Following the best ideas of the best professional investors holds a strong intuitive appeal. Not only do top fund management pros have track records that suggest they are smart stockpickers, but they can also draw on analytical muscle that vastly outguns anything available to the average private investor: from the brightest researchers; to the most sought after data; to cutting-edge algorithms. Indeed, if a stock has found its way into a leading fund manager’s top holdings, surely it’s fair to assume that its tyres have received a thorough kicking and the investment case is strong?
This is all well and good in theory, but it is also very easy to think of one recent, high-profile case for the opposition: Neil Woodford. The investment debacle involving the former star fund manager should certainly give pause for thought. Mr Woodford was the closest the UK had to an investment superstar when he set up his own firm. Sadly, after splitting from his old employer, Invesco, he swiftly assembled portfolios full-to-bursting with dud stock picks. His failure has been seized upon as yet more validation for advocates of the efficient market hypothesis, which at its most hardline insists such a thing as stockpicking skill cannot exist.
So where does the truth lie? Is there any reason to think intuition is right and that tracking fund managers' best ideas is worthwhile?
Trust your intuition
Simple intuition can often mislead investors. But while focusing on stocks that are popular with smart investors is not a silver bullet, there is actually evidence that following the top investment picks of fund managers can produce impressive results. Indeed, focusing on the ideas of only the 'best' fund managers may not do full justice to the phenomenon. Research suggests that even managers of poorly performing funds often demonstrate signs of stockpicking skill when it comes to their very largest holdings.
Global generalist investment trust Alliance Trust (ATST) has recently adopted a best ideas approach, taking the 20 top ideas from nine fund managers selected by investment consultancy Willis Towers Watson. “We know that about half of managers outperform and half of managers underperform before fees and costs,” says Craig Baker, head of the Alliance Trust investment committee and Willis Towers Watson's global chief investment officer. “But interestingly, a lot of managers perform quite well in their largest 10 or 20 positions, and then destroy value in the rest of the portfolio.”
One striking bit of research on this subject came out in 2005. The paper, titled 'Best Ideas', was published by MIT academic Randy Cohen, Christopher Polk of the London School of Economics, and Bernhard Silli of Goldman Sachs. The authors investigated the performance of US fund managers' best ideas, which were taken to be their most overweight positions.
The research was based on disclosed holdings over the 15 years to the end of 2005. The researchers’ most straightforward best ideas portfolio managed an average return of 1.26 per cent per month ahead of the risk-free return and produced 0.29 per cent of “four-factor alpha” – a measure of outperformance that attempts to take account of other known influences on returns such as momentum and value.
The stocks identified as “best ideas” did not overlap significantly between managers, but did tend to have a bias towards smaller companies and stocks showing good momentum; although momentum alone did not explain the outperformance the academics measured.
An understandable contradiction
A key quandary seems to be that if the average fund manager demonstrates skill in picking their top holdings, why is there so much evidence that the average actively managed fund cannot beat the market after fees?
On the face of it, this seems a huge contradiction. However, as is often the case with investing, human behaviour and the perverse incentives of the investment industry offer a credible explanation as to why the average manager finds ways to undermine his or her own abilities. The 'Best Ideas' paper offers a succinct explanation: “Our findings suggest that while the typical manager has a small number of good investment ideas… the remaining ideas in the typical managed portfolio add no alpha at all. Managers have understandable incentives to include these zero-alpha positions. Without them, the portfolio would contain only a few names, leading to increased volatility, price impact, illiquidity and regulatory/litigation risk.”
Mr Baker sees a similar cause for the seeming contradiction: “The reason this happens is because [a fund manager has] probably got 10 or 20 great ideas at any point in time. But as a fund manager, you don't run a 10 or 20-stock portfolio because the volatility would be so high you'll get fired at exactly the wrong time, when your style is out of favour. So you try to reduce risk relative to the benchmark for sensible business management reasons. By definition, the stocks you put in your portfolio to reduce risk are the largest stocks in the index, which, by definition, are the ones that have gone up the most. And so it's kind of an anti-value bet on average over the long term. So you get this phenomenon that lots of managers add value in their 10 to 20 best ideas and then destroy value in the rest.”
Career risk may be the key culprit pushing managers to adopt damaging overdiversification. But one other potential reason for big holdings performing better is that while fund managers have been shown to demonstrate skill when buying shares, they have been found to destroy value when selling.
A behavioural advantage
In a paper titled 'Selling Fast and Buying Slow: Heuristics and Trading Performance of Institutional Investors', academics Klakow Akepanidtaworn, Rick Di Mascio, Alex Imas and Lawrence Schmidt studied the daily holdings and trades of 783 portfolios, with an average value of about $573m. Between 2000 and 2016, they found the average annual performance of these funds could have been improved by 0.7 per cent had sell decisions simply been made at random. What an inditement!
