How to benchmark pension fund costs
It is, suggests Richard Stroud, chief executive of the UK’s Pension Trust, a question of being able to compare like for like. Take grocery shopping. UK Supermarket retailers Tesco and Asda compare their prices in order to attract customers. Why shouldn’t investment managers and third party administrators do the same, he argues. It is not a view that everyone agrees with. “You can’t compare third party providers with the Tesco versus Asda example, because you are buying a service, not commodity,” argues Allan Course, head of administration consulting at Watson Wyatt.
Still, there is some room for making cost comparisons, and many pension funds, notably those in the Netherlands, are already making them. They are benchmarking their administration and investment costs. “Pension funds need a benchmark. We do not have profit and loss which we can compare with other companies, so we need a benchmark to know which costs are high and which costs are low,” argues Henry Dikkema, group controller at the PGGM pension scheme.
Both PGGM, and ABP use CEM Benchmarking, the Toronto-based firm, to provide cost measurement. The fund provides benchmarking for both benefit administration, and investment. 60 pension plans from the US, Europe, Canada, and Australia participated in its 2005 survey. One third of the plans were from Europe and Australia
“It is important to consider your costs in the context of your peers. If you are trying to understand cost, one way to do it is to compare yourself to other pension funds of a similar size,” explains Hubert Lim, research director. He claims that CEM has a rigorous, proprietary process to validate all the data that is submitted to it. Pension plans complete a detailed survey that covers holdings, returns, investment costs, benchmarks, and other related data. The data submitted by any given fund is subject to omission and variance analysis, and is compared with data previously submitted and that in the overall universe. The omissions and variances are highlighted automatically when they are entered into the system, and are analysed by an experienced team of analysts. If there are any discrepancies, confirmation of data is required until the company is satisfied.
The firm compares investment costs for 24 asset classes broken down by management style. For expenditure spent in 2004, the global average for total costs was 37 basis points, mandate up of 31 basis points for investment costs, and 6 basis points for administrative costs.
Still, not everybody is convinced about the logic of benchmarking. “It is difficult to find the right benchmark. It really depends. For example, on the asset management side, it depends on what kind of management philosophy you have. There may be people who have 60% invested passively, and they have a different cost structure than those who are active,” points out Hartmut Leser, managing director of Feri Institutional Advisors in Germany.
Georg Inderst, a consultant, points out that some costs are harder to calculate than others. There are explicit and implicit costs. Implicit costs are things like the cost of delaying a trade or not being able to trade immediately. “These are big costs but they are hard to measure. At the moment, they are more of a concept,” he suggests. Still, other costs are mostly ignored. The time of the trustees away from their businesses, or the cost of premise for the pension fund, which is paid for by the sponsor, are some examples.
Still, pension funds say that benchmarking should become standard. The problem is, many fund are reluctant to make the information widely available. “Some funds are very reluctant to advertise what their costs are. We’ve produced a performance index, but it’s not been adopted by anyone else,” explains Peter Scales, chief executive of the London Pension Fund Authority. He says that there is a fear, particularly in local government funds, of being exposed and compared. “The government spends its life comparing local authorities and assessing them, and that makes people wary,” he says.