Financing in an era of shorter PPAs
What is the future for project finance when corporate PPAs and hedges make up a growing share of the market, and corporations are losing interest in committing to contracts even with 10-year terms? Banks finance hamburger chains without locked-in revenue streams. What happens when they have to do the same for power plants?
The following is an edited transcript of a panel discussion at the 30th annual global energy and finance conference in California in June. The panelists are Ted Brandt, CEO of Marathon Capital, Jonathan Kim, head of infrastructure finance for North America for Natixis, Alexander Krolick, head of infrastructure and energy finance for the Americas for Macquarie, and Himanshu Saxena, CEO of the Starwood Energy Group. The moderator is Ike Emehelu with Norton Rose Fulbright in New York.
MR. EMEHELU: Himanshu Saxena, which is riskier, an investment in a merchant gas-fired power plant in PJM or in a solar project with a 10-year power purchase agreement?
MR. SAXENA: Investing in a solar project with a 10-year PPA is riskier, but it depends on the electricity price. We have seen solar PPAs in California with fixed power prices of $25 a megawatt hour for 10 years. We have seen wind PPAs fixed for 10 years at $14 that don’t cover the variable costs of running a wind farm. On a cash basis, the assets are negative.
At the end of a 10-year PPA, maybe we have 20% or 30% of our invested capital back, and we would have to rely on the merchant cash flows after year 10 to get back the rest of our investment and earn a return. We would have to take a view on merchant cash flow starting sometime in 2029, for example, running for another 30 years.
In my mind, that is riskier than buying a gas-fired power plant in PJM, where you have certainty on capacity prices for the next three years and can hedge the next five years, if desired. You cannot hedge the electricity from a solar plant 10 years out. So depending on the PPA price and location, you could make a case that a merchant gas-fired power plant is less risky than a 10-year contracted solar plant with a low PPA price.
MR. EMEHELU: Alex Krolick, I can’t tell if you are agreeing or vehemently . . .
MR. KROLICK: I am sort of . . . debating in my head actually.
It depends on the market. We have been so focused on quasi-merchant gas plant financings in PJM that they feel almost commoditized at this point.
We are working on the first semi-contracted combined-cycle combustion turbine project in Mexico. It will be the first financing of that type in Latin America. We are comfortable with the risk profile. Market spreads are around $55. There is a real urgency on our part to get it done. If I had the choice of being in Mexico with a $20 PPA or I could capture $80 around the clock on a merchant basis, personally I would go for the $80 and try to do a large portion of that unhedged. The economics are compelling, and there is a first-mover advantage.
MR. EMEHELU: Jonathan Kim, merchant gas versus 10-year contracted solar PPA?
MR. KIM: I agree with Himanshu and Alex. The gas-fired assets, particularly in PJM, have been proven. PJM is more predictable versus a partially contracted solar project in California where the visibility at best is opaque. It is difficult to forecast prices in California, but at least you can take a view in a more established market like PJM.
The big challenge in Mexico is there is no established merchant market, but there is price data. We are financing some of these projects where solar is partially contracted with a sweep structure to reduce the amount of exposure to the merchant tail. It is a structure that has been proven in other markets. However, I don’t think we are going to be open to a purely merchant price without some underlying contracted revenue source.
MR. BRANDT: The devil is in the details. It is one thing to do a brand new H-class, lower heat-rate gas-fired power plant and another to buy a 10-year-old or 12-year-old plant.
The market you are in also makes a big difference. If you are doing a solar plant in PJM where there is a liquid market, there is more confidence about the price forecast. We see a lot of projections that have avoided costs at the end of 10 years. This gets into complications predicting the avoided cost of the local utility and regulatory risk that there will still be a utility purchase obligation 10 years from now. There are not a lot of checks and balances in places like North Carolina, and there is pretty much only one buyer in that market.
That said, I will tell you, as someone who raises capital for a living, that it is easier to raise capital for a contracted project with a 10-year PPA than for a merchant plant. You have to have contrarians like Himanshu; that is what makes markets.
Tax Equity Barrier
MR. EMEHELU: Himanshu, do you agree that raising debt for a solar project with 10 years of contracted revenue is easier even though, in your view, merchant gas might be less risky?
