Concentrated equity portfolios

Transcription

1 equity portfolios Do they work in the Australian market? Having established a belief in active management in Australian equities, we are often asked by Australian clients whether a concentrated approach to portfolio management is appropriate in the local market. This paper provides an overview of our views on this issue. First, we look at the rationale for using concentrated equity portfolios before exploring the historic experience of their use at the manager and client level. Next, we discuss the attributes we believe investors require to use these types of approaches, such as focusing on manager skill and appropriate governance and beliefs. Lastly, we reflect on how the Australian market is likely to evolve and focus on features of the market that may make it more difficult for some investors to access concentrated strategies, but can create an opportunity for others. Definition The term concentrated often suggests a low number of stocks in the portfolio, and this is typically true in broad markets like global equity but does not necessarily hold for the Australian market. The top 10 stocks (by market cap) in the S&P/ASX 200 comprise around 55% of the total index. It is possible for a portfolio with stocks to look quite similar to the benchmark in terms of risk and return, if it holds the larger index stocks at close to benchmark weight. The flip side to this is that managers with a relatively large number of stock positions can still have a high active share and tracking error because they simply do not hold many large cap stocks, or are significantly underweight in this part of the index. For the purposes of this paper we have defined a concentrated equity strategy as a strategy with both high tracking error and high active share. Tracking error is a statistical measure which provides an indication of the distribution of expected returns about a benchmark index. Active share (or active money) is a term used to describe the extent to which a portfolio is invested differently to the benchmark. We used an expected tracking error of 6% and active money score of 6 as our indicators of concentrated portfolios. From a top-down, portfolio level perspective, concentrated equity portfolios make sense. They target higher alpha, and the greater risk they take on to achieve this higher alpha is diversified by other managers and asset classes at the aggregate portfolio level. (This concept was discussed in a paper published by Towers Watson in June 2012, equity products: Why we generally prefer them to diversified products.) In the global context, we also tend to have a positive view of concentrated equity managers from a bottom-up perspective. Our manager research philosophy favours wellaligned investment businesses that can attract, retain and motivate skilled investors, because we believe that genuine skill is rare. Businesses that foster strong alignment between investment professionals and their clients are generally smaller and have more focused resources. By virtue of their size, they tend to target their efforts on only finding and investing in their highest conviction stock ideas. We prefer close-knit decision-making units, and this also favours smaller teams and businesses. Conversely, a diversified global equity manager that has the scale of resources to cover such a broad universe of stocks is often poorly aligned. With this in mind, it is hardly surprising that in a global context, we favour concentrated equity strategies. However, given that the Australian market is narrower and concentrated in a small number of stocks and sectors itself, we are often asked whether our preference for concentrated equity portfolios applies to the Australian market. Our starting point for building equity portfolios is to appoint managers where we have strong conviction in their investment skill. In the Australian equity context, our highest conviction managers (that is, those with a FREX skill rating of 1) are a fairly even mix of diversified (or core ) and concentrated managers. Why does this differ from the global context where the bias is more to concentrated managers? There are a number of reasons. In favour of diversified portfolios, the main Australian equity benchmarks contain stocks, so it is feasible that a relatively small investment team can cover most of the market to run a diversified portfolio successfully. Additionally, most listed companies are headquartered on the east coast of Australia, so there is rarely a need for more than one investment office. This contrasts with the global equity market, where there are over 2,000 stocks in the main indices, in widely spread geographical locations. The argument in favour of concentrated portfolios in Australia is however similar to the one for global accessing a manager s best ideas in a more concentrated form is likely to be more efficient than doing so in a diluted form through a more diversified portfolio. Further, combining these managers in a sensible manner should lead to an appealing outcome at the portfolio level.

