After a short interlude, secondary sales of infrastructure debt are set to become popular again. A little over five years after distress in the power sector caused large quantities of project finance debt to change hands, banks are again looking to unload their debt commitments in a troubled market. But additional leverage, the traditional way for secondary debt buyers to make attractive returns, is not as plentiful as it used to be.

As a result, banks' hopes for relief hinge upon an odd assortment of banks with low historical exposure to the asset class, a small number of dedicated project debt funds, and even some infrastructure funds, which are trying to make up for limited primary equity opportunities with a move into debt markets.

There is no sign of a fire sale, yet, but many project finance lenders are trying to move out from under their cheaply-priced infrastructure financing commitments. Because the banks' troubles have more to do with their own funding issues than any sector's difficulties, asset sales have been spread across sectors and projects.

The crunch and contracting

When banks need to contract their balance sheets, long-dated, low-yielding assets attract extra attention. But project finance debt, ever a problematic asset for portfolio managers, resists speedy attempts at contraction. It requires, as its specialists never tire of stressing, specialist credit analysis by dedicated teams.

These loans, partly because of the volumes of documentation and market studies they require, can offer a detailed picture of their sponsors' business plans, and do not often carry a public rating. Sponsors have traditionally been extremely wary of opening up, even as their demands on banks' balance sheets have expanded.

Sponsors have blocked attempts by banks to get project loans rated in preparation for their inclusion in securitization vehicles, have imposed restrictions on the information that lead arrangers can circulate in syndication memorandums, and maintain the right to veto the sale of debt participations to lenders they consider unsuitable. While strategic sponsors in the power and oil and gas industries originated these practices, and toll road operators have benefited from them, particularly with respect to traffic studies, infrastructure funds have been the most enthusiastic users of these restrictions.

Tank and unrest

Shining a spotlight on these restrictions is a suit that UBS has brought against Deutsche Rastatten Gruppe III, the owner of Autobahn Tank & Rast Holding, over the sale of its debt to unsuitable buyers. Tank & Rast, whose sponsors are Terra Firma and Deutsche Bank's RREEF, had blocked UBS from selling on debt to unnamed buyers, which have been reported to be Macquarie and JPMorgan Asset Management's IIF fund. UBS filed suit to prevent Tank & Rast from blocking the sale, calling its reasons "invalid and illegitimate".

Arrangers and sponsors have long argued over the ability of the former to sell on their debt commitments. Indeed, along with pricing flex provisions, it tends to be one of the hardest argued aspects of banks' commitment letters to sponsors. But rarely have these disputes been fought as brutally, a sign that the credit crunch has forced some lenders into drastic situations. Nevertheless, as one London-based syndication, banker noted, "I still can't understand why UBS didn't just sell the debt and wait for the lawsuit rather than starting one itself."

Infrastructure assets attracted substantial recent attention from institutional equity providers in the last two years. The market also attracted several new lenders, particularly European and US-based investment banks. Just as banks such as Morgan Stanley and Goldman Sachs became dominant players in US power when institutional, or leveraged loans, dominated that market, banks like UBS, which housed a US municipal finance operation and a small project bond capability, leapt into the infrastructure lending business.

UBS has now smartly exited the market, in tandem with a broader reduction in its investment banking operations. The offers from the two funds, if confirmed, would have been welcome at a time when appetite for leveraged debt has plummeted.
It's tempting to look for broader lessons for project finance lenders in the Tank & Rast dispute, though the deal is far from typical, and in fact is widely regarded as the high-water mark of the leveraged infrastructure boom. The borrower, which operates German filling stations and rest stops, is a monopoly, and petrol is an essential item for drivers, though a large proportion of the borrower's revenue is discretionary.

Terra Firma bought Tank & Rast for Eu1.1 billion ($1.65 billion) in 2004, and then sold a 50% stake to RREEF for Eu2.27 billion in late 2007. At the same time Barclays Capital, Royal Bank of Scotland, Société Générale and UBS refinanced Tank & Rast, with leverage reaching 10 times Ebitda. But the deal was richly-priced, with the Eu2.05 billion seven-year A debt priced at 175bp-195bp over Libor, and the Eu200 million B loan priced at 425bp.

An outlier, or first of a flood?

So the financing, by virtue of its generous pricing, and the troubles of UBS (there is no indication that the other three arrangers have considered similar sales) is not typical of the assets currently confounding banks. More likely candidates are the large port financings that closed in 2006 and 2007. These deals exhibit similar leverage, slightly less generous pricing, which is tied to debt/Ebitda levels, and involve substantial lender commitments, though not on the scale of Tank & Rast.