MR. SAXENA: I don’t think that debt is the issue. We went to the term loan B market last year to raise debt for a 2,000-megawatt merchant gas portfolio, and the offering was three times oversubscribed. There is no shortage of debt to finance merchant gas deals pricing in the 350-basis-point range. There is also no shortage of debt for contracted solar.
The challenge is to raise tax equity. I think finding tax equity for a merchant solar project would be hard. All of these deals are getting done around the needs of the tax equity players.
Developers who are signing seven- to 10-year hedges are not doing it because they believe in the value of those contracts, but because tax equity does not want to take any risks and, without tax equity, you can’t do these deals.
If I had tax capacity, I would be king. I would be getting returns in the 6% to 8% range on an after-tax basis, and I would be senior to the debt. Facebook just announced it did its first tax equity deal. Facebook has done something like 2,000 megawatts of PPAs. It just shows that smart people are saying that tax equity is over-priced.
Everybody is signing non-disclosure agreements to get a look at all the deals that are coming to market. A lot of capital is being raised from ESG investors. Capital Dynamics is raising a fund. Carlyle is raising a fund. There is probably $9 billion worth of new capital that is currently in the process of being raised solely to invest in contracted solar and wind farms.
There is not enough product to go around to satisfy this massive amount of capital, which is why the cost of equity to buy solar projects is being driven down to 7% and below. Why is it that cash equity returns are lower than tax equity returns when all the residual risks are sitting with cash equity? It is completely upside down.
MR. EMEHELU: We have now had a few years of experience with corporate PPAs. Alex Krolick, what lessons has the market learned about them?
MR. KROLICK: We reached financial close a few months ago on what I believe is the largest and longest corporate PPA ever done. It was a wind farm in Sweden with a 29-year offtake agreement with Norsk Hydro.
We are also focused on smaller startup load-serving entities and retail electricity suppliers in different jurisdictions around the world and trying to play off the spread between what a project needs to be paid and the electricity prices that these retailers are able to collect from C&I customers.
We have a strong commodities arm and provide credit enhancement for the revenue stream. That allows us to clip a ticket twice and provide a risk profile that is acceptable to the lenders on the project side. We have seen projects that were backed by casinos to automotive manufacturing to data providers and internet giants.
MR. EMEHELU: Jonathan Kim, how else are you seeing the market address the lack of standard longterm PPAs?
MR. KIM: We are not in many renewable energy single-asset deals because the spreads are too low, the fees are low, and it is a disintermediated market, meaning there is no need for an underwriter. Maybe Ted Brandt can raise the equity capital quickly and cheaply. Sponsors can do the same with debt, both fixed rate and floating rate, by going directly to lenders.
We are ESG driven, but it does not mean that we will sacrifice profitability, so we are focused more on merchant gas projects and holdco financings where the returns are more commensurate with the risk.
MR. EMEHELU: Ted Brandt, you have foreign clients who come looking for assets in the United states. What is the attraction given that the market is moving toward more merchant risk?
MR. BRANDT: Yields are higher here, at least on a nominal, after-inflation or real basis, by about 200 basis points over what they are seeing in Europe. They may also be higher than in Japan. I am not sure that you can generalize around all of Asia.
We are watching strategic German, UK and French companies coming over with lots of capital and bidding at low discount rates, because that is what they have been used to doing. ENGIE bought Infinity about a year ago. All they have being doing is hedges and corporate deals.
The strategics use their balance sheets to de-risk the projects and then almost all of them sell down 50% to 80% to a pension fund or other nondilutive passive capital. That’s the game. They are left with decent risk-adjusted returns as the developer. Maybe as it should be, most of the reward in this game is going to the developers.
MR. EMEHELU: Go ahead Alex.
MR. KROLICK: I am nodding emphatically.
MR. SAXENA: I was going to wear a t-shirt today that says, “Who needs returns when you have solar?” [Laughter]
MR. EMEHELU: I am going to trademark that, by the way.
Evolving Corporate PPAs
MR. SAXENA: You know you can’t. All of us are witnesses here. [Laughter]
We have been doing corporate PPAs for a while. We did one with Facebook, one with Target and most recently one with General Motors. Every week there is a new RFP from Lululemon to Facebook to General Motors to Walmart looking for renewable energy.
The companies putting out these solicitations are getting smarter as time passes. There used to be two differences between a corporate PPA and a utility PPA. The electricity is sold at the hub rather than the busbar, and corporate PPAs were a little shorter in duration, but beside that, the two types of contracts looked similar.
