No, really, it IS all about the traffic forecasts

OK, Medium, here we go. Long-time listener, first-time caller. The always-brilliant @munilass pointed me at this series of Streetsblog posts on the bankruptcy of the Indiana Toll Road. I have thoughts on this.

A couple of disclosures upfront:

I’m executive editor of IJGlobal, which is essentially a trade publication for the private infrastructure industry and incorporates Project Finance, for which I worked from 1999 until April 2014.

I am no longer resident in the United States, after returning to my homeland in the UK about four years ago.

My publication has a reporter, @sakshishar, who covers this market full time.

My views are not necessarily those of my employer.

What we mean by P3

Another thing I’d like to mention. The definition of “public-private partnership” is almost infinitely elastic. I’ve been involved in lengthy and rigorous email chains with colleagues about how to define PPP/P3, and I’ve written an article for IJGlobal (p19 of the linked, and I believe free to view, pdf) that outlines some the many wonderful ways that the label PPP can be applied to infrastructure procurement.

The US is probably the best example of how loosely the PPP/P3 label can be applied. Corporate grant funding for parks, pharmaceuticals companies providing discounted drugs to emerging markets, leases of operational toll roads, and new bridge construction — all called PPPs.

You can have a long and occasionally illuminating conversation about the wisdom of using the label so widely (my guess is the expansive definition suits the private infrastructure industry just fine), but you can’t say “this dead deal is a PPP, therefore PPPs are a bad idea.”

PPP can be applied to very old, and very new, ways of financing infrastructure. Turnpikes and railroads have attracted private capital for hundreds of years, as American transport professionals never tired of telling upstart European investors around the middle of the last decade.

There’s probably one central idea that links all infrastructure PPP: that money has a time value, and selling the lease of a toll road or paying a private company to build your bridge rather than borrowing the money yourself is worthwhile, if you think you can achieve better economy-wide results down the line by reusing the lease proceeds or the money you could have borrowed to build a bridge somewhere else.

You may have pure or impure motives for making that decision. Or, politics being the messy business it is, a combination of the two. But the performance of a financing does not necessarily tell you anything about the wisdom of that decision.

A few years down the line, and having had reference to the performance of whatever else you did with that money you got from the toll road sale or didn’t have to borrow to build a bridge, you might have an idea. The UK, which is getting close to the 20-year mark in PPP, is starting to produce some interesting data.

But the wisdom of former governor Mitch Daniels’ decision to sell the lease of the Indiana Toll Road and reinvest the proceeds in projects that won’t produce toll revenues, or for that matter Indiana’s decision to pay the private sector an availability payment for many decades to upgrade I-69, will take years to become apparent.

It’s become very fashionable to say that there are a multitude of ways to privately finance transportation infrastructure. But for our purposes there are two ways:

Raise financing against toll revenues on either an existing or prospective road, and be exposed to the risk that no-one uses it. The first PPPs in the US, including Indiana and the Chicago Skyway, were financings for the acquisition of the rights to toll revenues on that road for many decades. A second wave of PPPs comprised the construction of new toll roads, particularly in Texas, against the promise of future toll revenue.

Raise financing against a multi-year promise from government (usually state, but theoretically municipal) to compensate you for building that road and keeping it open. Florida, and more recently Indiana, have used these availability-payment structures where they didn’t want to borrow to build a road themselves, but weren’t sure whether they wanted to hand over the toll revenues on these roads.

Hydra-headed argument

Probably my biggest beef with the Streetsblog series is the somewhat loose way it describes infrastructure assets’ ownership. The series concentrates on Macquarie, because Macquarie has attracted the most attention, in the shape of the Chanos/McLean tag team.

Describing Macquarie Atlas Road’s structure as “hydra-like” is really just a way of saying that Atlas owns a lot of different bits of infrastructure assets, those assets all have their own project companies, special purpose vehicles, as the jargon has it, and that whatever goes wrong with one of those project companies will not affect one of the other ones.

This is good for the lenders to those other project companies, and good for government. The last thing that French government wants to hear is that a portfolio manager at a hapless German landesbank is now the owner of a chunk of its motorway network because San Diego real estate development did not go as planned and a toll road serving that region went bust.

So the ITR Concession Co. LLC project company is not a spin-off. It’s a project company. Its debts are not obligations of Atlas, or Cintra, or Macquarie or the state of Indiana. This is a good thing.

