There’s one thing that typically dominates the conversations I have with pension funds about their investment costs – alternative assets.
Total costs are always very sensitive to what happens in the alternative space. On average, across the 350 funds that benchmark their costs with CEM, 18% of assets are invested in hedge funds, private equity, real estate, infrastructure and other illiquid assets. These asset classes account for 54% of total costs. Private Equity in particular represents just 5% of assets, but 30% of costs.
The median global cost of investing in diversified private equity, including ‘carried interest’ (the performance fee), was 380 bps on the net asset value (NAV) for funds investing directly in Limited Partnerships (LP) alongside a General Partner in 2017. The median global cost for investing in infrastructure through LPs was 290 bps. These costs can be higher or lower depending on a number of factors.
On the face of it, these costs may appear extremely high and present a significant hurdle to overcome before the asset class can start to contribute meaningfully to the job of paying pensions when they are due.
But looking at cost in isolation is unhelpful. It must be viewed in the context of performance. To do that a good measure of the effectiveness of investment decision making is required, so we can understand if what we are paying is reasonable. This brings us to a fundamental issue – the widespread use of flawed benchmarks in private markets at a total portfolio level.
When funds use benchmarks that don’t properly reflect the characteristics of assets in which they are investing, then they have no sensible means to measure the outcomes of their active strategies - and therefore no means to understand if their costs are reasonable.
In the alternative space we often see funds using benchmarks such as a fixed hurdle rate, Consumer Price Index (CPI) plus a premium, or no benchmark at all. This is essentially an admission of defeat. CPI plus a premium may be a good proxy for growth in liabilities, but it should not be mistaken as a good yardstick when it comes to measuring the effectiveness investment strategy implementation.
Funds using unsuitable benchmarks run the risk of measuring more ‘noise’ than ‘signal’. Meaning those congratulating themselves on headline outperformance might do well to take a closer look.
The challenges in finding sensible benchmarks in private asset classes are numerous. But there are several key characteristics that funds can look out for to help ease the pain in pinpointing the right one. It should be a low cost, investable alternative that is fair but tough to beat. It should be a sensible yardstick for decision making throughout the investment process, a good proxy for risk and a means to provide accountability. Funds understand this, but still have difficulty finding benchmarks that work outside of public markets. Common challenges funds encounter include:
· Highly correlated but ultimately, uninvestable, peer-based benchmark options.
· Unsubstantiated use of premiums (with the premium itself also being uninvestable)
· Timing mismatches due to lagged reporting
· Mismatches between country mix, leverage and sub-types of asset.
This does not mean to say that alternatives should be avoided. On the contrary, they have produced excellent results for many funds. The point is that the measures of success must be right before investing. It is only through the right lens that we can truly start to focus on cost and value for money.
Good benchmarks do exist for most - if not all - alternative assets. CEM’s own research suggests that diversified private equity returns at a total portfolio level are highly correlated to small cap public equity, regionally weighted and with a lag (though the optimal lag is different for each fund). In our experience, this approach ticks all the boxes for a good benchmark and provides a sensible basis for a better understanding of net value added.
Talking about costs for alternative assets without addressing the adequacy of benchmarks puts the cart before the horse. The industry must address the issues with benchmarks, otherwise the cost debate will continue on the basis of supposition rather than data.
