Cost of Capital: 2013 Outlook
A group of industry veterans talked in early January about the current cost of capital in the tax equity, commercial bank, term loan B and mezzanine debt and project bond markets and what they foresee in 2013.
The panelists are John Eber, managing director and head of energy investments for JPMorgan Capital Corporation, Thomas Emmons, managing director and head of renewable energy finance for the Americas at Dutch bank Rabobank, Gerald Hanrahan, senior managing director of the power and infrastructure team at John Hancock Financial Services, Richard Randall, managing director and head of power and project finance at RBS Global Banking & Markets, and Jerry Smith, managing director and head of the tax equity desk at Credit Suisse. The moderator is Keith Martin with Chadbourne in Washington.
MR. MARTIN: John Eber, what volume of tax equity transactions did the market do last year and how does that compare to 2011?
MR. EBER: We saw approximately $5.3 billion in tax equity last year in the US renewable energy market. About $2.5 billion of that was invested in wind deals and about $2.8 billion in solar. That is down from about $6 billion the year before, although the solar market was up and the wind market was down. It does not surprise me that the wind market was down because of the uncertainty surrounding the production tax credit. Although Congress extended the credit in January, there was a risk it would not do so, with the result that projects that were at risk of not making it into service by year end were not built.
MR. MARTIN: What volume are you projecting for 2013?
MR. EBER: I am not really in the projection business, but solar should remain strong at the same or an even faster pace than in 2012. Wind will be a more difficult story. The delay in extending production tax credits means that it will take time for the deal pipeline to refill.
MR. MARTIN: In 2011, 55% of tax equity transactions in the wind market involved production tax credits. That was interesting because developers had the option of forgoing tax credits and taking the cash value from the US Treasury instead. The majority of the transactions did not take it. What percentage of wind deals in 2012 involved production tax credits rather than Treasury cash grants?
MR. EBER: We think 12 of the 16 deals we tracked in wind, or 75%, took production tax credits. That actually represented about 85% of the tax equity dollars raised in the wind market.
MR. MARTIN: And the reason for that was that given a choice between a tax subsidy tied to output and one tied to cost, people opted for output because wind turbines are becoming more efficient and their prices are falling?
MR. EBER: Yes, both. The cost of turbines is down. Most of the equipment we saw coming to market last year had increased capacity factors due to taller towers, longer blades and improved electronics. The efficiency of the equipment has been improving year over year, and we saw that quite noticeably in 2012.
MR. MARTIN: How many active tax equity investors would you say there are currently?
MR. EBER: We think 20 actively participated in deals last year. Of that number, 15 put money into solar and 10 into wind, so there was an overlap between the two.
MR. MARTIN: Jerry Smith, is that number of active tax equity investors consistent with what you saw last year?
MR. SMITH: Yes. There are probably different levels of activity within that range. Probably seven to 10 companies did multiple deals. Some of the others are relatively new or are returning to the market.
MR. MARTIN: John Eber, do you foresee more or fewer tax equity investors next year? For example, will the transition from Treasury cash grants to tax credits affect the number of tax equity investors?
MR. EBER: I do not think the transition will reduce the number of tax equity investors. A few investors will drop out as the grant disappears, but we have seen new investors return to the market each year as more companies find they have tax capacity. I expect that trend to continue.
MR. SMITH: I agree with that, but with one qualification. Although the number of participants and the total dollars they are willing to spend may be the same, those dollars will not cover as many transactions as we move from Treasury cash grant deals to deals that rely on tax credits. The dollars will not go nearly as far.
MR. MARTIN: So some investors will drop out during the year?
MR. SMITH: Correct. They will exhaust their capacity earlier in the year than they might have done if the Treasury were still paying cash grants.
MR. MARTIN: What are current yields? How much does tax equity cost?
MR. SMITH: Yields have been relatively stable over the past few years. In an unlevered transaction, yields are somewhere in the high single digits, maybe 7% to 10%, and then probably about mid-teens if there is debt at the project or partnership level.
MR. MARTIN: How much is the premium the project will have to pay for having leverage? It sounds like 500 to 700 basis points.
MR. SMITH: Anywhere from 500 and 800 basis points, depending on the deal.
MR. MARTIN: Eight hundred would be a larger spread than last year.
