Australia and Canada have the world’s most active institutional infrastructure investors. But they pursue different models, writes Georg Inderst
Australia and Canada are the two leading countries in institutional infrastructure investing. Australian pension funds have been pioneers in the field since the early 1990s, when their financial industry invented ‘infrastructure as an asset class’. Large Canadian pension plans spearheaded direct investments in infrastructure over the past decade. Today, pension funds in these two countries have the highest asset allocation in the world – approximately 5%, compared with a global average of 1%.
Lessons can be learned from the experience in these countries. A recent OECD working paper looks at factors such as infrastructure policies, the pension system, investment strategies and governance of pension funds.
With pension assets of about $1.5trn (€1.1trn) each, Australia and Canada are ranked fourth and fifth globally, behind the US, Japan and the UK. Both countries were, with the UK, early adopters of public-private-partnerships (PPP). There are other similarities between the two systems – in particular, trust-based pension systems, the absence of restrictive investment and solvency regulation, and a relatively stable political environment.
There are also some marked differences. Australia’s infrastructure industry started with a bang of large-scale privatisation in the 1990s, while Canada is still largely abstaining from privatisation. Unlike Australia, Canada has a well-functioning project-bond market. The occupational pension systems are at the opposite ends of the spectrum: voluntary defined benefit (DB) in Canada and compulsory defined contribution (DC) in Australia.
One of the main findings is simply that substantial infrastructure investments are possible in very different pension systems. It is often assumed that DC has prohibitive features for bulky, less liquid investments – for example, the large number of small individual accounts and daily fund valuations. In fact, Australian members can switch funds easily and quickly.
And yet, perhaps paradoxically, Australia’s industry-wide DC superannuation funds have produced a remarkable infrastructure investment history. Among the contributing factors, one can list the favourable demographics and economic growth, the use of default funds, and a widespread political and trustee commitment.
Illiquidity is often seen as a hurdle for mature and underfunded DB schemes. In this regard, the ‘maple revolutionaries’ of Canada found their own way. While pension funds have been ‘de-risking’ at the expense of listed equities, they were not forced into bonds by regulators (as, for example, in some European countries) but have diversified into alternative assets (estimated at 23%). Real estate and infrastructure are used especially, not only to diversify portfolios better but also for their long-term yield characteristics in liability-driven investing (LDI).
A second OECD finding from the two countries is that infrastructure investment vehicles can evolve and adjust according to investors’ needs. Listed infrastructure funds were initially most popular in Australia. These days, pension funds prefer outsourced, open-ended infrastructure funds, or ‘aligned asset managers’ at a comparatively low cost (the ‘new Australian model’). In contrast, the ‘Canadian model’ is about direct investing with in-house staff, aiming for better control and lower cost of investment.
A third crucial feature is the importance of a pension plan’s size when investing in less liquid assets. Both countries have a highly fragmented pensions scene. Private, unlisted infrastructure investing is primarily driven by large-scale funds – some of them with allocations of 10% or more – while smaller funds mostly invest little to nothing.
Fourth, asset owners need adequate resources, especially when investing directly: governance, management, operational, legal. Some Canadian plans admit that their own estimates of time and other inputs were too optimistic at the outset. Some smaller pension funds try to co-invest alongside larger ones, and there is much talk about different models of investment alliances and syndication.
In terms of performance, the experience has been mixed to fairly positive in both countries, but it varies considerably from investor to investor. Unfortunately, the data is still surprisingly poor. Many assets and products produce more or less the expected income and return profile. However, there have been disappointments during and after the financial crisis, and those can weigh heavily in highly concentrated portfolios.
Some of the lessons learned the hard way are:
• Overly optimistic demand projections and overvaluation of assets during the boom years (especially in the mid-2000s);
• Poor risk assessment (for example, demand risk of transport assets) and risk management (for example, excessive leverage);
• Market volatility of listed infrastructure funds (the ‘old Australian model’);
• Governance and fee issues with infrastructure funds;
• Pitfalls of investing in lesser-known overseas jurisdictions.
Important lessons can be learnt not only by investors but also policy makers. Governments in many countries would like to push more private capital into new (preferably domestic) infrastructure projects. However, most insurance companies and pension funds are mainly interested (or able) to invest in low-risk, operating assets.
Therefore, the idea of ‘asset recycling’ is popular in Australia – that is, the sale of old public assets to finance new projects. The administrations of both countries have consulted investors and the infrastructure industry in their PPP reform projects in recent times.
So far, the focus of Australian and Canadian pension funds has been primarily on equity. Although the interest on debt is rising, there is no such urgency for long-term investors to substitute banks as perceived in Europe. Project bonds play a much bigger role in North America, but such markets cannot be jump-started overnight. In Canada, insurance companies have been involved in long-term debt for decades while banks have traditionally only lent short.
From the outset, infrastructure has been a global asset class with surprisingly little home bias. Australia’s pension funds increasingly seek opportunities overseas. In Canada, PPP projects tend to be too small for large investors which invest most of their capital in larger ones in Europe and elsewhere. However, investors express concerns about the supply side, such as inconsistent infrastructure policies, the lack of a suitable project pipeline, retroactive changes of rules and regulations and political instability.
Following the example of institutional investors in these two countries, the demand for infrastructure assets is potentially very high. Many countries are competing for private capital, often with attractive sweeteners. For investors, however, it will remain difficult to bypass the traps of political opportunism and cyclical over-valuation of good assets.
Georg Inderst is an independent adviser