Factor Investing and Adaptive Skill: 10 Observations on Rules-Based Equity Strategies

Transcription

1 T H E J O U R N A L O F THEORY & PRACTICE FOR FUND MANAGERS SPRING 2016 Volume 25 Number 1 Factor Investing and Adaptive Skill: 10 Observations on Rules-Based Equity Strategies MIKE SEBASTIAN AND SUDHAKAR ATTALURI The Voices of Influence iijournals.com

2 Factor Investing and Adaptive Skill: 10 Observations on Rules-Based Equity Strategies Mike Sebastian and Sudhakar Attaluri Mike Sebastian is a partner at the Aon Center for Innovation and Analytics in Singapore. mike.sebastian@aonhewitt.com Sudhakar Attaluri is an associate partner at Aon Hewitt in Chicago, IL. sudhakar.attaluri@aonhewitt.com In recent years, some institutional investors have become disenchanted with the performance record of the traditional active equity investment industry. At the same time, investors have become increasingly focused on meeting their investment objectives in a low-return market environment. As a result, alternatives to traditional approaches have been sought out, and rulesbased investing (often called smart beta), with its allure of market-beating returns at low cost through often-simple strategies, has gained a great deal of interest. 1 In this article, we present 10 easyto-digest observations about rules-based investing and their role in an institutional portfolio, with a focus on the well-known fundamental and low-volatility strategies. We discuss what we believe to be the drivers of rules-based investing performance (exposure to factors such as value and small capitalization, among others) and how these strategies might fit into a portfolio, and we resolve the apparent contradiction of expecting a rules-based strategy to succeed when the average active manager with a bigger investment toolkit underperforms. We contrast the rewards of rules-based investing with those of exploiting adaptive skill that adjusts to suit complex and changing market conditions. Finally, we present a practical decision tree framework to help investors considering a switch among strategy types high-conviction active equity management, traditional active, and rules-based and equity alternatives such as long short hedge funds. Investors who have cost constraints, a particular aversion to active risk, or a desire to reduce equity beta, combined with a desire to have some chance for above-market returns, may find that rulesbased investing offers the most to like. The following 10 observations can help an investor make beneficial decisions on an equity portfolio structure in a world with rules-based investing tools, starting with their basic drivers of return. 10. RETURNS TO RULES-BASED INVESTING ARE MOSTLY DRIVEN BY EXPOSURE TO FACTORS Traditional indexes weight stocks by market capitalization, or the market value of outstanding stock. Rules-based portfolio strategies now often referred to as smart beta weight stocks by something other than market cap. The simplest non market cap way of weighting stocks is equal weighting, but this strategy requires holding equal weights in the very smallest stocks, which may be hard to trade. Therefore, most smart beta indexes use something else to weight stocks (i.e., a transparent, rules-based approach). These indexes deviate from traditional capweighted indexes by reweighting stocks in Spring 2016 The Journal of InvesTIng

3 a systematic manner based on well-defined factors such as relative price/earnings, relative volatility, momentum, quality, and other risk-based or market-segment criteria. Why avoid market capitalization? Critics of market cap weighting point out that if a stock becomes overvalued it becomes a bigger part of a market cap weighted index. Although an investor who holds stocks that become overvalued benefits from the price increase, if the market eventually fixes its mistake, the investor suffers a loss. A typical smart beta investor would have sold the stock, harvesting the gains as it rose and reducing exposure to its subsequent decline. Likewise, a smart beta investor would buy stocks as they fall, expecting a future reversion to fair value. This approach has a lot in common with traditional value investing. Because smart beta strategies do not include investing as much in stocks with the biggest market cap, they often tilt toward small cap. Another flavor, low-volatility strategies, weights stocks based on volatility or market risk, which is in turn based on the view that low-risk stocks may earn better riskadjusted (or even absolute) returns. Although there are many different types of smart beta strategies, historical evidence has shown that the value and small cap tilts account for much of their returns. 2 This evidence includes low-volatility strategies, although current research suggests that additional factors such as betting against beta (which favors low market risk) explain their returns as well. 3 Exhibit 1 shows the long-term value added of several popular smart beta portfolio construction methodologies after accounting for factor exposures in this case, the value and small cap factors. The gray area represents the boundary of statistical noise as shown, none of the strategies produced an alpha that can be statistically distinguished from zero supporting the idea that smart beta returns are explained by style. 4 Where historical but statistically insignificant alpha exists, it may be attributed to implementation methodology, additional (unexpected) risk factors in the portfolios, or simply to random noise. E x h i b i t 1 Style-Adjusted Value Added for Rules-Based Strategies Source: Clare, Motson, and Thomas [2013a; 2013b]. E x h i b i t 2 Average Returns by Style and Size CERTAIN RISK FACTORS HAVE OUTPERFORMED IN THE LONG RUN Although value strategies have generally performed better than the market across many stock markets worldwide over long periods of time, they have also shown extended periods of underperformance. 5 Exhibit 2 shows Source: Fama and French [2014]. the performance of U.S. stocks sorted into 25 groups by size (market capitalization) and valuation (measured by book-to-market ratio) over the period. As shown, small stocks generally outperformed large stocks, Factor Investing and Adaptive Skill: 10 Observations on Rules-Based Equity Strategies spring 2016