Same goes for Macquarie’s use of outside equity. Macquarie was much better at raising outside equity for its infrastructure assets than, say, Babcock & Brown. So Macquarie is still standing, because the infrastructure investments that it managed and went wrong did not gnaw away at its balance sheet, while Babcock & Brown went bust and is consigned to the memories of aging leveraged leasing nerds.

Whether ITR’s lenders should have extended so much debt, both upfront or the larger balance that accreted, is a good question. But Streetsblog does not make the case that this purely private mess (or, for that matter, the implosion of American Roads, the holding company for the Detroit-Windsor Tunnel, among other assets) is a matter for wider political attention.

What can go wrong

Streetsblog’s main argument is that the same practices, mostly over-optimistic traffic forecasts produced by a fund manager playing with other people’s money, that caused problems for ITR et al are going to cause a problem when governments are procuring new, rather than selling leases for existing, infrastructure.

My main response to this is that in the instances where traffic forecasts will be the victim of the private sector’s perverse financial incentives — mostly on roads where the private sector takes on the risk of toll revenues being lower than forecast — the public sector will lose only to the extent that the Federal government takes losses on a loan it extends to the project (typically it will end up losing a small fraction of the road’s upfront cost).

So when local government puts in some money to help build a privately run toll road, as happens in Texas and California, the amount it contributes is way less than the cost of the road, and in the event of a bankruptcy the state still hosts a nice new toll road that it only paid part of the cost of.

Critics might argue that it should not be putting any money into a privately-operated toll road. That’s a legitimate policy decision, but states’ ability to borrow for shiny new roads and bridges is usually constrained, and private financing allows them to reduce the amount they have to use to get these built. If something goes wrong with these roads, the state’s obligation is zero, unless it decides to buy back the road from the hapless banks’ portfolio managers at a discount.

Still, the federal government can, and does, lose money on failed toll roads when it’s made a TIFIA loan to that road. To which I can only say:

The owners of the road almost always get wiped out first.

All lenders lose some money at some point or other.

The lovely thing about having “hydra-like” lending policies (or what you might call “portfolio lending”) is that, just like Macquarie Atlas, your winners can make up for your losers. The US Department of Energy, which lost so much money on bankrupt solar panel maker Solyndra, is now pretty much certain it will make it up on all those borrowers who didn’t default.

In the instances where a government procuring a road is exposed to losses on privately financed infrastructure — when it decides to keep the revenue from tolls, and hopes to kick back less than that revenue in the form of availability payments to the road’s builder, but ends up having to pay more — it cannot blame the private sector for producing the forecasts for those revenues, because the private sector is supremely indifferent to the revenues that the road produces. The public sector alone relied upon those forecasts.

I could add that dubious traffic studies have also been used to finance non-profit toll roads, among them the Greenville road that Streetsblog references, and the Pocahontas Parkway, which managed to require a restructuring as both a non-profit road, and then, after Transurban bought it out of that restructuring, later had to be handed over to its lenders, including, yes, the federal government.

The federal government could also lose money when a state decides to stop making its availability payments to a private developer because it is not bringing in enough toll revenue, and that operator stops making payments on its TIFIA loan. Which would be a savage indictment of how Argentina is infecting state government in the US with its loose fiscal morals.

Forecasts really are the villains

But ITR, and American Roads, and potentially SH130 5&6, are not really the death knell for private financing of infrastructure. They should be, but probably won’t be, the death knell for unreliable forecasts.

In fact, I am astonished at the persistence of unreliable traffic forecasts, especially because their distant relative in project finance terms, wind production forecasts, have managed to improve from a fairly shaky start. At pretty much all phases of my career writing about toll roads there has been a toll road blow-up where revenues didn’t match forecasts. We’ve had this-time-it-will-be-different moments at least three times in Australia alone.

One day I’ll look at why wind forecasting managed to improve its reputation after a shaky start. The reasons will probably include a narrower band of expected losses or profits, and maybe the fact that, until recently, other people’s money was not as big a factor in renewable energy finance. In fact, if the new phenomenon of raising outside equity for renewables assets in yieldcos takes off further, we’ll have an interesting point of comparison.

It could be that litigation will impose the discipline on forecasting that financial markets could not. Securities litigation suggests not. It is unlikely to curb the use of PPP/P3/concession-based transport financing.