The Rise of Factor-Based Investing

Transcription

1 In recent years we have seen a proliferation of systematic or rules-based investment strategies brought to market under the smart beta umbrella. Smart beta incorporates elements of both active and passive investing, focusing on risk factors that have been identified as the underlying drivers of investment returns across different asset classes, time periods and market conditions. Many of these strategies are quite complex, but some are based on well-known phenomena such as the small cap and value effects. As with other innovations in finance, the emergence of smart beta ETFs is making these developments readily accessible to investors. In this report, we review some smart beta strategies that offer attractive portfolio benefits to investors. By Karim HK Ahamed, CFA, HPMP Partner & Investment Strategist, July 2016

2 Executive Summary Historically, investment managers have fallen into two distinct camps active or passive. Active managers presume that manager skill, or alpha, is a big driver of investment returns. However, true manager alpha is scarce, and many active managers struggle to keep pace with the broader market indices. Conversely, passive managers assume it is difficult, if not impossible, to beat the market and investors are better served by passive strategies that simply seek to match the returns of a desired benchmark. In recent years, there has been a plethora of investment strategies brought to market under the smart beta umbrella. The term smart beta incorporates elements of both active and passive investing to create systematic or rules-based strategies that aim to enhance portfolio returns, or reduce portfolio risk. In this report, we will review whether smart beta strategies are an innovative new tool for investors as advertised, or just another catchy marketing slogan. Advances in financial theory have enabled investment researchers to isolate common risk factors that drive portfolio returns. These factors represent the building blocks for smart beta strategies. Many of the newer smart beta strategies are quite complex, but some are based on familiar phenomena such as the small cap and value effects. Researchers have verified that these factors are predictable and replicable across different asset classes, time periods and market conditions. Our review of factor-based strategies leads us to conclude that smart beta strategies provide another tool for managers to improve the investment process. The factors we have highlighted in this report are well known, persistent and readily accessed by investors. Our research shows that combining two or more factor strategies offers better diversification and lower portfolio volatility than the stand-alone factors individually. A growing number of large institutions are now adopting a factor-based approach to portfolio allocation. As with other innovations in finance, the emergence of smart beta ETFs is making these developments accessible to smaller investors. The proliferation of funds coming to market suggests there will be a shake out and consolidation of the smart beta market at some point. By working with some of the top players, and focusing on factors which have been proven in the marketplace, we can capture the benefits of this new tool for our clients. 2

3 Source: R. Bruand, How Smart Beta Has Performed Amid the atility, MSCI blog post, In recent years, there has been a plethora of investment strategies brought to market under a host of names, including systematic alpha, alternative beta, engineered beta, strategic beta, factor strategies, risk premia, or most commonly, smart beta. In this report, we will review whether these smart beta investments are an innovative new tool for investors as advertised, or just another catchy marketing slogan for asset gathering by the firms concerned. First, some historical perspective. Traditionally, investment managers have fallen into two distinct camps active or passive. Active managers utilize skill and market expertise to take active portfolio bets designed to generate attractive returns versus a designated benchmark. They presume that manager skill, or alpha, is a big driver of investment performance. Empirical evidence shows that true manager alpha is scarce, and many active managers struggle to keep pace with the broader market indices. On the other hand, passive managers assume that it is difficult, if not impossible, to beat the markets and investors are better served by passive approaches that simply seek to match the benchmark return, by replicating the holdings of the index or benchmark portfolio and capturing the market return (beta). The Evolution of Smart Beta Developments in financial theory over the last 40 years have led to a better understanding of the drivers of portfolio returns. In the 1970s, it was widely assumed that alpha was the primary driver of returns. Advances in theory and the development of index-based strategies allowed managers to distinguish between beta, or marketbased returns, and alpha contributed by manager skill. By the 1990s, returns could be further broken out to reflect the contribution from the market (systematic beta) versus the contribution provided by the industry sector, country or regional weightings in portfolios. More recently, return streams have been dissected further, highlighting the important contribution of strategy and style factors to overall returns. 3

