Is “merchant” still a dirty word?
Many bankers lost their jobs in the wake of the Enron collapse, and many independent power plants were put on the market. Lenders grumbled that the forecasts of energy prices on which they relied were wide of the mark and that the worst-case projections from consultants were not the worst case. Contracted assets found a ready market. Merchant plants were harder to sell. Is it really possible that only a few years later lenders are ready again to finance new merchant power plants?
Six market veterans discussed this topic at a Chadbourne conference in June. The six are Joseph Esteves, managing director for finance at LS Power Development, LLC, a leading US independent power developer, Steve Cheng, a managing director of Credit Suisse, William Sutherland, vice president of project finance for Manulife Financial, Douglas Sherman, an under- writer with CSG Investment, Inc., which is affiliated with Beal Bank, Markus Christen, a former member of senior management of Credit Suisse and now a private investor with MC Capital Partners LLC, and Charles Wilson, director of business unit finance for Duke Energy Corporation. The moderator is Rohit Chaudhry, a project finance partner in the Chadbourne office in Washington.
MR. CHAUDHRY: Steve Cheng, let me start with you. What financial structures allowed lenders to get comfortable with merchant risk in the merchant plant financings that were done in the late 1990s?
MR. CHENG: I don’t think there is a difference between the structures that were used then as opposed to now. The difference is in how assets are valued today and what lenders rely on before agreeing to lend. For example, lenders were more likely in the past to accept a value for a new-build project of $600 to $700 per kilowatt. Today, they are valuing assets at a fraction of the actual construction cost. There is a big difference between lending to a project that the bank thinks is worth $600 and one that the bank is only willing to accept is worth $300 to $400 a kilowatt. The other thing that has changed is financings were done in the past based on market studies and projections. There is a much more developed and liquid market for power today than there was in the past. Lenders no longer have to rely solely on what a consultant says should happen. They can look at what the futures market itself says the power will be worth.
MR. CHAUDHRY: Doug Sherman, what did lenders expect the last time around in terms of leverage, cash sweeps and debt service coverage ratios?
MR. SHERMAN: The most common financing structure at that time was a mini-perm loan supported by a tolling agreement. Leverage in such structures was as high as 85% to 90%. Lenders were generally comfortable with interest-only-type structures and were deferring repayment of principal. Today, the market has reverted to a full cash flow-sweep type of structure. Coverage ratios remain in the 1.2 to 1.5 range.
MR. CHAUDHRY: Markus Christen, anything to add to how merchant deals were structured the last time around?
MR. CHRISTEN: A lot of the structuring is driven by perception in the market. I financed two wind farms in the late 1980s and early 1990s. No one else wanted to touch them. Today, wind farms are the darling of the banking industry. My point is everything goes in cycles. Perception in the market defines what is possible and what kind of structures you will use. Initially, only a few brave lenders are willing to do it. Next, everyone is falling over each other to do deals. Next, something blows up, people get burned, and no one wants to do it.
The phrase “merchant plant” can have various meanings. You need to dig deeper in your analysis. What risks is the lender really taking? The word “merchant” means that the project is selling into the market rather than under a long- term contract. It really matters what type of fuel the plant uses and how well developed a market there is in the area where the plant is located.
There are more hedging products today that can be employed as a risk mitigant.
It remains very difficult today to finance new combined- cycle gas-fired power plants in markets where gas is at the margin. There are too many such plants already in certain markets. Lenders will wait before taking that risk.
MR. CHAUNDRY: I want to get a sense from each of you what you think was the main reason why the merchant plants failed the last time. I don’t want an elaborate answer — just the main reason. Steve Cheng, let’s start with you.
MR. CHENG: The old financings were over leveraged. The problem was too much debt and not enough economics to support the debt.
MR. ESTEVES: A number of projects suffered from too high leverage and not being able to withstand the normal types of cycles that one should expect in a commodity market. However, in projects that were the true disaster cases, people may not have done as much diligence as they should have or else they just ignored things because of all the excitement around building new generating plants. I am referring to things like transmission access and even ability to secure fuel at attractive prices. The toughest problems have been where projects literally cannot sell power so that they are not covering their fixed expenses.
MR. SUTHERLAND: I think the problem was too much liquidity in the financial markets leading to an over build and too much capacity. The banks just piled in.
MR. SHERMAN: The developers led the charge, and the bankers were more than happy to feed them the liquidity that they needed. There was irrational exuberance. Size mattered. For example, Calpine was out there with an announced goal of 76,000 megawatts and anybody who announced a plant would get an immediate stock market bump, with the result that everyone was vying to announce as many plants as possible in a short time period. The other problem was too great a concentration on deploying gas-fired power plants. Everyone assumed coal plants would be retired from service. Instead, such plants have been running at as high as 95% capacity factors.
