FTAdviser.com

Advertising

It is one of the great dilemmas of investment. How do you choose a fund manager? In statistical terms, all you have are the fees (and other expenses) and the past performance. We know that, other things being equal, lower fees give the investor a head start. But when you look back at the data, it is tempting to believe that superior performance greatly outweighs the fee advantage.

But the main trouble with looking at past performance is that, it is not simply unreliable, it can be counter-productive. A study by the academics Andrea Frazzini and Owen Lamont of the University of Chicago found that high flows into US mutual funds predict low returns in future. Specifically, mutual funds with high flows tend to underperform the broad average by three percentage points a year.

A number of factors may be at work here. One is fashion. Managers may build up a good reputation through buying growth stocks (or, in recent years, value). Just as investors notice their great record and buy into the funds, the fashion changes. Another possibility is that managers may have a small-cap bias that allows them to outperform when their fund size is small; as the fund rises in value, they are forced to buy large-cap stocks, erasing their advantage. A third factor may be fees; successful managers may be tempted to increase their annual charge, reducing the net return to investors.

The effect on the average retail investor can be devastating. Not only do they tend to buy shares at the top of the market (such as 1999-00), they back managers at the end of their hot streaks. Morningstar has analysed the data from the US mutual fund market. Over the past 10 years, the average technology fund has delivered an annual return of 3.25 per cent. But when the flows are broken down, the average investor has lost 5.04 per cent a year. In Latin American funds, the return of the average investor has been more than six percentage points behind that of the average fund.

What is to be done about this? The temptation is to be contrarian, but that runs up against the fundamental problem that such a strategy requires investors to back the worst-performing fund managers of the past. Sometimes that may simply be rewarding incompetence. Even the argument that one would be picking managers whose style was about to come back into fashion is far from watertight; since poor performers will often be fired before their luck turns.

Another answer might be to select funds that have done well but have not benefited from rapid asset flows. By definition, of course, such a strategy will only work for a minority of investors. And a further option would be to go for sectors that have performed poorly in the past and may thus benefit from a reversion to the mean. That does seem sensible in theory but can run into problems in practice; Japan has been disappointing bargain-hunters for a long time.

The problem lies in the seductive nature of past data. Of course, looking back, we can find funds that have outperformed substantially, just as we now know that a cumulative bet on Cardiff to reach the FA cup, John McCain to win the Republican nomination and the Coen Brothers to win the best directing Oscar would have paid off handsomely. But we didn’t know in advance those things would happen. Worse still, we tend to remember our vague intuitions that did prove correct and forget the many other occasions when we were flat wrong.

In the absence of a crystal ball, it is silly to ignore the investment advantage that low fees give us. And that means, on most occasions, steering clients towards index trackers and exchange-traded funds.