When it comes to top holdings, once a stock has started to underperform it may quickly lose its status in the top flight, effectively meaning it is sold as far as a 'best ideas' portfolio is concerned. However, in reality, a manager may well stick with a position in the stock while it continues to perform poorly, thus providing a longer-term drag on the portfolio as a whole. Indeed, to find long-term dogs in any portfolio, the best place to look is likely to be among the smaller positions that at one time may have been much more significant holdings.
The revamped Alliance Trust has been structured specifically to try to benefit from this type of best-ideas-only sell discipline. Mr Baker says: “If you've got a 20-stock limit and you find a new idea, you've got to sell something. So it just puts more of a sell discipline on you. If you've got a diversified portfolio, you can always put in another stock. So [stock limits give] a behavioural advantage.”
Frances Gerhold oversees the JP Morgan Global Growth & Income fund (JGGI), an investment trust that backs in the best ideas of the firm’s 80 equity analysts. She also sees a value in the focus provided by a best ideas approach. “The challenge you have is that you have lots of input, but how do you actually whittle that down and kind of focus on what are the best ideas?” In the case of the JP Morgan Global Growth & Income trust, a stock ranking framework is used to identify best ideas rather than the cash-invested-based ranking provided by individual fund managers.
These insights about the value of concentrated and focused bets chime with research that shows a common characteristic of funds that do outperform is that their holdings tend to vary significantly from the benchmark – so-called 'active share' – and show high concentration in their top holdings. Ironically, as these ideas have gained popularity, evidence has also started to emerge that forced adoption of this kind of strategy can increase the problems of badly performing funds.
The best of the rest
While it is all well and good to point to the outperformance of fund managers’ best ideas per se, there is something distinctly unappealing about looking for investment ideas among the top holdings of a fund manager with a rubbish track record. Any unease about following the ideas of struggling fund managers is arguably justified.
Another interesting bit of research into fund manager stock picks pitted the skills of top-performing managers against those of underperforming managers. Russ Wermers of the University of Maryland, Tong Yao of the University of Iowa and Jane Zhao of PanAgora Asset Management devised a way of rating shares based on the performance records of the fund managers holding them. The team constructed portfolios where each stock got a positive weighting based on how popular it was with top-performing fund managers offset by a negative weighting for its popularity with poorly performing fund managers. So stocks that were popular with top fund managers and overlooked by poorly performing managers would get a higher weighting in the portfolios and vice versa.
The researchers described their experiment as “a magnifying glass on the collective stockpicking wisdom of fund managers”. Based on 27 years of data to the end of 2006, the researchers found the top 10 per cent of stocks outperformed the bottom by over 4 per cent a year. The stocks that ranked highly among top-performing fund managers were also found to be more likely to deliver positive earnings surprises in the future.
This would seem to suggest that, even if the average fund manager can show skill in selecting their top holdings, the top holdings of the best performing fund managers are likely to be of more interest.
Theory into practice
As the Woodford debacle reminds us, following fund manager best ideas can prove calamitous. Nevertheless, there is enough evidence to suggest the top holdings of top managers can provide a valuable resource for private investors looking for potential stock ideas. We also think this is a good source of ideas for our team of specialist companies writers to scrutinise. As part of our new weekly Ideas Farm pages in the magazine, starting next week, we’ll be highlighting top fund manager best ideas from a variety of sectors and markets.
Our Best Ideas lists are based on portfolio data from Morningstar. Fund disclosure dates vary, which means the data on some holdings can be over two months old, while in other cases it is only a few weeks old. The good news is that research suggests the performance of portfolios based on lagged information is normally not too different from performance based on more timely information. Indeed, in the case of the Wermers, Yao, Zhao research already cited, the performance was “only slightly weakened” when lagged to reflect public disclosure of holdings. More to the point, these ideas are there to provide a strong kick-off point for research rather than an end in themselves.
For larger companies, global and UK, our best ideas lists represent funds’ top three most overweight positions relative to the index. For specialist areas, such as sustainable best ideas and smaller companies, the lists are based on the top five holdings. The funds being tracked are those that have shown strong performance over the past year and are held in good regard by the industry – including our own Top 100 Funds list. We currently have seven lists we’ll be rotating through each week on the new Ideas Farm pages, which are: Global, UK Large Cap, UK Small Cap, Sustainable UK, Sustainable Global, Tech and Healthcare.
In this introduction to the new ideas pages, we've kicked off with a list of Global best ideas, but also include the other best ideas lists below. These lists may be relatively slow to change given many top managers have reassuringly long holding periods, but we’ll aim to highlight new stocks as they appear on the list as well as those that drop out. We hope readers find this useful.