4 Chart 1: The Evolution of Smart Beta Source: Smart Beta ETF Strategies: Expanding the Investor Toolbox, Invesco PowerShares, April 2016 The launch of index-based passive investing strategies led to the creation of style-based investing based on portfolio tilts such as value over growth or small cap over large cap. From here, indexing evolved to incorporate alternatives to traditional market capitalization-based weighting, such as equal-weighting of each holding in the portfolio, or weighting holdings based on fundamental or accounting factors. The realization that investment returns were driven by common risk factors led to the creation of single factor portfolios focused on additional factors such as dividend yield, profitability, buyback activity and share price momentum. The evolution of investment practice to reflect advances in financial theory can be discerned from the above chart. By now, it is widely accepted that investment returns reflect a combination of the asset class return plus contributions from discrete style factors and strategy factors, with a limited contribution from alpha derived from manager skill. As we see below, each of the asset class return, style factor return and strategy factor return is underpinned by discrete factors. Empirical research has verified that these factors are predictable drivers of returns, and their effects have been replicated over extended periods, in different regions and under varying market conditions. These discrete factors represent the building blocks for smart beta investing strategies. 4

5 MACRO STYLE ALPHA Chart 2: Sources of Portfolio Risk and Return In fact, academic research has demonstrated that much of the return from active management is attributable to portfolio tilts aimed at capturing the premiums associated with well-known risk factors such as size, value, momentum, volatility, yield and quality. Smart beta methodology can be applied to a continuum of investment strategies of varying complexity. For example, an enhanced indexing strategy may simply adjust the portfolio weights so that each security in say, the S&P 500 index is weighted equally, as opposed to the usual weighting based on the stock s market capitalization relative to the overall index. A slightly more complex application would be to select only those stocks that meet a value criterion or size threshold as a way to capture the well-known premiums associated with value stocks or small cap stocks. A further refinement would be to use statistical tools to select only those stocks that score highly across two or more style factors. Finally, the manager could use big data tools such as a review of key words in a company s SEC filings to identify stocks that may have significant upside in the coming months. Source: S. Shores, Smart Beta: Defining the Opportunity and Solutions, BlackRock, February 2015 The term smart beta describes a variety of systematic or rules-based investment strategies seeking to outperform a specific benchmark, or to reduce portfolio risk in relation to the benchmark. As shown in Chart 2, smart beta combines elements of both active and passive investing, and it can be applied across numerous asset categories. As with passive investing, the goal is to derive returns by investing in all or some of the securities in a desired benchmark, such as the S&P 500 index. At the same time, the strategy employs active bets or portfolio tilts as a way to exceed the passive return from simply holding the index. Chart 3: The Alpha/Beta Continuum Chart 3 shows that the smart beta approach can be used to invest in stocks, bonds, commodities and many other categories of assets. But regardless of the asset class, when the investment returns are dissected, it becomes apparent that performance is driven by some combination of these common criteria. A portfolio that can harness these factors in a systematic fashion can help the manager exceed the return on the benchmark, or reduce risk relative to the benchmark, more consistently and at lower cost than traditional actively managed strategies. Source: P. Tindall, Understanding Smart Beta, Towers Watson, August