MR. CHRISTEN: There was complete disregard of market fundamentals as bankers fell over themselves to win mandates. The bankers were relying on the consultants, and every consultant assumed he was working on the one merchant plant that would actually be built and all the others would be canceled. In fact, all the plants were built.
MR. CHAUDHRY: Did the lenders really take a hit in the last merchant wave or did they come out more or less whole because, when the assets were sold, the sales proceeds covered the debt? the loan and the hedge runs out, the lender will be exposed to a plant that has reverted to pure merchant status.
At CSG, we will take merchant risk, but based on an analysis of the core value of the project as a merchant plant. We strip it down. We focus on how the plant will behave on a merit-order-dispatch basis at a specific site. We look at all the locational factors, gas price, the ages of competing plants, and the shape of the capital structure.
MR. CHAUDHRY: Steve Cheng, do you see many lenders today who are willing to finance a new power plant on a purely merchant basis without a hedge?
MR. CHENG: People are looking at it. The only deal to date that I know of that was done as a merchant plant right out of the box was the restructuring that Credit Suisse did for Boston Generating. There were no financial hedges. Boston Generating had the benefit of a couple reliability-must-run contracts that provided a foundation for revenue stability, but - as a percentage of total revenue - the must-run output was a small percentage. At some point, a purely merchant deal will get done, but all the deals that have been done in the last year to 18 months have employed some type and some amount of hedging: financial hedging, tolling agreements or power purchase agreements. At some point, there will be another purely merchant deal.
MR. CHAUDHRY: To get a sense of the size of the market, how many deals have you seen in the last 18 months that fit Markus Christen’s definition of merchant - no power purchase agreement but with a hedge?
MR. CHRISTEN: I don’t think you can answer that in the abstract. Some lenders took a hit. Some equity investors bought the distressed assets and made a ton of money. The question is how nervous did a particular lender get, and when did he sell. No doubt some banks sold at 50¢ or 60¢ on the dollar. Some bankers probably lost their shirts. Others did not. Some new entrants in the market made a ton of money.
MR.CHENG: There have been more than a dozen such deals.
New Merchant Financings?
MR. CHAUDHRY: Using Markus Christen’s definition of a merchant plant as a plant without a long-term contract, but with some form of hedge against price risk, are lenders starting to do merchant deals again? Doug Sherman?
MR. SHERMAN: Anyone who has entered into a term B structure where there is a hedge with Morgan Stanley or Goldman Sachs is effectively taking merchant risk on the back end. In all of these structures, there is a sweep of 100% of cash flow. However, in the event there is not enough cash to repay
MR. CHAUDHRY: How many do you expect in the next year or two?
MR. CHENG: I expect about the same number.
MR. CHAUNDRY: Joe Esteves, are there particular markets in the United States where it should be easier to finance a merchant plant without a hedge?
MR. ESTEVES: LS Power just closed a financing where we put a hedge in place right before closing, but we had a financing commitment to close without the hedge. If the question is whether projects can be financed on a purely merchant basis without a hedge, then our experience demonstrates that they can.
Lenders are a lot more cautious about the value they place on a plant. They are no longer relying solely on forecasts as was mentioned earlier. Someone said the market is relying less on consultants. The truth is there are more roles today for consultants because now you need a commodity hedge consultant and the bankers need a valuation expert.
It helps that the markets are more transparent today. A lender can look at where prices are trading several years out. Even if the project is not signing a contract today to sell the output, there is a sense that it could sell at the future prices if it wanted. As markets become more liquid, there will be less need for hedges.
MR. CHAUDHRY: Markus Christen, do you want to add to that?
MR. CHRISTEN: I think it is important to ask what kind of deals are being done today on a merchant or quasi-merchant basis. They tend to be acquisitions of existing assets rather than greenfield projects. Would anyone finance a new gas-fired power plant on a merchant basis? The answer is probably no.
MR. CHAUDHRY: Why is there a distinction between acquisition financing and greenfield financing? Is it because existing assets are trading at a discount to cost?
MR. CHRISTEN: Yes, that is a big factor.
MR. CHAUDHRY: Bill Sutherland, you have been working on wind farms. Do you see a difference between doing a merchant wind deal versus a thermal deal?