MR. SMITH: Don’t read too much into that. It might be a different person saying what the spread is. The spread has not changed in the last year.
MR. MARTIN: John Eber, it has seemed for the last couple years like the cost of tax equity for benchmark wind deals — projects developed by large, balance-sheet wind developers — has hovered around 8%. Would you say that is where it remains as we start 2013?
MR. EBER: We have not seen a lot a lot of change, and nothing has happened to suggest significant changes in the near term.
MR. MARTIN: Many people expect interest rates to increase this year — at least the bond market is signaling that already. Are interest rates a factor in the yields?
MR. SMITH: Tax equity yields are driven by demand and supply for tax equity rather than interest rates. However, they can be a cap on tax equity yields, since tax equity is competing with other sources of capital as a source of financing.
MR. MARTIN: Is there a difference in the cost of tax equity for the following types of facilities and, if so, how wide are the bands? Where does utility-scale solar PV price in relation to wind?
MR. EBER: For quality projects, we do not see a significant difference between the cost of tax equity for wind and solar.
MR. MARTIN: When you say “solar,” are you referring just to utility-scale PV or also residential and commercial rooftop installations?
MR. EBER: Residential is a different market than utility scale and is priced differently. That may change over time as the market becomes more comfortable with residential solar.
MR. SMITH: Utility-scale wind and solar have been around for a while. Residential has a different credit profile.
MR. MARTIN: How does tax equity for geothermal compare in cost to wind?
MR. EBER: It is difficult to say. Too little geothermal is getting done to say what is market.
MR. MARTIN: What about biomass?
MR. EBER: We are not seeing much activity in the biomass market, so that is a difficult one as well.
MR. MARTIN: What evolution if any do you foresee in deal structures this year?
MR. SMITH: There are three main tax equity structures in use today in the market. The traditional partnership flip is used primarily in the wind space, but can be used in solar as well. The solar market also uses sale leasebacks and lease passthroughs or, as some people call them, inverted leases. The three structures are pretty well developed and are flexible enough to be applied in most situations. There might be some evolution around the edges of the existing structures, but I do not see a need to move to a different structure.
MR. EBER: The only difference in the market going forward is we will not see many more renewable energy deals involving Treasury cash grants.
MR. MARTIN: Moving to bank debt and starting with Tom Emmons, the North American project finance market was a $40 billion market in 2011 and predictions were that it would end up at only $20 billion in 2012. Any idea where it finished?
MR. EMMONS: The databases vary, but the ones on which we rely show $24 to $25 billion in project debt in 2012. The breakdown was somewhere around $20 billion for bank debt and $4 to $5 billion for bonds.
MR. MARTIN: Rich Randall, same numbers?
MR. RANDALL: We came out around $23 to 24 billion with the same breakdown.
MR. MARTIN: How many active banks were there in 2012, and how many do you expect in 2013?
MR. EMMONS: It is hard to calculate. All in, there are probably 70 banks who are involved in project finance, but some of them are small players. In 2011, probably 50 banks committed more than $100 million and maybe 40 banks committed more than $200 million. In 2012, those numbers were somewhat lower, but then the volumes were down by 40% so you would expect that.
MR. MARTIN: That is still a pretty healthy market. Richard Randall, same numbers?
MR. RANDALL: We see things a little differently. We only saw about 25 active banks in 2012. I concur with Tom Emmons about the 2011 figures.
We saw a lot of banks last year either exit permanently or go on hold because of the European capital adequacy issues. We are now seeing banks return to the market. We put out a syndication strategy for a client last week on a deal. There are 40 banks that we see open for business in 2013 to which we might syndicate.
MR. MARTIN: Yesterday, the European regulators took a more generous approach to what qualifies as liquid capital for Basel III. Do you see that helping to bring more European banks back into the market or will euro troubles continue to drag them down?
MR. RANDALL: Our initial take on the ruling is that it will affect investment grade liquidity facilities at the corporate level. Project finance lending will still have to adapt to Basel III. The rules for risk-weighted assets and the capital that will have to be applied toward project finance loans appear set in stone. Such loans will become a bit heavier for banks to carry.
MR. MARTIN: Tom Emmons, I heard you say at midyear last year that loan volume was down across the board for all project finance banks, except the Japanese and Canadian banks. By lending at the same volume as before, they gained market share. Did that remain true through year end?
