As this blog notes from time to time, a lot of governments are looking for new private investors to invest in infrastructure projects. They are particularly interested in attracting institutional investors, that is insurers and pension funds, since their liabilities – insurance policies and pensions – last a long time, much like the return on an infrastructure project that trickles in over 25 years or more. The trouble is (as noted below) persuading them that such projects are not too risky. Now comes some research that re-affirms that they are not.
Sequoia Investment Management is an infrastructure asset management company. Put simply, it looks after loans given to infrastructure projects on behalf of the lenders. As such, it has a vested interest in drumming up interest in such lending. Even so, its recent report on the robustness of infrastructure debt is worth a read.
Sequoia have taken data collected by Moody’s, the credit ratings agency. Just as Moody’s, Standard & Poor’s and Fitch issue credit ratings to countries and companies, so they do for financial instruments for infrastructure projects. The data in question concerns default rates – that is, how often a loan instalment fails to be paid back and on time – for debt on infrastructure projects* of various kinds. A basic measure of how risky infrastructure projects are. Sequoia selected the data for European projects – 936 in all – and compared it to the data for European loans to companies, i.e. corporate loans. And the results are startling.
Even though ‘infrastructure’ here covers a broad range of projects, from schools and hospitals to power transmission lines, the default rates start out low and only get lower. The light blue dotted line shows default rates for A-rated corporate debt, the highest range of ratings in Moody’s system**. Despite the fact that this debt is rated in the same range as the infrastructure loans, within seven years we see infrastructure projects outperforming those corporate loans. For corporate loans rated Baa3 – still in the ‘investment-grade’ range deemed relatively low risk and unsepculative – a clear gap is visible by year five.
When you narrow ‘infrastructure’ down to public-private partnerships (the ‘PFI/PPP’ line) , which are procured by governments, the default rate over time lessens more. When narrowed down even further to ‘availability’, in which payments come from the public sector and are not linked to demand or tariffs, the default rate lessens even more. Within ten years, infrastructure projects show a rate of default so low as to be negligible. The relatively high (but still low in absolute terms) default rate for infrastructure in the early years reflects the fact these projects have to be built, unlike most companies that raise corporate debt which are up and running already. And those early years should be of little concern to the institutional investors who need to take a long view.
Institutional investors who run away from infrastructure because of the construction risk should take note: this project debt is less risky over time than corporate debt, and corporate debt in turn is, all things being equal, a lower risk investment than equity (share capital). Yet those same investors are quite happy to invest in the equity of public companies listed on the stock exchange. This makes no sense, although that doesn’t stop the refuseniks doing it.
The Sequoia report tells us not only that projects are a safer investment in the long run than the average large, relatively safe company (plenty of corporates are not rated investment grade) , it also tells us that the debt on such projects is relatively low risk for their credit rating. Which means that the way ratings agencies assess them and assign ratings to them is flawed. Two loans with the same credit rating, one corporate, one infrastructure project, can’t be expected to behave in the same way and one is considerably riskier than the other. Given how much investors look to ratings agencies to help reach investment decisions, I would say they need to do a better job.
*Note for geeks: That is to say, standalone projects with special purpose vehicles typically financed on a limited recourse basis; not infrastructure corporates.
**I.e. the range from Aaa (colloquially ‘triple-A’) to A3, comprising seven notches in all, equivalent to AAA to A- in the Standard & Poor’s system.