For instance, the $955 million financing for AIG Highstar's acquisition of the P&O Ports North America assets was priced at 150bp over Libor, for 13.5x leverage, but could drop to as low as 75bp if that figure dropped to 5x. The $1.88 billion in debt that Ontario Teachers closed for the OOIL financing had a similar pricing structure, which started at 140bp. The ports market was already exhibiting signs of fatigue when Citigroup closed the syndication of a $2.5 billion package for Goldman Sachs' acquisition of 49% of ports operator Carrix in September 2007. That deal, also priced initially at 150bp over Libor, attracted nine banks, but it sparked considerable unfavourable comment in the syndication market. RREEF is believed to have made an equity contribution to bolster weak cashflow at its Maher ports assets.

But these deals, despite a worsening economic outlook, are not yet changing hands at levels that indicate distress. Toll road financings and some of the larger Middle Eastern deals, particularly corporate and quasi-corporate commitments of names such as Qatar Petroleum, are favoured, but change hands at minimal discounts, and lenders can claim, with reason that these sales allow them to recycle capital. Such sponsors have more pragmatic attitudes to such sales, particularly if the buyer is another bank. Before the crunch, some borrowers had succeeded in confining debt participants to OECD-registered banks, and any bank with an interest in maintaining good client relationships still bears their wishes in mind.

The alternatives

Banks are not just favoured buyers of debt, they are still the best-placed institutions to buy it, providing their balance sheets are strong enough. The larger European public finance banks, and the oil-flush Gulf banks, are still in a position to carry out large block purchases, and when debt is priced at below 100bp, as many toll road and GCC deals are, a 2% to 5% discount can make a marked difference to the bank's yields on this debt.

Some banks have approached ECAs in an attempt to gauge their interest in helping them recycle capital, according to one manager at an export credit agency active in commercial lending. Most ECAs, despite the inducements, have been wary of taking the bait, because such purchases have a negligible relationship to their mandates, which entail support for equity investments and equipment exports.

Hedge funds have been able to raise debt to buy distressed mortgage-backed bonds, often from the banks that are selling them these bonds. But hedge funds are not raising their own debt on cheap enough terms to buy anything but the deepest-discounted assets. While hedge funds, particularly in the US, provided a useful way for banks to cut their losses in merchant power, those deals were distressed enough to whet the hedge funds appetites.

UK and US lenders adopted divergent strategies for dealing with the collapse in those power markets. In the UK, where a regulatory-driven dive in power prices was long predicted, and where assets were close to the headquarters of banks active in the market, banks tended to hold on to their distressed loans. Some banks even looked to add to their holdings, culminating in the short-lived CGE distressed asset-buying vehicle.

In the US, lenders provisioned in a much more conservative fashion, and were much more willing to sell up quickly. As a result, hedge fund lenders typically emerged as the equity-owners of gencos following their bankruptcy processes. In the UK, on the other hand, some project finance banks recorded substantial gains from the flotation of defaulted power plant Drax.

The new owners of US plants did not manage to list their equity and several of them may suffer underneath high-priced subordinated financings, but strategic foreign buyers are still scouting the market, and hedge funds have captured most of the increase in asset values as the US power market recovered. Moreover, the B loans with which these power companies were recapitalised do trade fairly easily.

Funds remain niche players

For those unrated assets not suitable for liquid trading, there are a small number of institutions than can do the credit analysis. Some of the larger life companies, while they have limited interest in floating rate debt, could participate. TCW's open-ended collateralised project loan obligation is one buyer, and has few imitators. Hastings runs a high-yield fund, of which infrastructure debt is one component, but it prefers primary deals. TCW closed a $1.5 billion Global Project Fund III in 2007, which complements its equity and mezzanine-focussed Energy Fund XIV.

According to Blair Thomas, group managing director for energy and infrastructure, "Banks have some good assets for sale, though we think that they were priced too cheaply initially, and discounts will need to reflect that."

Thomas says that TCW has built up a model for credit default swaps on project finance debt that has been adapted from the ISDA template. He says that the time and effort that has gone into the development is considerable, which would indicate that few banks have devoted much effort to infrastructure assets in recent years, even though synthetic risk transfer would be a neat way to circumvent sponsor restrictions.

Thomas, pointing out that the project funds' investors are actively looking for debt exposures, is sceptical that infrastructure funds would be long-term competitors for assets. "It's an indication of how bad things are in primary equity markets that they're looking at debt, but I don't see how they can make the returns their investors were promised this way"

One source at JP Morgan's IIF fund, without confirming that it wanted to buy any Tank & Rast debt, did say that it was looking at subordinated and mezzanine debt investments, a category into which some of the Tank & Rast debt would fall.
Banks' best hopes so far have been to repackage project loans into securitizations that can be used for repo funding from the European Central Bank. So far this year, BES, with its Eu1.1 billion Lusitano Project Finance 1, and NIB Capital, with its Eu963 billion Adriana Infrastructure CLO 2008, have availed themselves of the opportunity. But these deals have involved opening projects books to the ratings agencies, and sponsors have been wary of allowing this.

Bankers are also sceptical that the funds will emerge as major players, although their confidence presupposes that their debt commitments stay strong, and their balance sheets suffer no more major damage. Neither is guaranteed.