10 years after Lehman: Infrastructure
Infrastructure went into the financial crisis emulating the leveraged private-equity model, but is now evolving into a more diverse, long-term asset class. Christopher O’Dea reports
Infrastructure funds were gaining a foothold in institutional asset allocations in the years before the global financial crisis. They tended to be closed-ended, limited-life structures used commonly at the time for private markets. The fund management companies behind them had tapped the underwriting desks and loan departments of financial institutions that had previously raised debt capital or made loans to utilities, road-building companies and other large-scale projects – and they presented these new recruits as infrastructure investment teams.
As it turned out, investing capital was rather different from placing a debt issue or syndicating a loan. While capital was for the most part deployed to companies engaged in providing essential services, the collapse of financial markets and the ensuing recession revealed that too many of those companies either had short-term revenue deals that disappeared during the slowdown, or had taken on so much debt that there was no cushion against adversity.
The long-term income streams that had been expected to flow into pension fund coffers was cut short, and investors set about assessing what had gone wrong. It raised a pointed question: Can the fund management industry deliver the long-term cash flows investors want from infrastructure by using a private-equity fund model that was designed for short-term, leveraged returns?
Conversations with investors, investment managers and consultants suggest the industry is coming up with investment offerings that meet the varying requirements of pension funds. But rather a binary choice between the private-equity fund model and a new optimum structure, what has emerged in the past 10 years is a build-to-suit approach depending on individual strategies and investor requirements.
This flexible approach reflects lessons learnt since 2008, when numerous infrastructure investments foundered under heavy debt loads. The first is that the key performance factor at the asset level is not necessarily the amount of debt that an asset is supporting, but the security of the revenue stream available to support the debt service.
Taking that concept up to the fund level, the second aspect is that it is not the vehicle structure that matters most, but rather the ability of the vehicle’s revenue contracts to support the total debt being used in a given strategy – a lesson that taught managers to eschew debt at the fund level.
Investors have also learned the importance of robust operational management of infrastructure. That, in turn, has focused attention on the ability and capacity of investment managers to supervise the assets in their portfolios, and the ability and capacity of investors to monitor and evaluate their managers.
Investors clearly like the steady income from infrastructure, and in some cases, consultants say, they have sought to continue to receive income from assets beyond the expiry of fund structures – rather than have the assets sold and the capital returned. There is no simple solution on that front, but the discussions highlight the continuing search for the optimal way to access infrastructure returns.
Problems in the infrastructure sector after 2008 tended to arise in funds where the amount or structure of leverage – or sometimes both – was out of sync with the revenue-generating capacity of the underlying assets.
“The first generation of funds, raised in the early 2000s, did take their example from the private equity world, in terms of structure,” says Stephen Dowd, partner and head of infrastructure investments at CBRE Caledon Capital Management. “In the crisis, it was mainly funds that had not thought properly about leverage that tended to get caught out.”
Dowd would know. Before joining Caledon (which was acquired by CBRE last year) in 2014, he was a senior vice president in the infrastructure team at Ontario Teachers’ Pension Plan. During the crisis, “trouble occurred where financial structures that were appropriate for utilities or other long-term contracted assets that have very predictable cash flows were put on top of assets that had less predictable cash flows that vary more in response to changes in economic activity”, he says. “Where there was too much leverage and too much variability on the income line, the result was a liquidity squeeze or some issues around over-leverage.”
Performance suffered and so did the reputation of the asset class, but Dowd notes that questions of leverage and fund structure are separate issues, and cautions against conflating the two. With respect to leverage, “infrastructure investors definitely have learned the lessons”, he says. “You can see that in the greater subtlety of investment strategies, and especially in the greater subtlety around leverage strategies for different types of assets. Managers now, for the most part, are much more subtle and sophisticated – and the lenders are as well – about which assets can handle high leverage and which assets need more moderate levels of leverage. It’s definitely a lesson learned, and it’s definitely been well absorbed in the business.”
The issue of leverage highlights a paradox, Dowd says. While core real estate tends to have low or no leverage, “it’s the exact opposite when it comes to infrastructure”. Core assets with the most predictable cash flows are the ones banks are most willing to lend to. “So banks will lend longer-term, and they will be able to raise capital in debt markets on a very long-term basis because of the certainty around the cash flows,” he says.
What matters is the security of income. “The key thing is the contract,” says Dowd. “If you’ve got a lot of certain cash flows; if you’ve got them under a 15 or 20-year contracts; if you’ve got them under a regulatory formula that allows for a certain amount of leverage, you’ll still see very high leverage amounts on very stable assets.”
Private equity-style funds play a valuable role in the infrastructure ecosystem, Dowd contends. “Their mission is to take physically what looks like an infrastructure asset but doesn’t necessarily have the financial structure around it, or the contractual structure and the commercial elements that make it interesting for infrastructure investors, and de-risk these assets,” he says. “Once the private equity manager has done its job and made a certain level of return – which has risks associated with it, and therefore a certain target return associated with it – investors who are looking for lower risks then buy, at presumably a lower return.”
Investor understanding of the risk-reward characteristics of infrastructure assets has increased since 2008, in tandem with a greater appreciation of the need to rigorously assess the where and how any given strategy plans to use debt capital. Investment managers have responded by more precisely defining where a given strategy sits on the risk-return spectrum, a process of evolution similar to what occurred in traditional commercial property sectors.
“Originally we had a bifurcation of the market into core and value[-add], to use property terminology,” says Bronte Somes, head of infrastructure equity Europe at UBS Asset Management. “We’ve moved now to a more refined segmentation of the market into super core, core, core-plus and then the more PE-style end of the market.” She says this is partly down to “the amount of capital that has flowed into the market”. Somes adds: “Managers have looked to distinguish themselves, and they’ve done that by finding their niche on the risk-return curve.”