MR. EMMONS: Yes. The movements are gradual. Basically the Europeans lost market share. That was picked up by the Japanese, the Canadians and some regional and super-regional American banks.
MR. MARTIN: Rich Randall, what is a term loan B transaction?
MR. RANDALL: It is essentially a bank loan in structure, but with an institutional lender. The interest rate floats. The structure, terms and conditions are very similar to bank debt, except for the buyers of the debt who are usually private equity funds, insurance companies and other institutions.
B loans have been sub-investment grade issuances of debt. The credit is in the B to BB level. Investors in B loans are looking for yield.
The tenors are about seven years. The scheduled amortization is very light initially — usually only about 1% annually — and then there is a cash sweep, but most of the principal is not due until a balloon payment at maturity.
B loans emerged basically as a leveraged buyout instrument and that is still the primary use of the term loan B market. The market operates in conjunction with the high-yield bond market.
MR. MARTIN: So it is a way to finance a more risky project than one might be able to finance in the bank market.
MR. RANDALL: Correct. B loans were used in the project finance market mainly for merchant assets. The product evolved after Enron went bankrupt and merchant deals ran into trouble in the period 2000 though 2002. Banks exited that market. It was too risky. Money was lost in bankruptcies. The B loan market picked up to fill that liquidity gap.
MR. MARTIN: If the North American project finance market last year was $20 billion and there was another $4 to $5 billion in bonds, was part of the $4 to $5 billion made up of B loans?
MR. RANDALL: I don’t have that off the top of my head, but I assume it was. There were probably $3 to $4 billion in B loans.
MR. MARTIN: Tom Emmons, in 2012, base interest rates on bank debt were 225 to 275 basis points over LIBOR, trending up toward 300. Upfront fees were on average 275 basis points. Are these accurate numbers for 2012? Where do you see them headed in 2013?
MR. EMMONS: Yes. The margins and fees have been stable over the last couple years. I expect them to remain flat or to rise slightly in 2013. It depends on the demand. Rates are a function of supply and demand, but with a floor. Let me come back to the floor.
High demand could come from wind if wind comes alive again, and there are lots of developers who have had projects on hold that they are bringing back to market. There could also be greater demand in 2013 for upstream oil and gas projects. Shale gas development is creating more demand. Some new combined-cycle gas-fired power plants will be built to replace coal with cheaper gas.
All of this could mean higher demand, in which case margins and total compensation could go up.
The supply is pretty elastic. It is not immediate, but it matches over time. I think there is a floor. At the current tenors, this pricing probably just passes the hurdle rates for the European banks based on their costs of capital and liquidity.
MR. MARTIN: Rich Randall, any sense where B loans price?
MR. RANDALL: For BB plants with some merchant risk, they are probably LIBOR plus 400 and above. That would be in a market where bank debt is pricing at LIBOR plus 250 for a more traditional project with a power purchase agreement.
We see quite a bit of appetite from banks and institutional investors. The two compete with each other on price.
Banks are returning to the market this year. We will see the number of active banks increase to around 40. I think we will see pressure on margins from the increasing supply of available capital both in the bank market as well as the private placement market.
I agree with Tom — there is probably a 250-basis-point floor tied to the cost of funding — but I think the private placement market will continue to provide downward pressure on pricing.
MR. MARTIN: Does the fact that there is a healthy term loan B market suggest that it is possible today to finance purely merchant plants?
MR. RANDALL: Not purely merchant. The electricity price must be hedged to guarantee that interest and some level of principal amortization will be paid. The interesting thing about the B loan market is that it does not love construction risk. The banks are much more efficient at financing construction risk. The B loan market does not allow for delayed draws, so you have some negative arbitrage with which a construction borrower would have to deal.
That makes for a disconnected market. For that reason, we are toying with whether one can create hybrid structures where you have banks and B loans combining in order to eliminate the negative arbitrage during construction. It is unclear right now whether that can be done because banks are more conservative in the structures and the balloon payments that they can live with compared to the terms with which the B loan market can live.
MR. MARTIN: How long a hedge is required to finance a merchant plant?
