Unpacking the CalPERS decision

The California Public Employees' Retirement System ( CalPERS ), which is under-allocated to infrastructure but has an outperforming programme, has recenlty disclosed that its preferred way of investing in the asset class going forward will be through separate accounts.

The why of it is not new (or surprising), with CalPERS citing better economic terms and governance as the main drivers behind its decision. It's also important to put the California pension fund's move in the context of its own high-profile admission this summer, alongside fellow pension the California State Teachers' Retirement System ( CalSTRS ), that it had difficulty tracking the amount it paid in manager expenses. But its decision matters because it again puts the spotlight on infrastructure managers' ongoing existential crisis. 

CalPERS, which only has 0.7 percent ($2.2 billion) of its $288 billion portfolio allocated to infrastructure, against a long-term target allocation of 2 percent ($5.76 billion), is telling fund managers that if they want a piece of that extra $3.76 billion, they'll have to tailor their products to its needs. It's also bluntly telling general partners (GPs) that it's not really that interested in their preferred product – the commingled, blind-pool fund.

Earlier this year, the pension decided to slash external fund managers from 212 to about 100 over the next five years, but deliberately excluded infrastructure fund managers from this cull. In fact, it wants to increase the number of infrastructure fund managers it employs from a current six to a target of 10. It just wants to do that on its own terms.

The tug of war over fees is certainly not new, as attested by the rise of limited partner (LP) direct investments, co-investment programmes, separate accounts, direct equity partnerships with developers, or tailored programmes with the likes of Townsend or Partners Group. But when an announcement like CalPERS' comes along, it's worth revisiting just how pressured GPs are nowadays on everything from investment structures to asset class benchmarks

In one of the most interesting exchanges during our latest European fund management roundtable, Antin Infrastructure Partners chief executive Alain Rauscher complained that infrastructure assets are increasingly being packaged to offer a dividend to direct LP investors. 
This is dangerous, Rauscher argued, because it makes LPs think of infrastructure in terms of a coupon – something that offers them a better dividend than treasuries during this historically low interest rate period – but makes LPs “have no policy in terms of internal rate of return (IRR) and yield […] decoupling the protection of value that comes from using IRR as a benchmark”

Value creation, which IRR helps measure, is key for a GP's ability to charge carried interest. The more 'private equity-like' an infrastructure GP is, the more value creation and carried interest matter. If the notion of infrastructure investing is decoupled from the notion of value creation, or the latter is minimised, it's easy to see what's at stake here. 

It's worth highlighting at this stage that there are plenty of infrastructure managers prepared to reinforce less traditional views on infrastructure investing. As we wrote in our December/January issue, some renewable fund managers are more than willing to devise inventive ways to help LPs book their equity funds through their fixed income quotas, wittingly or unwittingly strengthening the notion that infrastructure is nothing more than a bond-like investment.

It's also worth noting that traditional GPs like Antin or EQT , through less orthodox investments in funeral homes, laboratories, cooking oil and laundry businesses, are also challenging the notion of what a typical infrastructure investment looks like.

In the end, there might not be a right or wrong way to invest in infrastructure. And it's certainly healthy to have differing views vying for supremacy. The trend that will almost certainly not go away, though, is LPs' forceful bid for more control over how they invest in this asset class. 

That and commingled, blind-pool funds. While the pressures we described are real, you need only look at the $27 billion set to be collected next year by Global Infrastructure Partners and Brookfield for their third infrastructure funds to know the model is in rude health. ?

ED KRAPELS' PERSPECTIVE

On the previous page, Bruno Alves argues that infrastructure fund managers are going through an ongoing existential crisis, as he explores decisions by CalPERS and others to invest in infrastructure using separate accounts as the preferred way of investing in this asset class.

He concludes with the observation that two of the largest infrastructure funds – Global Infrastructure Partners and Brookfield Asset Management – will book $27 billion for their new funds in 2016.

Hence the conundrum for many institutional investors: they want to increase allocations to this asset class, but the vast majority of the opportunities to do so come as invitations to be LPs in the same old fund structures. The problem with those structures, CalPERS and others have found, is difficulty in tracking, and therefore justifying, the amount paid to fund managers.

The funds have responded by devising a range of variations such as co-investment programmes to create the appearance of direct investment. Even so, the deal flow still originates with the fund. Most funds have a strong bias in favour of quick results, so they prefer to engage in buying existing infrastructure assets. The vast majority of infrastructure deals, therefore, are in secondary markets, and the return to LPs is ultimately determined more by the acumen of the GPs as 'infrastructure traders'. In this cycle, so far, nothing has gotten built: it's all about buying and selling existing stuff. 

The few funds that are willing to build new infrastructure facilities must find the developers who can actually get things started, permitted and built. For developers, therefore, the funds are a means of access to capital, but it comes at a cost. Naturally, the question arises, why shouldn't the developer and the LPs deal with each other directly?

It seems that the simplest way to break the existential crisis (and with a tumbling stock market, crisis seems to be an apt word) is for institutional investors to invest directly in infrastructure asset development and take responsibility for monitoring and managing that investment. That's not easy to do when your minimum cheque size is $200 million and your biggest problem is recruiting and retaining talented staff. 

Platform companies offer one solution to the problem. To a private equity firm, a platform company is the initial entity the firm acquires in an industry, which then becomes the foundation for other acquisitions in that industry. In an M&A-oriented discussion, this makes sense. But M&A should not be the only business a platform company is in.

The platform company can add more value if it's primarily in the development business, creating viable projects from scratch. In the electric infrastructure arena, for example, projects that are designed from scratch target internal rates of return of between 10 and 20 percent and a long string of steady returns, sometimes lasting decades because the underlying asset is so durable

Nothing valuable is easy, so the institutional investor that decides to pursue these higher returns must do something that it has traditionally eschewed: development risk. As Stanford professor Ashby Monk has noted, however, platform companies can be “access points” to the private capital markets where most infrastructure is incubated, developed, financed and built.

Platform companies are indeed access points, and some of them are eager to help institutional investors get better returns from infrastructure investment. But the quid pro quo is simple: the platform companies need to be able to access the institutional investors for development capital, in order to give them access to the better returns that typically go to the early investor. ?

 

Join us at the Berlin summit 2016 where this and many more topics will be discussed.