Years ago, when Burton Malkiel was first making the case against active fund managers, the answer seemed clear. Stock markets, it was thought, were efficient: all information is immediately embedded into share prices and so fund managers can't know more than the market.
Today, though, this answer is inadequate. Markets don't seem entirely efficient. In fact, Heiko Jacobs and Sebastian Muller have listed no less than 138 deviations from efficiency, such as momentum and defensive effects and the ability of various accounting measures to predict returns, such as accruals or Joseph Piotroski's F-scores.
Why, then, don't fund managers exploit these? It’s not because funds are themselves the market. Unit trusts in the UK all companies and equity income sectors last year managed just £220bn of funds, which is only one-eighth of the total capitalisation of the All-Share index. It’s perfectly possible in theory for unit trusts to out-perform in aggregate by profiting at the expense of (say) overseas or retail investors.
But they don’t. Why?
One possible reason is that some of those 138 apparent deviations from market efficiency are, as Duke University's Campbell Harvey says, "likely false": they exist only in one dataset and haven't been widely replicated.
But there's more to it than this. At least two ways of beating the market are robust: defensive and momentum stocks have outperformed for long periods in different markets. There are three reasons why unit trust managers haven't been able to exploit these.
One is that they don't short sell stocks, which means they can only profit from half of mis-pricings: under-pricing rather than over-pricing. They are like batsmen who are only allowed to score runs on the off side.
But long-only investing in defensives and momentum has beaten the market, so this isn't the whole story.
One problem is that these strategies carry benchmark risk. A fund manager who buys defensive stocks runs the risk of underperforming if the market rises sharply. Momentum stocks also carry this sort of risk. Victoria Dobrynskaya has shown that they have high downside beta: they can under-perform a falling market.
To a fund manager, this risk matters. He is judged by relative performance, so underperforming risks getting him sacked.
A second problem is liquidity. Randolph Cohen and colleagues have shown that even average fund managers have stock picks that do beat the market; this is consistent with there being several deviations from full market efficiency. But something stops these good ideas leading to good fund performance.
To see what this something is, imagine you're running a £500m fund and want to hold 20 stocks. Your average holding will then be £25m, which is equivalent to more than 8 per cent of the typical UK small-cap stock. This is a massive position which you cannot acquire without moving the market against you, and which you cannot sell easily either. You therefore have a problem. Either you hold bigger stocks, which limits your investment opportunities - especially as larger companies are better researched and so less likely to be mispriced. Or you hold more than 20 stocks, which dilutes away the good returns you make on your handful of good ideas. Either way, you lose performance.
Take Fundsmith’s equity fund. Its manager, Terry Smith, boasts of holding "a small number of high quality, resilient, global growth companies that are good value and which we intend to hold for a long time". He holds only 29 stocks. But he manages £8.7bn, which means his average holding is £300m. Because of this he can only hold large stocks: the average market capitalisation of his shares is £66.1bn. This is feasible for someone investing in global equities and who intends to invest for the long-run so need not worry about liquidity risk. But nobody could adopt such a strategy with such a big fund if they were confined to UK stocks or if they were worried about being unable to sell quickly.
Now, here's the thing. When we believed in the efficient market theory we thought it was theoretically impossible for any fund to beat the market consistently. But given the number of deviations from efficiency, we now know it is at least theoretically possible. If a fund is small enough relative to its investment universe to have concentrated holdings and if it has a fund manager with a good enough reputation not to have to worry about benchmark risk, it might beat the market.
The problem is, however, that the risk of the fund not doing so is borne more heavily by the investor than by the manager himself; if the fund falls, investors lose but the manager still gets his fee. As the FCA says: "the investor bears virtually all the risk". For this reason, investors must be very sceptical of any fund's claim to be able to beat the market.
