The luck of the manager
Whether good fund managers are genuinely skillful or simply luckier that others is always going to be highly subjective. Ask an outperforming active manager and they’re likely to attribute it to skill; ask a poorly performing one and it’s more likely to be a poor run. The Pensions Institute’s series of lectures ‘Fund manager performance: skill or luck?’ gave, among others, two academics time to disclose the findings of recent research.
And the conclusions? Some of those active managers with an unwavering belief in their abilities are perfectly justified; many more, however, have luck rather than any inherent skill to thank. Allan Timmerman, professor at the University of California, presented new research on star performance among fund managers and the question of pure luck versus genuine skill in US mutual funds.
His findings concluded the performance of the top 1–10% of funds is not a result of sampling variability, otherwise known as luck, neither is the performance of the bottom 1–10% of funds down to bad luck. In other words, those managers on top of the pile have every reason to feign superior skills while poor performers claiming a bad run are fooling themselves (and possibly their clients). However, the research also concludes that the most extreme funds’ performance may in fact be down to luck.
According to Timmerman, performance distribution varies according to the investment objective. Strategies sampled included aggressive growth, growth, growth and income and balanced or income. With aggressive growth he found the worst performance may be down to chance although superior performance could not be attributed to it, genuine skill exists. As is the case with superior performance within the growth funds. With growth and income the performance of the top 1% can be explained as a fluke and for balanced and income it is genuine skill producing outperformance.
David Blake, professor of economics at the UK’s Birkbeck University, spoke on performance clustering and incentives in the UK pension fund industry. By using data from over 300 funds Blake’s presentation quantified a link between fund inflows and past relative performance. Of greater concern is that underperforming US mutual funds appear to take on risk in the second half of an assessment period if they have underperformed in the first half.
Blake also compared US funds with UK funds and found the latter underperformed the market average but to a lesser extent than the US funds. US funds have a far wider dispersion of returns than their UK counterparts. According to the research, large funds tend to underperform smaller ones although Blake stressed there is no systematic link between fund size and total portfolio excess return. Of interest was the link between size and past performance – 15% of the quartile containing the smallest funds were in the quartile of worst performing funds, 32% of the quartile containing the largest funds were also in the quartile of worst performing funds.
Other findings included a narrow dispersion of returns around the median manager. There is also underperformance by the median fund manager relative to the market and an outperformance relative to the peergroup. Blake attributed this to a lack of incentives from the fee structure, a tendency to measure performance relatively rather than absolutely and to a concentration in the industry- in the UK the top five fund managers have 80% of the market, in the US, the same figure is 14%.
Roger Urwin, Watson Wyatt’s head of investment consulting, took the existence of skilled managers as given and spoke about the best means of attaining sustainable alpha. Highly skilled active managers exist but are rare creatures requiring time and resources to track. “There is skill out there, it’s just very difficult to find it though,” he said.
A common error when selecting managers is mistaking luck for skill. We are all subject to in-built biases in the governance process and managers tend to be chosen according to non-financial as well as financial criteria. A strong brand can lead to a somewhat self destructive bias – a good name attracts assets and greater assets under management can dilute performance, so the argument goes.
Urwin criticised the beauty parade as a poor process and one that gives too much sway to what he called the ‘human side’. Past performance is often used for selection despite being a notoriously hopeless gauge of future performance, and numerous short cuts – disposing of a manager if a star leaves the team, for example – are used in manager selection when they are totally inappropriate. Solid quantitative research is more predictive and, the choice of mandate is vital and manager correlation is something often overlooked.
So although the notion of genuine skill exists, there is apparently an element of luck to fund management. For those erring towards the second category, they can heed Urwin’s remark, a quote attributed to the golfer Gary Player who is reputed to have said: “the harder I practice, the luckier I get”.