2 A logical starting point is to ask whether concentrated equity products have actually outperformed in Australia. We investigated the historical experience to see whether they have performed better than their more diversified peers in the past, and the results in Figure 01 show that they have 1. We note that the performance is shown gross of fees. This is in itself unsurprising, since concentrated strategies contain fewer constraints, ensuring best ideas are expressed in the end portfolio. Similar results were found in a paper by Yale professors, Cremers and Petajisto (2009) 2 in the US equity manager universe. Figure 01. Individual strategies cumulative excess return growth of $1,000 (gross of fees) $1,350 $1,300 $1,250 $1,200 $1,150 $1,100 $1,050 $1,000 Jul-05 Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 However, if we look at the performance profile more closely, there are periods where diversified products as a group outperform the concentrated products (see Figure 02). It appears to have occurred in the period following the financial crisis of Why might this be the case? Do certain market conditions favour diversified investing over concentrated, and why is this so? To answer this, we look to the alpha generation of Australian managers. Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Apr-13 Figure 02. Individual strategies rolling one year (concentrated minus diversified) median excess returns per annum (gross of fees) 6% 4% 2% -2% -4% -6% Jun-2006 Dec-2006 Jun-2007 Dec-2007 Jun-2008 Dec-2008 outperforms Jun-2009 Dec-2009 outperforms Jun-2010 Dec-2010 minus [LHS] Jun-2011 Dec-2011 Jun-2012 Dec-2012 ASX 300 [RHS] Are concentrated managers picking winners or avoiding losers? We find that, generally, the risk profile of a concentrated Australian equity product is more reflective of the stocks that are not held rather than those that are held. Research by Inalytics expands on this further, finding that underweight positions (which include stocks that are held and not held by the manager) had contributed some 4 of the alpha generated by Australian equity managers. This appears to be something quite unique to the Australian market, and compares with global equity managers whose underweights actually detract value (-1) over the same period. This suggests that concentrated managers in Australia should be considered differently relative to their global peers, because one of the keys to successful concentrated investing in Australia is to avoid the losers. Jun We examined the risk and return characteristics of concentrated and core Australian equity strategies, looking at a broad sample of managers and products across the Australian equities universe (179 products in total) both past and present was analysed. We attempted to control for survivorship bias but acknowledge it will still be present and will therefore skew the results to be higher than they might be otherwise. The period from July 2005 to June 2013 was deemed most suitable for our examination as it provided a sufficient number of strategies within both the core and concentrated strategy universe. However, this relatively short period also included the significantly volatile period of the global financial crisis (GFC) and so may not be entirely representative of normal investment market conditions. Including the GFC was useful to demonstrate the performance of the different strategies during such a period. We categorised the manager universe into core and concentrated. As discussed above, deciding which manager fits in which category is more complicated than the number of stocks in the portfolio. We used our manager research team, who have a strong qualitative understanding of the managers to decide the most appropriate category. The S&P/ASX 300 was used as the benchmark for the analysis. As a direct measure of performance, we have performed this analysis on a gross of fees basis. We note that it is a reasonable assumption that the fees for concentrated portfolios will be higher than core portfolios which clients should also take into consideration. Managers and products that were selected in the sample were not limited to high conviction managers favoured by Towers Watson research. Managers and products that were closed at the time of our paper were included in our historical sample to reduce survivorship bias. 2. Martijn Cremers and Antti Petajisto. How Active Is Your Fund Manager? A New Measure That Predicts Performance (7 August 2006). 2 towerswatson.com