4 and value stocks (especially small value) generally outperformed growth. Historical data on the relationship between beta/ idiosyncratic volatility and return is shown in Exhibit 3. Here, the analysis uses U.S. stocks (sorted into five groups of increasing historical five-year beta/idiosyncratic volatility) to address questions of whether the lowvolatility stocks outperform the high-volatility stocks. Finance theory suggests that higher risk should result in higher return in the long run; instead, we see lower returns in higher beta/volatility stocks. Other factors, among them momentum (past winners keep winning) and quality ( better run companies outperform), are available for investment. By some counts, over 300 different factors have been identified in the finance literature. 6 Value, small, and low volatility, however, are the primary factor exposures currently found in many of the most common institutional smart beta strategies. Potentially attractive factors have high absolute and risk-adjusted historical returns, low correlation with other parts of the portfolio, and a large body of credible evidence (research) showing they exist and why. Factor investing is nothing new. Some combination of broad equity market, bond duration, credit, and liquidity risk factors are found in nearly every institutional portfolio. (Indeed, value, small, low volatility, and other factors are included in broad equity portfolios as well.) Traditional factors, however, primarily deliver investors a return premium for bearing risk, whereas the story for value and other more exotic factors is somewhat more complicated. E x h i b i t 3 Average Annual Returns by Beta/Volatility FACTOR OUTPERFORMANCE IS DRIVEN BY MARKET MISTAKES AND RISK The simplest way to explain the value premium is this: Investors become too optimistic about favored stocks and bid up their prices too much, only to see them later fall back toward fair value. Value-oriented investment strategies earn excess returns by exploiting this situation. Although this is a market mistake, or inefficiency, value stocks might also have higher returns because they are riskier. For example, such stocks might be less liquid or more prone to financial distress. Compensation through higher expected returns for taking on more risk is a risk premium. The low-volatility effect (i.e., low-volatility stocks outperform high-volatility stocks) is related to value and other risk factors and probably is also driven in part by behavioral factors such as investment managers focus on benchmarkrelative performance, which may deter them from investing in low-volatility, high tracking-error stocks. 7 The distinction between the two reasons for historical outperformance risk premium and inefficiency has implications for how investors might access the extra returns and whether they will persist in the future. Risk premiums are attractive only if they are big enough relative to their risk or sufficiently diversifying when combined with other sources of risk and return. Many risk premiums, however, can be obtained without exceptional skill and are not necessarily at risk of disappearing permanently. 8 Inefficiencies, on the other hand, are attractive because they potentially translate to extra returns (and potentially without much extra market risk) if investors can exploit them. It generally takes skill, however, to exploit inefficiencies, which can more easily disappear if enough people recognize and act on them. 7. MANY RISK FACTOR RETURN PREMIUMS HAVE SURVIVED FOR A LONG TIME BUT THEIR VITAL SIGNS NEED ONGOING MONITORING Source: Fiore and Saha [2015]. The value, small, and low-volatility effects have existed for the past several decades despite general awareness of their existence among the investing public. This may be because such effects are a form of compensation for real risks, are deeply rooted in how investors behave, or have not yet been fully exploited by investors. If factor-based strategies continue to attract assets, it may Spring 2016 The Journal of InvesTIng

