Analyst WILLIAM SOBCZAK explains how trusts that opt for a highly concentrated portfolio can succeed in achieving above-average performance. (Note: This article was first published in December 2018 and has not been updated since).
Since the launch of the first index fund in 1976, passive investing has proven to be a successful investment strategy for both institutional and retail investors. The first of its kind, the Vanguard 500 Index fund, has delivered an annualised rate of return of 10.01%, totalling to a return of over 1,500% since 1989. While good in absolute terms, in relative terms because of fees it has underperformed the index, with the S&P 500 delivering an annualised return of 10.12% over the same period. Although there is only a small difference between the two annually, we calculate that over the 42 years this equates to underperformance of about 53%.*
On the other hand, active management hasn’t, if one looks at the performance of the average fund, covered itself with glory either in terms of outperforming benchmarks. According to the most recent S&P Indices vs Active Management (SPIVA) report, which offers information on the passive vs active debate in the US over the course of 2017, 63.1% of large-cap managers, 44.4% of mid-cap managers, and 47.7% of small-cap managers underperformed the S&P 500, the S&P MidCap 400, and the S&P SmallCap 600 respectively. Over a five-year period, the numbers look even worse for supporters of active management: 84.23% of large-cap managers, 85.06% of mid-cap managers, and 91.17% of small-cap managers lagged their respective benchmarks.
So while outperformance of a benchmark is possible, the numbers above suggest that active managers are mediocre, and that those who can achieve outperformance over the long term are therefore difficult to identify. The interesting question, then, is what marks this small sub-set out from the great majority? What are the small minority of active managers who are outperforming their benchmarks doing differently?
Go big or go home
The overwhelming evidence is that fund managers who are willing to back their convictions with punchy bets are the ones that tend to outperform by the highest margin. There have been numerous studies looking at the different attributes that contribute to manager outperformance, ranging from simplistic momentum strategies (Jegadeesh and Timan, 1993) to the quality of the managers’ education (Chevalier and Ellison, 1999). A growing body of more recent work suggests that the concentration of a manager’s portfolio has a significant effect on the relative performance delivered by the manager.
Using active share and tracking error, Antti Petajisto (2013) uncovered the evidence that the most active managers were the ones most likely to outperform their benchmarks, even after taking account of fees. It was in his previous work with Martijn Cremers in 2006 that the term ‘closet’ tracker/indexing was popularised, and as can be seen below, four types of active management were examined.
We can take away two conclusions from the study. Firstly, that the most active stock pickers were able to add the most value for investors, beating their benchmark on average by about 1.26% a year (after fees) and secondly, that while ‘closet’ indexers essentially matched their benchmark index performance before fees, that inevitably meant that they consistently underperformed after fees.
Despite these findings, closet indexing is a relatively common strategy in both bull and bear markets. More recently Cremers has suggested that fund managers should aim for an active share above 80% if a large-cap manager and above 90% if a small-cap manager. However, he estimated in 2016 that only 30% of US mutual fund assets are held in funds with an active share of at least 80%, and only about 10% of funds with an active share of at least 90%.
Supporting the work of Petajisto (2013) and Cremers (2006), Cohen, Polk, and Silli (2009) examined the performance of stocks that represent the managers ‘best ideas’. They discovered that the most bullish ideas were the ones that consistently delivered the greatest returns. Yet, because of the way that the investment industry works, and in particular how managers are incentivised, they believe that having a highly concentrated portfolio is not the ‘optimal’ course for many managers to pursue. The tendency in practice is for them to introduce stocks into their portfolio about which they have much less conviction.
Cohen, Polk, and Silli summarise four key reasons that managers may over-diversify in this way:
- Regulations can often make it difficult for investment funds to be highly concentrated.
- Following the work of Berk and Green (2004), manager compensation is often tied to the size of the fund. Managers are therefore incentivised to continue broadening their investment portfolio even if they do not have alpha-generating ideas.
- Since the performance of the portfolio is such an important determinant for the manager’s wealth, large investments in a small number of holdings could, in the wrong circumstances, put a manager’s job security at risk.
- Investors tend to judge investment decisions in irrational ways. For example, Morningstar’s well-known rating system makes it difficult for a highly concentrated fund to secure a top rating, however good its return. Morningstar’s methodology heavily penalizes idiosyncratic risk.
Following the research from Petajisto, in 2015 Martijn Cremers took a slightly different approach to understanding the relationship between concentrated portfolios and performance. This time he assessed only high active share portfolios and looked at how important investment holding periods were in determining performance.
Fascinatingly, he found that it was only the investment managers who had a patient perspective, defined as a holding duration of more than two years, who outperformed, to the tune of more than 2% per year. In comparison, funds which traded frequently generally underperformed, whether or not they had a high active share. Cremers found that both closet indexers and low active share funds underperformed on average even if they adopted a patient strategy.
The theory in action
The findings from these studies are particularly relevant for investment trusts, where there is a well-known argument that the capital structure of a closed-end vehicle gives managers the ability to take a longer-term view on available investment opportunities. Being able to build up a revenue reserve also allows the manager of an investment trust to deliver smooth and consistent income, not having to make short-term decisions as a result of fund flows, and to plan for the future from a longer-term perspective.
