Join our community of smart investors
OPINION

Skill in doubt

Skill in doubt
November 5, 2012
Skill in doubt

Skill, as any golfer knows, is replicable. Even an incompetent hacker will occasionally hit a good shot but only someone with genuine skill can do so repeatedly. And here's the problem. In normal times, share prices often have momentum and market conditions one year are much like conditions the next. This means it's possible for a lucky fund manager to have persistent success and appear skilful when in fact all that's happened is that his luck in one period has carried over into the next.

And this is where the crisis has been so helpful. In generating more volatility, it has given fund managers a tougher test. A manager who could perform well in the slump of 2008, equity recovery of 2009 and in the more normal times since might well have genuine skill.

So, have they? To test this, I looked at the performance, relative to the sector, of the actively managed UK All Companies funds in Trustnet's database, which have five years of returns. That's 215 funds.

One thing leaps out here - there's little persistence in relative performance. Across the 215 funds, the correlation between relative returns in the year to October 2008 (when shares slumped) and in the year to October 2009 (when shares recovered) is minus 0.54. Most funds that outperformed in 2008 underperformed in 2009, and vice versa.

Granted, there's a positive correlation (0.3) between relative performance in the past 12 months and in the previous 12. But this could be because market conditions have been much the same this year as last, so a strategy that did well one year would have done well in the other. There's a negative correlation (-0.18) between relative performance in the past 12 months and relative performance in 2008.

All this is consistent with fund managers generally having no ability to beat the market. Yes, some were bearish in 2008 and beat the market. But most of these failed to position themselves for the recovery. And most of those that did well in the recovery had underperformed in the slump. Perma-bears sometimes do well, as do perma-bulls. But neither is much evidence of an ability to beat the market consistently.

But there are eight funds that managed to beat the market in each one of the past five 12-month periods: Axa Framlington UK Select Opportunities; Cazenove UK Opportunities; CF Lindsell Train UK Equities; Liontrust Special Situations; Royal London UK Mid-Cap Growth; SVM UK Growth; Threadneedle UK Mid 250; and Unicorn Outstanding British Companies.

Do these have genuine skill? Not necessarily. Imagine fund managers were just monkeys picking stocks at random and that they all had a 50-50 chance of beating the market in any 12-month period. We would then expect 3.1 per cent to do so in five successive periods: 0.5 to the power five is 0.031. And 3.1 per cent of 215 is 6.7. This is very close to the eight that have actually achieved this feat. It might therefore be that even this apparently impressive achievement is just luck.

But just because something might be due to luck doesn't prove that it is. Research by Laura Veldkamp and colleagues at Stern School of Business in New York has found that there's a small proportion of US unit trust managers that do consistently well. And how they achieve this varies across time. In good times, the best funds beat the market by being better stockpickers than others. But in recessions they do so by timing the market - by shifting into low-beta shares just before the market falls.

Granted, most studies show that funds can't time the market. But, says Ms Veldkamp, such studies have looked at all periods, in most of which market timing doesn't work. But, in recessions, it can do so because bad macroeconomic conditions can swamp the merits of otherwise good stocks.

What's more, she finds, the skills of market timing and stockpicking go together. Paying very close attention to companies' fundamentals can alert the best fund managers to potential downturns and thus encourage them to become more defensive just before the market falls.

Does it follow from this that investors should try to spot skilful fund managers?

Not necessarily. For one thing, the pay-offs to skill are not huge. Ms Veldkamp and colleagues estimate that the best quarter of US fund managers earn risk-adjusted returns of less than a percentage point per year more than average - and that a quarter of this is eaten up in higher fees. And of the eight funds in the UK to have beaten the market in each of the past five years, the average outperformance in those years was 7.4 percentage points. For the typical investor, this is less than you'd make from a good stock pick.

The smallness of this outperformance should not be surprising. Funds have to hold lots of stocks and the laws of mathematics mean that this limits how much they can outperform. Across the 215 funds in our sample, the standard deviation of returns relative to the sector has averaged 7.4 percentage points in the past five years. This means we'd expect the 34th-best fund to outperform by 7.4 percentage points.

What's more, the chances of us spotting a skilful fund in advance are not huge. Ms Veldkamp and colleagues find that the best funds tend to be younger, smaller and to hold fewer stocks. While this is a useful guide, it's not perfect. And this gives us another nasty mathematical fact to deal with - that if you multiply a reasonable chance of spotting a good fund by a reasonable chance of it outperforming, you're left with quite a low chance. For example, if you have a 70 per cent chance of spotting a fund which has a 75 per cent chance of beating the market, your chance of buying a market-beating fund is only a little better than 50-50.

Perhaps, then, the question of whether fund managers have genuine skill is less important for investors than generally supposed.