The offer is at least being taken seriously. The sponsors of the Astoria power project have in front of them a proposal from Calyon to refinance along conventional project finance lines. If Calyon is successful it will be the first time that a B loan financing has been replaced by a more restrictive project finance structure and sold to a more traditional group of lenders.

The buyers of the Edison Mission Energy global IPP portfolio - Mitsui and International Power - have also been working with a group of five project finance lenders to finance their purchase. Morgan Stanley, retained by acquisition vehicle IPM Eagle as an adviser, had initially thought that a deal dependent upon dividends from project companies would be best paced in the B loan market.

The two instances, which have both emerged in the last month, would seem to confirm a growing impression that the B loan market is close to its high-water mark. And they may give comfort to project finance lenders that they will again be able to compete. For the difference is more that just one of semantics - several European banks will be basing staffing decisions for New York operations on what they think the B loan market will do in 2005.

Staffing or stuffing?

But the process will likely be opaque - the examples of project deals that have emerged so far are still slightly bespoke, and the examples of assets best suited to the classic structure that went to the B loan market are more commonplace. But the increased confidence of project lenders is apparent. "They're definitely looking to book assets right now," says one investment banker active in power. "I sense a real hunger."

How much this hunger is down to improved conditions in the market, and how much is down to wanting to maintain staffing levels at US-based lenders is unclear. For instance, several market observers contacted by Project Finance, have suggested that New York bankers' reluctance to look at US PPP deals their European offices are willing to book might be because the margins do not support US-style staffing levels. "Staffing levels are based on booking deals with 250bp margins, whereas European-style deals come with margins a percentage point lower," notes one. That may be a factor in the renewed enthusiasm for power lending.

The mainstay of the power project finance market in the US has been the renewables sector. In the last 12 months there have been three deals from Shell, and a handful of financings for smaller developers. Few of these will make a lender's budget, especially when, as in the recent case of the Three Winds deal, they are so widely syndicated. Three Winds (search under "Whitewater"  for details), a $123 million deal led by Fortis Capital, offered final takes of $9 million to participants.

Moreover, the deal is believed to have paid 150bp over Libor, a margin that few power bankers in the US would believe to be generous. In the circumstances, the banks might be forgiven for making more aggressive pitches for power business, especially given their cost of capital, if not their risk tolerance, is better than the average hedge fund's.

Calpine - the B loan cheerleader

And banks are still regarded as the best place to finance construction risk. Even Calpine's CFO, Bob Kelly, a man not noted for his fondness for project finance lenders, accepts that they are still a natural place for a developer to hunt for capital (see the cover story of Project Finance's September 2004 issue). Calpine has done several deals for gas-fired projects in the market, on the back of strong contracts, and for as limited a period as possible.

But Calpine, in part for operational reasons, will not use project banks to finance plants post-construction. Indeed, the ink was barely dry on the financing for its Rocky Mountain plant, procured from six lead arrangers, before it announced that it would refinance the project, as well as the earlier Riverside financing, through a B loan raised through CSFB. CSFB notes that the B market allowed the combined entity to raise $170 million more on the assets than the separate project financings.

Astoria - a true exotic

CSFB also maintains that the B market was the only place where the Astoria Energy project could raise debt. Astoria, a 552MW plant located in Queens, New York City, had a 10-year power purchase agreement with local utility Consolidated Edison (ConEd). The agreement was potentially lucrative, and enabled the developer SCS Energy to gain a foothold in a transmission-constrained market. But it did not provide enough revenue to put towards building a plant in an expensive location.

So SCS gave up some of the equity to EIF, Caisse des Depots and SNC Lavalin, and paid the price demanded by B loan lenders - 525bp over Libor for a first lien and 875bp for a second lien. According to sources at the arranger, the first-lien piece, $500 million, corresponded to the likely revenue under the contract with ConEd, while the second lien was sized according to the potential recovery value of a completed project. Calyon either believes that a bond solution could cover both of these pieces, or will be able work around the two-tier structure.

To get an idea of the peculiarity of the Astoria project, it is necessary to get an idea of the cramped and constrained nature of the plant's site, and have a familiarity with the New York labour market. Astoria has a per kilowatt cost of $200 dollars, and had sponsors that were not willing to meet some of that cost with equity. While project developers have often been accused of putting plants where costs were low and permitting was easy rather than where power is needed, the Astoria project would always be an extreme example of the opposite trend.

