In June 2003, in the first issue of our Implemented Consulting Survey, we attempted to dispel some of the myths about multi-manager investing and to set out some realistic expectations. It was not about achieving instant stardom in the performance tables, we said, but:

“ …… Rather, it is to deliver performance that is consistently competitive over the short term (i.e. above the median for single manager products), adding up to meaningful outperformance over the longer term, but all the while at lower risk than those managers. In other words, they are trying to eliminate the peaks and troughs of investment performance.”

At the time we commented that it was “Early days, but the signs are encouraging.” That was based on 3 year data to June 2003, which showed that the median multi-manager had not only outperformed the median single manager in the all-important Australian shares and international shares sectors, but had done so at considerably lower risk.

Now, almost five years down the track, the picture is quite different. When we look at the evidence we see that only part of the two-pronged risk and return objective has been achieved. As a group, multi-manager products have certainly produced less volatile returns than their single manager counterparts, but that has actually come at some cost in terms of relative performance.

Again, our conclusions relate to Australian and international shares, the most important building blocks of a diversified portfolio. These are the ‘purest’ tests of the multi-manager argument.

In Table 1, we look at short-, medium- and longer-term statistics for the Australian shares sector. The table shows that over 1, 3, 5 and 7 years, the median multi-manager has delivered returns that would rank no better than 3rd quartile in the single manager universe.

So while the risk (as measured by tracking error) has been roughly half that of the single manager median, the performance has been worse than we – and the multi-managers themselves – would have expected. That said, they have matched or slightly exceeded the benchmark index over all periods, but that is before fees are taken into account. Net of fees, the value added has been marginal at best.

The picture is slightly different for international shares, as shown in Table 2. Here, over each of our chosen time periods, the median multi-manager has beaten the single manager median by a small margin. However, the performance has still fallen short of upper quartile which, over the longer term, is where would expect to see them. (It is worth noting here that, in international shares, both the median multi-manager and the median single manager have added value in excess of their fees.)

The issuers of multi-manager products could argue (and some do) that it is hard for them to add a lot of value during bull markets, and that they really come into their own when markets are flat or in decline. That may be so, but the evidence to date is not compelling.

The 7 year data covers an entire market cycle, including both bull and bear markets, yet the multi-managers’ performance still lags behind the single managers in Australian shares, and is only marginally ahead in international shares.

The data for calendar 2007, which includes the post sub-prime downturn in the December quarter, is equally unconvincing. In Australian shares, the median multi-manager underperformed the single manager median by a full 2% over the year, while in international shares the returns were level pegging.

So it is yet to be seen whether the supposedly favourable climate for multi-managers will help them prove their claims. If we do experience a bear market of any duration, and if the multi-managers fail to add value relative to the single manager alternatives, then we may need to reassess the merits of their approach.

It is certainly true to say that, with a multi-manager product, the range of outcomes is narrower than for the single manager alternatives. So the investor is spared the risk of experiencing the worst return in the market but, by the same token, they are also spared the upside risk of experiencing the best return.

That can be important, especially over a short and volatile period such as the 12 months to December 2007 where the range of performance in Australian shares was 27.6% (best of 33.7% to worst of 6.1%) among single managers, but only 11.8% (best of 24.1% to worst of 12.3%) among multi-managers.

Asset consultants, who ultimately drive most of the multi-manager products available in the market, pride themselves on being able to select and combine the ‘best’ managers. Over time you would expect them to outperform the single manager median, which represents the universe of managers they have to choose from. So far they have not done so. All they have done is combine managers to deliver average performance and eliminate the single manager risk.

At the end of the day, it may be that clients will accept average performance as long as the volatility is relatively low. But with so much focus on the performance side of the risk/return argument, the multi-manager marketers may have their work cut out.