3 While absolute alpha is the name of the game, institutional investors will frequently assess the performance of the underlying asset classes, and even at the individual product level within the sectors. These investors will often assess each manager s efficiency by looking at risk-adjusted returns. portfolios will generally run higher tracking errors. Figure 03. Individual strategies rolling three year median tracking error (gross of fees) 8% 7% 6% 5% Figure 03 shows that tracking error has been significantly higher for concentrated portfolios in comparison to diversified portfolios over the majority of rolling time periods. This is to be expected given our definition of concentrated portfolios was to include managers with high expected tracking error. 4% 3% 2% 1% Figures 04.1 and 04.2 show the difference in excess return distribution between diversified and concentrated portfolios. Consistent with Figure 02, it shows that a larger dispersion exists within the concentrated portfolios when compared to core portfolios. This result is indicative of our belief that concentrated portfolios are more skill based with a higher risk and return profile which allows for a wider spectrum of results. We attempted to control for survivorship bias but acknowledge it will still be present and will therefore skew the results towards higher outperformance than would otherwise be the case. Jun-08 Figure Three year excess returns distribution diversified 1 8% 6% 4% 2% -2% -4% -6% -8% Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12 Jun-13 95th percentile 75th percentile Median 25th percentile 5th percentile As a result of this higher tracking error, the riskadjusted returns (as measured by information ratio) for diversified products are, on average, slightly higher than those achieved by concentrated products as a group, according to our analysis. We note that the difference is quite small, as shown in Figure 05. Figure Three year excess returns distribution concentrated 1 8% 6% 4% It is worth noting that this analysis is gross of fees. So a logical question that might follow is, if diversified products have a higher information ratio, and generally cost less in terms of fees, why not build an equity sector allocation from a number of diversified products? We believe that the key downside of this approach is that the expected alpha is significantly lower, and at the overall portfolio level the objective of pursuing active management should be to maximise alpha. We believe that this approach leads to inefficiencies, through over-diversification and portfolio redundancy. In other words, investors pay fees for active management, only to get close to index performance outcomes. Figure 05. Individual strategies rolling three year median information ratio (gross of fees) Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12 Jun-13 Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12 Jun-13 95th percentile 75th percentile 2% -2% -4% -6% -8% Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12 Jun-13 Median 25th percentile 5th percentile Australian equity portfolios 3

4 Do diversified portfolios of concentrated managers lead to stronger outcomes at the portfolio level? Given the track record of concentrated products in the Australian market, does the assertion hold that they lead to stronger outcomes at the overall portfolio level? We attempted to address this by generating a series of randomly selected portfolios (albeit still within the Australian equity asset class); one group was selected from diversified products, the other from concentrated products. We chose from as large a universe of products as possible, so we did not filter by whether we have a positive view of their skill. As shown in Figure 06, we found that the excess returns of the portfolios of concentrated products was higher than the portfolios of diversified products. The median tracking error of the multi-manager portfolios of concentrated products was lower than that of the individual concentrated products, which strongly suggests that the additional risk at the product level is being diversified away at the portfolio level, even within the single asset class. We find that the median risk-adjusted return for the portfolios of concentrated products is fairly similar, if slightly lower, than the portfolios of diversified products, as shown in Figure 07. Figure 06. Multi-manager portfolios rolling three year median excess return (gross) % % % Jun-2008 Dec-2008 Jun-2009 Dec-2009 Jun-2010 Dec-2010 Jun-2011 Dec-2011 Jun-2012 Dec-2012 Jun-2013 Figure 07. Multi-manager portfolios rolling three year median information ratio (gross) Jun-2008 Dec-2008 Jun-2009 Dec-2009 Jun-2010 Dec-2010 Jun-2011 Dec-2011 Our analysis of the Australian market is therefore consistent with Towers Watson s global findings; at the portfolio level, the concentrated products yield a similar information ratio to a portfolio of diversified managers, but higher absolute returns. Those who forget the past are condemned to repeat it To provide a practical perspective on how investors actually use concentrated Australian equity products we spoke to our clients and other market participants about their experience with these portfolios. The past experience with concentrated Australian equity portfolios in Australia has been mixed. Between 2001 and 2005, the medium-term performance of these concentrated strategies was generally quite strong. This, and relatively low cross-sectional volatility in the market, attracted many institutional investors to concentrated strategies. Figure 08. Individual strategies rolling three year per annum median excess return (gross of fees) % % % % % Jun-2012 Dec-2012 Jun Jun-2008 Dec-2008 Jun-2009 Dec-2009 Jun-2010 Dec-2010 Jun-2011 Dec-2011 Jun-2012 Dec-2012 Jun towerswatson.com

