No sooner had I written about Cardiff Airport(below) than news broke about the piece of infrastructure adjoining it, the M4 motorway, or rather the prospect of a new road being built to relieve congestion on it. Reports seem confused right now as to whether a prospective M4 relief road would be tolled or not.
But regardless of whether or not that particular road is tolled, I can say one thing with certainty. No new toll road would look anything like the only tolled relief road in the United Kingdom, the M6 Toll. Well, not financially anyway.
As I write, the M6 Toll’s future hangs in the balance. Its owners, Australian mega-investment bank Macquarie (via the usual web of companies) are seeking to restructure the capital structure of the project. In project finance, restructuring is what you do when you have to admit that your project isn’t going to make quite as much money as you hoped and so you are unable to meet the requirements of the lenders (banks in this case) to the project.
On one side of the table, a large group of banks led by Commerzbank, Santander, Crédit Agricole CIB and Espirito Santo Investment, advised by Deloitte. On the other, Macquarie and an advisory team from KPMG. Not their infrastructure team, you understand; the sort they’d use for a straight refinancing, but their restructuring team. This deal will be the infrastructure equivalent of telling Mummy and Daddy you’ve pissed your student loan up the wall and can they bail you out, but with the proviso that Mummy and Daddy gave you permission to do so.
Still, it could be worse. It could be the KPMG insolvency team.
The problem with the road is not, as the uninformed Mickey Clarke told BBC Radio Wales, that the M6 Toll “hasn’t ever made any money”, although it’s true that traffic has been below forecasts. Toll roads usually lose money in the early years when they have a lot of debt to pay off and traffic is still ramping up. The problem is that its debt is too big to be serviced by the traffic, and looks set to stay that way. Briefly, this is why.
The M6 Toll has been in operation since 2003. In 2006, debt at the project company was refinanced, i.e. bought out with more debt. This was the top of the market for infrastructure finance and indeed most finance, when the most reckless pre-crisis deals were being done. And this was such a deal.
As usual, a swap was put in place to turn the variable interest rate payable on the debt (the interest rate is always benchmarked to an inter-bank lending rate, which in the UK is Libor and which changes every day) into a fixed rate. This was a special accreting swap: which means that the fixed interest rate starts out artificially low – and the interest that would otherwise have been paid is added to the principle of the debt. Instead of amortising (going down as it’s paid off), the debt accretes as the rate stays low (1 per cent plus the fixed margin that replaces Libor) at for the first five years. Then it starts to go up every year for a further 15 years. This began in 2011.
The idea behind this was that traffic revenues would go up fast enough to pay off all this interest. Instead, they have remained low (although they did go up in the fourth quarter of last year, according to Macquarie, while traffic went down). A road with a design capacity of 100,000 cars reported average daily traffic of 37,000 in the last quarter of 2012. It’s not just these banks that are to blame for this absurdly over-optimistic financing: it’s Standard & Poor’s, who at the time, gave it an investment-grade credit rating (meaning the debt was affordable and looked likely to be repaid). Meanwhile, Macquarie used the £1.03 billion debt pile to pay itself a dividend of £392 million (similar to what it’s done in recent years with Thames Water).
Result? According to the annual results of M6 Toll owner Macquarie Atlas Roads (HQ: Bermuda), that £1.03 billion has become £1.21 billion. The debt has a nine year term, meaning that unless refinanced it has to be paid off in 2015. But even if refinanced, that interest rate swap runs for 30 years. So, more debt, getting more expensive, and no sign that there’ll be enough revenue to service it in a few years.
No bank in their right mind would want to start lending to the M6 Toll now: it’s practically toxic. As it is, the existing banks will have to bite the bullet and accept changes to their financing terms, and an extension to the term of the debt. Hardly a standard-bearer for the private investment the government wants to get into the road network.
So far, so bad. But it’s so bad, in fact, that it would never happen again. And that’s my point: we shouldn’t (unlike Mr Clarke) use the M6 Toll as a reason not to build a toll road, because only a lunatic would repeat the M6 Toll.
I’m not just talking about the financing either. The M6 Toll and its owners get all their revenue from tolls. Obvious when you think about it, but it didn’t have to be like that: they could give the toll revenue to the Department for Transport and get a fixed fee for looking after the road (which is what happens on Polish motorways built and operated by the private sector).
And that’s likely what would happen with a new toll road. Post-crisis and post-toll roads in other countries like Spain and Australia going bankrupt, investors much prefer to avoid taking the risk on the traffic across a road. They much prefer taking availability risk, i.e. they get paid as long as the road is ‘available’*. Traffic goes down in times of economic hardship, especially if tolls are involved. And especially in a country that isn’t used to toll roads and, judging by the M6 Toll, doesn’t like them. So the people who’d end up funding the costs of running a toll road would almost certainly be the UK taxpayer.
More on UK roads when I get the time…
*This is how a number of PFI road schemes are funded in the UK, including the M25 widening. The M25 is still looked after by a private consortium and will be until 2039. Another payment mechanism, as used on the A69 Carlisle-Newcastle scheme, is shadow tolling, under which the government pays the operator a fee per kilometre travelled per vehicle. But this, too, is a form of traffic risk, and therefore unpalatable.