M25 – Lessons and challenges
The UK M25 design, build, finance and operate (DBFO) project reached financial close on May 20. By Paul Davies, partner, and John Carr, director, PricewaterhouseCoopers LLP.
To view the digital edition of this report please click here.
As a reminder – the £6.2bn contract includes widening the M25 to four lanes between Junctions 16 and 23 and between Junctions 27 and 30, the refurbishment of the A1(M) Hatfield Tunnel and the operation and maintenance of the M25 and the Dartford Crossing for 30 years. This article looks at the history of the M25 PPP, why it was decided to undertake it is a PFI project and how it reached financial close in these difficult markets. We also ask what lessons we can take from the M25 that will prove useful for future projects. In a nutshell, the lessons include:
* PFIs can still represent good value for money in the current markets;
* Consultation can help build a market for a project and determine the limits of risk transfer;
* Larger projects may need to be tranched to ensure sufficient contracting and finance capacity, and good value for money;
* Offers of government funding can be a catalyst to bringing in the finance markets;
* Rumours of the death of long-term lending are grossly exaggerated, but the market will continue to be challenging for some time to come.
The scope of the project
The M25 is one of the busiest motorways in Europe and is the key strategic orbital route around London. It forms the hub of the UK's motorway network. Up to 200,000 vehicles per day use the most heavily trafficked sections. It is significantly congested at peak times, during road maintenance and when there are incidents. This causes delays and increases the risk of accidents. As a result of a Highways Agency study carried out in 2001, it was decided to widen most of the rest of the motorway that had not already been widened to four lanes – the exception to this was the section between Junctions 3 and 5, where traffic flows did not justify it.
Why a public-private partnership/PFI deal?
A key driver for the Highways Agency in deciding whether to procure any particular project on a PFI basis is what procurement route would deliver best value for money. The Highways Agency is an experienced procurer of PFI projects (other than the M25, it now has 10 road DBFO projects, the TCC and NRTS traffic information projects, all procured on a PFI basis) and has a history of successfully delivering those projects. As a result, the market, both contractors and financiers, already understood the model well, which meant there was a well developed, competitive market ready and willing to bid for the new contract.
The Highways Agency conducted extensive analysis as to what projects in its programme might deliver best value for money if procured on a PFI basis under the new rules for investment appraisal, which in particular applied to the value for money analysis. Its conclusion was that the best candidates were three motorway widening projects that were in its programme; the M25, the M6 from Birmingham to Manchester, and the M1 north of Birmingham.
However, both the M6 and the M1 projects required new land to enable widening, the uncertainty of which would not sit well with a PFI procurement. So attention focused on the M25, where the widening could be undertaken without new land, although even this would present some significant engineering challenges (for example, how would an extra lane be fitted through the Chalfont viaduct, a Victorian listed structure? Some departures from standard would also be required (essentially discontinuous hard shoulders).
Considerable work was done analysing whether it was likely that procuring the widening of the M25 on a PFI basis was likely to deliver best value for money. Three essential judgements were that there would be significant economies of scale if the widening were undertaken as a single project; that the only realistic way of procuring the widening (expected capital cost in the region of £2bn) through a single project was as a PFI, based on integrated bidding groups; and that there would be benefits from combining the widening with operations and maintenance.
The maintenance of the M25 was undertaken by virtue of it being part of the Highways Agency's Maintenance Area 5, which also included a number of so-called "stubs and tails" (a number of motorways and trunk roads that intersected with the M25). It was judged that the benefits of integration that this approach gave should be preserved in a DBFO project.
So it was decided that from a procurement and operational point of view the best value option appeared to be to procure the widening of the M25 through a 30-year DBFO contract that would also include the maintenance of the M25 itself, the "stubs and tails", and the Dartford crossing. The latter is actually the A282, not the M25, whose operation and maintenance was previously procured through a separate contract.