5 However, during the financial crisis of , a number of these concentrated strategies suffered significant drawdowns at various times. As a group, concentrated strategies underperformed their more diversified peers. This led many investors to question their belief in concentrated strategies and their tolerance for shorter-term performance volatility. This was compounded by the fact that concentrated strategies are typically more expensive than diversified strategies. Our discussions with both our clients and investment managers we work with indicate that the demand for concentrated products diminished significantly after the financial crisis. It is perhaps unsurprising that many investors divested these strategies during this period, if not shortly after. The repercussions of these decisions have a direct impact on the performance of investor portfolios. As noted in Towers Watson s paper, The cost of trigger-happy investing 3, most trustee boards are psychologically biased to action in the midst of poor performance by their investment managers. It could be argued that many of the termination decisions, during the financial crisis of 2008 as well as other times, were focused disproportionately on short-term factors (such as immediate performance) and not on fundamental questions surrounding the investor s belief in the manager s skill over the long term. Studies conducted by academics 4 and by Towers Watson 3 have explored the issue of manager selection and termination. The results indicate through both quantitative and qualitative research that the inopportune churn of managers in a portfolio actually destroys value. The analysis demonstrated that on average, previously poor performing managers (who are terminated) see a turnaround in their results whilst managers with stronger outcomes (who are hired) subsequently struggle to beat the benchmark. Figure 09 illustrates this effect. For many of our clients, the experience with concentrated equity portfolios has been confronting. For those that found the performance volatility too difficult, it has served as a lesson learned; they have acknowledged their product level preferences, and built their portfolios accordingly. This has in some instances, led to much needed capacity becoming available with previously closed highly skilled managers, which has provided an opportunity for investors who have the requisite beliefs and governance structure to withstand the shorter-term volatility. Figure 09. Possible value destruction from hiring and firing 3 years prior to change 1 Terminated managers have underperformed for 3 years Performance vs benchmark 4 Terminated managers see performance rebound } 3 years after change Average value lost in process Date of change or manager Hired managers have 2 Hired managers performance performed strongly 3 is close to benchmark for prior 3 years over next 3 years 3. Penfold, Robin. The cost of trigger-happy investing. Towers Watson publication May Goyal, Amit, and Sunil Wahal. The Selection and Termination of Investment Management Firms by Plan Sponsors. The Journal of Finance LXIII, no. 4 (August 2008): Australian equity portfolios 5

6 Ask not whether a concentrated product is suitable for your portfolio, rather ask whether you have conviction in the manager s skill. The importance of underlying manager skill Regardless of whether a concentrated or diversified approach is used, if active management is to be used, we believe it is critically important to use high quality, highly skilled managers. At Towers Watson, we build conviction in a manager skill within our framework of beliefs, which is summarised in Figure 10. We believe that a great investment manager is one with exceptional people who can successfully implement great investment ideas. The sustainability of this competitive advantage is underpinned by a business structure that aligns the organisation s interest with the investors and their clients, allows investment flexibility (through opportunity set) and supported by a stable platform at both the firm and team level. We do acknowledge that our beliefs lead to a bias towards certain types of products (such as concentrated global equity products). Figure 10. Towers Watson success factors Competitive advantage Sustainability of competitive advantage Investment professionals Approach/insight generation Portfolio management Firm and team stability Opportunity set Alignment Ingredient for success: the importance of alignment An investment manager s preference to build concentrated portfolios tends to be innate to how they think about investing; their investment philosophy, how they express conviction and how they manage risk. We interviewed a number of these managers, and from these discussions it was evident that concentrated investing influences the way investment managers have structured their businesses and aligned their staff as well as the portfolio itself. A concentrated strategy is frequently a manager s highest conviction expression of their investment philosophy; an investment philosophy underpins how a manager believes they will beat the market. At times, certain investment styles or approaches may be out of favour and/or may underperform for extended periods of time. This means that both the business structure and the remuneration of key investment personnel must be sufficiently robust to cater for times when the concentrated strategy is out of favour, whilst rewarding staff when it is performing well. We have found these conditions seem to heavily favour more boutique structures, where the investors either have majority ownership of the business and/or participate significantly in the success of the strategy. The founders of boutiques are often refugees from large funds management organisations. An oft cited reason for leaving to set up a boutique is the pressure portfolio managers can be put under by non-investment executives when going through a period of poor performance. 6 towerswatson.com