MR. RANDALL: The market seems to be at around five years post construction, so that is essentially a 7-year hedge. We are just starting to see rumblings of maybe 10 years in certain markets. Electricity is fairly illiquid. Seven years all in will accommodate a construction loan plus a 4 1/2- or 5-year term loan.
MR. MARTIN: Tom Emmons, bank loan tenors seemed to shorten last year to seven to 10 years with mini-perm features. Where do you see them headed this year?
MR. EMMONS: The shortening really was a sea change as a consequence of the re-pricing of capital and liquidity problems, particularly among European banks. Borrowers have accepted mini-perm features. The shortening of tenors is creating opportunities for institutional lenders, and they have been stepping up. I think it is a permanent shift.
MR. MARTIN: How are debt service coverage ratios set, and what are they currently for contracted wind and solar projects?
MR. EMMONS: They are based on a judgment about the stability of cash flow available for debt service. They are typically 1.45 times debt service for wind. They are probably 1.35 for solar.
MR. MARTIN: Rich Randall, what are they for new natural gas-fired power plants?
MR. RANDALL: They are 1.4 if you are talking about a traditionally amortizing loan done in the bank market. B loans use a different structure since they essentially require payments of interest only. Those coverage ratios are 2 to 2.5 times debt service, and you have a sweep of excess cash flow for principal amortization.
MR. MARTIN: You mentioned the possibility that banks will team up with institutional lenders to do a combined bank and B loan financing. What structural issues does that create?
MR. RANDALL: It has not really been done yet, but such deals are coming. There will be inter-creditor issues. The debt would be pari passu. The banks would provide the delay draw feature to permit some construction financing. There will be debt sizing issues since the traditional B loan investor is going to allow for higher leverage. The banks will press for lower leverage. Trying to find that equilibrium will be a challenge.
MR. MARTIN: Are banks back to underwriting or are the larger transactions still being done strictly as club deals?
MR. EMMONS: We have not seen much underwriting. The retail market is still questionable, and there is still a lot of volatility in the bank market. We see mostly clubbing.
MR. RANDALL: We see a move to club underwriting where maybe banks lead a range or will underwrite a portion of the transaction, but it certainly would not be a fully underwritten deal.
MR. MARTIN: David Albert, the Carlyle Group, Energy Capital Partners and others have formed mezzanine debt funds. Do you sense a greater interest in mezzanine debt than before among borrowers and, if so, what is driving that interest?
MR. ALBERT: There are a couple drivers. The European debt crisis has caused some of the European project finance banks to pull back. Some are coming back into the market, but compared to pre-crisis levels, there is a smaller supply of bank capital available and a more conservative approach by banks to lending than in the past.
The other driver is that developers are looking for capital that is not as dilutive as traditional private equity. We see a greater desire by developers to retain as much ownership as possible and not have to give away control and 80% of the profits, if you will, in order to raise private equity.
Our capital is more expensive than bank debt, but it is flexible and less expensive than private equity, and we are not seeking control or governance rights.
MR. MARTIN: What is the spread typically between senior bank debt and mezzanine debt?
MR. ALBERT: It is hard to give a simple answer because mezzanine debt is not a standardized product like bank debt or even a project bond. Most of the deals that we have done to date have had a unit tranche where senior debt and mezzanine debt are drawn together. The cost of debt on an all-in basis in these transactions is somewhere in the low double digits.
However, mezzanine debt can also be used at the corporate level. How that is structured can vary widely. For example, we can take a high coupon with very little equity kicker, or we can take a single-digit fixed-coupon return with much more of an equity kicker.
The transaction terms run the gamut, and that is also true for fees and everything else. The pricing is more holistic and tied ultimately to the risk profile.
MR. MARTIN: Tom Emmons, you said bank debt is pricing at maybe 225 to 300 basis points above LIBOR. What is that as an interest rate? Six percent? Lower?
MR. EMMONS: You would have to swap LIBOR, so that adds to cost, but somewhere in the low 5% range stepping up over time to a little above 6%.
MR. MARTIN: So David Albert, you are in the high single digits as a combined cost of senior debt and mezzanine debt, or in the same range for standalone mezzanine debt at a fixed rate and with an equity kicker, correct?
MR. ALBERT: It depends on the risk profile. We are willing to take construction risk. We are willing to be subordinated. We are willing to take on more challenging risk profiles.