6 Chart 4: Examples of Smart Beta Strategies Approach Complexity Alpha Generation Methodology Active/Passive Enhanced Indexing Low Targeted exposure by adjusting existing benchmarks/indices based on desired factors such as volatility, value and quality Create a non-market capitalization weighted benchmark Passive Rules-Based Strategies Medium Select and weigh investments based on a set of fixed rules focused on specified risk factors such as value and market capitalization size Active security selection using a structured rules-based approach Active Fundamental or Factor Based Strategies Medium Identify securities that are attractively priced relative to future prospects based on fundamental factors such as valuation, earnings, quality and momentum Use covariance matrices and/or multiple regression to determine the weight of factor exposures Active Metadata Strategies High Invest opportunistically using a highly quantitative or big-data driven approach that incorporates a combination of fundamental, technical and event-driven models Employ a combination of statistical techniques and models to collect and analyze vast amounts of data Active Source: Man, Machine & the Market, Goldman Sachs Asset Management, June 2015 When it comes to implementation, combining factors has been shown to be more effective than using single factor strategies. As illustrated in Chart 5 on the following page, combining two or more factors helps to smooth out the variability in returns experienced with single factors. For example, value investing is a proven driver of attractive long term returns, but investors in the strategy need to be prepared for sustained periods of underperformance when the investing style is out of favor. There are two basic methods for combining strategies. The first is to create a blend of individual factor strategies, such as an equally-weighted combination of four independent single factor strategies. Advantages to this approach are simplicity and the flexibility to weight each factor individually to obtain a desired portfolio tilt. Drawbacks to this approach are more volatile returns, a tendency for uncorrelated factors to cancel each other out, and the potential for higher transaction costs if one factor model is selling a given security while another factor model in that portfolio is buying the same security at the same time. The other method is to screen across all factors in the same portfolio and netting out buys and sells of the same security within the portfolio. Although it is more complicated to implement, this integrated approach helps to reduce redundant transaction costs and provide better smoothing of portfolio variability, as it tends to select securities that score well in the aggregate across multiple dimensions, whereas the single factor technique picks securities that score highest on a single factor, even if they score badly on the other factors. To illustrate how investors can use factor-based strategies as a tool to customize portfolios, Chart 6 on the next page shows how risk exposure can be dialed up or down with the aid of smart beta factors. An investor looking to reduce risk can focus on high quality or low volatility strategies, whereas someone looking to add risk for upside can focus on small cap or high beta strategies. 6

7 Chart 5: Diversification and Smoothing Impact of Blending Single Style Factors Source: K. Teloeken, Investment Style Risk Premiums as a Source for Excess Returns, Allianz Global Investors, 2015 Chart 6: Dialing Risk Up or Down Using Smart Beta Source: Smart Beta Implementation, Smart Beta Research Center, PowerShares, March

8 Chart 7: Which Factor Combinations Work Best Factor Size Value Minimum atility Dividend Yield Average Size 50.00% 68.18% 42.93% 55.00% 35.85% 53.48% 51% Value 73.77% 54.10% 46.20% 50.00% 46.21% 72.87% 57% 64.75% 64.39% 75.41% 77.50% 82.76% 79.07% 74% Minimum atility 54.10% 45.45% 50.54% 49.18% 59.31% 64.34% 54% 42.62% 50.76% 65.22% 71.67% 59.43% 68.22% 60% Dividend Yield 56.56% 71.21% 55.43% 69.17% 60.69% 52.87% 61% Source: J. Smith, The Bayesian View to Multi-Factor ETF Investing, ETF Trends, Certain combinations work best together when combining factors. Chart 7 above is an analysis of rolling one-year returns on the MSCI All Countries World (ACWI) index from Jan 1995 to March It shows that on a stand-alone basis, a small company strategy had only a 50% chance of outperforming the ACWI, while a momentum factor strategy had a 75% likelihood of outperforming the benchmark, the highest of all the factors. However, when momentum is combined with quality (profitability), the likelihood of outperforming the index jumped to 83%. Combining value and small cap factors offered a 74% chance of beating the benchmark, whereas combining value and quality reduced the probability to 46%. This makes sense, as value and small size are complementary factors, while value and quality are opposing attributes that cancel each other out. To illustrate this point further, Chart 8 on the next page shows common risk factors as building blocks for various investment categories. Since the desired type and level of portfolio exposure can be obtained from common factors both within an asset class and across different asset categories, it helps to explain why a multi-factor approach to factor investing works well in practice. In earlier reports, we have mentioned the influence of the business cycle on portfolio returns. In the same way, we find that certain factor attributes also perform particularly well or badly at different stages of the business cycle. As illustrated in Chart 9, adding, or eliminating, specific factors at the appropriate juncture in the business cycle as part of the tactical asset allocation process can help a manager to incrementally add to returns and/or avoid losses by fine-tuning the portfolio to reflect changes in the market environment. Thus far, we have kept the discussion about smart beta techniques quite broad, although we provided some examples of the application of factor strategies to equities. We want to emphasize that the risk factors we have outlined are common to multiple asset categories. To illustrate this key point, Chart 10 is another variant of the period table we have used before, this time showing annual performance rankings by risk factor across various asset classes over the last fifteen calendar years. We see that rankings vary greatly, with no discernible trend in rankings from year to year. Once again, we conclude that broad exposure to a variety of factors improves the chances of capturing upside potential while helping to control downside risk. 8

