For all the angst about leverage levels and asset pricing in the infrastructure market – highlighted late last year in a risk analysis pitch by Standard & Poors (S&P) – investor and lender hunger for stable infrastructure-flavour cashflows has not been satiated. In 2006, 684 private equity funds worldwide achieved financial close, raising over $432 billion in aggregate commitments. Of that, according to S&P, $100-$150 billion is earmarked for infrastructure.

With so many new infra buyers, asset pricing is on the up. PAI Partners recently sold its France-based Saur water group to a CDC-led consortium for Eu1.72 billion – a 60% increase on the Eu1.03 billion it paid when buying Saur from Bouygues in 2004, and coextensive with just a 33% increase in Ebitda over the same period.

Most recently, and despite generous debt margins and a short tenor – 175bp and 200bp on the senior debt and 400bp on the junior, both over seven years – the Macquarie-sponsored Arqiva broadcast mast towers deal (£3.4 billion in senior and junior debt in total) has struggled in sub-underwriting and general syndication because of concerns over asset pricing – Macquarie itself estimates that the £2.5 billion paid for National Grid Wireless by Macquarie/Arqiva equates to a multiple of 21x historical earnings and 17x with synergies factored in.

As Mike Wilkins, managing director of infrastructure at S&P, notes: "With debt-to-Ebitda multiples in recent deals ranging from 12x to 30x, it is clear that, as a result of rampant demand, the infrastructure sector is in danger of suffering from the dual curse of overvaluation and excessive leverage – the classic symptoms of an asset bubble."

Willing lenders

S&P is right to ring the alarm. Lending generally has become more lax as bank liquidity has increased. Take any banker aside in the Middle East and they will admit that the bank competition for deals has been at the expense of due diligence.

Furthermore, asset pricing concerns did not stop HSBC coming in as a lead arranger (along with RBS, Barclays and Dresdner) on Arqiva, or sole underwriting £1.5 billion in seven-year senior debt and a £350 million capex facility for Macquarie's follow-up to Arqiva – the $1.9 billion Airwave emergency services telecoms acquisition from O2: Arguably HSBC has got more current exposure to Macquarie-sponsored telecoms debt than Macquarie Bank – and appears unfazed by that.

In addition, infrastructure debt margins and structures are being employed for assets that arguably do not have infrastructure cashflow characteristics. The recent Saur deal – with debt priced at infrastructure risk levels – is an asset backed by medium 12-year term contracts and not, for example, a 60-year concession.

High valuations and asset pricing in the infrastructure market at 10x-plus Ebitda is not new. In 2005 Babcock & Brown Infrastructure (BBI) paid 13x Ebitda for PD Ports and 13.7x Ebitda for Cross Sound Cable. It is also not always the case that 10x Ebitda is paid – BBI only paid 8.75x Ebitda for WestNet Rail in western Australia in 2006.

However, it is high asset pricing born from the expectation of very high re-leveraging post acquisition that is causing jitters. And those types of deal – the highly leveraged Bristol and London City airport deals in the UK, the Indiana and Chicago toll roads in the US and M6 refinancing in the UK – are predominantly sponsored by Macquarie or one of its many infra funds: Only the City airport deal – sponsored by AIG and Global Infrastructure Partners (Credit Suisse and GE Capital's infra fund) – is non-Macquarie.

Concerns about Macquarie's asset pricing were first voiced over deals like the Chicago Skyway 99-year lease concession in the US where the winning Cintra/Macquarie bid was $1.1 billion more than the nearest other bidder. Those voices were silenced – and probably rightly so given the tenor of the concession – by a respectable IRR forecast of 12-13% partially achieved through refinancing via imported financial engineering from Macquarie's past infra deals in Australia.

For the refinancing a zero coupon issue was expected by the market. What appeared was a synthetic zero coupon bond issue featuring accreting fixed interest swaps on floating rate notes covered by a triple-A monoline wrap that is extendible out to 53 years should a Series C refinancing prove necessary in the future.

Furthermore, the deal released $373 million of capital back to shareholders less than seven months after the acquisition closed and pre-funded an $80 million capital improvement programme.

The deal was a financial engineering coup and one that has since been repeated on the Indiana Toll Road and M6 refinancing in the UK. In effect, initial interest payments are set at below market rates with principal increasing over time. Less cash flow is diverted to creditors leaving more to be paid to investors as dividends over the short term.