The utility PPAs have not changed — they still look the same as 10 years ago — but corporate PPAs have evolved to shift more risk to the project owners.
We saw a 10-year PPA recently for a project in Texas with a tracking account to track the extent to which the prices paid for electricity under the contract exceed current market prices during the contract term. Any negative balance in the account must be repaid to the corporate customer at the end of the contract term. The reason the payment is at year 10 is that is what the tax equity demands for risk management. If there is a large negative tracking account at year 10, the tax equity is out and the owner is stuck with that risk.
Whether it is tracking accounts, caps and floors, exit or termination rights, corporate PPAs are evolving. Microsoft has come up with a new corporate PPA that is shape-neutral and hub-neutral. This evolution is not helping what the cash equity already sees as an imbalance between risk and reward. The corporates know that they have bargaining power, so we see an evolution that is making corporate PPAs less and less financeable.
Credit is another issue. One large entity that has been one of the most prolific buyers of renewable energy makes a special-purpose entity the electricity purchaser and offers a guarantee from the parent for one year. For one year of revenue! Where is the credit? If the market goes down by $30 a megawatt hour, they effectively are liquidating their damages with one year of revenue. We built the project for 25 to 30 years with what is supposed to be 12 years of contracted revenue.
MR. EMEHELU: Let’s go back to a question the program suggests this panel will address.
If you are in this business, you should have a better sense of where future electricity prices will be than a corporate buyer who is not in the business. Therefore, isn’t it fair that we should reduce the emphasis on contracted revenue and outsourcing of price risk?
MR. KIM: I think the question was, “Why can’t we finance projects like we finance a fast food place, for example, McDonald’s?” Does anybody know how many McDonalds are in the US? There are 13,905 at the end of 2018.
The difference between a retail outlet like McDonald’s and generators is the customer interface. There is a business that gets financing just like McDonald’s, and it is called a utility. Utilities have the direct retail interface. If there is a desire to get that type of financing, it is not going to be 80% debt to 20% equity. It will look more like McDonald’s where it is 40% or 50% leverage with some ratings put on it. Nobody knows what the price of a Big Mac will be next year, let alone in 10 years, and if that is the paradigm we are in, then the financing is going to change. There is a level of debt that you can get for merchant solar. It is not going to be a happy number for most people.
MR. EMEHELU: If I want to open a McDonald’s franchise, I would put down, what — 40% equity? — for an operating asset? I get a seven-year loan, no questions asked. If I come to you with a solar facility and I am willing to put down 25% in equity, how would you structure the debt?
MR. BRANDT: The moot point, and Himanshu said this, is that the tax equity won’t come in. I don’t know any tax equity that would take that bet even though they are in a first-lien position. You do not get to the debt question until you can raise the tax equity.
MR. KROLICK: I agree. A lot of thought is going into how to structure such a loan. What is the minimum structure that you need to have in order to make tax equity comfortable?
MR. BRANDT: It is being tested now in Texas. We have some assets in the market with a seven-year hedge. We will get that deal done.
MR. EMEHELU: Since everyone is deriding tax equity, let’s see if any tax equity investor in the audience wants to defend . . .
MR. BRANDT: Let’s stipulate that there is a hedge running to 2026.
MR. EMEHELU: . . . you are agreeing that you are responsible for this problem.
MR. SAXENA: Just for the record, we love tax equity, we need you, we are your best customers, we love you. [Laughter]
MR. KROLICK: We are hosting an after party for tax equity investors. [Laughter]
MR. KIM: I wish banks were treated that way, you know, all we hear is, “Come along or shut up.” [Laughter]
MR. KROLICK: The banks are paying for the after party. [Laughter]
MR. EMEHELU: Another thing we see are build-own-transfer transactions where the utility ends up with the project at the end of construction as a way to deal with the lack of power contracts.
MR. BRANDT: We started seeing this about four years ago. I think that the big winners have been NextEra.
A utility will do an RFP that says, “We are interested in projects that can supply electricity at the following nodes.” NextEra would say, “We have sites in all of those places, but what we are not willing to build-own-transfer on everything. What we are willing to do is build-own-transfer on four of them, and then we want PPAs on the other six.” NextEra has been sweeping auctions, not 100%, but doing very well with that strategy. It is not exactly losing money on the build-own-transfers. There is a developer margin there, but it appears to be doing this primarily to get PPAs.
