For the majority of pension funds, taking on a multi-manager is an all-encompassing decision that can risk making the trustees look like they are just following the latest fad without boosting returns.
As a result, trustees or those in charge of the pension scheme are being more cautious than multi-managers expected five years ago. Stephen Delo, chairman of the Association for Institutional Multi-Manager Investing and managing director of one of its members, Escher, says in 2000 there had been wide predictions that £20bn (€29bn) of European pension fund assets would switch to multi-managers within three years. But five years later Delo says that, even in the absence of comprehensive data, he would be surprised if the figure was as high as current estimates of £10bn. Russell/Mellon, for example, which provides information on managers and performance, says it did not split out multi-managers’ performances or mandate wins.
Delo says the reason for the relatively slow take-up in Europe, especially when compared with the growth of the multi-manager concept in the US, South Africa and Australia, was partially the switch in interest to hedge funds but mainly the fact that most multi-managers “have delivered lacklustre returns with some being accused of being closet index trackers”.

An improvement in returns is one of the main reasons for most pension funds in any asset management switch. Emmanuel Koller, pension fund manager of the CHF400m (€257m) British American Tobacco Switzerland scheme, says the target for its multi-managers was to outperform the benchmark by 2% per year, after costs.
Northern Trust manages BAT Switzerland’s entire portfolio in five asset classes – Swiss equities, global equities, Swiss bonds, global bonds in Swiss Francs and global bonds in other currencies. BAT uses nine managers for these five asset classes, Koller says. But since the switch from specialist managers to Northern Trust was made in July 2001 the portfolio has underperformed by 0.2%, as at end-2004, Koller says. But within that three-and-a-half-year period, Koller says the performance had shown marked differences. “I believe good manager selection will beat the benchmark. We were happy in 2001 and 2002 as the markets declined and Northern Trust showed relative good performance but when the markets have been up it has been too difficult to get close to the benchmark.”
A large number of multi-manager mandates, however, are for specialist briefs rather than for the whole fund. The Finanssektorens Pensionskasse (FSP) industry-wide scheme in Denmark, which manages about €2bn, gave Russell Investment Group a near-€80m brief to manage a hedge funds portfolio last year.
And the near-€9bn Pensionskassernes Administration (PKA), which represents eight sector funds in Denmark, replaced fund manager Denver Investment Advisors with multi-manager giant Russell Investment Group last year to manage a €110m US small-cap equities brief. Although Peter Melchior, director at PKA, declined to comment on the reasons behind the switch he did not rule out further mandates being passed to multi-managers.
For the UK-based £130m Monarch Airlines Retirement Plan the decision to move from Gartmore to Northern Trust as manager of its £60m in UK, continental European, Pacific basin and Japanese equities was also performance led. “The decision was made to seek improved returns,” says pensions manager Annette Eversden. “Gartmore had been underperforming and currently Northern Trust is in line with the indices, although some have outperformed and some have underperformed.”
She says that after taking consultant Mercers’ advice on the criteria for judgement, the trustees had seen presentations from three or four multi-managers. The main criteria for choosing Northern Trust were on past performance over the last three years and the level of expertise offered. One of the unsuccessful pitches had come from a very new multi-manager, which the board felt did not have the required track record, she adds. But as the mandate is still only about a year old, she says the trustees would be keeping a very close eye on Northern Trust, seeing the team regularly at their quarterly meetings.

Some in the industry see as slightly ironic that a key criteria for choosing multi-managers is on their past performance. Earnshaw says past performance made up only about 15% of its judgement on whether to pick an underlying manager. He says softer value judgements, for example on the manager’s levels of conviction about the fund and the organisation’s strengths and size, were more important.
For both Monarch and BAT the move to multi-management was driven by the need to improve performance by gaining access to the best managers. Delo says the growth in demand for multi-management was primarily driven by smaller pension funds trying to get access to specialist managers rather than using generalists that offer top performers under one fund management roof. For most schemes with less than £30m in assets the choice was often a multi-manager or index trackers, he added.
But even when a pension scheme is of a sufficient scale to gain
access to boutique fund managers that can show the best returns, Delo says the difficulty was in the retirement plans choosing the managers. “Trustees are not good at selecting managers, even if they have investment consultants on board. They can be haphazard and only change if managers are
underperforming, which can be
the worst time to move. Multi-managers bring sophistication to boards,” he says.
However, Delo says that choosing a multi-manager was a strategic issue rather than just a tactical manager-selection process and required the trustees sitting with their actuarial consultants to model different scenarios for the scheme’s assets and liabilities.
Once an asset allocation strategy for equities or bonds and other classes had been decided the trustees needed to decide whether to use an active or passive manager. Only once an active strategy had been chosen could the trustees look at whether they had the skills in-house or needed external fund managers picked with the aid of consultants or multi-managers.

For trustees, an increasingly important reason to outsource manager selection was being driven by governmental and regulatory concerns at their relative lack of experience and the increasing risk that trustees could be held accountable by scheme sponsors for failing to put in place best practice – as shown by the Paul Myners review of the UK institutional fund management market.
But as with Monarch and BAT, trustees relied on the consultants to advise whether multi-manager was the best potential solution and to organise a beauty parade and suggest the criteria for deciding. Trustees, Delo says, do not often look at the detailed reasons for choosing the multi-managers. The consultants are widely seen as very knowledgeable, and those, such as Mercers, recommend that trustees examine the multi-manager’s research process for average number of underlying fund managers in a portfolio to identify the risk/reward profile, the level of qualitative and quantitative judgements, the number of managers and research resources and their experience.
For trustees choosing between multi-managers the average number of underlying fund managers used is often a clear sign of differentiation. Delo says that typically a multi-manager portfolio used between one and nine fund managers, with a clear trade-off in risk and return, ie, the larger number showing greater diversification and less chance of underperforming significantly but, equally, greater similarity to the index.
The impact of costs was not seen by pension funds as a key criteria in choosing a multi-manager. BAT says the total expense ratio of its multi-manager was about 1% but any savings went on the scheme’s underperformance. And pension funds say they were unsure how the fees were split between the underlying fund managers and multi-managers in order to put pressure on its overall costs.

Pension funds were also relatively unconcerned about potential conflicts of interest between consultants and multi-managers and how it would execute a change in the multi-manager, either to a rival or directly to a specialist or passive fund manager. Most managers, like Escher, do not have a lock-in, although some, like Northern Trust, charge an anti-dilution levy of up to 10% if there would be costs to other pension schemes in its multi-manager portfolios.
There are also market risks to be taken into account with moving from one manager to another – ie, the cost of being out of the market or the transition cost, which can be about 0.15% with some managers, such as SEI, for smaller, straightforward portfolios.
But when picking a multi-manager such criteria can be important factors in helping trustees decide on a switch.