9 Chart 8: Risk Factor Decomposition of Selected Asset Classes Source: The Rise of Factor Investing, Smart Beta Factor Roadshow, BlackRock, Chart 9: Key Measures and Their Impact on Economic Growth Cycle Source: Making It Work: Factor Pathways, State Street Global Advisors, March

10 Chart 10: Historical Returns by Factor Highest Lowest Value Value Curve Value Low Value Curve Value Value Value Value Value Low Curve Low Low Value Value Curve Curve Curve Low Value Low Value Value Value Value Curve Value Low Curve Low Value Low Low Value Value Low Value Curve Low Curve Low Curve Value Curve Value Value Low Curve Value Curve Value Low Source: The Rise of Factor Investing, Smart Beta Factor Roadshow, BlackRock, Value Curve Value Value Value 10

11 Chart 11: Total Smart Beta Assets and Funds Source: A Smart Approach to Smart Beta ETFs, Morgan Stanley, Implementation Exchange traded funds (ETFs) are widely viewed as the most likely vehicles for implementing a smart beta strategy. Although hedge funds have made use of smart beta techniques for many years, virtually all of the recent growth in smart beta investments has occurred in ETFs, as shown in Chart 11. BlackRock forecasts that smart beta assets in ETFs will quadruple to $1 trillion by To understand why, we note two points. First, smart beta strategies are predicated on delivering a return in excess of that from a passive strategy based on a benchmark such as the S&P 500 index and ETFs are the predominant vehicle for passive investing. Second, as explained in our previous report on the evolution of ETFs, the creation/redemption structure specific to ETFs confers significant benefits, such as the ability to provide ready liquidity to investors in turbulent markets without the need for damaging redemptions to raise funds. (If you would like a copy of the earlier report on ETFs, please contact your HPM advisor.) Factor-based strategies are sometimes criticized as just a more expensive form of traditional passive indexing with a catchy marketing slogan to drive asset gathering by the sponsors. To examine these points, let us first look at Chart 12 on the following page from Vanguard Investments, which compares the characteristics of their Total Stock Market Index fund (a passive strategy based on the Russell 3000 index) and S&P 500 index fund to those of actively managed equity funds, single factor ETFs and smart beta ETFs. 11

