Consultants who use first-hand knowledge to deliver first class advice.
At JANA we showcase our insights and thought leadership.
Our work is centered on achieving the best outcomes for our clients and their beneficiaries.
Our people make all the difference.
Relevant tools, forms and documents.
This paper provides an overview of multi-asset style risk premia portfolios, also known as alternative beta and risk premia. It reviews style risk premia as a concept, the benefits and concerns with style investing and potential ways investors may utilise style risk premia strategies within their overall investment portfolios.
This paper provides an overview of multi-asset style risk premia portfolios, also known as alternative beta and risk premia. It reviews style risk premia as a concept, the benefits and concerns with style investing and potential ways investors may utilise style risk premia strategis within their overall investment portfolios.
The concept of style investing is well known, particularly in equity investments. Often you hear investment commentators discuss the ‘value’ or ‘quality’ style in describing relative performance within equities. Style investing (such as value, where the expectation is that under-priced securities will outperform overpriced securities) has been around for decades, made famous by Graham Dodd in the 1930s (Benjamin Graham and David Dodd published their book, Security Analysis in 1934. The book advocated focusing on the value of a company and the quality of its management). While style investing has been well known, investors’ ability to access various styles of investing has been indirect and confined to long only, actively managed portfolios and more recently smart beta portfolios.
Academics often refer to different styles as factors. Factors are a set of characteristics that explain a security’s risk and return and are generally categorised into macroeconomic drivers and those caused by risk transfer, behavioural biases and structural impediments. Macroeconomic factors are generally non-diversifiable in so far as investors are unable to immunise these risks completely from their portfolios. Examples of macroeconomic factors include economic growth and inflation. The drivers of style risk premia are described in the table below:
Traditional asset classes such as equities and bonds expose investors to varying degrees of macroeconomic factors, as well as additional style risk factors that are more dynamic in nature. Exposure to risk factors may provide either a negative or positive expected return or risk premia.
The concept of investment returns being driven by the return of the market and other factors is a recently accepted concept, at least in academic circles. Since the 1960s, investment theory has relied on the capital asset pricing model (CAPM) theory that an investment portfolio’s risk and return is a function of its market risk or beta, leading to the rise of passive investing with the first index fund offered in 1971 (Kate Ancell, “The Origin of the First Index Fund”, The University of Chicago Booth School of Business, 2012.). In the early 1990s, Fama and French (Eugene Fama and Kenneth French, “The Cross-Section of Expected Stock Returns”, Journal of Finance, June 1992.) extended the basic concept of CAPM to include value and size (smaller capitalisation firms outperform larger firms), alongside market risk as additional factors explaining an equity portfolio return. Later that decade, Carhart (Mark Carhart, “On Persistence in Mutual Fund Performance”, Journal of Finance, March 1997.) extended the Fama and French model to include momentum (securities with increasing price momentum outperform securities with low price momentum) as a fourth factor. Within academic circles, acceptance of these style factors has not been universal as this would imply that investors are not completely rational and hence markets are not efficient, a central tenant of CAPM.
Academics have published reports on hundreds of factors, leading to one well known description as a ‘factor zoo’. Many of these factors have arguably been the result of data mining (manipulating the data to get the answer you want), localised factors unable to be replicated across regions or asset classes and a lack of economic or behavioural rational explanations for their existence and continued persistence.
While spurious risk factors are and continue to be a concern, there is a consensus that a small number of style factors across regions and asset classes persist with differing explanations on why these style factors exist. As discussed previously, the existence of style risk premia typically results from either risk transfer, structural impediments or behavioural biases.
The investment community has tended to focus on a handful of style risk premia that have shown historical persistence across, and within, asset classes, coupled with an economic rationale for their existence and continued persistence. This latter point of persistence continues to be debated. While rational explanations can be reconciled back to requiring a return for taking on additional risk, behavioural explanations challenge certain investors as market anomalies should not exist in a CAPM framework of rational investors. For this reason, many investors place a lower confidence on these style risk factors, despite one of the best-known style risk factors (momentum) having one of the longest and strongest track records across and within asset classes (Y. Lemperiere, C. Deremble, P. Seager, M. Potters, J.P Bouchaud, “Two centuries of trend following”, Capital Fund Management, May 2014.).
