The bank debt market in most countries continues to be very competitive, with both loan margins and financing structures coming under severe pressure. There is also increasing competition from monoline bond insurers – especially on larger projects, and for refinancings of operational projects.

Monoline triple-A credit enhancement is especially suitable for accessing institutional appetite for long-term fixed income bonds (sometimes inflation-linked). This capital markets liquidity continues to grow apace, fuelled by growing demand from pension funds for long-term assets to match their long-term liabilities. Although the volume of long-term government bond issuance has started to increase in recent months, demand still outstrips supply and the opportunity to invest in assets with long-term predictable "utility-type" cash-flows and an underlying quasi-public sector logic, also combined with a triple-A wrap, continues to appeal to investors.

However, the monoline wrap can be used equally well, and sometimes to even better effect, to credit enhance bank debt. This approach can be especially suitable for new projects with construction risk, or for projects where sponsors want to have the flexibility to refinance the debt at some point in the future when the project performance has been established (e.g. projects with volume (traffic) related revenue streams) and the credit spread is therefore deserving of a reduction to match the reduced risk.

With a traditional wrapped bond these two aspects (construction and refinancing) can bring unnecessary inefficiencies:

* Construction phase: bond investors usually want to make all of their investment in a bond in one go. But this requirement does not sit comfortably with the actual features of construction cash-flow, where obviously the main funding is only required over the construction period (typically two to three years) to pay for construction work as it is done. Therefore, with a bond there is a negative carrying cost, which arises from the requirement for the (temporarily) surplus bond proceeds to be re-invested on a short-term basis which normally produces a lower rate of interest than is payable on the bond itself, thereby creating an inefficiency.
*  Refinancing: given the institutional bond investors' requirement to have long-term assets to match their liability profiles, they certainly do not want their investments to be prepaid earlier than scheduled. Therefore, under bond structures there are normally prepayment penalties (e.g. under the "Spens" clause in the UK market) which ensure that the investor still receives his future profit element over the remaining life of the loan.

However, banks are not so rigid on this point (indeed it is sometimes helpful to demonstrate to credit committees that the original lending decision has been validated by other parties!), and significant prepayment penalties – at least on larger transactions – are relatively unusual these days. (It should be noted, however, that breakage costs for cancelling any hedging arrangements will always need to be covered, and therefore careful consideration of the most flexible form of hedging should be made at the initial structuring stage. It is not always necessary to hedge the full loan for the full term.)

So, as long as the requirements to have the risk of future interest rate (and possibly inflation) changes satisfactorily hedged are met, the cost and or source of the underlying funding can be changed without too much problem.

Notes: (1) as a practical matter, given the current still relatively low historic long-term fixed rates in both the Sterling and Euro markets, the motivation for sponsors to refinance such bonds in the future may be considered relatively low. However, we have witnessed credit spreads for PPPs continuing to fall over time, driven by favourable lender/investor experience and increasing liquidity. So perhaps one should not totally discount this trend continuing. (2) Also in the UK the government has a standard requirement to receive a 50% share in any refinancing gains, and logically this can also be expected to reduce the likelihood of refinancing – at least in situations where the prospective refinancing benefits are marginal).

"De-risking"

Another development in financial structuring which has particular usefulness, at least in some circumstances, is the concept of "de-risking" – the ability of the public sector grantor of the PPP contract to reduce the amount of risk transfer in the contract (either at the outset, or at some point during the project life).

An example of this comes from Norway, where although the road projects being procured will be user-pay toll roads – such as the E18 roads – the form of PPP contract used by the National Roads Directorate bases the payment mechanism on availability of the road rather than traffic volume. This is sensible in view of the relatively low traffic volumes forecast on most of these roads, which would make them very difficult to bank on a conventional basis. Therefore, sponsors and lenders can base their (lower) risk-assessment on the aspects of the project's operations and maintenance that are more controllable by the project company. This naturally results in a lower risk premium and ultimately better value for money for the public sector.

Of course, the public accounting treatment of the project debt raised needs to be considered here: under Eurostat guidance, and of real importance in those Euro zone countries struggling to keep their public sector debt within the constraints imposed by the Maastricht Treaty, is the need for the public sector to transfer the construction risk and either the revenue or operating risk to the private sector project company. But this should be readily achievable in most cases.

