In a report published on January 16 2013, the UK National Audit Office (NAO) expressed concern that if the British government gave large construction risk guarantees for new infrastructure projects substantial liabilities could arise for the British taxpayer.
This is an ongoing debate in the UK, but it highlights an issue of global relevance – numerous governments are now pushing for the growth of financing, particularly from institutional investors such as pension funds, of national infrastructure spending plans, while investors are increasingly looking at long-term assets like infrastructure.
Recent research by EDHEC-Risk Institute in the context of the NATIXIS Research Chair on infrastructure debt investment argues that construction risk guarantees are simply not necessary if scientific portfolio construction methodologies are applied to infrastructure investing.
In effect, they are likely to be damaging not only from a public welfare perspective but also from an asset management one.
Moral hazard arises from public sector guarantees – large projects that receive blanket (95 per cent to 100 per cent) guarantees of the debt financing create multi-billion pound liabilities for the taxpayer – take Metronet for example. Giving such ‘extremely naïve’ guarantees, in the words of the NAO, is a failure to recognise that construction risk is mostly a function of who is exposed to it.
Construction risk is either the result of unforeseen ‘exogenous’ conditions, such as the weather, or that of ‘endogenous’ incentives created by contracts allocating risks to different parties. Contracts that create incentives to control cost overruns can reduce and sometimes eliminate construction risk.
A construction firm that is given incentives to control costs, has plenty of experience of how much things cost to build, and is large enough to diversify project-specific construction risk (it is involved in numerous projects in multiple locations), is a good candidate to take construction risk.
This is what happens with standard project financing, from £50m school projects to £4bn pipeline projects. Construction risk is reduced by risk transfer in project financing as opposed to being increased by public sector risk guarantees. Using new data, it has been found that the median cost overrun in project finance is zero, against 20 per cent for traditional procurement.
However, while construction risk is to a large extent project-specific and thus diversifiable, it is still rewarded because it is a systematic moment in the life-cycle of infrastructure projects: construction risk typically attracts higher credit spread until projects are built and operational.
From a portfolio construction perspective, the construction stage of infrastructure projects can be thought of as a separate but related investment opportunity in infrastructure debt.
The empirical question becomes one of whether it creates value to add construction-period debt to the infrastructure investment opportunity set. From recent credit rating research (Moody’s 2012) that project finance creates predictable credit risk transitions as a function of its term to maturity: for a period of 10 years, the average project finance term loan migrates from Ba (or BB) to an A rating as the project is built, ramps up and becomes fully operational.
Moreover, if project finance construction risk is largely idiosyncratic, defaults triggered by construction risk should be mostly uncorrelated.
This predictable credit risk transition path suggests the opportunity to diversify infrastructure debt portfolios across the project life-cycle: combining assets with different risk/return profiles and low correlations creates diversification benefits.
In other words, investing in a portfolio of infrastructure debt that does not include some construction risk amounts to choosing to receive lower returns while possibly taking more risk.
In more technical terms, it is equivalent to investing below the efficient portfolio frontier. Adding some construction risk to an infrastructure debt portfolio would therefore increase returns and reduce portfolio risk thanks to diversification.
Investors in infrastructure debt should actively seek to invest in construction risk.
Moreover, if construction risk can be used to build efficient infrastructure debt portfolios, there is little need to push it out of sight and into new public sector liabilities.
Frédéric Blanc-Brude is research director at EDHEC Risk Institute-Asia