12 Chart 12: How Active is Smart Beta? Source: C. Philips et al, An Evaluation of Smart Beta and Other Rule-Based Active Strategies, Vanguard Research, August 2015 This chart demonstrates that at the passive end of the spectrum, the Vanguard strategies have very low active share, or deviation from the referenced index of the 3,000 largest US stocks. At the other extreme, the actively managed strategies deviate significantly from the Russell 3000 index by taking active portfolio bets that tilt portfolios away from the index. As expected, single factor ETFs show greater variability, as they can use a smaller number of stocks to gain exposure to a single factor. Smart beta ETFs show less variability than single factor ETFs, as larger pools are required when combining multiple factors, resulting in broader portfolios whose holdings conform more closely to the Russell 3000 benchmark. It is clear from the chart that factor-based ETFs have an active share profile quite distinct from both classic actively managed strategies as well as traditional passive indexing strategies. On the issue that factor-based strategies are simply a marketing tool for investment firms, we are aware of reports that academic researchers have identified up to 300 separate factors thought to influence investment returns. There is some evidence that investment firms have launched smart beta ETFs based on factors that could not be verified or replicated by others in the investment community. Rigorous analysis by AQR Capital Management, Research Affiliates and others has whittled down the number of verifiable factors considerably when their researchers were unable to replicate the return drivers in different markets or under different economic conditions or time periods. As a result, these researchers concluded that many of the 300 factors cited were often the result of data mining until a desired outcome was achieved. The factors we have highlighted in this report are well known, persistent and readily accessible. 12

13 Another objection that has been raised about smart beta strategies is that they are expensive in comparison to traditional ETFs that offer very cheap and liquid exposure to myriad investment strategies. Expense ratios on ETFs employing factor-based strategies were much higher initially, perhaps because of the smaller asset base and higher upfront development costs. As more money has flowed into the category, expenses can be amortized over a larger asset base and fund expense ratios have come down. Firms such as BlackRock and Goldman Sachs have made a concerted effort to price their factor-based ETFs at a level comparable to their traditional ETFs, taking away much of this argument. The proliferation of funds coming to market, many of which have been created under less rigorous criteria, suggests that there will be a shake out and consolidation of the smart beta market at some point. By working with some of the top providers and focusing on factors which have been proven to provide persistent and replicable results outside of the academic research, we believe we can capture the benefits from this new tool to help improve returns and/or reduce risks for our clients. Conclusion Smart beta is another stage in the evolution of portfolio theory. Some wellestablished factor strategies have been tested over many years, principally by hedge funds seeking an investment edge. A growing number of pension plans, endowments and foundations are adopting a factor-based approach to portfolio allocation, in place of the traditional asset class-based approach. As we have seen with other innovations in finance, the emergence of smart beta ETFs helps to make factor-based investing accessible for smaller investors. Many of the newer smart beta strategies are quite complex, but some are based on familiar phenomena such as the small cap and value effects. Despite being widely known and used by so many investors, these effects have persisted over long periods without losing their effectiveness, either because the factors are so strong they cannot be arbitraged away, or because investor behavior gets in the way. For example, an investor employing a momentum strategy or a value strategy has to be prepared for extended periods of underperformance when the strategy is out of favor and investors often lack the patience or discipline to hold on to a severely underperforming strategy long enough for it to turn around. Source: Aaron Bacall, cartoonstock.com. Reprinted with permission. (Our special thanks to AQR Capital Management, BlackRock and Goldman Sachs for their help with our research for this report.) 13

14 Important Disclosures This document contains general information that is not applicable to everyone. The information contained herein should not be construed as personalized investment advice. Past performance is no guarantee of future results. There is no guarantee that the views and opinions expressed in this document will come to pass. Information contained herein has been obtained from sources believed to be reliable but is not guaranteed accurate. All information is as of July 20, 2016, political and economic changes after that date may impact the accuracy of the information herein and HPM Partners LLC ( HPM ) is not obligated to update this document. Investing in the financial markets involves risk, including the risk of loss of the principal amount invested, and may not be appropriate for everyone. The information presented is subject to change without notice and should not be considered as a solicitation to buy or sell any security. HPM is an SEC-registered investment adviser with offices in New York, Illinois, Ohio, Michigan and California. For information pertaining to the registration status of HPM, please contact us or refer to the Investment Adviser Public Disclosure website ( For additional information about HPM, including fees and services, send for our disclosure statement as set forth on the Form ADV using the contact information herein. Please read the disclosure statement carefully before you invest or send money. 18