MR. SUTHERLAND: In fact, we did a merchant wind deal last December, closing a completely exposed, unhedged project in Alberta, and we gained comfort in that because a wind farm, unlike a gas plant, has a very low operating cost. It is dispatched in all events. It has a profit margin under pretty well any conceivable market price scenario. There are few lenders actually looking at merchant wind, with the result that we have been able to be conservative in the amount of leverage put into such transactions. The point is there are opportunities for lenders to finance merchant plants today in the wind sector.
MR. CHRISTEN: To second that, we are doing a merchant wind deal in Texas right now.
MR. CHAUDHRY: What type of leverage is on offer typically in a merchant wind deal?
MR. SUTHERLAND: It depends on the capacity factor that a particular wind farm is expected to achieve. Projects in some of the better wind regimes can achieve much higher leverage than the projects in other locations. It also depends on the market into which the project will sell. The leverage in the Alberta project was 40%.
MR. CHAUDHRY: Doug Sherman, you said you will be willing to look at purely merchant deals. What kind of leverage do you think such a project can achieve?
MR. SHERMAN: Somewhere between 50% and 70% for the right thermal power plant and a little less for a wind farm. Like Bill Sutherland, we put out a term sheet on a merchant wind deal. This one was in the United States.
MR. CHAUDHRY: How do you get comfortable that problems like overcapacity that were the downfall of many merchant plants the last time around will not repeat this time? MR. SHERMAN: We take a conservative approach. We look at the mean-reverting gas prices. We spend considerable time on a discounted cash flow analysis. We look at where the plant is in terms of system-wide heat rates. We look at comparable sales prices for where similar assets are trading in the market in addition to the cash flow analysis.
MR. CHAUDHRY: Joe Esteves, you spent a lot of time this year on financing the Plum Point project, a large greenfield coal-fired power plant in Arkansas. If I am not mistaken, Plum Point had no offtake contracts when the financing closed. Can you talk about how that deal was structured?
MR. ESTEVES: It is probably not the simplest structure that the market has seen. We think there is an important timing advantage to getting to market soon. We are once again in a building boom. Everyone is talking about coal just like everyone was talking about gas earlier. It was important to get started quickly on construction and not wait until the plant was fully contracted.
Having the plant under construction also helps with potential offtakers. It can be challenging to persuade someone to sign a long-term contract when you are talking about a wait of another four years before the plant is ready to start generating electricity. Offtakers suspect that other options may become available in the meantime, and they would rather wait to see what develops. It is more difficult today to have everything secured in advance.
Turning to our financing structure, most of the potential offtakers for the electricity from Plum Point were municipal utilities and electric cooperatives. We ended up selling about 37% of the plant capacity to two muni groups and a coop. The project is 665 megawatts. We had long-term contracts for about 100 megawatts, and the rest was essentially merchant.
We needed to do something to assure the lenders that there would be stable revenue for at least some period of time.
We also ended up putting in place options on gas to hedge a coal unit, which was interesting. There were a number of hedging alternatives available – everything from short-term physical sales to financial deals on electricity. We ended up with a hedge on gas because it was very important to us to be able to unwind in a transparent fashion. We expect to sign long-term contracts to sell electricity immediately after construction. We wanted to make sure we could unwind the hedge without taking a big economic hit. The bid-ask spread in some types of hedges can be wide. Gas hedges trade in a more liquid market than electricity hedges. Thus, for both unwind reasons and bid-ask spread reasons, we ended up with put options on gas.
MR. CHAUDHRY: Could you give a brief explanation for the audience of what a financial hedge is?
MR. ESTEVES: I think they come in different varieties, but the ones that seem to be most prevalent in the last few deals that were done come in the form of heat rate coal options. Think of it as a traditional call option on any type of commodity. In this case, the strike price on the option is set up to mirror the true cost of running the power plant. Thus, in essence, the buyer of the call option is looking for opportunities where market prices of electricity exceed the strike price or exceed the operating cost of the unit.
There are several institutions that are willing to act as counterparties in such hedges, including Credit Suisse, J. Aron and Morgan Stanley. Everyone is taking a page from the derivatives that already exist in other commodity markets and applying the same learning to this market.
It is important that the hedge be one that settles financially rather than through physical delivery. As you might imagine, there are lots of nuances. In an ideal world, you try to mimic the true cost of the facility by taking a specified heat rate and perhaps an indexed gas price. If you can’t secure gas at that index or if you are not actually selling electricity from your plant at the electricity index on which the call you sold is based, then you are not perfectly hedged, but you may be able to satisfy yourself that it is a manageable risk.
You can you write contracts on anything. For example, you can set up a contract that says the counterparty to the hedge will pay the developer the price at X hub to the extent it exceeds a specified heat rate times an indexed gas price.
