This is an old problem in economics: two parties can reduce their costs by making a relationship-specific investment (with no alternative use), but this creates the possibility of opportunistic behaviour ex-post and the solution is to sign a long-term contract before the investment is made. With infrastructure, economies of scale require single, durable and immobile investments in relationship specific assets and investment can only take place if long-term contracts are in place.
Infrastructure assets are not real assets (as opposed to commodities or real estate) because they are only worth what contracts creating commitment to fund, build, operate and use a piece of infrastructure in exchange for an income stream say they are. Contractual governance is the main factor driving the value of such investments.
At the primary level, investing in infrastructure means investing in long-term contracts for the provision of specific capital goods and their related usage. Moreover, most infrastructure delivers a public service ultimately underwritten by the public sector. For this reason, infrastructure is only ever conceded, either explicitly or not. In other words, the source of enterprise value springs from the existence of a relationship between the public sector and a firm allowing the latter to provide a public service according to an agreed business model.
Of course, not all infrastructure is solely a function of public policy. However, the value of investments in highly specific, durable and immobile assets remains a function of explicit or implicit long term-contracts.
Investors can start thinking of infrastructure as investing in longterm contracts, in which different parties agree to take different risks and to receive their related payoffs.
A new asset class
Speaking of asset classes makes sense if the addition of a new class to an opportunity set improves the available risk-return tradeoff. An asset class also has to be investable if its characteristics are to be captured.
Most research on infrastructure investment refers to a list of industrial sub-sectors as the asset class. Assuming one can agree on a definition of what infrastructure assets consist of, a basket of these assets, weighted by their true value, would be a true representation of the infrastructure asset class.
But even if it exists in theory, this asset class cannot be invested as such: infrastructure remains too fragmented for investors to contemplate holding a basket of investable infrastructure assets. As a consequence, there is no passive infrastructure investment strategy since actual investments are likely to include exposures to the class and to firm-specific risk.