But it has also been a response to investor demand. Institutions with value-added exposure are now looking to add core or super-core managers to balance their exposures, and vice versa. The more refined labelling of strategies reflects underlying project economics, Somes says.
In general, super-core refers to the most highly-regulated assets and long-term contracted assets, which face little volume or demand risk. Examples of super-core assets would include UK and European regulated water and gas/electricity distribution. Core assets include district heating assets, transport or communications assets such as a cellular towers business, “which have mid-to long weighted-average life in terms of the contract, but does have some re-contracting risk”, Somes says.
Core-plus assets would be those with shorter-term customer commitments, of three to five years, and potentially merchant exposure to fluctuations in power prices or demand. “Then we’re moving into the private-equity or value-add space, where assets typically require a combination of a turnaround in the business or the need to grow by making bolt-on acquisitions or moving in a different direction,” Somes says.
One change that has occurred across sector Somes observes is tighter control of leverage. The ratio of net debt to enterprise value (EV) has ranged from 50% to 60% over the past 18 months, but varies across assets. More highly regulated assets carry relatively high leverage, about 73% net debt to EV, while some renewables assets with government subsidies can support 90% leverage. “It very much depends on the asset,” she says.
Overall, the points at which leverage is taken on are being limited. “A positive lesson learned in the crisis was not to have debt at the fund level in addition to the asset level and the holding company level,” Somes says. But in the past 18 months, she adds, “we have seen leverage creep up on acquisition financing”.
Although prices on many assets have moved higher, “they are being supported in the current market because low rates are driving a very different environment”, she says. “For one thing, managers have spent the last five years locking in lower cost debt financing and seeking to refinance as much as they could. Lower rates are creating some head room in a senior debt structure, and there is still a gap between the cost of debt and the cost of equity that justifies putting those debt structures in place.”
The debate about the role of PE-style infrastructure funds raises more fundamental questions about the appropriate way for pension investors to access the assets class, says Peter Hofbauer, head of infrastructure at Hermes Investment Management. Prior to 2008, “there was still a question mark around whether infrastructure was or wasn’t a separate asset class,” he says. Once accepted as a separate asset class, institutions not only allocated capital to infrastructure, but sought different methods of access, ranging from direct investment to separate accounts, commingled funds or for the largest investors, in-house teams.
While commingled funds were the dominant form, investors are “realising that, for certain assets within the infrastructure spectrum, the private equity approach isn’t necessarily suitable,” Hofbauer says. “Private equity focuses on transitional capital. It looks for an exit, and in that period, to maximise returns, and it tends to do that by way of a control investment.”
Hofbauer contends that the main objective of most infrastructure investing – to lock in long-term cash flows – conflicts with that approach on several fronts. “We don’t feel transitional capital is suitable for businesses and assets where investors need to play more of stewardship role,” he says. The key issue is that many infrastructure businesses provide essential services.
“Those businesses have broader stakeholder obligations than just financial investors – they have a social contract,” he says. “They have an obligation, not just like every other business to workers, but they genuinely have intergenerational aspects to them,” he adds. “They have quality-of-life aspects and a range of other stakeholders such as government and regulators which need to be taken into account.”
It is those characteristics that underpin the objective that most institutions have for their infrastructure allocations – to invest in real assets that can provide a real rate of return over a long duration by generating substantial cash flow with low volatility. While network businesses providing public services are the most likely to provide these characteristics, Hofbauer says, there is one more dimension in which such assets differ fundamentally from those that can be acquired and managed via the control transactions and governance models typically employed by PE funds – public service networks tend to be so large that “no single investor has the capacity, generally, to buy the asset”.
In Hermes’ view, infrastructure investing needs to move beyond fund structure to include development of governance methods suited to managing such very large assets in which pension fund investors with long time horizons are just one of several owners of the asset.
Australian and Canadian pension funds have undertaken such transactions in the case of toll roads and similar large projects. “By definition, if no single investor has the capacity to buy the asset, you are moving away from a control transaction, where one party controls the business and can implement and effect the required changes, to an ownership and governance model which is based on consortium type approach,” Hofbauer says.
There is no doubt that pension funds are eager to lock in the long-term cash flows that properly financed and de-risked infrastructure assets can provide. There are several options beyond the traditional PE-style fund. “Open-end funds seem to be a better fit on a few different levels because the open-end structure provides long term consistent exposure to the asset class similar to the ODCE funds in commercial property,” says Jan Mende, senior vice-president in Callan’s Real Assets Consulting group.
But at present, the choice of open-ended funds remains limited. Blackstone recently entered the sector with an open-end fund, Mende notes, and “some other managers are working on open-end structures”. The criticism that open-end funds deliver lower returns than closed-end funds misses the mark, she says. “That may be true, but [open-ended funds] serve a different purpose – it’s to hold the assets and collect the income.”
In fact, the desire for income may lead to innovation in infrastructure investing. Some managers have reported that some of their investors want to hold on to assets for longer terms than the life of the fund that owns the asset, Mende says.
“It will be interesting to see if some of those closed-end funds somehow have an exit that turns into an open-end vehicle, and it if that will be available for new investors to join,” she adds. Once an asset is stabilised and producing yield, some investors want to collect the dividends.
But that could be a tall order. “Some investors can’t stay forever, and if managers want to raise new funds they have to return capital to investors to demonstrate they can achieve their strategy,” she says.
“I can’t really say how it will play out, but the next few years should be interesting as closed-end funds reach their sunset and work through extensions,” she says. What’s clear is that the evolution of infrastructure investing is moving into a new phase.