The dust has – mostly – settled. Bankers have resumed attending syndication meetings. The backlog of financings that dates back to September 2008, and Lehman Brothers' implosion, has cleared. Bidding has resumed at pre-crisis levels for social infrastructure assets in markets such as the UK and Canada.

This recovery mirrors a more general recovery in corporate debt and equity markets, and like that recovery, is still vulnerable to shocks. The Nakheel debt standstill, and a threatened rash of downgrades to government-related issuers, indicates that banks' assets are far from solid. Debt against assets with exposure to traffic, trade and tourism will take time to recover.

But the recovery in credit markets has taken hold to a sufficient extent that it is possible to assess the responses of governments and multilaterals to the crisis. At the macroeconomic level, governments and central banks' responses to the crisis were broadly similar, and involved funnelling liquidity to banks both healthy and unhealthy. All projects with floating rate exposures, even those that closed with eye-watering margins in late 2008, are benefiting from the cheap money binge.

But governments' infrastructure-specific responses have been more diverse. Responses reflect a combination of ideological preferences, fiscal conditions and the availability of existing mechanisms to support infrastructure lending. Several governments, most notably that of the United States, have adapted existing programmes to the new lending environment, or seen demand for them increase.

Direct competition

Of all the programmes that involve direct lending to infrastructure projects, none are as tolerated by commercial banks, or beloved of sponsors, as the Federal Highway Administration's TIFIA loan programme. TIFIA debt, to recap briefly, involves the US government providing projects with 35-year subordinated debt at a single-digit margin to US treasuries.

TIFIA debt is now near-essential on a privately-operated transport project in the US, and has appeared on each of the last five road deals to come to market – SH130, the Capital Beltway, I-595, the Port of Miami Tunnel and the North Tarrant managed lanes project. Only one of them, Port of Miami, even flirted with not using a TIFIA tranche, before being driven back into the arms of the TIFIA joint programme office by bond market conditions.

But the product had a protracted adolescence, as bankers parsed the impact of TIFIA's insistence on sharing security over assets in the event of a default. Between 1998, when the programme was established, and the closing of Cintra's SH130, ten years later, the department closed two loans for PPPs, for South Bay Expressway and Pocahontas Parkway, two projects with troubled histories.

TIFIA debt is now "need to have, not nice to have", in the words of one sponsor, because it fills a useful gap in projects' capital structure. With long grace periods, subordination, at least in terms of cashflow, and a willingness to bend its own soft rules, TIFIA has become impossible to ignore.
TIFIA has won sponsor acquiescence by being generous in its calculation of project costs, of which it is limited to funding 33%, by agreeing to be used as a refinancing, so long as this covers some new construction, and by agreeing to a potential refinancing of other debt without demanding TIFIA repayment. On the North Tarrant financing, which priced as Project Finance went to press, it has cast aside its soft prohibition against being the largest lender on a project, since its $650 million loan is much larger than a proposed $400 million private activity bond, though less than 33%, when Texas' $570 million contribution and $420 million in sponsor equity is counted.

That financing was meant to be a bank/TIFIA financing, and banks are sceptical that TIFIA can take a back seat to bond investors in a potential workout situation the way it can with banks. They also question the means by which North Tarrant, a relatively untried managed lanes project, gained a rating high enough to permit both bond and TIFIA to be investment grade. "Of course, I'm bound to say this," says one banker who was working until recently on North Tarrant, "but banks really do play an important role in making sure that project structures are robust. And again, I'm bound to say this, but I'm not sure ratings agencies can do this as effectively."

This complaining says a lot about banks' fears that they will be permanently excluded from business once conditions improve. The European Investment Bank (EIB) attracted similar criticism before the crunch. In 2009 the EIB shot to first place in Project Finance's poll of sponsors' favourite development financing institutions on the back of its willingness to plug gaps in large infrastructure projects such as the UK's M25 and Manchester Waste, Portugal's Central and Pinhol Interior, and Poland's A2. Bankers hope that it will prop up several of Europe's more ambitious rail schemes.