Many of the style risk factor explanations involve both rational and irrational explanations for their existence. Equity value is one such style risk factor where the debate on the explanation for its existence is contested. Some academics believe value is the result of risk transfer, a return premium for value stocks being inherently riskier than growth stocks. Others contest it is the result of behavioural biases as investors typically extrapolate past earnings growth rates into the future and overpay for high growth stocks. While the explanation is probably a function of both rational and irrational factors, its past and likely future persistence as a style risk premia has an economic basis for its existence.
Main Style Risk Factors
The style risk factors that are well known and generally used by active managers are summarised in the following table.
Some of these style risk premia are arguably sub-sets of broader styles. Low volatility style premia in equities is viewed by some investors as a derivative of value and quality styles with low volatility stocks often viewed as safe and unexciting value stocks, while high volatility stocks are seen as glamour or growth stocks attracting high price multiples. Carry is another example of having potential cross-over with value as higher yielding securities are often the result of low prices for underlying securities.
Style risk factor investing is implemented systematically through a rules based approach that requires dynamic portfolio rebalancing in order to maintain exposure to particular style factors. A simple example is a value strategy which overweights the cheapest securities based on fundamental metrics, such as price-to-book ratio, and underweights the most expensive securities using the same value-based metric. Periodically the value focused portfolio will be rebalanced to make sure it maintains its value factor exposure, as distinct from a portfolio based on a market capitalisation index who’s weighting to a security is adjusted as its price (and therefore market capitalisation) changes.
There are a range of style risk factor-based offerings providing varying degrees of concentration to individual risk factors within their investment strategies:
The diagram below captures the spectrum of implementation options, highlighting the movement in exposure to the risk premia, complexity, investment capacity and market exposure across the options:
While smart beta portfolios share common ground with style risk premia portfolios in targeting exposure to style risk factors, their construction and approach are very different. Smart beta portfolios are long only portfolios focused typically on one style risk premia in one asset class, although there is a movement to multi-factor smart beta strategies. In contrast, style risk premia portfolios diversify across multiple style risk premia (value, quality, carry, momentum) and across multiple asset classes (equities, fixed interest, commodities and currencies) while maintaining market neutrality to major asset classes. This results in very different risk exposures, volatility of returns and performance drivers. See table below.
Historically all returns generated by investment managers were considered alpha. With the advent of CAPM, alpha morphed into all excess returns above the market return or index, recognising a large function of risk and return was driven by systematic macroeconomic risk factors or beta.
Today, alpha is considered as the return of a fund or assets not explained by common risk factors. As more common factors are identified, alpha becomes beta and is commoditised, resulting in lower fees for investors and the concept of alpha being redefined.
Alpha is generated by information or ideas not easily replicable and includes security selection and market timing. Beta on the other hand is driven by common non-diversifiable or systematic macroeconomic risks that each asset class contains varying degrees of exposure. Static risk characteristics such as value and momentum are considered style risk factors or ‘alternative betas’ that are found across and within asset classes.
Traditional market capitalisation-based passive investing, despite holding the same assets as style risk premia investing, exposes investors to changing risk exposures and portfolio characteristics over time. Whereas style risk premia portfolios undertake dynamic trading and rebalancing to maintain constant risk exposures and portfolio characteristics. (Note this has important implications for valuation of style risk premia portfolios, as discussed in section 7).
As style risk premia investing uses shorting, leverage and constant security trading, the experience, skill and judgement of the manager is vital to effectively risk manage these products. Defining the strategy (risk premia inclusion, weighting, rebalancing) and implementation through risk and cost management are important factors. For this reason, fees for style risk premia investment portfolios sit somewhere between active and passive investment management.
Style risk premia portfolios differ from traditional active management strategies in that they generate returns not from bottom-up security selection, but instead through capturing risk premiums or style factors that are persistent drivers of return, resulting from risk transfer and/or exploiting structural and behavioural anomalies.
These style factors are found within equities, fixed income, currencies and commodities and across asset classes. As stated previously, style risk premia strategies are implemented in a non-directional long-short manner, minimising equity and fixed interest market exposures with the intention to provide returns that are not correlated to traditional equity and fixed interest portfolios.
Although different firms have different definitions, style risk factors are broadly classified into value, carry, momentum and quality/defensive. Each of these style risk factors are generally lowly correlated to each other. The following table from BlackRock shows the theoretical average correlations among risk factors in a multi-asset class portfolio.
When combined with an equal risk weighted allocation, the resulting portfolio provides strong diversification benefits that helps smooth out the volatility of returns, reduces draw down risk and decreases reliance upon single style risk premia as drivers of return, creating a more balanced risk and return portfolio.