The initial assessment of whether the project would represent best value as a PFI was made before the credit crunch. In the event, of course, the increase in lenders' pricing did reduce (but not destroy) the value for money case for the PFI route; but almost as fast as bank pricing was increasing, underlying interest rates were falling, so that the overall cost of finance was less affected than one would imagine, and value for money was preserved.
The value of market consultation
However, the decision to procure the widening of the M25 through a single PFI project was subject to the Highways Agency being satisfied that the market had sufficient resources to deliver the benefits that analysis suggested should be possible from such an approach.
Would contractors have the resources and capability to deliver on such a basis? Would the financial markets have the capacity to provide funding for the project on terms that would not be significantly different to a smaller DBFO project? Could the market price the whole project as one, and could the later upgrade sections of the project be priced at the outset at a level that would represent good value for money? Extensive market consultation was thus undertaken, including an Industry Day in May 2005.
From a funding perspective, a clear message emerged from these consultations. Capacity itself was not an issue, but there were two essential prerequisites for a wide community of funders to be interested in the project. First, they should not be exposed to demand risk, and second, that the Highways Agency should move away from the position adopted on all its previous road DBFO projects whereby no compensation was paid in the event of termination due to contractor default. The number of funders that were prepared to accept this was relatively limited, and would not be sufficient to fund the £2bn that it was then the intention to raise.
Contractors were also quite clear about their capacity to forward-price the work under the contract and whether such risk transfer would represent value for money. In short, the long construction period of the M25 clearly meant that it should be funded and contracted in stages.
The later upgraded sections
Consequently, as well as for operational reasons, the widening was divided into four sections. The intention was that the major schemes be staggered because if it were undertaken all at the same time, the disruption and consequential congestion would have been totally unacceptable, which coincided with the results of the market consultation.
The contract therefore requires that Sections 1 and 4 (Junctions 16 to 23 and Junctions 27 to 30 respectively) be undertaken as soon as the contract was signed, but that capacity would not be added to Sections 2 and 5 (Junctions 5 to 7 and Junctions 23 to 27 respectively) until the widening of Sections 1 and 4 had been completed.
It is expected that the widening of the first two sections will have been completed by the time the Olympic Games start in 2012. Section 3 (Junctions 1 to 3) has already been widened outside the contract as it was part of a scheme to improve the layout of Junction 2 and connecting roads, although the widened sections would come within the long-term responsibility of the DBFO company. This raised issues both as to how the later upgraded sections should be priced, and what should be done about the necessary finance for this later widening.
In respect of pricing, it was decided that it was most unlikely that a classic PFI fixed construction price would give best value, given that the widening would not even start until three to four years after contract signature. A number of possibilities were considered, but the chosen route, partly informed by the market consultation, was that the price should be built up by the DBFO company itself from its experience of the actual outturn costs of widening the initial upgraded sections. This meant that the tender price would only cover the widening of these initial sections.
So that the Highways Agency's hands are not tied, the contract gives them the right, but not the obligation, to require the DBFO company to increase the capacity of these later sections by using hard shoulder running. There is an obligation on the DBFO company to undertake the widening if asked so to do, subject to it being able to raise the necessary finance and demonstrating value for money.
The Highways Agency was, of course, concerned that the widening of the later sections could be jeopardised if the necessary finance was not available at the time and it required bidders to submit funding plans that showed how they intended to raise the funding necessary for all the widening.
However, it recognised that there could be adverse value for money implications of having funding for the later sections committed from the outset – for example, there would be significant commitment fees to pay on funding that would not be utilised for three years. Bidders were therefore invited also to submit funding plans that only included the funding necessary for the initial sections. In practice, these were the plans actually used to reach financial close. This proved a prescient decision in the light of subsequent tightening of capacity in the funding markets.
The M25 structure therefore showed that for larger projects with a long construction period, it may be beneficial only to raise funding and contract for the earlier years of a project, although in practice this may well constitute the majority of the overall project expenditure.
