Asset Classes

Fees kill alpha from hedge funds

Close-up computer monitor with trading software. Multiple exposure photography.

The hedge fund portfolios for nearly 400 large institutional investors do not deliver on their promises of added return or risk mitigation and could be replicated at much lower cost by simple debt/equity blends, research by CEM Benchmarking, the Canadian-based provider of independent cost and performance analysis, has found.

The analysis draws on 17 years (2000-16) of CEM hedge fund data from 382 investors, mostly pension funds but some buffer funds and sovereign wealth funds.

On average, the hedge fund portfolios of these funds performed poorly, due in large part to the hefty fees paid to service providers.

The analysis shows that, before costs, the hedge fund portfolios added 1.45 per cent, relative to a custom-made CEM equity/debt benchmark; however, because hedge fund costs are so significant, there was negative alpha after costs. Across all styles in the CEM database, costs in 2016 were 2.72 per cent; that included 2.2 per cent for direct investing and 3.26 per cent for fund of funds. This diminished the hedge fund portfolio value add to   –0.54 per cent for direct and –2.11 per cent for fund of funds.

One of the authors of the report, Alex Beath, senior research analyst at CEM, says it was important for CEM to construct a benchmark to measure the outperformance of the hedge fund portfolios. Funds used two types of benchmarks for hedge funds in 2016: cash-based indices and specialty hedge fund indices. Both are flawed, Beath says.

Cash-based benchmarks, such as Libor + 4 per cent, have a correlation with hedge fund returns of about 7 per cent, are not investable, and are easy to beat.

Similarly, specialty hedge fund benchmarks are flawed for a number of reasons, not the least of which is that they are based on self-reported hedge fund returns that are not investable, or synthetic hedge fund replication, which is easily outperformed.

In selecting benchmarks, Beath says, there are a number of principles that should be used, including that the benchmark should be investable.

“An investable benchmark is what you could have had, a real alternative that was possible, and ideally implementable at low cost,” he explains.

The benchmark should also have similar risks to the investment program and fairly reflect available returns.

“Benchmarks that are too easy or too hard to beat may give undue credit for investment skill, or not give credit where it is due,” Beath says. “If a benchmark says it should produce a certain return and you put that into your asset allocation model and it’s the wrong information, it could have big consequences.”

CEM created a simple debt/equity benchmark to improve and standardise performance comparisons.

It found, on a gross basis, about two-thirds of the funds’ portfolios outperformed. But when costs were considered, only one-third outperformed.

“We’re not saying hedge funds have no skill; before costs they do,” Beath says. “But it’s the costs! They’re not serving their clients. If costs do come down, it could be worth it, but the way returns and costs are being shared right now is not in the best interests of investors.”

The investor portfolios that were analysed showed a variety of exposures to hedge funds and managers and ranged from five mandates to 50.

“When funds are putting together their portfolios, our benchmark indicates the diversifying elements of each hedge fund are cancelled out,” Beath says. “The nuance of a particular strategy is cancelled out.”

Join the discussion