Style risk factors offer attractive characteristics as part of a diversified portfolio. These benefits include:
ReturnsSimilar to many hedge fund strategies, the return targets of many of the style risk premia products are defined in terms of returns per unit of risk (also called the Sharpe Ratio). For style risk premia strategies, these targets range from 0.7 to 1.0. Hence, for a portfolio that targets 10% p.a. volatility, the targeted return would be 7-10% p.a. above cash.
In practice, many of the available products do not have live track records long enough to verify if the performance target is reasonable, with the longest live track record portfolio commencing in September 2012. This manager’s performance shown in the charts below targets a risk reward ratio of 0.8 with a targeted volatility of 8% p.a. and has achieved a risk reward ratio of around 1.0 on an after-fees basis.
JANA’s own view is that a more conservative long-term risk reward ratio of 0.5 is achievable based on our experience of hedge funds and active management strategies. This still provides an attractive expected return, especially considering our view that we are entering a lower return environment.
Style risk premia portfolios are lowly correlated to traditional equities, bonds and hedge fund indices, providing good diversification benefits. The following table shows the correlation of returns to major equity, fixed interest and hedge funds indices generated by three managers that JANA has approved as part of its research process.
Style risk premia portfolios are also moderately correlated to each other, indicating there would be risk and return diversification benefits of appointing more than one manager. The table below shows the correlations of the three JANA approved managers to each other.
The following chart shows a composite of live and back-tested performance of the three style risk premia managers’ maximum drawdowns versus MSCI World ex Australia (Managers’ target between 6-10% annualised volatility based on monthly data.).
Similar to our analysis of returns, we view these numbers with scepticism and believe they provide a starting point for assessment of likely correlation and magnitude of potential drawn downs rather than data on which to base future expectations. Maximum drawdowns (peak to trough) based on the data provided have been moderate, circa 10% and much lower than equities, illustrating the diversification benefits versus equities. JANA believes a ‘rule of thumb’ of 2 times the invested portfolio target volatility for expected maximum drawdowns to be a more conservative expectation.
Fees for style risk premia portfolios are lower than most hedge fund strategies and comparable to many active long only managers. The funds JANA has approved for investment offer base fee only structures starting from 0.55% to 1.00% p.a. and scale down depending on the quantum of assets invested and volatility targeted. The underlying investments of style risk premia portfolios are highly liquid, enabling these investments to offer a high degree of liquidity. Liquidity terms range from daily to monthly for redemptions with one day to one week’s notice period. Many style risk premia portfolios offer increased transparency relative to hedge funds with portfolio holdings, return and factor exposures provided to investors.
While style risk premia portfolios offer attractive benefits as part of a balanced portfolio, they are not without their limitations:
Philosophy and Complexity
Style risk premia portfolios are relatively new. Investors need to form an opinion on whether they believe all or some of the style risk factors are likely to persist going forward as many are the result of inefficiencies in capital markets. This in turn requires an understanding of their own philosophy regarding what drives returns in capital markets and whether they believe in fully efficient markets (all investments should be passive) or some lesser view of efficiency in capital markets (allowing some form of active management and the existence of market anomalies). JANA’s own view is that alpha and exposure to different style risk premia (alternative beta) can play an important role for portfolio diversification.
Style risk premia portfolios are also complex. The use of hedge fund techniques that include leverage, short-selling and constant trading requires a higher level of understanding of the risk management techniques used by these managers. As style risk premia portfolios are still relatively new, there is no consensus on how these portfolios should be designed or implemented. Differences in how a portfolio is designed and implemented has resulted in a wide dispersion of returns among style risk premia managers, despite being an alternative beta. Understanding the details and nuances of how a manager designs and implements their portfolios and contrasting different managers requires a greater degree of due diligence and understanding than for typical long-only investment portfolios.
While alternative beta portfolios are highly risk diversified relative to a typical balanced fund, they are not, nor should they be viewed as products which will deliver a smooth and consistent return profile in achieving their risk and return objectives. As the charts in section 7.1 detail, style risk premia funds will have drawdowns and rolling negative returns, albeit expectations are that these will be less extreme than equities.
Liquidity and Crowding Effects
As style risk premia portfolios are designed to be market neutral, they utilise short-selling and leverage to remove market or beta risks. In removing these market risks, they are exposed to increased risks resulting from crowding effects and resulting liquidity shocks. These are risks that ca not be completely isolated from these strategies, but the risks may be minimised through effective collateral and risk management, as well adopting a number of manager exposures that design and implement style risk premia portfolios differently.
