The funding environment for infrastructure PPP projects is much less favourable than it was two years ago. Whilst the increase in margins since pre-financial crisis continues to a large extent to be offset by the reduction in base rates, the real issue is that capacity for long tenor debt is extremely limited. When the terms on which debt is available no longer match those of the underlying project, the issue of refinancing risk arises. If only shorter term debt is available, assumptions must be made at financial close that the project company will in the future be able to refinance its debt within certain parameters of timing, price and other terms, and there is a risk that those assumptions will not be adhered to in some respect.
There are various manifestations of refinancing risk – the cost of finance may be higher than assumed, or it may be unavailable, or only available on terms that are not compatible with the existing transaction structure or documentation. In the absence of measures to mitigate or reallocate the risk, it naturally falls in the first instance on the project company/ sponsor (inability to refinance will result in default, and a refinancing on more onerous terms will affect the sponsor's return and may necessitate the injection of additional equity), and on the incumbent lenders (who must weigh up their potential exposure on contractor default termination against the prospect of continuing to fund the project perhaps at a loss). Depending on the circumstances it may also fall on the public sector: although it is a key principle in most PPPs that if the project has to be retendered due to project company default, the effects should not be adverse for the authority, in some circumstances that objective may not be fully attainable (for example where there is no liquid market). As between the lender and the sponsor, it is standard for the finance documents to include provisions which reallocate the risk between them, for example cash sweep provisions which require cash to be retained by the project company which might otherwise be available to pay distributions, thus improving financial cover ratios and making the project company a more attractive proposition for any potential refinancing lender. The causes of refinancing risk include credit issues such as poor performance by the project company as well as changes in market conditions.
Before the financial crisis the issue of refinancing risk was rarely considered: the tender documents required committed long term financing consistent with the financial model from the outset, and because this was readily available, the risk did not arise. However the contraction of liquidity which followed the onset of the credit crisis threatened to derail projects that were nearing financial close such as the UK Highways Agency's DBFM project for the expansion and maintenance of the M25. In that case, adherence to the requirement for long tenor debt resulted in the lenders requiring sufficient margin step-ups and other enhancements for them to feel comfortable that if the project company did not refinance in the medium term as intended, the unitary charge would still be sufficient to support their cost of funding for the full term of the loan. In effect, the authority bore the premium associated with a full assumption of refinancing risk by the lenders.
It was recognised that the future condition of the capital markets was highly uncertain and that if the situation improved the conventional 50/50 allocation of refinancing gains would be inappropriate, so the authority's proportion was increased, and the authority was given the right to initiate a refinancing to reflect the importance to it of obtaining cheaper funding if and when it became available.
Since then, it has become apparent that for a number of reasons significant levels of liquidity for long tenor bank debt are unlikely to be restored in the foreseeable future, although banks are likely to remain an important source of short term capital, and the agenda has shifted towards accessing institutional investors as a possible alternative source of long term finance on the assumption (yet to be fully tested) that private finance will still be capable of providing value for money on a risk-weighted basis. Given the fundamental nature of this change, it has generated questions and issues at a number of levels. However, refinancing risk has remained integral to the debate at several of these levels.
One approach which many commentators have naturally focused on is to seek to identify short term impediments to institutional investor participation. Construction risk is a particular area of concern as the construction phase involves a higher degree of risk than the operating phase, and even amongst those investors who see a high degree of correspondence between their risk appetite and that of project finance banks historically, many argue that institutional investors need to develop their risk management resources and expertise before they take on a significant exposure to construction risk.
A variety of possible models exist for addressing these concerns about construction risk, principally:
(i) mitigation through a portfolio approach (i.e. allowing a limited proportion of construction phase assets in a portfolio of predominantly operating phase assets)
(ii) credit enhancement, either through guarantees or other credit support, or the provision of subordinated debt
(iii) bank to institutional investor solutions, i.e. bank finance during the construction phase refinanced by institutional debt in the operating phase
None of these solutions is without significant complexities. In relation to the first, the main issue is conditions in the secondary market. On the supply side, the cost of funding of many project finance banks is higher than the interest rate under the loan, however they are reluctant to realise a loss by selling at a discount, and in the case of hedging banks there may also be concerns about orphaning a swap exposure. Portfolios have been sold in individual cases where the seller has for specific reasons not been subject to the sale constraints, including to institutional investors – e.g. the Bank of Ireland's sale of a portfolio of UK infrastructure loans to Aviva Investors in June this year was specifically sanctioned by the regulator. In the longer term, as the costs of Basel III start to be introduced and escalate, project companies may also start to feel the pressure, as in most cases banks are entitled to pass these costs on to the borrower. However the exit route for the borrower in these circumstances would normally be to refinance the debt, and the resulting refinancing gain would have to be shared with the authority.