There could be a management team that has its own skin in the game. It is putting in some capital, but there is still a large hole in the capital structure, and maybe the team can only get a certain amount of bank debt or cannot get any bank debt. We are willing to come into situations like that and, in those situations, our all-in return is obviously going to be much higher, and it will be a higher mix of equity.
We have been an anchor investor in some term loan B deals where we have taken a very low double-digit fixed return with no equity upside.
The point is that mezzanine debt is a flexible product with a wide range of potential risk and return profiles associated with it.
MR. MARTIN: Mezzanine debt displaces equity, and it is cheaper than equity. You said there is often an equity kicker. How is the equity kicker structured?
MR. ALBERT: If we are talking about oil and gas assets, it is structured as either a net profits interest or an overriding royalty interest. If we are talking about a power project, it usually takes the form of warrants that can either be struck as penny warrants or have a higher strike price. Again, there is a fair amount of flexibility in how we structure such mechanisms.
MR. MARTIN: Are there upfront fees and, if so, how much?
MR. ALBERT: You can either structure upfront fees or advance less than the full amount of the loan so that the fee takes the form of original issue discount of a couple points. It is not dissimilar to what you see in the bank market or the term loan B market.
The range can be anywhere from a point and a half to three points. It depends on the transaction.
MR. MARTIN: For how long a term do you lend?
MR. ALBERT: The tenor on mezzanine debt will usually be longer than for the more senior debt in the capital structure. The senior lenders will usually require it.
That said, the range of our debt can be anywhere from shorter dated meaning a couple of years to longer than the term loan B market. The key adjective is flexibility in terms of how we structure our investments.
MR. MARTIN: I can see a pattern here. I was going to ask what other differences are there between mezzanine debt and senior bank debt, but it all comes back to flexibility. Are there specific differences that are worth flagging?
MR. ALBERT: We are a source of capital that can provide significantly more leverage than the banks or the term loan B market. Because of that, not surprisingly, our cost is greater, but for borrowers who are looking for capital without the highly dilutive impact giving up control to a private equity fund, we fill a need.
We feel very much like commercial bank debt, only the leverage will be a little greater than what you would see in a commercial bank deal. Or we can play a role where we are much more equity like and charge higher interest with the interest paid in kind rather than in cash. We are taking on a different risk profile than the senior lenders. Our return profile is also different.
MR. MARTIN: How should one calculate how much mezzanine debt a project can support? Are there debt service coverage ratios and, if so, what are they currently?
MR. ALBERT: There are, but again, we are willing to provide construction financing and there is a great deal of flexibility in general to how our product is structured. One of the prior speakers talked about 2 to 2.5 times the interest payments in the term loan B market because there is no principal amortization.
Most of our deals do not have fixed amortization schedules. They tend to have a cash sweep.
The debt service coverage ratio is a function of the underlying asset that we are financing. For example, with an oil and gas loan, especially upstream, you are dealing with an asset that depletes over time as you lift those hydrocarbons from the ground.
The profile and the coverage on an asset like that will be different than a power plant with a 35-year life.
The difference will be even more dramatic when you are comparing that to a power plant with a long-term offtake agreement. Then you can get much closer to what you see in a commercial bank loan in the 1.5 times debt service range.
It all comes down to how risky are the cash flows? We are willing to finance merchant risk on the power plant side. The debt service coverage ratio will be higher in a situation like that than it would be for a contracted asset.
MR. MARTIN: Jerry Hanrahan, you heard from Tom Emmons and Rich Randall that bank participation was down slightly from 2011 to 2012. Many people expect to see the project bond market fill the gap. We heard that the volume of institutional debt in 2012 was $4 billion to $5 billion.
Do you agree with those numbers? What volume do you expect in 2013?
MR. HANRAHAN: A lot of that may be term loan B debt. The project bond market has really been fairly shallow in the project finance area.
In 2011, a lot of the European banks that were having difficulty talked to us about 2012 being the year of the project bond and said we would see a lot of deals as the bank market lost its appetite. That did not really come to pass.
The reduction in European bank lending was filled primarily by Asian and Canadian and some US regional banks. It is interesting to hear the predictions about this year.