5 put pressure on the associated premiums. Low volatility oriented assets in particular have increased dramatically since the global financial crisis. Investors who choose to try to exploit these factors especially with narrower strategies that focus on them individually must keep an eye on their continued health. This may be challenging, even more so in an extended period of underperformance. Investors should demand ongoing reporting on the prospects for the relevant risk premiums from their managers who may need to adjust their strategies to adapt to market conditions. Managers should even be willing to shut down strategies if the risk premiums they exploit show warning signs of disappearing permanently. 6. FACTOR STRATEGIES GO IN AND OUT OF FAVOR; BE DYNAMIC OR IN FOR THE LONG HAUL Regardless of performance drivers, risk factor strategies as with other investment types go in and out of favor for extended periods of time. This presents a risk, but also an opportunity, as skilled investors may be able to add value by being dynamic in their approach. A dynamic approach can be implemented by using active equity managers who incorporate value, low volatility, and other factors into a broader strategy, or by the asset owner or multi-asset manager using style-focused strategies to express market views. Investors who do not wish to engage in a dynamic approach should be prepared to simply invest for the long term, sticking with strategies even as they experience periods in which they are out of favor. 5. THERE ARE SEVERAL POSSIBLE APPROACHES TO FACTOR INVESTING; A TYPICAL INSTITUTIONAL PORTFOLIO PROBABLY ALREADY EMPLOYS AT LEAST SOME OF THEM Smart beta is a way of applying factor-investing techniques to a broad stock portfolio in a systematic manner. There are, however, several other ways to attempt to profit from factors: Style index funds that focus on subsets of the broad market (but are market cap-weighted) have been available for many years. Quantitative active products often employ factor approaches expressed through skill-based and adaptive model building. Many fundamental active products also use similar approaches to select securities and have significant factor exposures in their return patterns. Opportunistic deep-value investing allows investors to seek to exploit significant mispricing of certain assets. 9 In general, entire asset classes can experience misvaluation, from which investors can potentially profit using tactical views. Investors need to balance the costs and benefits of each potential approach. 4. VIRTUALLY ALL INVESTMENTS ARE ACTIVE IN SOME WAY; THUS, INVESTORS SHOULD SKIP THE DEBATE, MEASURE THEIR ACTIVE RISK, AND MAKE SURE THEY ARE BEING COMPENSATED FOR IT An ongoing debate exists as to whether smart beta strategies are active, passive, or something in between. In practice, virtually all investments are active in some way. An investor s position in a market cap-weighted equity index fund is an active bet on the equity risk premium. A style index fund, or a smart beta portfolio, is an active bet on a collection of risk factors. A traditional active equity manager combines all of the aforementioned features as well as an active bet on a particular manager s skill. Investors should measure the overall active risk in their investment programs, whatever the sources. They should then seek to answer two questions. First: Am I comfortable with the overall level of risk in the program and its sources? Second: Do I expect to be compensated appropriately through risk premiums, factor returns, and profits from skill for those risks? 3. NOT ALL SMART BETA IS CREATED EQUAL; CAREFULLY REVIEW THE RISK EXPOSURES OF THE SPECIFIC STRATEGIES CONSIDERED Investors may choose from a wide and diverse set of smart beta strategies that generally deliver a combination of factor exposures. The type and size of such exposures, however, vary by strategy, and other factors (whether intended or not) may be present as well. Likewise, strategies differ in rebalancing methodology, Factor Investing and Adaptive Skill: 10 Observations on Rules-Based Equity Strategies spring 2016

