In search of the perfect benchmark
One area of pension fund management on which both managers and trustees must remain fluid in their thinking is the thorny one of benchmarks. As a fund matures, and in some cases heads towards more specialist investment, the original measure of performance may no longer be relevant.
“For many years we measured ourselves against the WM 50, which provided us with a peer group benchmark,” says Roy Peters of Aerion, the new name for the BG Pension Funds Management in London, once part of the former British Gas. “As our schemes have matured we have moved to a customised benchmark. This followed a number of asset reviews over the past few years, which showed that we had a different asset mix to most funds in the WM 50, and persuaded us to move to a new performance measure indicator. In our case the peer group measure was becoming less and less relevant, and so we changed this year.”
The main difference Peters finds is that the new benchmark is rebalanced each quarter. “With the old benchmark we got a certain amount of drift. We would have to wait four to six weeks after the end of each quarter to see how we were performing. The situation is much clearer when we are tracking equity indices. For example, under the peer system it is never clear if that group is overweight in a particular stock or sector.”
In Aerion’s case UK equities are tracked via the FTSE100 and their US counterparts on the S&P500. But using local indices such as these brings its own problems. The last year has seen a number of new companies arriving on the indices, and in some cases moving in and out, particularly in Asia.
“There is no difficulty settling on an index. In our case we discussed it with the trustees and our consultants,” Peters adds. “It means, of course, that we are more aware of how we are performing day-by-day, although that in itself brings problems because there are still difficulties with the construction of the indices. We need more consistency from the providers, as we still see distortion.”
Stephan Chauderna at Nestlé Pension Funds in Vevey, Switzerland, believes that the free float issue, raised at the beginning of the year, must be taken into consideration, but says providers are moving in the right direction. “We have a family of indices which we recommend around our group, and these obviously provide a benchmark for our world-wide investments. They also help to reduce the problem of any distortion evident in a particular market or index. We do measure ourselves against a peer group, but very much as a secondary performance indicator.”
This is particularly relevant in Switzerland, where the Swiss Market Index is affected by the massive imbalance in sectors on the SWX. Some funds use the Pictet LPP2, the classic measure of fund performance in Switzerland, but this is notoriously inefficient. Indeed, such has been the criticism from managers that Pictet introduced a new index in September.
There are other issues, however, that managers have to take into consideration when choosing an index to incorporate into their benchmark calculations, as Jon Bruintjes of Gasunie in the Netherlands explains. “One of the most important aspects of an index from our point of view is liquidity. When we do an equity swap, obviously we want to operate within our benchmark indices. This means that we need liquidity and particularly in the options market. To that end we use the FTSE100 as our European benchmark, S&P in North America and the Hang Seng in Hong Kong.”
In the Netherlands, funds can make peer comparisons through figures produced by the Central Bureau of Statistics. “Managers do use these as a secondary measure of their performance,” says Jan Huits of Buck Consultants in Gouda. “Typically they will also look at the performance of the largest fund in their industry and seek to match that.”
More often comparisons will be made with the domestic WM benchmark, sometimes for what some fund managers see as the wrong reasons. “Our holdings have a different profile to those in the WM benchmark,” says Dr A C M Groeneveld of SPF Beheer in Utrecht. With 30,000 members (25,000 pensioners) and assets of Dfl20.4bn (e9.2bn), this railway employees’ fund is one of the country’s largest schemes. “Nevertheless, as a result of political considerations our performance is compared in this way, whether that is fair or not. The situation is the same for all funds in the Netherlands.”
SPF Beheer has its own internal benchmark, however, which compares investments by type. “We have our own particular approach to benchmarks, and how we choose indices to base them on,” says Groeneveld. “We look at exactly how the index is made up, and try, so far as possible, to find one which reflects our own holdings proportionally sector by sector. It is impossible to find one exactly the same, but we settle on the one that is closest.”
Two years ago, the fund managers and trustees chose the MSCI Europe for its European equity investments, and the Merrill Lynch EMU Direct Government Index for its bond holdings in Europe. “To date, we have been very happy with these particular indices and feel that they truly reflect our holdings and help us to track our performance at regular intervals.”
Mette Kirketerp, head of asset allocation and risk management at Copenhagen-based PKA, the largest of the labour market funds in Denmark, agrees that benchmarks should reflect holdings but feels many other issues have to be considered. She highlights some of these when asked about the value of external or peer benchmarks. “I can only speak for Denmark, but here a useful external benchmark does not really exist. Although results are published each year, and we can see how each fund has performed on the surface, these statistics do not tell the whole story, as important figures are missing. There is no risk profile, nor any indication of liability and exposure which would help the observer to decide on how the yield should be assessed.”
When looking at the choice of indices for internal benchmarks, she reiterates what is needed to reflect the broader view of individual mandates. “For fixed income mandates in Denmark it is fairly straightforward, since all our holdings are domestic, for equities, however, it is a different story. We choose indices by zone, to reflect our holdings in different markets. Once again domestic holdings are perhaps higher than one might expect, but that is because reforms have been some time coming. Overseas we try to look at the broader picture. For example, in the UK we feel that the FTSE 100 is too narrow for our liking, and does not really reflect our mandate or holdings in this area.” As a global investor Kirketerp says PKA would like to have a global index to measure performance, but is not happy with any on offer. “If we could pull together the indices we use into one statistic that would be ideal, but we are still trying to work out how that can be done.”
When deciding how indices are chosen she returns to the idea of the “broader picture”. “When we sit down to look at a region we look at a variety of indicators, some reflecting risk, some value, others growth. Each of these may give a very different view of how the investments are performing. Which we choose in the end depends upon the mandate, the region, and how we assess the risk. So far as changing the benchmark is concerned, this does not happen very often, mainly when we amend the mandate or appoint a new manager and want to re-assess our level of exposure and risk.”
While the ideal benchmark may be a figment of some statisticians’ imaginations, there is little doubt that as Europe’s pension funds mature and widen their portfolios there is a move away from peer comparison towards specialist benchmarks.