MR. EMEHELU: What has been your experience with using a hedge as a substitute for a PPA?
MR. SAXENA: Hedges have more risk than traditional PPAs. They range in risk from fixed-shape settling at the hub, versus delivered shape settling at the hub, versus delivered shape settled at the node. Hedges work well if they are done right, if the shape, volumes and location are right.
We have seen some hedges go very wrong. If you are building a wind farm in the heart of the Texas panhandle where you have a massive amount of congestion and you are settling your hedge at ERCOT North, you may as well settle it in Germany. It is so far away that there is no correlation between the price of power in the panhandle and the price in ERCOT North.
There have seen some basis blowouts of $12 to $14 a megawatt hour. If the price of the hedge is $14 and your basis is $14, guess how much money you are going to make on that hedge.
What happens in our sector is some things work and then people extend them to the far extreme, so that the same technology that seemed to work really well in select situations ends up getting bastardized across projects that should never have it.
A hedge is a tool. It is a tool that can be beneficial if appropriately structured for certain projects in the right locations, but if it is misused, you will get cuts.
MR. EMEHELU: I am going to push this question one more time. Assuming that tax equity can be found for a merchant solar deal, how would you size the debt?
MR. KROLICK: We run into this more frequently with merchant tails in contracted projects than projects that are fully merchant from day one. The question is how to monetize the post-hedge revenue. There are three ways to attack the problem, but they all lead to the same question: what is your balloon at the end of the contracted period?
The predominant approach is that we all talk, debate and disagree about what merchant forecasts to use. We apply a haircut to them. We come up with a higher sizing ratio, and we argue about how many years of credit the lender will give. I would love the lenders to give me credit up to the useful life, but the reality is we probably end up in the three- to five-year range after a 12-year hedge.
Another way that people are looking at it is to look at the projected levelized power price when the balloon comes due that can, with a reasonable debt service coverage ratio, amortize the balloon down to zero.
The third way to do it is really to use shorthand, which is to come up with a debt-per-KW metric similar to what we have in PJM, and let’s just use that because it is simple and we are familiar with it even though I think it is being misused.
We are watching this play out in real time. There is no consensus that I am aware of among the banks.
MR. EMEHELU: Jonathan Kim, Natixis has done merchant projects in Latin America. Is there anything we can learn from your experience?
MR. KIM: The difference between Latin America, particularly Mexico, and the US is that Latin America is a growing market so prices are still high versus the US, which is potentially a deteriorating market in terms of price. You have a much better chance of debt repayment in Mexico than you have in PJM, for example. It is more difficult to do merchant in an established market where there is continuing downward pressure on electricity prices.
MR. EMEHELU: Are there any audience questions? Keith Martin.
MR. MARTIN: Jonathan Kim, you said that lenders cannot finance generators like they would a McDonald’s franchise because a generator sells wholesale to a single customer while a McDonald’s store sells at retail to thousands of customers. But is the analogy to McDonald’s a better one where the generator sells on a merchant basis into the spot market?
MR. KIM: I think that is a different paradigm. McDonald’s sells — I am not a McDonald’s junky, by the way. [Laughter] Each McDonald’s sells on average 167,000 Big Macs a year, and presumably it is not to one person. [Laughter] If there are that many customers, then there are choices in terms of going to a different customer whereas a generator is usually trying to sell all of its power to a single entity or into a single market. The prices in that market fluctuate. Lenders take the single offtaker risk or the more uncertain revenue stream into account.
MR. KROLICK: Burgers don’t have basis risk. [Laughter]
MR. KIM: Basis risk is a risk that absolutely nobody wants to take. No one is offering a hedge to cover it. The project owner ends up with it, so in a hedged project where there is at least a floor under the electricity price, a lender also has to take basis risk into account.
MR. BRANDT: I would take the other side and say we will see merchant solar. Solar is a zero variable cost asset. If I were raising capital for it, I would want to have eight or nine different assets and to think about financing a portfolio as opposed to a single asset.
MR. EMEHELU: More private equity rather than project finance?
MR. KIM: A little less debt. You are basically financing with more equity.
