Geysers—Calpine's geothermal assets—could mean for Calpine's credit profile. First, our assessment of Calpine's credit quality, i.e. the $2.7 billion deleveraging, does not assume the use of any proceeds from the potential sale of the Geysers. If Calpine divests the Geysers, or leverages those assets, or sells a minority ownership, we would expect incremental deleveraging.

More to the point, the Geysers are as unique a generation portfolio as you can get. One look at the daily renewables production on the California Independent System Operator website explains why. At 725 MW, the Geysers are the largest geothermal assets in the world. They are also the most stable form of renewable production. They also represent 15% of Calpine's EBITDA. Given our belief that participation in renewable generation is one of two main ways an IPP can respond to the technological headwinds, we'll have a few questions for the company if the Geysers are entirely divested. However, we think that sale of a minority stake or issuing project level debt in the assets is still possible.

Talen Energy Corp.: Relatively Smaller And Highly Leveraged

Unlike NRG and Vistra, Talen Energy Corp. has little retail presence and relies heavily on (volatile) energy margins for its profitability. Talen is also the smallest of the IPPs we discuss here, with a total capacity of 14.7 GW and 2018 company reported adjusted EBITDA of $756 million. While we view PJM, where the bulk of Talen's business is, to be a more favorable market than ERCOT due to the stability the capacity market construct provides, PJM generally also has less upside than ERCOT, as prices do not spike as frequently in the summer months and low cost gas from the Marcellus and Utica shales continues to depress regional power prices.

PJM represents 75% of Talen's overall capacity and yearly generation, ERCOT about 9%, and the Northeast gas-fired assets another 10%. Talen has announced the retirement of Colstrip units 1 and 2 in Montana effective Dec. 31, 2019. While Units 3 and 4 will continue to operate, the Montana segment will be a smaller component of the portfolio, which we view as credit neutral to positive. Chart 14 gives the breakdown of Talen's portfolio by capacity and generation.

Chart 14

Talen's generation fleet is also older that its peers', and generally has less capacity than them. In addition, Talen owns no renewable generation, although it recently announced a 100MW solar joint-venture with Pattern Renewables 2 LP. About 38% of Talen's total generation and nearly 45%-50% of its gross margin comes from its nuclear plant, Susquehanna (compared to, say, a 15% EBITDA contribution from the Geysers to Calpine). Although we view this plant as pivotal to Talen's continued profitability, heavy reliance on a single asset presents a higher degree of risk in the case of unforced outages or potential adverse public policy decisions regarding nuclear. While there has been talk of Pennsylvania passing a bill to provide ZECs to nuclear generation, we believe the large gas lobby in the state may make passing such legislation more difficult than it was in New York, Illinois, Ohio, or New Jersey. The passage of ZECs would clearly be a credit positive for Talen. Chart 16 shows Talen's heavy reliance on Susquehanna for its total realized energy margin.

Coal-fired generation continues to represent a noticeable share of Talen's fleet, constituting around 27% of total generation. We view this as a credit weakness given the push towards greener energy and the possibility of a carbon tax. Talen, however, completed a coal to gas conversion on its Brunner plant and could consider further conversions for some of its older coal plants, which could be credit positive if these plants are able to run at lower heat rates and lower overall dispatch cost.

Chart 15

Similar to Calpine, Talen also has much higher leverage than NRG and Vistra. While Talen has made efforts to reduce leverage since the leveraged buyout by Riverstone Holdings LLC in 2016, and although Talen has removed near term maturities through 2025, adjusted debt to EBITDA remains elevated at more than 6.0x, and we expect leverage to remain high in our forecast period.

We think that what separates Talen from the front running IPPs is its inability to generatete sufficient cash flow to help it quickly reduce debt. In addition, while cost savings have exceeded our expectations, Talen's management has not articulated a sustainable growth plan.

To improve its credit profile, Talen would need to continue to reduce leverage while growing its overall portfolio and diversifying into lower-carbon generation. The company would also need to continue to demonstrate more prudent financial policy than it has in the past.

Chart 16

We also think that the lack of significant retail business makes Talen more vulnerable to the volatile energy markets. Energy margins constitute about 60% of its aggregate gross margin (see chart 16). This compares poorly to peers like NRG, which has a significant retail exposure that hedges against its wholesale business.

Chart 17

Cash Flow And Retail Power Are Key

With 50% of its EBITDA now from retail power, NRG has migrated to a mostly matched book business model that could buttress its overall EBITDA and cash flow against lower wholesale prices, at least in the short to medium term. Vistra, at 25% of EBITDA from retail power, is net long generation, but has enough retail load that it should also exhibit some resilience.

The bottom line is that NRG and Vistra have to continue to put up cash flows that are somewhat resilient to wholesale price declines. They must maintain the discipline to keep capital spending low, continue to focus on capital efficiency and cost savings, stick to conservative debt/EBITDA targets, and optimize their fleets. We note that there has been some chatter in the markets that if the integrated model delivers sustainable cash flows over the next few quarters and the stock price of these IPPs doesn't respond, then private market interest could increase. If that should happen our ratings would factor financial sponsor motives and intent.

Historically, IPP performance had closely tracked the financial performance of refineries and the most gas-leveraged exploration and production companies. In a quick follow up over the next month or so, we will compare and contrast IPP's with refiners, who face similar volatile commodity cycles yet have investment grade rated companies. We will further refine our views on the sector, so to speak. However, the performance of IPP's lately has diverged from that of refineries due to the evolution of the IPP business model to a balanced portfolio of generation and retail load, as well as idiosyncratic factors at each company, such as shareholder activism at NRG and the acquisition of Dynegy by Vistra. There are indeed a few jokers left in the deck. Stay tuned.

Related Research

• Vistra Energy Corp. Outlook Revised To Positive On Cost Savings And Stable Retail Business; Ratings Affirmed, Sept. 4, 2019
• NRG Energy Inc. Outlook Revised To Positive On Accelerated Debt Paydowns And Strong Retail Performance; Ratings Affirmed, Sept. 4, 2019
• Unregulated Power: S&P Global Ratings' Evolving View Of Retail Power, May 14, 2019.