The institutional market for infrastructure has weathered a storm of challenges and changes over the past 18 months: tumultuous write-downs, brand-name fund cancellations, discounted secondary market sales, delayed dividend yields, unexpected levels of correlation with major market indices for listed portfolios, sizeable volatility in traffic for volume-based transportation assets, fundamental questions about valuation, and placement agent scandals. Not surprisingly, pension CIOs started to pull back from making new fund commitments in early 2008. Equity has been sluggish to flow. Recent auctions for the Gatwick and Melbourne airports failed to attract much bidder interest. First-time fund managers whose funds have not closed are expressing displeasure as prospects become bleaker.
Clearly, the market is undergoing fundamental changes. Generally, pension CIOs have become a much more introspective and wary lot. They have also become more sophisticated. After being hit hard by the crisis they have had little choice but to pull back and re-evaluate their allocations, strategies and fees from a first principles standpoint.
This article pulls together observations from across the market and suggests that we are entering a new era of institutional investment in infrastructure – the era of infrastructure á la carte – with new partnerships, coalitions and channels of capital flow being formed. As part of this trend, pension managers have started to pick and choose from a broad array of investment channels: direct deals, co-investments, fund managers, funds of funds, listed companies and managed accounts.
Each of these channels has some important characteristics, outlined here:
• Directs. The direct channel is effective for basic buy-and-hold structures with low teen internal rates of return, where private equity fees are prohibitively expensive. The large Canadian pensions continue to lead the way with their direct investment programmers and are the envy of many of their US and international peers. Their organisations are evolving to include internal sourcing, structuring and asset management functions. Perpetuating this trend, Macquarie and other fund managers have put greater emphasis on sourcing and advisory at a time when fundraising has been difficult and pensions are excited about reducing costs by pursuing direct investments themselves.
• Co-investments. The co-investment channel is effective for large multi-billion dollar transactions that exceed concentration limits of any one investor and for pensions who lack internal resources. There are several discussions in the market about forming new investment clubs, consortia and coalitions. For example, OMERS is leading talks about a global investment alliance, PCA is convening discussions about multi-pension pooled vehicles, and there has been a series of meetings hosted at Stanford University on this topic. The desire for club-style arrangements is especially strong amongst pensions who lack internal resources to go direct themselves, but who see the obvious benefits of reduced fees and long-term alignment of interests with like-minded institutions worldwide that share common goals, time frame, and risk-return expectations.
• Fund managers. The fund manager channel is useful for assets that have a value-added component and in markets where information asymmetries and access necessitate deep local knowledge and embedded relationships. However, the world of fund managers is being thinned out rather quickly. Many first-time fund managers who launched funds in 2005-08 have been unsuccessful in convincing the market of their value-add. The recent market turmoil has not helped. Of the funds that reached financial close in 2005-08, some were overzealous with leverage and pricing and will suffer write-downs. Emerging from the initial rush to launch infrastructure funds, there will be a small group of fund managers who break away from the pack to deliver healthy net returns and who will thus have continued access to the market.
• Fund of funds. The fund of funds channel can be attractive for small to mid-size pensions who lack internal resources to research and select fund managers themselves and for larger players who lack an infrastructure allocation but want to enter the space incrementally. Two fund of funds have recently reached financial close are H21 and Pantheon. But on the whole the model has thus far had limited demand, which is partly due to the structural problems with the market: double-layer fees not necessarily supported by return expectations; fundraising challenges in the primary infrastructure market; and a limited universe of managers with proven track-records.
• Listed companies. The listed company channel has benefits of liquidity and lower resource intensity. Some of the more sophisticated pensions have developed capabilities to compare relative risk-and-return of listed opportunities relative to unlisted opportunities. This approach has been especially attractive during the recent downturn when listed prices have been generally depressed. More take-privates are likely too.
• Managed accounts. The managed account has become increasingly attractive as a means to pursue a multi-channel strategy, both avoiding the need to internalise staffing and preserving flexibility. Pensions are exploring how to hire consultants on a discretionary or non-discretionary basis to manage custom accounts that could include direct, co-invest, fund and listed positions.
The multi-channel strategy has already been embraced by several of the larger and more experienced Canadian and European pensions. They are building up direct investment capabilities, participating in clubs and consortia, assessing the relative risk-and-return of listed versus unlisted opportunities and continuing to assess top-decile fund managers with demonstrated track-records and proprietary deal flow access. As the á la carte strategy for infrastructure matures and is adopted more widely, there will be a corresponding need for a) new data products and b) new business services.
• Data products. Data providers who track fundraising activity may consider widening the purview of data they collect and offer to the market. In the old world dominated by fund managers, it made sense to collect data on funds in the market, fundraising status, size of commitments, and date of financial close as an indicator of overall market activity. But in the new multi-channel world, data products will need to be widened to capture the relative growth of other channels too. Everyone gasped when in the first half of 2008 fundraising data showed a dramatic drop in activity by unlisted infrastructure funds. However, what the data didn’t capture is the extent to which the market is changing structurally and the pace of growth of the new relationships and coalitions that will support the new channels of capital flow.
• Business services. The multi-channel market is going to require new business services. Pensions that go direct will need buy- and sell-side advisors to help with sourcing and structuring of direct investments and co-investments. They will also need independent directors who have deep operational expertise. These are both areas where first-time managers who were unsuccessful with their fundraising can turn to next. The need for general education of pension plan trustees and investment teams is still tremendous, too. Consultants would do well to publish guidebooks mapping out risks and resource demands associated with each of the investment channels.
As the multi-channel market evolves, there are new research needs as well. Some questions that I would encourage my fellow academics and practitioners to take up in their writings are as follows:
• What are the tradeoffs between co-investing alongside fund managers versus co-investing alongside like-minded pensions who have built-up direct investment capacity, such as Alberta Investment Management Corporation, APG, Canada Pension Plan, Ontario Teachers’, Queensland Investment Corporation, and Victoria Funds Management?
• What is the role of emerging markets infrastructure in a portfolio – if the concession or regulated structure effectively blocks the investor from sharing in the GDP growth story, and there are considerable hedging costs, then what benefit is there other than mere diversification?
• How can we do a better job of comparing relative levels of risk across assets and structures? Market participants are quite skilled at communicating return expectations, but language and ability to consistently assess and compare political, market, operational, re-financing, terminal value and other risks remains clumsy and inconsistent.
• What policy and fiduciary issues arise from direct investment by pension funds in a public asset? How does the inherent tension between the pensions’ desire to earn returns and the public sector’s desire to minimise costs of services to the public get resolved? Are public pension funds the best stewards of public infrastructure, and why?
• What are the historical returns on private infrastructure transactions? Can fund managers make available more performance data to the market, including IRRs and asset level cash flows and distributions for different types of infrastructure assets?
As the market returns to normalcy, all these questions will shape the dialogue. Equity and debt are once again beginning to flow, but the structural changes discussed in this article are likely here to stay. As the multi-channel strategy takes hold, there will be more emphasis on direct investment, club deals and managed accounts, and closer oversight by wary pension CIOs of the specific risk-and-return characteristics of individual infrastructure investments. Fund managers will play a continued role, but will need to demonstrate their value-add relative to the other channels. Overall, these are positive changes and signal the development of a more sophisticated and robust market in the long run.
A copy of this article first appeared on in the November 2009 issue of Infrastructure Investor magazine.
Many thanks to David Altshuler, Philip Borel, Bob Hellman, Raffi Mardirosian, Shailesh Pathak and Cezary Podkul for helpful comments on earlier drafts of this article. The usual disclaimer applies; all errors are the author’s.