But while this is an obvious lesson for larger projects, it is only feasible if the structure of the deal ensures that the authority is not overly beholden to the incumbent contractor when the later sections arrive, which effectively means that the authority must have the capacity to fund those sections itself and contract separately if necessary.
Senior debt funding competition
Although the consultation with the funding markets had suggested that there was sufficient capacity to provide the necessary funding, the Highways Agency judged that it was unrealistic to expect three bidders all to come with classic PFI committed funding. This would in effect have required the funding markets to find £6bn of committed finance. It was therefore decided that there should be a senior debt funding competition to be run by the preferred bidder. This was in the days when funding competitions were the norm.
However, the Highways Agency judged that it was essential that bidders submitted a funding plan, and delivered evidence that those funding plans were realistic and deliverable. Otherwise there was a risk that it would choose a winner of the competition that would then not be able to deliver the bid that had won it.
Bidders were therefore required to include within their bids support from a potential provider of senior debt – the senior funding adviser. This support did not need to cover the full amount of the debt required, but would in effect show that the funder intended to provide an amount of senior debt on the terms and conditions as set out in the funding plan. In return, the senior funding adviser would be given a right to match up to 50% in the funding competition.
In these credit-restricted times, this approach is even more necessary for larger deals. There simply isn't the market capacity, or even the appetite, to underpin committed finance offers from competing bidders.
Tenders were originally received in the autumn of 2007, just as the first signs of the credit crunch were emerging. But tenderers' funding plans were essentially on pre-credit crunch terms. Following evaluation of the tenders, the tenderers were invited to submit limited retenders to clarify some uncertainties in their tenders. Following analysis of these limited retenders, the Connect Plus consortium (Balfour Beatty, Skanska, Egis and Atkins) was appointed preferred bidder in June 2008.
Although there were many other things to be done, the critical path from that point to contract signature and financial close was clearly going to be dominated by raising the funding.
Raising the funding
The funding competition was run by Connect Plus and its financial adviser HSBC, with the Highways Agency and its financial adviser PricewaterhouseCoopers LLP being closely involved and PUK giving additional input. However, it quickly became clear that the tightening of the financial markets meant that there would be real issues about capacity. The funding competition really became a bookbuilding process, with Connect Plus and the Highways Agency working together to achieve this. The funding competitions seen over the last few years, often between bank and bond alternatives, were simply not a realistic option given the state and capacity of the market.
In particular, when it became clear that there was a real risk that there would be insufficient capacity in the commercial banking markets, the Department for Transport (DfT) decided that it was prepared to provide up to £500m of senior debt itself as a co-funder. This is very much in line with what TIFU, the Treasury's new infrastructure financing unit, is now prepared to do, but this decision predated the creation of TIFU.
Although in practice this co-funding was not needed, the willingness of DfT to provide it was crucial to the successful outcome. It demonstrated to the funding markets that the government was committed to the project proceeding and that lack of capacity would not be an impediment to the project proceeding on traditional PFI lines.
The uncertainty of whether such government co-funding was required remained right until the end. Although the deal was in fact oversubscribed, it was not clear until the final week that this would be at pricing that could be accepted, that it wouldn't be overly reliant on outliers whose pricing didn't reflect the rest of the market, and that even banks with attractive terms might not meet the tight deadlines. The terms of co-funding were therefore negotiated and documented in full in preparation.
It is clear that the help brought to the project and in the current market will need to be a common feature for all of the larger PFIs. While co-funding was not required, the logic of establishing TIFU for future deals is absolutely clear.
The important contribution of the European Investment Bank (EIB) to the deal should also be mentioned, as it was supportive throughout and prepared to make significant amounts available to the project, both by way of direct lending and facilities guaranteed by the commercial banks.