But the A2 financing, which included a struggle between commercial lenders and the EIB over allocations, suggests a return to pre-crunch wariness. Says one banker "I think we're very happy when the EIB comes in to provide capacity on the large and difficult transport projects, but I find it hard to see the justification for its involvement in smaller social infrastructure projects, except that policy demands that the EIB spread the wealth across the EU."

The EIB, on the Portuguese road deals, already strayed outside its brief to support designated Trans European Networks, and with its support for the UK's wind development programme, will gain a foothold in a new sector. However, by mandating RBS, Lloyds and BNP Paribas-Fortis as intermediaries on this £1.4 billion ($2.3 billion) it is likely to mute any complaints. Such lending does not enjoy the margins of uncovered deals but tends, by hogging much less capital, to still be profitable for banks.

New mandates

Lending programmes have found it harder to get up and running from scratch. Probably the briefest appearance was British Columbia's offer to provide debt financing for the Port Mann project. The offer followed the failure of banks to settle on a pricing formula for a commercial debt tranche, but fell through after a deal with the project's sponsor, Macquarie, fell through. The Port Mann bridge is now being procured and built as a design-build public works contract.

The choice of a direct lending says a lot about British Columbia's approach to PPP, in that it refused to countenance any support mechanism that took it past its value-for-money threshold, and that it had the resources to put up a debt tranche of as much as C$1 billion. A subsequent initiative, called wide equity, which substitutes higher completion payments and sponsor equity for commercial debt, has led to howls of protest from banks and even some sponsors, though it looks unlikely to become a permanent market feature.

The province of Ontario, with a larger pipeline and more complicated history with PPP, opted for something that looked very much like a fudge. Its first post-crunch financing, the long-delayed Niagara hospital project, included a bank tranche, a bond tranche, the standard completion payment, and a series of milestone payments that it said should be refinanced in the bond market, but which it may be difficult to withdraw from the project's capital structure absent an unlikely ratings upgrade. By the time of the second of Ontario's deals, Bridgepoint, the milestone payments were unnecessary. A combination of bank and bond debt, and generous ratings, means that post-crunch pricing has returned to a viable level in the province.

Canada's Federal government has announced two programmes to support infrastructure development. The first is a C$1.2 billion fund to be run by the newly launched PPP Canada, which looks likely to concentrate on smaller, and less well-served corners of the market. The Federal government has also temporarily allowed Export Development Canada to operate in the domestic market, though it is barred from financing hospitals, schools and other social infrastructure. It could however, crop up financing power deals, where it has a track record of working in the US, Middle East and Latin America, and where its approach of joining deals as an arranger rarely ruffles bank feathers.

The UK Treasury's Infrastructure Finance Unit (TIFU) remains the least understood of the government lending programmes. Its creation reflected the fact that the most active banks in the country's PFI market, a combination of UK banks and European public finance specialists, all had a horrendous crisis. The unit, which came into being in March 2009, and closed its first deal, a £120 million loan for Greater Manchester Waste, a month later. It has not closed a deal since, though it has since been linked, unconvincingly, with the Crossrail project.

Government lending can serve as a "policeman", in the words of one lender, preventing banks from setting pricing levels above the Treasury's expectations. In this reading the M25, whose size looked like presenting a challenge to lenders, panicked the UK's department for transport into pressing for the creation of a dedicated lender. The Treasury stresses that TIFU, the belated result of this lobbying effort would act only to supplement lagging bank appetite, and the abrupt withdrawal of Bank of Ireland capacity on Greater Manchester fit that description.

While TIFU's presence was ultimately not required on the M25, it acted as a watchdog and prodding cooler heads among the banking community to prevail. Several banks would be happy to se the back of TIFU. "I think competition has returned to the market, and can't see how the Treasury will be able to justify funnelling untold billions to marquee projects in the current fiscal environment, especially if there's now a functioning lending market."