Some people are saying that with Basel III kicking in, that will move more lending to the project bond market and to people like us. I hope that is the case, but we have not seen it yet. Last year, we saw very little volume in project bonds. Topaz was the only significant deal, and that was a large solar transaction brought to market by Mid-American Energy Holdings. Otherwise, it was a fairly slack year for project bonds.
We did a few direct deals, but there were not many widely syndicated or 144A-type transactions.
MR. MARTIN: Your colleague, John Anderson, told me in mid-2012 that he thought there were 15 active institutions in the project bond market. I am not sure how one would get to that number if there was one deal. How many institutions do you see ready and willing to buy project bonds?
MR. HANRAHAN: The number is probably a little higher, probably 20 to 25, but it is a tiered market. You have probably eight or nine larger institutional investors like us, who tend to be the anchor investors. Then you have another tier of mid-range and smaller players, probably 15 to 20, who fill out deals.
MR. MARTIN: Project bonds are priced off of 10-year Treasury bonds, I believe. What is the current spread? Where do you see both Treasury bond yields and spreads headed in 2013?
MR. HANRAHAN: Since 2010, Treasury rates and spreads have been moving in concert with each other and largely offsetting each other, with the result that we have had relatively constant yields.
We have been guiding people to focus less on the spread per se and more on the coupon. We end up with coupons in the 5 1/2% range for a project with a solid investment grade, things like gas-fired power plants or transmission deals. We end up with more of a premium heading up toward a 6% coupon for the riskier or weaker deals, maybe some of the renewables.
That is the current pricing range, and I expect it to continue.
We priced a couple of renewable deals recently in that range. That translates into spreads to Treasuries anywhere from the low 300 to maybe around 400 basis points, depending on the riskiness of the deal.
MR. MARTIN: The tenors are much longer than bank debt. A rule of thumb used to be that a project bond could be issued for a term as long as one year shy of the term of the power purchase agreement. Is that still true?
MR. HANRAHAN: In most cases, we see deals where the tenor of the debt matches the tenor of the underlying contract. Sometimes we see a tail on the power contract of one or two years, but the market has reached the point where the two tenors — the debt and the underlying contract — typically match each other.
MR. MARTIN: An upfront payment is required on a bank loan of perhaps 275 basis points. There is none on a project bond, but there is an upfront payment in effect in the form of original issue discount, correct?
MR. HANRAHAN: Not necessarily. The deals to which you are referring are the widely syndicated deals brought to us by arrangers or sometimes in the 144A market. Those usually do not require upfront fees and you are not getting original issue discount. We pay par. All of the economics are contained in the spread. Sometimes there are fees in some of the smaller direct deals we do, but not when they come syndicated.
MR. MARTIN: Project bonds require two rating agencies to rate the debt. Is that correct?
MR. HANRAHAN: Not really. That is if you are doing a 144A transaction. Those have rating requirements, but there are no ratings requirements for a deal done as a direct placement.
We get that question a lot from arrangers and issuers,
asking what our rating requirements are. At least for us, we do not require an external rating. Many issuers and arrangers will get one even for a private placement because it helps ease the syndication process. It does not have to be two. One is fine.
MR. MARTIN: Another significant difference between project bonds and bank debt is the make-whole payments that are required if the project bonds are repaid ahead of schedule. What is a make-whole payment, and how is it calculated?
MR. HANRAHAN: It is calculated based off of spread to current Treasuries at the time of the prepayment. We are putting up debt at a fixed rate. We have matching liabilities to fund the debt. The make-whole payment is protection for prepayment risk against our liabilities. The remaining payments that would have been made on the debt are discounted back at the then-current Treasury rate plus a spread, typically 50 basis points.
MR. MARTIN: Are project bonds available for both construction debt and term debt?
MR. HANRAHAN: Yes. A frequent misconception about the bond market is that it does not take construction risk. It does, although it does not have the depth or the flexibility of the bank market when it comes to construction lending. There have been many large-scale construction deals done in the project bond market with construction periods of up to 24 to 30 months. A recent example is the Neptune undersea transmission cable.
MR. MARTIN: When you provide construction debt, are there construction draws or does the money have to be taken down all at once?
MR. HANRAHAN: There can be draws. There are typically draws every two to three months and, in those cases, there
will be fees. You basically get a commitment fee on undrawn capital.