The financial structure
After a tremendous effort, financial close was achieved on May 20. The key elements of the funding package were:
* A 16-strong commercial bank lender club providing £700m of commercial bank debt and a £215m EIB guarantee, which falls away following the completion of the widening
* A £185m unguaranteed EIB structured finance facility
* Shareholder equity of £200m, resulting in gearing of 85/15
* An average debt service cover ratio (ADSCR) of 1.4x required by lenders to ensure Connect is robust to a suite of possible downsides
* The ADSCR results in equity IRR in the base case higher than required by Connect Plus's shareholders, and higher than would have been acceptable to the Highways Agency. PwC therefore introduced a rebate mechanism whereby some of the cash not needed for debt service is rebated back to the government. On base case assumptions, £170m will be rebated back to the Highways Agency over the life of the contract
* Once the widening phase is complete, the contract provides for a minimum amount of compensation that would be paid to Connect Plus in the event of termination due to default by Connect Plus amounting to 60% of the total senior debt outstanding at the termination date, which effectively acts as a floor to the standard SOPC4 market value compensation provisions.
Margins, pricing and maturity
One of the biggest issues in the negotiations with the banks was maturity and this is an issue that will continue to be faced on PFI deals as many banks have been increasingly reluctant to lend long-term and the capital markets have not re-emerged to fill this gap.
Banks' reluctance to lend long-term is due to a combination of factors, each of which has a different level of significance for individual banks, but includes concerns over the capital they need to allocate to such long-term assets and the fear that their cost of raising funds might increase further, leaving them with a mismatch between long-dated loans and a more expensive cost of finance.
But for an authority to adopt a significantly shorter tenor would have undermined the whole-life cost approach of undertaking the project on a PFI basis. The structure therefore needed on the one hand a long-term loan commitment, while at the same time having elements within it that allowed a sufficient number of banks to determine that in practice this would be a loan of a far shorter tenor. So while the debt has a legal tenor of 27 years, two elements were introduced to heavily incentivise an earlier refinancing – a margin ratchet and a cash sweep.
The margins are: years one to seven 250bp; years eight to 10 300bp; years 11 to 27 350bp. There are analogous step-ups in the margins on the EIB debt. So it is pretty clear that Connect is incentivised to refinance before these ratchets kick in. There is also a cash sweep, as follows: years eight to 19, 50% of cashflow available post-scheduled debt service will be used to prepay outstanding and; years 20 to 27, 100% of cashflow is dedicated to debt repayment
The shareholders are taking the risk on the cash sweep in that they will not receive their desired base case returns if the cash sweep actually happens. It is noteworthy that many banks did not want a 100% cash sweep, as this would leave little incentive on the shareholders to manage the project well. There is certainly no uniform message among banks as to the desirability and amounts of sweep preferred.
In addition to the EIB, the banking group included Lloyds, Barclays, BBVA, SMBC, KfW, WestLB, HSBC, Bayerische, Dexia, RBS, Calyon, Helaba, Mitsubishi, NAB, Natixis and Société Générale. Given that only a few months previously it was doubtful that the whole amount of the financing could be raised, this was very encouraging, both for the deal and for the market. In particular, it involved a number of banks that previously were thought to have exited the market, but now seem to have taken the strategic decision to support project finance in general and PFI in particular.
So on the bright side, it does show that large deals of strategic significance can still attract finance; rumours that the market had switched to short-term lending or that there was simply insufficient capacity to fund large deals were ill-founded. But at the same time, the M25 success does not mean all is rosy again for a number of reasons:
* Without a resurgence in the capital markets within project finance, capacity will continue to be a problem for larger PFIs;
* While ostensibly banks can still commit for the long term, pricing ratchets and cash sweeps mean that in practice they are shorter-term lenders;
* These margins are inevitably eating into value for money and if sustained in the long term will erode PFI's attraction relative to other procurement routes;
* Banks are still unclear exactly how much capital they have available for project lending – deals of far smaller sizes may still struggle to attract finance on attractive terms;
* These constraints probably mean larger PFIs will need to be tranched, so that finance is raised in bite-sized chunks.
There are some clear lessons here for future large PPPs.