Labour and infrastructure

Via twitter, I picked up this interesting piece on how Labour's stance towards business could be made a bit more coherent. I thought I would bung up some thoughts on the points made about infrastructure, and what might be achievable here.

First up, I think we have to be clear about who wants what, and when. It's not quite as simple as saying that infrastructure needs finance and investors want to invest in/own infrastructure. Politicians want non-government finance to support the construction of new infrastructure. Pension funds and other institutional investors looking for a bond substitute generally want to own infrastructure that has already been built. Politicians are focused on greenfield, pension funds are more interested in brownfield.

This is because, again generalising, the risk return characteristics of infrastructure assets change over time. In the construction phase there are lot of unpredictable risks, so investors are going to want a premium in return for shouldering them. In the operational phase everything is more predictable (though political and reputational are maybe bigger factors than assumed) and the pay-off is lower. This means that, as an investment, infrastructure looks different depending on when the investment is to take place. As one report I read put it, infrastructure assets begin like venture capital/private equity, and become more like bonds.

It also varies by the sector that the infrastructure asset relates to. If you look at the risk-return characteristics of various types of infrastructure then you may find that existent social infrastructure (schools & hospitals) provides a nice fit if you are a pension fund looking for something bond-like. It's perhaps not surprising that the first investments of the Pensions Infrastructure Platform were in the secondary PFI market. The funds bought into existing assets (or project companies tied to the assets - the ownership of the asset usually resides with the public sector) rather than funding any new construction. Other markets see more pension fund investment in greenfield, but from what I've seen most UK investment is brownfield.

You can argue that the existence of a institutional investor market for brownfield infrastructure provides an out for those that are willing to finance greenfield. So the fact there is a secondary market to sell into allows initial investors to sell up, move on and hopefully finance the next greenfield project. But that is a slightly different pitch. Alternatively you could try and attract pension fund capital looking for an inflation hedge into greenfield investments by doing something to the financing to change the risk/return characteristics, like providing an underpin. But that ups the cost of financing, and might look worse than PFI!

I think it's also important to note that, as with pension fund investment more generally, most infrastructure is undertaken through intermediaries, rather than directly. This means that the time horizons of the investor are actually those of an asset manager. This is important, as there has been some unease about the structure of some infrastructure funds, which are built with a view to the manager being able to sell on assets within a few years rather than considering that a pension fund might want to own it for decades. This is changing. Some of the largest pension funds invest directly, and some asset managers have developed so-called "evergreen" funds. But the picture in unlisted funds is broadly like private equity more generally. The holding period for a particular asset might be half a dozen years or so.

Finally, we need to be aware that ESG criteria are not well embedded in infrastructure. Of course you might consider that some infrastructure investment is inherently "responsible" since it can involve the construction of wind farms and other renewables. But, as I wrote previously' labour issues can struggle to get attention in responsible investment overall, and this definitely extends to infrastructure (see DCT Gdansk for details). This is true for asset owners as much as asset managers sometimes. More specifically, I do not believe that in the UK infrastructure investment currently considers the need to respect workers' rights. It was notable that even the Labour-commissioned Armitt Review failed to address this.

Incidentally, I think this is particularly tricky in relation to greenfield. If you are building a new rail link I think you can make some decent projections about the potential environmental impact. You know what the immediate physical impact will be, and can model carbon footprint etc. But how do you do the same for labour issues? How can you tell in advance if the employer will genuinely respect workplace rights, or, on the other hand, if they will exploit precarious work by keeping workers on temporary contracts? How do you track this if the economic employer changes when the asset is sold on post-construction?

I don't mean to sound negative above, I'm just setting out what I've discovered in my own work around infrastructure. So I would suggest a few general, if obvious, guidelines. First, and most important, pension fund investment in infrastructure should only take place if it is in the interests of the beneficiaries. Second, government should consider whether pension fund investment in infrastructure is the cheapest financing option - I do wonder if we risk creating "pension fund PFI", another off balance sheet model that is more expansive that government borrowing. We do not want to crowd out the public sector. Third, a Labour approach to this issue must promote rigorous ESG standards, particularly with respect to workplace issues. And if we get back in power there should be union representation on the National Infrastructure Commission.