Soft power

Victoria's liquidity backstop for the Aquasure syndication was much less controversial. Unlike the M25, where TIFU debt was seen, perhaps unfairly, as a competitor, the Australian state provided a promise to sponsors that it would make good on any under-subscription in syndication, but disclaimed any interest in setting pricing.

For Aquasure, which syndicated at the end of October, the state asked sponsors to submit best-efforts bids, because in September 2008, when it approached sponsors, full-underwritten bids were impossible to find. The offer, which benefited from the relatively strong fiscal position of the state, was a delicate way of maintaining liquidity without making lenders anxious.

The A$3.7 billion ($3.34 billion) syndication ended up 50% oversubscribed and banks were scaled back. NAB and Westpac bought in 32 banks, including lenders that backed a competing consortium. The process may not have saved the state money, but it brought a large project to close much faster than waiting for underwriting appetite to reappear.

The state of Queensland came up with a proposal, the supported debt model (SDM), that in rationale owed much to a pre-crisis UK initiative – the credit guarantee finance (CGF) model.

Under the CGF the UK government could take advantage of its cheaper cost of funds, but with banks or monolines providing guarantees. Under the SDM, funding during the construction phase is 100% financed by the private sector. However, during the operating phase the state treasury provides 70% of the senior debt (considered the risk free component) and the private sector provides the balance of funding as sub-debt (20%) and equity (10%).

SDM is designed to allow commercial banks to concentrate on construction financing, at tenors where they were most comfortable, and then refinance most of their commitments with government-provided low-cost debt. The South-East Queensland schools project was the debut, and so far, only, use of the model.

SDM has issues with intercreditor arrangements that resemble those on TIFIA financing, although in this instance commercial banks would be subordinate, subject to some protections. It also encounters similar conflict-of-interest issues to the UK's CGF – with the state acting as both procurement authority and lender there is the potential for the sponsors and sub-lenders to be pushed in unwelcome directions.

The UK's CGF eventually failed to get traction, mainly because the Treasury had to book the debt twice – once for the issuance of gilts and once for the project debt – but also because it was based on guarantees from sickly monolines and PFI lenders. Supported debt does not have these issues. However, given Queensland's sporadic deal-flow and large proportion of deals with traffic risk, it will be some time before the SDM's ease of adaptation is tested.

Banks are happier with government initiatives to reduce refinancing risk. Short tenors continue to be an issue in the Australian market, and were one reason for the timing of the supported debt model. Under the refinancing guarantee, banks submit pricing both with and without a refinancing guarantee, and government can decide if it is cheaper to provide a guarantee or countenance the higher pricing.

The policy has echoes of the fix that BAM, Vinci and CFE settled upon with Infrabel on the $1.07 billion Liefkenshoek project, where the public sector agreed to shoulder the burden of increased funding costs in return for a share of refinancing benefits. Funding costs, in an era of cheap liquidity and persistently low interbank rates, have become less of a live issue, but refinancing anxiety persists, especially in markets outside Canada, where vibrant bond markets have yet to return.

Direct state guarantees are still on offer. The French government is extending its Dailly protections for PPP lenders to construction, and even demand, risk on larger and more complex projects such as Reunion Tram-Train. The guarantee has a ceiling of Eu10 billion, but is only available for the next two years, like the US government's grants to renewables projects. Sponsors will be wary of factoring such generous temporary packages into their business planning, however. The US has rolled out debt guarantees programmes for renewables projects, and has so far succeeded in creating a vibrant financial and legal advisory services market (for more see inset next page).

Emerging economics

Emerging markets governments have lacked the room for manoeuvre of more advanced economies, and have been more inclined, where fiscal circumstances permit, to procure new infrastructure directly, or, in the case of China, to direct banks to lend to favoured sectors. India's banking sector emerged from the crisis relatively unscathed, though State Bank of India continues to have an outsize role in the country's project finance market.