In relation to the second model, many commentators have argued in favour of a solution based on enhanced credit/performance support from the building contractor. However this is likely to be difficult to achieve in an environment where contractors' balance sheets are under increased pressure, and from a policy perspective any intervention designed to promote this approach would have to be weighed against potential adverse effects in terms of market access and the costs associated with the additional support. In relation to support directly at borrower level, IUK has recently provided some details of the approach it will be following under the UK guarantees scheme to be enabled by the Infrastructure (Financial Assistance) Bill which is currently being fast-tracked through Parliament. Although there is a considerable amount of flexibility in terms of the products to be made available, the most natural application is likely to be providing a monoline-style comprehensive credit wrap with a view to achieving a high investment grade credit rating and thereby accessing the wider capital markets, rather than targeting specific risks. It is also possible that the scheme could be used to provide a construction phase guarantee for the benefit of institutional lenders. IUK has explained that the Treasury has been in correspondence with the European Commission and is confident that no state aid is involved as the pricing will be set to satisfy the market investor test, however potential beneficiaries are likely to want to explore the position in detail as the risk of an adverse finding would fall on them rather than the aid provider.
The third model has a number of attractions, particularly because it makes use of short term bank liquidity. However the issue of refinancing risk has generally been seen as the impediment. It might be possible to build a safety margin into the unitary charge (to be recovered through the refinancing mechanism to the extent of any surplus), but in the absence of committed (and collateralised) fixed price takeout funding there is always a risk that the refinancing will fall outside the assumed parameters leaving a residual risk with the sponsors and/or with the authority. Applying a conventional value for money analysis to the allocation of this risk it could be argued that it is essentially outside the sponsors' control (with the exception of certain elements affecting credit risk), and to that extent allocating it to the sponsors is unlikely to represent value for money. This may also be a case where there are limits on the extent to which a real transfer of risk to the private sector is possible – for example in the case of extreme disruptions to the market at the time of refinancing.
It is worth noting in passing that under the Regulatory Asset Base (RAB) model changes in the prevailing generic cost of finance are taken into account on periodic reviews and thus essentially borne not by the utility but by consumers. This is not intended to be an argument in favour of an extension of the RAB model to conventional PPP assets, merely an illustration of a different approach to the allocation of refinancing risk. The simplest way of transferring residual refinancing risk to the authority in the case of an availability based PPP would be a mechanism allowing for an uplift in the unitary charge. The authority can also be given an option to provide backstop financing at a given price.
A hybrid between the second and third models might also be possible, in which banks (or the Treasury) might provide subordinated debt during the construction phase alongside senior institutional debt, with the subordinated debt being refinanced or alternatively repaid before the senior debt in the operating phase. Such an approach would either reduce the level of refinancing risk in the project thus making it more manageable, or (in the case of early repayment), eliminate it altogether.
Yet another approach is to challenge the assumption that institutions cannot provide committed fixed price takeout financing. Given some institutions' focus on long term liability matching, it does not seem unrealistic to suppose that they may be able to acquire assets on a forward basis – i.e. to commit to fund at a future point in time on the basis of a fixed margin. This is the approach the Dutch pension fund asset manager, APG, has been able to take in relation to the N33 road project which recently achieved commercial close, where APG has committed to make fixed price takeout financing available for 70% of the senior debt. In order to access competitive pricing from pension funds the government offered to apply RPI indexation to up to 70% of the portion of the unitary charge corresponding to senior debt service and bids with different indexed and fixed rate components were compared with each other based on prevailing swap rates. Bidders were also required to submit bids based solely on bank debt, in order to demonstrate that any financing plan they submitted involving pension fund debt produced a lower cost of funding. The index linked portion was capped in order to retain the oversight expertise of the project finance banks. APG's refinancing obligations are conditional solely upon satisfaction of the completion test in the project agreement by the prescribed longstop date, and APG has a degree of involvement during the construction phase as contingent creditor. The senior debt component of termination compensation with respect to the index linked portion in non-contractor default scenarios is based on partial rather than full Spens. Many commentators have argued that government benefits from a natural hedge against inflation risk and its acceptance of inflation risk in this structure therefore does not involve any assumption of additional risk. An additional benefit of applying indexation to the unitary charge in this way is that the regulatory capital costs of swaps are rising and by aligning the unitary charge with the borrower's basis of funding it obviates the need for an interest rate swap.
Most of the proposed solutions above involve some kind of government intervention or other action. Another level of the debate involves consideration of wider issues beyond the immediate objective of achieving an increase in institutional investor participation. In particular, what are the merits of different possible vehicles/structures for institutional investor funding and what are their implications in terms of possible options available to government? Two considerations that ought to be relevant in this context are sustainability and efficiency. Pre-credit crisis experience of the monoline wrapped project bond market suggests that products which enable access to the capital markets via the provision of a comprehensive credit wrap tend to be associated with the development of risk management resources and expertise in the credit wrap provider, rather than in the purchasers of the bonds themselves. Based on this precedent, it seems likely that the benefits of the UK guarantees scheme (which will only be open for a two year window) in cultivating institutional investors as a sustainable source of funding for infrastructure will be limited. From an efficiency perspective, the fact that the main justification for promoting institutional investors is one of natural alignment with their risk appetite (at least in the operating phase) would tend to suggest that models in which investors can develop the capacity to take project risk (such as the Pensions Infrastructure Platform (PIP)) should be capable of providing the most efficient solution in the long run, perhaps in collaboration with banks in the construction phase. It was recently announced that the PIP has secured the critical mass of founding investors needed to move to the next stage of development, and it is to be hoped that institutional investors' interest in the various project bond initiatives will not distract them from also seeing the advantages of the model of which it is an example.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