6 fees, trading costs, and other dimensions. Although most adhere closely to a set of fixed, transparent portfolio construction rules, some have a greater degree of manager discretion and often cross the line into traditional active management. Smart beta investors are generally buying a specific portfolio strategy not the perceived skill of a team or a complex investment process so these specifics need to be closely examined in the investor s due diligence process. 2. ADAPTIVE SKILL IS THE BEST WAY TO OBTAIN POTENTIALLY ABOVE-MARKET RETURNS The historical outperformance of factor strategies versus traditional market cap weighted indexes is due at least in part to market inefficiencies. Exploiting market inefficiencies is an endeavor that generally requires significant investment skill, including the ability to use adaptive skill, meaning skill-based strategies that can adapt to complex and changing market conditions, unlike static strategies. Broader, less constrained mandates, which may incorporate value-oriented as well as other strategies, may enhance the probability of achieving success with a skilled active manager. Factor investing, like other forms of equity investing, may benefit in particular from relaxing the long-only constraint. As the average traditional active manager generally underperforms the benchmark, how can a narrower, rules-based smart beta strategy outperform it? First, the average active manager simply may not be able to successfully exploit factor premiums, even when they are widely known and can be harvested with mechanical strategies. Average managers investment philosophies may lead them in a different direction from value and other factor-related strategies. They may not be able to capture factor premiums as efficiently as even a simple strategy. Alternatively, they may diminish any gains from factor premiums through other unsuccessful trades. Second, adaptive investment management skill is expensive in terms of higher fees and trading costs. The cost gap between a cheaper (but well-founded) and a more expensive strategy can make the difference between outperformance and underperformance. Historical evidence suggests that a well-chosen and structured set of exposures to attractive risk factors may be expected to outperform the average, higher-cost traditional active manager. 1. THE RIGHT PORTFOLIO DEPENDS ON INVESTOR SUITABILITY FACTORS THAT INCLUDE RETURN OBJECTIVES, TOLERANCE FOR RISK AND COST, AND OVERSIGHT RESOURCES The most impactful portfolio changes an investor can make will almost always be in the area of overall asset allocation and ensuring optimal diversification. Investors should begin by reviewing their circumstances and objectives, including suitability for highly active investments, and ensuring that broad high-conviction asset categories such as liquid and illiquid alternatives are represented appropriately for their portfolio s size, time horizon, and governance resources. Within the equity portfolio, investors have the tools to suit a wide variety of investor characteristics: Market cap-weighted index funds seek to provide low-cost market returns with minimal implementation risk. Efficiency-minded investors who wish to minimize cost and complexity while seeking to outperform the average equity investor will be well served here. Smart beta and other risk factor approaches are attractive tools for investors with the governance resources to skillfully select and monitor investment strategies and the desire to seek added value versus market cap-weighted indexes, but who want or need to reduce implementation cost or reliance on traditional active equity management. This may include large, sophisticated investors who, by virtue of their size, operate within capacity constraints. Smart beta may also fill a factor-exposure need in a portfolio; for example, a need to eliminate an unwanted style bias. Lastly, factor strategies are a potential tool to express views on various market segments, allowing dynamic application of medium-term views. Traditional active strategies have come under fire recently, but skilled managers have the potential to add value, especially in an unconstrained or concentrated mandate. As such, investors with a tolerance for costs, the access to manager selection skill, and the patience to stay with highconviction strategies for the long term may find traditional active strategies attractive. (Keep an eye, however, on the value a manager adds after Spring 2016 The Journal of InvesTIng

7 E x h i b i t 4 Decision Tools A High conviction refers to the highest rated, as opposed to typical, active strategy. Such strategies are typically broader in mandate, less constrained, and often more concentrated, taking higher active risk relative to the benchmark. B Oversight requirements refer to the selection and monitoring of traditional active managers and of specific rules-based strategies. Factor Investing and Adaptive Skill: 10 Observations on Rules-Based Equity Strategies spring 2016