Lack of Valuation Anchors
Due to the relatively high turnover of style risk premia portfolios required to maintain a constant style risk exposure, valuation is much less informative in understanding the strategy’s historical and future performance. For example, a portfolio of securities built to gain exposure to the ‘momentum’ risk premia at a given valuation point will most likely not hold the same securities six months later. As such, comparing valuations has a very low predictive power in explaining returns. The inability to use valuations is a consternation to many investors. JANA’s own view is that lack of valuation for timing and risk management is mitigated by many style risk premia portfolios investing across multiple style risk premia on an equal risk basis, limiting the concentration risk of one style risk premia driving risk and return.
Manager Selection and Diversity
Style risk premia portfolios rely upon strong risk management, implementation and execution capabilities. How these alternative betas are selected, weighted, risk managed and traded will result in large performance dispersion across strategies. Reviewing performance over short time periods will likely lead to incorrect conclusions on the skill of the manager as the breadth of investment universe, risk factors included, and volatility targeted will likely be the larger driver of short term performance dispersion than implementation skill versus similar funds. For this reason, we recommend investors have a long-term horizon focusing on managers with proven alpha investment capabilities, resources and ongoing commitment to the evolution of the product. We also recommend clients diversify across a few managers to limit manager concentration risk for the very reasons listed above.
Due to its underlying characteristic of low correlation to major asset classes, style risk premia portfolios usually reside within an ‘alternatives’ asset class. For investors that demarcate between growth and defensive alternatives, the answer depends on overall tolerance for volatility, whether these portfolios form part of a diversified portfolio/asset class or a standalone allocation and their expected role within the overall portfolio.
Style risk premia portfolios are offered with varying levels of targeted volatility with most offerings ranging between 6% p.a. and 12% p.a. (Volatility is calculated using monthly annualised numbers). This compares to historical volatility of equities of around 13-15% p.a. and bonds of 3-5% p.a.
As mentioned previously, we consider style risk premia portfolios to have maximum drawdowns significantly less than equities and to not occur concurrently with drawdowns of either equities or bonds, excluding those caused by pronounced liquidity shocks. Therefore, without any other considerations, style risk premia portfolios that target volatility less than 7-8% p.a. could conceivably be included as part of a ‘defensive alternatives’ asset class.
However, the decision to allocate style risk premia to either the growth or defensive asset bucket includes not only volatility, but also correlation with existing investments and the expected role of defensive alternatives within the broader portfolio. For investors likely to allocate style risk premia on a standalone basis to their ‘defensive alternatives’ asset class, volatility is an important consideration. But if the allocation is part of a diversified portfolio of alternatives, then consideration of their correlation or diversification to the existing portfolio and contribution to overall volatility should be reviewed. This may feasibly result in higher volatility offerings being incorporated in the ‘defensive alternatives’ asset class if the contribution to overall risk and return is beneficial.
Lastly as already stated, style risk premia funds should not be expected to provide a smooth and consistent return profile. While their objective is to target returns above cash, they will experience periods of negative returns when viewed on a monthly and even yearly basis. If investors’ objectives or expectations for ‘defensive alternatives’ is for consistent, positive cash like returns, then we would advise allocating style risk premia funds as part of a diversified alternatives portfolio and not on a standalone basis.
Another potential applicability for style risk premia portfolios is within pension and transition to retirement options as equity substitutes and/or complements. These products offer, depending on targeted volatility, potentially equity like returns over time with lower variability of returns and lower maximum draw down risk.
Style risk premia strategies lie between the spectrum of passive and active investing and are generally lowly correlated to each other and equity and bond markets. As a result of the systematic nature of these strategies, they typically provide higher transparency and liquidity and lower fees than traditional investment diversifiers such as multi-strategy and fund of hedge fund portfolios.
While style risk premia portfolios offer attractive benefits as part of a balanced portfolio, they are not without their limitations. These include complexity, liquidity and crowding effects, cyclicality of performance and a lack of valuation anchors to predict or time exposure.
Depending on the objective of the investor, style risk premia portfolios may form part of the ‘growth’ or ‘defensive’ ‘alternatives’ asset class as well as having a likely applicability for pension and transition to retirement options that focus on capital stability.
Overabundance of objectives?
2021 Outlook: optimism and opportunities in a near normal ...
REITs revisited in 2020 – the listed real estate alternative