But although a coup, it is hard to see how the model is sustainable over the long term – particularly if interest rates rise and refinancing the growing back-ended debt is no longer an option. Such a scenario is the antithesis of what pension funds – the market Macquarie and many infra funds are selling to – want for their long-term portfolios.

Valuations

Much of the problem is accounting-based – in effect the lack of clarity over how highly leveraged acquisitions, particularly Macquarie's, are valued.

Macquarie uses accounting that allows changes in the value of investments to be counted as revenue – in effect generating profits based on future asset pricing and refinancing at low interest rates, and disbursing those profits today (for more details search MIG + Emperor on www.projectfinancemagazine.com).

Macquarie also uses holding company structures to ring-fence debt – the non-recourse finance norm. But the holding companies also enable Macquarie to avoid limits on the amount of debt carried by regulated companies. For example, in 2005 a Macquarie-led consortium acquired French toll road operator APRR for Eu7.1 billion at a total leverage of around 11x cashflow. The holding company structure allowed Macquarie to avoid difficulties with the French government's maximum permitted leverage of around seven times cash flow.

None of these practices are illegal – but they are far from the norm and even further from the kind of conservative approach that appeals to pension funds.

According to a recent report by Edward Chancellor and Lauren Silva in Breakingviews, Macquarie funds tend to generate revenues by changing the assumptions in their valuation models. This often involves reducing the discount rate applied to future cash flows. The discount rate also includes a risk premium, which reflects the uncertainty of future revenues. By starting with a large premium and reducing it in successive years, Macquarie produces a reliable stream of revaluation gains. The logic is that as infrastructure businesses mature, their earnings become less volatile and more predictable and therefore the risk premium should also fall.

In recent years, risk premiums across all asset classes, including leveraged infrastructure businesses, have dramatically declined. At some stage they are likely to start rising again and Macquarie's investors appear potentially vulnerable to such an outcome: According to a footnote in Macquarie Infrastructure Group's (MIG's) accounts, a 0.5% rise in the risk premium would wipe over A$1 billion off the value of its investments.

Macquarie's aggressive engineering and valuation has occasionally made debt placement difficult. The refinancing in 2005 of Macquarie/Cintra's Bristol Airport deal, although eventually placed, has not been repeated. The deal upped the Ebitda ratio from 6x on the original financing to 12x creating a £121 million gain for Macquarie at very cheap debt pricing – 70bp-100bp over Libor in step-ups over the eight-year term.

The loan stalled after some banks complained about low fees, low margins and the lack of a default cover ratio. The structure was tweaked to include full underwriting fees for participants, a default cover ratio and an ownership covenant from Cintra.

Despite not being repeated – at least not in the airport sector – Bristol was both a leverage taboo, which Macquarie successfully broke albeit with difficulty, and a glimpse of the low margin debt that European banks now regularly offer despite taking on higher risks.

Macquarie has also had some significant successes in selling down debt – notably Chicago Skyway and the £4 billion Thames Water acquisition debt, which sold so strongly in senior syndication that no further sell-down was necessary.

And to date 99% of what Macquarie has done has been profitable for both investors and lenders – its is simply a case of matching the correct leverage for the type of asset and many of Macquarie's are monopolies and can therefore sustain high debt as long as Macquarie has the ability to charge more for their services. When that has been bad local users political restrictions have come into play and put an end to highly leveraged asset plays: The Thames deal is a case in point – not only is Thames a monopoly but Macquarie's ability to leverage the deal was restricted by UK water regulator Ofwat.

The Macquarie-model is aggressive, but it is not a model that is being widely adopted by the news entrants into the infra fund market. Consequently the prospect of an infrastructure dotcom-like death is unlikely – Macquarie-style deals are a very small part of the total infrastructure market and cashflow- rather than asset price-based long-term investors make up the vast majority of the growing infra fund market.

It also seems certain that Macquarie's big fee earning days will slow. If interest rates continue to rise MIG will start tapping the unlisted market to fuel growth, a sector that is dominated by pension funds looking for long-term stability at fixed rates and with no appetite for the kind of high leverage structures Macquarie favours, or the internal fee generating practises the bank has earnt so much from in recent years by selling assets from one of its funds to another.

$m value of new funds, by region

Source: The Rise of Infrastructure Funds by Ryan J. Orr, Ph.D/
Andrew Cantor/Connor Kidd IV/Josh Rafaelli