MR. MARTIN: In what sense is using your own equity “financing”? [Laughter]
MR. BRANDT: It is an old-fashioned idea. [Laughter]
MR. SAXENA: You can look to Mexico to get to the answer. The capital structure for a merchant solar project in Mexico is a 50-50 debt-equity ratio. If you have a project with a long-term offtake contract with the national utility, CFE, you are looking at something close to 80% debt and 20% equity.
Mexico is an open market. It is not like PJM where you have capacity pricing upsetting the pricing of energy and vice versa. The closest to a parallel market in the US is Texas. I agree with Ted Brandt that lenders should be financing merchant solar projects in Texas. They won’t finance it with 80% debt, but something closer to 40% to 50% debt. The overall cost of capital will be expensive.
MR. KROLICK: I sometimes struggle with the analogy of merchant solar to river hydro. River hydro has a similar risk profile and close to a zero marginal cost to operate. A portfolio of uncorrelated assets would be better to remove some of the resource risk, but at the end of the day, it will be a question of leverage and location.
MR. WOODRUFF: Mark Woodruff with I Squared Capital. To what extent do you believe corporations looking to enter into PPAs are motivated by trying to lock in electricity prices versus branding and virtue signaling?
MR. BRANDT: We have a unit that has been working with corporates, and I would say the second is the bigger driver, but the decision to buy is not made by a single person. The chief environmental officer will be an MBA with a bit of tree-hugger mentality. He reports to the CFO, and the CFO is absolutely looking at dollars and cents, historical costs and projected costs. Ultimately, you have to satisfy both constituents. The deal has to be economic, and it has to help the company reach sustainability goals.
MR. SAXENA: Most corporate buyers have an advisor who will run an NPV analysis.
MR. MARTIN: Ike, let me ask another question. All of you make it sound like if the tax equity would agree to finance based on an uncontracted revenue stream, then the banks would lend. The tax equity investors are just another group of banks. They are willing to accept part of the their payback in a form of tax benefits and the rest in cash. Why are they so different? Are they really the stumbling block to financing merchant projects?
MR. EMEHELU: I am really hoping that a tax equity investor here will volunteer.
MR. BRANDT: I would answer with a little bit of a smart-aleck remark that the DC Solar deal did not exactly help the tax equity market with this kind of thing. That was a rental fleet that turned out to be a total fraud, and about $700 million in tax equity investments had to be written off. I think it would have to be sponsor-specific. If NextEra wanted to do this, it could raise the tax equity.
MR. KIM: The nature of a tax equity investor is also different from a bank. For example, Natixis is driven by client relationships and by a desire to be excellent in a particular sector. I don’t think tax equity says, “I want be excellent in power, and all I am going to do is deploy tax equity in this sector.”
MR. BRANDT: The focus is on absolute risk-adjusted returns.
MR. KIM: They have a range of possible places to invest and risk-reward outcomes. They are relative investors.
MR. SAXENA: Keith, I can tell you this. We have a lending relationship and a tax equity relationship with Citigroup. When I go out to lunch with the lending guys, they pay for lunch. [Laughter] When I go out with the tax equity guys, I pay for lunch. [Laughter] So you know where the power sits. [Laughter]
MR. MARTIN: I have one more question for Himanshu. You said, if I heard you correctly, that you get about 30% of your capital back by the end of a 10-year power contract, and you have to rely on the merchant tail for the other 70%. Are those the right numbers?
MR. SAXENA: It depends on the project. It used to be that you got your capital back during the life of the PPA, and that doesn’t happen anymore. In every new deal that we are seeing, you get 30%, in some deals it’s 40%, but you are not getting your capital back during the PPA period because the PPAs are short and the PPA price is low. So any equity investor is taking a merchant energy price risk.
MR. KROLICK: The deciding factor on every M&A transaction that we have seen recently in the renewables space is people’s views of out-year electricity prices.
We recently sold a portfolio of behind-the-meter batteries in southern California. We had a 10-year contract for the capacity. The deciding factor in the auction was the willingness of the equity investor to believe in post-contract revenue. People on the equity side are taking a view about this sort of thing every day, and the lenders are trying to play catchup.
MR. MARTIN: And the most optimistic forecast is the one that wins the bid?
MR. KROLICK: Absolutely.
MR. BRANDT: The cost of capital also counts.
MR. KROLICK: A little bit. [Laughter]