MR. CHAUDHRY: Bill Sutherland, are people talking about hedge products in wind deals as well?
MR. SUTHERLAND: There have been several done recently with hedges on them. Hedges must be structured in a manner that takes into account the variability of the wind resource. We typically characterize the resource base case as a P50 case, meaning that there is an equal chance the project will generate more or less electricity than has been projected.
Hedges typically are based on a P95 or P99 output case. In a P99 case, there is only a 1% chance that the project will underperform. That means there is assured production to meet the requirements under the hedge, and the balance of production remains unhedged. However, when you take into account that production tax credits are essentially a contracted revenue source, and there may also be a contract to sell renewable energy credits from the project at fixed prices, these contracted sources of revenue plus the hedge leave little real market exposure.
MR. CHAUDHRY: Doug Sherman, what is the typical term for a hedge?
MR. SHERMAN: Most of the hedges I have seen run five years in duration.
MR. CHAUDHRY: And how does this then affect the structure for the project debt?
MR. SHERMAN: In some cases, lenders have tried to match the term of the debt to the hedge and have a sweep of 100% of cash flow. In other cases, I have seen two years of excess debt beyond the hedge period — for example, a 7-year loan based on a 5-year hedge. As I noted earlier, the lender is still taking merchant risk for the period after the hedge expires.
MR. CHAUDHRY: So lenders are generally taking two years of merchant risk. What do these hedge products generally do for the leverage in a deal? How much more debt can a developer hope to secure with a hedge than if his plant is financed on a purely merchant basis?
MR. SHERMAN: In some cases where we have looked at a plant on a purely merchant basis, we may be able to get to 70% or 75% leverage. In some of the hedged projects, I have seen leverage go as high as 80% to 85%. So I would say that a hedge allows for as much as 10% additional leverage.
MR. CHAUDHRY: Do others on the panel agree with the 20% figure? (The other panelists nodded their heads affirmatively.) Charlie Wilson from Duke, I believe you have a comment?
Plus Ça Change?
MR. WILSON: Thanks for putting me on the spot. I think you have to ask the question Markus Christen asked earlier. When you use the word “merchant,” be clear what it is you are talking about. Is it a purely merchant plant or is there some financial hedging involved? The ability to put hedges in place that absorb market risk is extremely limited. Are there tolling agreements? People sometimes lose sight in such deals that what they are really doing is converting commodity risk into credit risk. Projects that had tolling agreements with Williams and Enron turned out to be unhedged in practice.
What went wrong the last time is a complicated litany: too much liquidity from the financial markets, inaccurate projections of electricity demand, stillborn deregulation, lack of uniformity in terms of the market models state by state, poor market design in places like California, and lack of capacity markets.
There was an illusion the last time around of a liquid trading market. Enron, when it existed, made everyone think that there was a very deep and liquid market to hedge commodity risk because Enron was able to conceal the fact that it was just recycling and circling and nothing was really getting done at the end of the day. When Enron collapsed, people, like Duke, who had built their merchant models on the same basis, which was largely a trading-centric model, found out there was inadequate financial liquidity.
Power is not like other commodities like oil and metals. Yet that was the basis for this whole trading-centric idea - the view that you can more efficiently hedge at a large portfolio level with a large trading organization that trades in a very deep market. This ignores the fact that electricity is locational. You can’t get to the model on which the earlier boom was based when you don’t have a uniform market structure.
PJM is a good example. You have literally 1,200 different pricing locations. It is impossible for Wall Street to come up with a way to generate enough liquidity at each pricing point to allow adequate hedging. People use the gas market analog; gas has enough history and liquidity that it allows people to trade around particular nodes relatively efficiently. However, this is not the case with power, and it may never be the case with power.
My advice is don’t get caught up in the euphoria of the private equity and highly-leveraged transaction folks. To us, that is a short-term, opportunistic financing model. It has been driven by depressed asset values. It has been driven by the extraordinary liquidity that shifted into the market from hedge funds and private equity looking for ways to invest the extraordinary amount of capital they have amassed.
We think those people will make money, probably a lot of money, on the assets we and others have sold them, but we do not think that is a sustainable model for the next round. Those guys are not going to own those plants for very long. My view is the next time around, it will be back to the future. You will see more inside-the-fence plants. Plants in regions with active capacity markets will be easier to finance.
MR. CHAUDHRY: I was planning to end with your comments, Charlie Wilson, but I want to turn back to Joe Esteves and ask whether he disagrees with anything that was just said.
MR. ESTEVES: I think the only thing I disagree with is the implication that he was ill-prepared and put on the spot. [Laughter.]