The new coal paradigm

In fact, forthcoming financings may make Astoria a pathfinder, even though the projects in question will have little in common with the Queens project. The reason is that the majority of planned additions to the US generating fleet over the next several years could be coal-fired, rather than gas-fired. Coal plants have been only occasional project finance candidates, the most recent Greenfield deal being the Springerville project lease.

Springerville was a 400MW expansion by a western co-op, the Tri-State Generation and Transmission Association, led by Credit Suisse First Boston in November 2003. That deal was notable as the first time GE Capital, the lease equity provider, entered a deal before construction was complete. GE has now repeated the commitment by agreeing a $940 million pre-completion sale-leaseback for Calpine's Fox Energy Center, which will come online between January and December 2005.

But Springerville largely resembles a structured municipal power credit, debt built on the back of the owner's credit. Tri-State, as well as Unisource, the lowly-rated previous owner, and the Salt River Project, are the offtakers, but Tri-State's credit dominates. As a structure, it may be a tough one to adapt to future financings.

There are municipals among those looking to develop new coal capacity, and they may find a leveraged lease structure attractive. More likely, several of them will opt for tax-exempt bonds, and the utilities on the list may get rate-base treatment and fund their commitments through mortgage bonds.

The new coal callers

But there are several coal producers, as well as several smaller developers that will lack these opportunities and are currently looking for power purchase agreements. Peabody Energy, for instance is well advanced in its plans for plants in Kentucky (Thoroughbred) and Illinois (Prairie State), and for the latter has signed a letter of intent for an engineering, procurement and construction contract (EPC) with Fluor. It is Springerville's coal supplier, and has ambitious plans for mine mouth projects.

Peabody says that construction on the Prairie State plant will begin once it gains both the relevant permits and contracts sufficient to cover at least 80% of the proposed plant's output. Thoroughbred, which has a construction permit, has spent much of the year fighting an air quality suit in Muhlenburg County Circuit Court.

Permitting issues aside, Peabody's work in gaining power purchase agreements will be the crucial determinant of how it could gain financing, and how other competitors might reach the market. A long-term contract would be highly attractive to project lenders, which would also be largely happy taking on construction risk, providing the technology was suitable.

A shorter contract, however, would make the plant the preserve of the B loan lenders. A precedent exists in the Coleto Creek financing, which closed in July 2004. Coleto Creek is a 632MW coal-fired plant located at Fannin, in Goliad County, Texas. It was built in 1980, and is a relatively efficient coal-fired baseload plant that dispatches into a power pool (ERCOT) that is dominated by gas capacity.

The owners, Sempra Energy Partners and Carlyle/ Riverstone, had contracts with a weighted average life of 4.2 years for 93% of the Coleto's capacity. Some of these were with Sempra affiliates, and most are with A-rated counterparties. But given the length of the contracts, and the easy covenants available with the B loan lenders, a project deal did not make sense.

The debt, made up of a $150 million eight-year second lien loan, a $205 million seven-year first lien loan, and $95 million in letters of credit, was through Citigroup, JP Morgan, and Goldman Sachs. That the three were prepared to divide a relatively small mandate gives an idea as to how important B loan business has become, as well has how much even well-capitalised private equity and utility buyers, as opposed to struggling merchants and start-ups, have come to use the market.

B still sharp

But bankers familiar with the market say that even if there is a chance that the coal-fired projects access the project market, B loans will be an important component of the exit financing for bankrupt merchants. The NEGT assets (search "NEGT" for details), which wait for bankruptcy court approval for a sale, will be candidates. And the banks, which have been working with operations such as North American Energy Services and Competitive Power Ventures to optimise foreclosed assets, may be close to reaching sales values that would cover their provisioning.

Such assets will have recovery values that are more realistic than the money spent on their construction. They are also unlikely to have any long-term contracts attached, and often will be in very unforgiving, overbuilt, merchant markets. Here, KGen, a financing for Matil Patterson's $475 million purchase of Duke Energy's south-eastern portfolio, had a small legacy contract with Georgia Power, part of Southern Comppany. But the plants, which are located in Arkansas, Georgia, Kentucky and Mississippi and have a generating capacity of 5,325MW, have no other contracts.

The financing, again led by CSFB, went to the B loan players. The structure will predominate so long as distressed merchant assets do not become overnight attractive candidates for contracts. The main question, one to be settled by the ultimate purchasers, will be which investment bank grabs the most business.