7 What governance and beliefs are required to use concentrated equity portfolios? We believe there are a number of characteristics that influence the degree to whether concentrated equity strategies can be successfully implemented by institutional investors. These factors include: Access to resources: The number and quality of available internal resources (for example, investment team) and external resources (for example, consultants) will determine the extent to which investors can locate, research, select and monitor strategies. Tolerance for short-term underperformance: A low tolerance may lead to the termination of managers during periods of underperformance. Size of assets under management (AUM): We believe there is an AUM sweet spot for implementation of concentrated strategies. A small fund size will make it difficult to create a sufficiently diversified portfolio. However, high AUM could also limit the ability to implement concentrated strategies due to the: Greater capacity constraints faced by concentrated managers, Limited number of open and skilled concentrated managers in the Australian market, and Increased governance burden involved in researching and monitoring a larger number of strategies. Ability and willingness to pay higher fees: strategies tend to charge higher fees than diversified active strategies primarily due to their higher expected risk/return characteristics and greater capacity constraints. As such, an investor needs to have the requisite fee budget to spend on this part of the portfolio. The future of concentrated equity portfolios in Australia Towers Watson believes that while it is perfectly logical for individual investors to use active management to attempt to enhance their returns, this cannot add value in aggregate for all investors. Not all institutional investors should pursue active management, and likewise not everyone can and should use concentrated equity portfolios. However, for those with an appropriate set of beliefs, and suitable governance structure, concentrated equity strategies may present an attractive opportunity which should allow them to benefit from market inefficiencies that result from the sheer weight of money flowing out of actively managed strategies. Despite the potential opportunities, investing in active management, and particularly in concentrated equity strategies, will likely place additional governance burden on Investment Committees and Boards, and it is crucial to recognise the trade-offs involved. equity strategies by their very nature are capacity constrained. This poses an intriguing problem given the forecast growth in the superannuation assets over the next 20 years. In the Appendix to this paper, we show that by making a few simple assumptions, we can conclude that the demand for concentrated Australian equity portfolios is likely to significantly outstrip the capacity of the market to supply high quality concentrated equity portfolios. As the average investor grows ever larger, we may see the use of concentrated managers wane simply as a result of lack of capacity. Australian equity portfolios 7

8 Superannuation funds will ultimately have to make a difficult decision on how they manage their Australian equity portfolios moving forward. In the coming years, we expect to see large institutional investors moving more towards higher passive allocations to Australian equities at the expense of active strategies (both diversified and concentrated). We are also likely to see a shift away from Australian equities in favour of global equities. This will be primarily due to capacity issues and fee pressures resulting from increased regulation and competition. These funds will maintain allocations to concentrated equity strategies albeit at a smaller allocation and with a select few skilful managers (in a more partnership style relationship). Interestingly, this in itself could actually create a greater alpha opportunity for those who retain larger allocations to concentrated equity strategies as the market becomes less efficient. However smaller funds may have trouble accessing skilful managers particularly as the demand for skilful managers outweighs supply and some managers will be in a position to choose the investors they want to partner with. This leaves us with what we see as the sweet spot for concentrated equities (that is, medium sized institutional investors who are neither too big nor too small to allocate to concentrated strategies with an appropriate set of beliefs and governance structure to support investing in concentrated strategies). We may also see large investors acquiring boutique equity managers or building larger internal teams to manage their Australian equity portfolios (that is, direct investing) to minimise some of the capacity issues we have highlighted. Ultimately, all investors will face similar pressures around governance, fees and tolerance for underperformance with the determining factor on whether or not to implement concentrated Australian equity strategies coming down to the beliefs of those responsible. Conclusion The key conclusions of our paper are as follows: The theoretical arguments in favour of concentrated equity portfolios are sound, and are broadly confirmed by actual experience in the Australian market. However, given the fact that the Australian market is already relatively concentrated, we believe that there is a place for both concentrated as well as more diversified portfolios in a typical Australian investor s portfolio. Looking into the future, it is clear that there will be significant challenges in constructing an efficient Australian equities portfolio as the average superannuation fund continues to grow at a faster pace than the Australian equity market coupled with the potential for further consolidation in the industry. Towers Watson believes that not all institutional investors should pursue active management, and likewise not everyone can and should use concentrated equity portfolios. However, for those with an appropriate set of beliefs, and suitable governance structure, concentrated equity strategies are likely to present an attractive opportunity. Whenever active management is used, it is critically important to utilise high quality managers with genuine skill and a business model that is sustainably aligned with the interests of its clients. 8 towerswatson.com