Finally, and some would say most importantly, since investment trusts have independent boards, the manager is answering to people who should, at least in theory, have investors’ long-term goals in mind rather than the short-term swings in sentiment that open-ended fund managers are open to. It follows that investment trusts whose managers adopt a patient investment strategy and hold highly concentrated portfolios should – if these academic theories are correct – be more likely to deliver superior performance over time.
According to our analysis, theory certainly seems to be filtering down into practice. Within the equity investment trust universe, the average number of holdings has steadily reduced over the past five years, with the average trust having 82 holdings, compared to 91 in 2013. This represents a decrease of over 10%, which we think is significant.
Indeed, this matches with the anecdotal evidence we pick up from meeting managers and boards of an increasing desire for a more concentrated approach. Several managers we have met recently have told us that they expect to run more concentrated portfolios than they have in the past. Examples include JPMORGAN AMERICAN, HENDERSON EUROTRUST, JPMORGAN CLAVERHOUSE, and JUPITER UK GROWTH (formerly Jupiter Primadona).
While overall it seems that investment trust portfolios are becoming more concentrated, one can see significant variations between constituents of the various sectors. For example, the UK equity income and flexible investment sectors have seen the average number of holdings over the past five years decline by at least ten stocks, whereas, in the UK equity and bond income sector and European sectors, the average number of holdings has increased by more than 20 stocks.
As the academic literature illustrates, alongside greater portfolio concentration, lower turnover portfolios also tend to outperform. Over the past four years, we can clearly observe a shift towards more patient investment strategies in investment trusts, with the mean turnover of portfolios reducing by about 20%.
Just as with the portfolio concentration statistics, turnover rates vary greatly among sectors. The sector with the lowest portfolio turnover over the past year was the Japanese sector (including smaller companies) at 16.37%. By contrast the greatest turnover was in the global equity income sector, which on average had a turnover of over 40%.
Applying the theory
Having examined the investment trust universe, we have composed a list of trusts that can be defined as concentrated in terms of either the total number of holdings or the proportion of their assets which is invested in their largest holdings. We have also included trusts which pursue an explicitly ‘concentrated’ approach, even though the total number of holdings may seem anything but.
Managing two trusts with fewer than 30 holdings, Nick Train, co-founder of fund management boutique Lindsell Train, has the most overtly concentrated approach. Over the past ten years FINSBURY GROWTH & INCOME has delivered returns of 297.4% in comparison to the 96.6% returns from the FTSE All-Share. Train believes that the Lindsell Train business model encourages patient investing. A small and compact team ensures that employees aren’t compelled to continually offer new ideas in the hope of gaining greater recognition.
Among trusts that don’t necessarily have the lowest concentration or turnover in absolute terms, some stand out as having a high-conviction approach for different reasons. This is in the sense that the top ten holdings amount to a large proportion of the portfolio. For example, JUPITER EUROPEAN OPPORTUNITIES has just 42 holdings, which is barely half the sector average of 70, and of these the top holdings account for over 70% of net asset value. The manager, Alexander Darwall, has an enviable long-term track. Over ten years this trust has tripled the returns of the benchmark, the FTSE World Europe.
Within the global sector, SCOTTISH MORTGAGE, while having 95 holdings, is heavily concentrated in terms of its top ten holdings representing 53.6% of NAV. Run by James Anderson and Tom Slater, the portfolio revolves around investing in innovative companies that the managers believe can revolutionise established industries. This has led the trust to become the largest UK listed equity investment trust with assets of more than £6bn.
Although at first sight one might not think that it fits the bill, ALLIANCE TRUST offers, despite having more than 190 individual holdings, what Willis Tower Watson calls “concentrated diversification”. The trust utilises the best ideas of eight different managers. Each portfolio usually has only 20 names and a very high active share, effectively making the main portfolio an aggregate of a number of more-or-less independent concentrated portfolios. A similar mandate for institutions has demonstrated significant outperformance since it was launched in 2015. By putting eight such portfolios together, shareholders in Alliance Trust benefit from all the advantages of concentrated portfolios, but with the benefit of some additional diversification.
Another concentrated trust in the AIC global sector is the INDEPENDENT INVESTMENT TRUST (IIT). With a portfolio of 23 stocks relative to the peer group average of 80, IIT has performed well over both the short and long term. The company aims to provide strong returns over extended periods through investing in UK and international securities and, on occasions, futures. This allows the manager an unusually high degree of freedom. Over one, three and five years the trust has generated double the returns of the MSCI World index.
* These and all other calculations in this article were correct as at 9 July 2018.
WILLIAM SOBCZAK joined Kepler Partners in February 2018 as an investment trust analyst. He is a graduate of the University of Western Australia with a BSc in Psychology. Kepler Trust Intelligence offers investors a library of high-quality, up-to-date investment strategy articles and fund analysis all written in-house by experienced analysts, on two separate retail and professional investor sites.