Brazil's BNDES cemented its dominant position in that country's infrastructure market, leaving opportunities for commercial banks only where a deal, by using foreign content or involving concession payments, rendered itself ineligible. The use of foreign boilers limited, for instance, BNDES' contributions to MPX Energia's Pecem I financing, but for the follow-up, Itaqui, it will be covering nearly 60% of that project's cost with debt. On the other hand, a commercial lender, Bradesco, and an export credit agency with no track record of closing toll road financings, JBIC, came together with the Inter-American Development Bank to finance the $1.47 billion RodoAnel Oeste concession in Brazil. JBIC's financing, untied to either Japanese sponsor equity or content, may not be repeated, though Bradesco's R760 million subordinated loan, the first use of such a facility in the country, deserves to be replicated on projects with uncertain traffic demand projections. (For more, see p.28 this issue.)

Mexico's government is taking steps for its state-owned development bank, Banobras, to lend directly to sponsors building new road concessions, even where these involve upfront payments back to central government. In Mexico, where bank appetite is not close to returning to pre-crisis levels, such a move might be justified, though it will discourage banks from expending substantial resources on bids in Mexico, and several New York bankers focusing on Latin America have indicated that will shift focus to Brazil. (For more on the programme, as well as a novel new product from MBIA and Opic, see Mexico feature p.50 this issue.)

At least one reason for the lack of eye-catching infrastructure finance initiatives in the least developed emerging markets is that development finance institutions had never ceded much ground to commercial banks. The Eu113 million project financing for the Rabai project in Kenya did suffer from a mid-2008 spike in pricing, as well as a series of challenges to its license. But with a lending group consisting of DEG, Proparco, FMO, Emerging Africa Infrastructure Fund (EAIF), and European Financing Partners, its October 2008 close was not unduly troubled by Lehman Brothers' collapse.

This dynamic has not stopped the International Finance Corporation from putting forward a response. Its Infrastructure Crisis Facility has spawned a debt pool, with commitments from development banks Proparco, DEG and the EIB, which together have allowed it to grow to $4 billion. Cordiant Capital, a veteran B loan investor, will manage the fund.

The debt pool has already made at least one disbursement, a loan of indeterminate size to Carrix for a Vietnamese port concession. It had, at least until recently, a strong rationale, in the form of weak B loan market liquidity. The $2.3 billion Panama Canal financing featured minimal commercial bank participation and probably took place at the low point for project lenders, though the borrower's aggressive insistence on long tenors would have made the deal a stretch for banks in the best of times.

Financings like RodoAnel indicate that for the proverbial right projects, in the proverbial right markets, B loan appetite is back. The mutilaterals that have committed to the debt pool now have to demonstrate that their contributions cannot be made just as effectively as A lenders.

Fix and make up

The final category of response is hardest to categorise, since it takes place on an ad hoc basis during the course of negotiations and rarely acquires the trappings of a programme. Ontario's enhanced milestone payment, for instance, just about became institutionalised, even if it had a short shelf-life. The governments of the Gulf Cooperation Council region insist on value for money, but rarely have to worry about accountability. They do not have to justify their modification of power and water purchase agreements to fractious legislatures.

There's a rationale for trying to formalise a fix, especially where a project's financing might be open to state aid or other legal challenges. The series of tweaks to the concession structure on several Portuguese roads, which were responses to lender concerns, have resulted in their rejection by the country's Accounts Court, because these did not allow for all bidders to respond at best and final offer stage.

The country's roads agency Estradas de Portugal, is likely to cover the risk to lenders of the concessions' cancellation if it cannot turn round the court. Lenders, by and large, say they prefer this sort of sweeping cover to the efforts of procurement authorities, usually of the Anglo-Saxon bent, to indemnify themselves. "I'm mystified as to why we need a change of law reserve on a UK PFI deal," says one lender. But this sort of flexibility has its advantages, and lenders are forgiving, provided government can show a consistency at least in principle. "Tweaks are a normal part of putting together a workable deal," adds the lender, "it's just the circumstances that change."