9 Appendix: Can the market absorb the likely future demand for concentrated mandates in Australian equities? We have explored the capacity of the market to meet the demand for concentrated equities over the next 20 years by using a simplified projection below. Let s look at the following scenario: The superannuation industry is forecast to be $6 trillion by 2030, of which approximately 3 will be SMSFs if the same ratio is maintained as it is today. This means the super fund industry, excluding the SMSFs industry, will equate to approximately $4.2 trillion by According to APRA March 2014 statistics there are 352 super funds (excluding SMSFs and small APRA funds). Given the regulatory burden of managing a fund and the ongoing consolidation in the industry we have assumed the number of super funds will roughly halve to 200 by The average default strategy had an asset allocation of 28% to Australian equities. For the purposes of this analysis we have been conservative and assumed a 25% allocation to Australian equities at the overall fund level. Of the 25% allocation to Australian equities, we assume a 1 allocation to concentrated Australian equity strategies (that is, 4 of the Australian equity portfolio). Whilst noting these numbers will vary greatly across funds, we looked at the impact of the above base case assumptions on the ability of super funds in 2030 to allocate to concentrated equity strategies. Figure 11. Industry growth 4.5 As shown in Figure 11, the average super fund will have to allocate approximately $2 billion to concentrated equity managers in 2030 a 75 increase from today s levels ($270 million). Now compare this to the Australian equity market which, based on the 10 year annualised average, we expect to grow to approximately $4.3 trillion from a starting point of $1.5 trillion today a 19 increase. So in effect, the average super fund s allocation to concentrated equity managers will grow at least three times more quickly than the Australian equity market by 2030 which is likely to create significant capacity issues in the market. For concentrated equity managers investing in the large to mid-cap universe of the ASX, we would have concerns if more than about $5 billion in FUM was being managed today. This number will be closer to $12 billion by 2030 using the growth figures above, however we highly doubt that the average super fund will be able to allocate $2 billion to skilful concentrated equity managers in an efficient and pragmatic manner. Keep in mind these numbers are for the average super fund, so for larger funds in the industry the results will be much more adverse and may even preclude them from investing in concentrated equity strategies altogether. One way of avoiding capacity issues is to allocate to more managers, however this will increase governance requirements, availability of open and skilled managers, and likely increase portfolio redundancy (that is, neutralisation of active bets for which high fees are being paid). It is also questionable whether there will be a sufficient number of skilled managers available. 4 $4.2 trillion $5.25 billion $ trillion $ billion $2.1 billion $1.1 trillion $1 trillion 1 0 $670 million $270 million 75 growth $270 billion $108 billion $420 billion Total Super Fund Industry (excl SMSF) $ Alloca on to Australian Equi es $ Alloca on to Australian Equi es Per fund alloca on to Australian Equi es Per fund alloca on to Australian Equi es Australian equity portfolios 9

10 Contact If you would like to discuss any of the areas covered in more detail, please contact your usual Towers Watson consultant, or: Jessica Melville Investment Consultant, Manager Research jessica.melville@towerswatson.com About Towers Watson Towers Watson is a leading global professional services company that helps organisations improve performance through effective people, risk and financial management. With 15,000 associates around the world, we offer consulting, technology and solutions in the areas of benefits, talent management, rewards, and risk and capital management. Learn more at towerswatson.com The information in this publication is general information only and does not take into account your particular objectives, financial circumstances or needs. It is not personal advice. You should consider obtaining professional advice about your particular circumstances before making any financial or investment decisions based on the information contained in this document. Towers Watson Australia Pty Ltd ABN , AFSL TW-AU June 2015 Copyright 2015 Towers Watson. All rights reserved. towerswatson.com /towerswatson