Fig 1: Three Party Agreement

In France, this logic has been used for many years in financing public infrastructure projects such as tram systems, police stations, schools and hospitals efficiently. A commonly used structure is to have a three-party agreement between the public sector grantor (A), the private contractor/operator (B) and the bank (C) (please refer to Figure 1).

Under this simplified structure, after the construction phase has been satisfactorily completed and the project accepted by the grantor A, the bank C has effective direct recourse to the grantor through an irrevocable assignment by B of the stream of availability payments it will receive from A. This enables the bank to base its pricing on the credit quality of A (not B) which invariably will be higher as it will normally be a central or local government owned entity.

This also has the benefit to the bank in terms of BIS risk-weighting: the risk weight of a loan to central government is 0% (i.e. the bank does not have to set aside any capital to support its exposure) and for a loan to local government it is 20% (compared to 100% if the risk is directly on the private sector PPP company itself). This results in a lower cost of capital for the bank's exposure and therefore this can be reflected in a lower financing cost to the project overall.

Even when the new BIS rules on bank capital requirements (Basel II) are introduced (from 2007 onwards), this structure should still provide attractive financing economies: Under Basel II it is the actual credit rating of a Borrower (or more precisely the loan itself) which determines the amount of capital the lending bank has to set aside to support it. Most infrastructure project finance loans under traditional financing structures will, especially after the (higher risk) construction phase has been completed, achieve investment grade (i.e. a minimum rating from S&P or Fitch of BBB- or the equivalent Baa3 from Moody's).

Depending on the status of the lending bank's internal credit risk assessment model the risk weighting it has to apply to such loans will vary (and will normally be lower than the 100% under Basel I).

Risk transfer from public sector to private sector is still achieved: construction risk lies squarely with the PPP company, and the operational risk of deductions from the availability payment for poor performance is dealt with through a direct indemnity from the PPP Company (and its shareholders – who are most likely to be associated with the construction contractor and operator). Through this mechanism the grantor is still fully protected, but the bank's risk is insulated from the actual operational performance.

Variations on this simple theme have been used to achieve very effective financing costs for projects in Germany (such as the Offenbach Schools), Italy (such as the general contractor road projects for phases of the major highway upgrading for the Autostrade Salerno-Reggia Calabria) and even in the UK for the refinancing of the London Underground PPP contract awarded to Tube Lines (which is carrying out the refurbishment of the Jubilee, Northern and Piccadilly lines).

Secondary market

Another feature of the market that has been growing strongly is the secondary market for PPP equity. Various specialist funds are competing for stakes in projects being sold by the original sponsors, thus enabling the construction contractors to recycle their (often scarce) capital back into new project opportunities.

Leading investment banks such as Goldman Sachs and HSBC have announced the establishment of infrastructure funds which will provide an opportunity for institutional capital to participate in the growing PPP and wider infrastructure market developments. This influx of liquidity will assist in delivering the hundreds of billions of funding required to fulfil the stated investment requirements of European countries over the next 10 years.

On the senior debt side, secondary activity is low as banks generally want to hold good performing assets (often priced attractively compared to lower current market levels). This demand for loan assets is also exacerbated by the competitive presence of the European Investment Bank (EIB), which often provides 50% or more of the senior debt required at very low pricing and therefore is under pressure from public sector entities to maximise its involvement – thereby leading to an understandable reduced funding opportunity for the commercial banks.

However, to improve returns and achieve some capital efficiencies, banks are increasingly turning to sophisticated financial structuring techniques such as synthetic securitisation. In 2004 DEPFA, and more recently Sumitomo Mitsui and NIB Capital together, have successfully placed synthetic CLO of pools of UK PFI assets assisted by the German development bank, KfW. Again, this structure creates an opportunity for the banks to recycle part of their capital back into further PPP lending, and provides another way in which institutional investors can participate in the market.

We are likely to see more of this over time, but the effect of Basel II on bank capital will also be a contributory factor to banks' capital allocation strategies for PPP assets. There is a possibility that the good credit quality of operational PFI projects could result in lower capital risk weightings, with consequently lower and very competitive loan pricing, and if this is achieved banks can look forward to retaining a substantial involvement in the market.

Author: Paul Leatherdale, Managing Director,
Head of Infrastructure Finance Unit DEPFA BANK plc, 1 Commons Street, Dublin 1, Ireland. Tel:+35317922371. Fax:+353 1 792 2164. Email: paul.leatherdale@depfa.com Website:www.depfa.com