8 accounting for any persistent factor exposures he or she may have.) Expanding high-conviction active equity management into the equity alternatives space further increases the potential for alpha and reduced equity market exposure by loosening the long-only constraint. 10 Investors with maximum suitability for active investing may find the greatest fit with a combination of hedge fund and highconviction traditional mandates. Cost-conscious investors should be aware that alternative investments come with materially greater fees and potential additional risks associated with the use of derivatives, leverage, and illiquidity, among others. For investors considering enhancements to their equity portfolios, Exhibit 4 lays out a set of decision tools to map their options. The tools relate to three potential starting points: (1) enhancements to a portion of the equity portfolio that is currently invested in a market cap weighted index fund; (2) in a typical active manager; or (3) in a high-conviction active portfolio. The Exhibit considers shifts among these categories and to two of the most common rules-based (smart beta) strategy types. It is strategic in nature and does not consider current views the investor may hold regarding the near- or medium-term prospects for value, low volatility, or other factor strategies, or active management in general. Investors should begin a review of the equity portfolio with a consideration of overall return objectives, risk tolerance, cost constraints, and their oversight resources. With these considerations in mind, investors should ensure that all moves are in the direction of increased efficiency and closer alignment with their circumstances and objectives. We believe that most investors are best served by a combination of low-cost, market-like returns through market cap weighted index funds and efficient deployment of active risk. Generally, this means moving away from typical active equity management and toward high-conviction traditional active management and equity alternatives. Profiting from exposure to potentially attractive return factors through rules-based strategies is also a useful approach. This is particularly true for investors with cost constraints, an aversion to significant active risk from concentrated portfolios, or a desire to reduce market exposure at relatively low cost in the medium or long term through low-volatility investments. Rulesbased or other factor-based strategies may also be used as part of a strategy to express views on various styles. In all cases, success in producing better outcomes than those of the equity market requires access to skill either in active manager selection or in identifying and monitoring attractive risk factors. Investors without access to such skill should construct an efficient, low-cost portfolio that seeks to capture market returns. Those who can exploit that skill may improve their odds of meeting overall investment objectives with careful portfolio construction and a long-term focus. ENDNOTES 1 Although we, like some others, are not particular fans of the less-than-descriptive term smart beta, we recognize its wide use in the industry. We use the terms rules-based investing and smart beta interchangeably in this article. 2 See Clare, Motson, and Thomas [2013a]. 3 See, for example, Frazzini and Pedersen [2014]. 4 The gray area specifically represents the boundary of a t-statistic of 2.0 or less for alpha relative to a portfolio of the Fama and French size (SMB) and value (HML) factors. 5 See Fama and French [2014]. 6 See Harvey, Liu, and Zhu [2014]. 7 See, for example, Baker, Bradley, and Wurgler [2011]. 8 Empirical evidence shows that certain risk premiums associated with systematic sources of risk, although cyclical over short horizons, have generally been positive over long time periods. Their cyclicality may in fact be one of the reasons they have not been arbitraged away. See, for example, Bender et al. [2013]. 9 See, for example, Scotto [2014]. 10 Although smart beta strategies can be implemented via long-only or long short approaches, in our view, equity hedge funds are likely the most effective at market timing, risk premium timing, and stock selection investment skills that are not easily captured by rules-based indexes. REFERENCES Baker, M., B. Bradley, and J. Wurgler. Benchmarks as Limits to Arbitrage: Understanding the Low-Volatility Anomaly. Financial Analysts Journal, Vol. 67, No. 1 (2011), pp Bender, J., R. Briand, D. Melas, and R.A. Subramanian. Foundations of Factor Investing. MSCI Research Insight, December 2013, Foundations_of_Factor_Investing.pdf. Spring 2016 The Journal of InvesTIng

9 Clare, A., N. Motson, and S. Thomas. An Evaluation of Alternative Equity Indices. Part 1: Heuristic and Optimised Weighting Schemes. Working paper, Cass Business School, 2013a.. An Evaluation of Alternative Equity Indices. Part 2: Fundamental Weighting Schemes. Working paper, Cass Business School, 2013b. Fama, F., and K.R. French. A Five-Factor Asset Pricing Model. Fama-Miller Working Paper, Fiore, C., and A. Saha. A Tale of Two Anomalies: Higher Returns of Low-Risk Stocks and Return Seasonality. The Financial Review, Vol. 50, No. 2 (May 2015), pp Frazzini, A., and L.H. Pedersen. Betting Against Beta. Journal of Financial Economics, Vol. 111, No. 1 (January 2014), pp Harvey, C.R., Y. Liu, and H. Zhu. and the Cross- Section of Expected Returns. SSRN abstracts, 2014, papers.ssrn.com/sol3/papers.cfm?abstract_id= Scotto, M. Opportunistic Deep Value Investing: A Multi-Asset Class Approach. Aon Hewitt, 2014, consulting/2014_opportunistic-deep-value-investing_a- Multi-Asset-Class-Approach_WP.pdf. To order reprints of this article, please contact Dewey Palmieri at dpalmieri@iijournals.com or Factor Investing and Adaptive Skill: 10 Observations on Rules-Based Equity Strategies spring 2016