Cass Business School and Invesco PowerShares explore dynamics of smart beta

Nov 18th, 2015 | By | Category: ETF and Index News

(John Maher, CFA contributed to this article)

Cass Business School (Cass), in co-operation with Invesco PowerShares, a leading global provider of exchange-traded funds, has released a series of four research papers, examining the origins, sources of returns, opportunities and due diligence challenges of smart beta.

PowerShares and Cass Business School explore dynamics of smart beta investing

The results of the Cass Business School smart beta study confirmed the robustness of smart beta as a superior investment style when compared to traditional market cap-weighted approaches.

In a recent survey on smart beta investing, conducted by index provider FTSE Russell and aimed at US financial advisors, the most commonly cited reason for not using smart beta within client portfolios was a lack of product knowledge (42% of respondents). Independent research and educational initiatives between industry and academia, such as the PowerShares/Cass collaboration, are therefore imperative to enhance investor understanding of smart beta, which, in turn, will ultimately lead to more investors having the necessary confidence to deploy smart beta tools more comprehensively.

Bryon Lake, Head of Invesco PowerShares – EMEA, commented: “Smart beta is becoming an increasingly popular component in investors’ portfolios and these products are becoming ever more mainstream. We were pleased to link with Cass Business School and believe that these papers will help inform and educate investors in how smart beta can be used effectively within portfolios.”

Paper 1: The origins of smart beta.

The first paper looks at the origins of smart beta. The establishment of the Dow Jones Industrial Average in 1896 may be considered to be the birth of the modern stock market index; however, it was not until decades later, with the introduction of Harry Markowitz’s ‘Modern Portfolio Theory’ in the 1950s and, Eugene Fama’s ‘Efficient Market Hypothesis’ (EMH) in 1970, that the notion of so-called ‘passive’ index-tracking investing became increasingly mainstream.

Although there are considered to be various forms of the EMH, each with differing degrees to which the theory may apply in reality, the basis of the notion implies that asset prices fully reflect all available information. One of the direct implications is that it is impossible to consistently ‘beat the market’ on a risk-adjusted basis. To earn returns superior to the market, an investor would need to hold a portfolio that is riskier than the market. This relationship between the risk-return profile of a portfolio of assets relative to the risk-return profile of the market portfolio became known as ‘beta risk’.

Although the EMH was a ground-breaking theory, evidence soon began to emerge that questioned its key predictions. Experimentation with alternative rules-based investment strategies seemed to produce consistent risk-adjusted returns above what could be expected from exposure to ‘beta risk’. This sparked a body of research into discovering ‘alternative betas’, the birth of smart beta, as we know it today.

Paper 2: Sources of return in smart beta strategies.

The second paper from Cass examines the sources of return of eight major rules-based smart beta strategies, back-testing the results to 1969, and investigating the statistical significance of each policy to further confirm their reliability. The figures used for the test were the returns on the 500 largest US stocks derived from the Centre for Research in Security Prices. Some of the weighting strategies being tested against a traditional market cap-weighted approach included equal-weighted, diversity-weighted, volatility-weighted, equal risk-weighted, minimum variance, maximum diversification, fundamentally-weighted, and risk efficiency-weighted approaches.

The results were striking with every alternatively-weighted strategy outperforming the market cap-weighted approach in terms of mean return per annum. Furthermore, most approaches reduced the risk of their portfolios with equal-weighted, diversity-weighted, and risk-efficiency strategies being exceptions. Overall, every strategy produced larger Sharpe ratios than that of the market cap-weighted profile, indicating superior risk-adjusted returns.

Responding to potential critics of the back-tested data who may question the relevancy of the results in practice due to the absence of transaction costs, Cass generated the turnover statistics related to the annual rebalancing for each strategy. The strategy with the highest turnover was the maximum diversification strategy where, on average, 47.9% of the portfolio was required to be sold each year. Cass then estimated the corresponding transaction costs (e.g. bid-ask spreads) required to lower the superior risk-adjusted return of each strategy to that of the market cap-weighted strategy. The lowest average transaction cost estimated was 1.6% for the maximum diversification strategy, ranging upwards to 10.7% for a fundamentally weighted approach. To provide a frame of reference, the quoted bid-ask spread on the Vanguard S&P 500 Growth ETF, as of 18 November 2015, was 0.07%. That is to say, all implied bid-ask spreads were implausibly high, suggesting that each strategy maintained significant merit in a real-world application that considered trading costs.

The paper thereafter regresses the returns of each strategy against the Fama-French three factor model, enhanced with the additional momentum factor. The aim of this step being to separate the sources of smart beta return into ‘market’, ‘size’, ‘value’, ‘momentum’, and ‘residual’ factors.

Every strategy responded favourably to a ‘value’ tilt with the minimum-variance portfolio adding a considerable 1.62% per annum relative to a market cap weighting.

The findings also illustrated that almost every strategy enhanced their return through adding a ‘size’ factor to their profile (the exception being a minimum variance portfolio which sacrificed 0.01% return per annum). The approach to most successfully harness the size factor was an equally weighted strategy which gained a further 0.35% per annum over market cap weightings through this exposure.

Most strategies responded negatively to the momentum factor with only the minimum variance portfolio and maximum diversification portfolio showing marginal gains. Every strategy maintained a positive ‘residual’ factor, suggesting there were yet to be identified factors contributing to the superior return of their approaches.

Paper 3: Combining investable smart beta strategies for portfolio optimisation.

The third Cass paper looks at nine smart beta strategies that have been transformed into investable indices (those chosen by Cass for the study were all S&P Dow Jones Indices), examining whether a combination of these indices may raise the risk-adjusted return of an investor’s portfolio compared to a single-factor strategy. The strategies tested included well-known varieties such as equal-weighting, small cap, quality, low volatility, value, momentum, dividend yield, growth and low beta investing.

The paper compares the maximum drawdown figures of each strategy with that of a market cap-weighted approach. All performed relatively similarly to the market cap-weighted approach apart from the low volatility strategy whose maximum drawdown of 35.4% was significantly less than the maximum drawdown of 50.9% returned from a market cap-weighted strategy.

The paper thereafter investigates the strength of combining strategies within a portfolio. A portfolio containing an equal dollar investment in each of the nine strategies as well as a ‘risk balanced’ portfolio, whereby the weighted volatility of each of the nine components is equal, were created. The results showed that both of these combination approaches produced superior absolute and risk-adjusted returns when compared to the traditional market cap-weighting approach; however, the results of both were very similar to a simple ‘equal-weighted’ smart beta strategy. This suggests that smart beta approaches may successfully be combined but, in these examples, do not provide significant benefit over a single smart beta strategy.

Given that this was a relatively simple method of combining strategies, the paper sought out to address whether there may be a combination that does provide a significant benefit to the investor. The paper investigated the outcome of allocating to each of the single strategies based on the direction and magnitude of each index’s price movement over the past six months, otherwise known as momentum. Two strategies were trialled in this regard. In the first instance, a portfolio was formed of five of the original smart beta indices, selecting those strategies that were exhibiting the strongest momentum signals, and assigning an equal weighting to each. The portfolio was rebalanced on a monthly basis. The second method entailed an equal investment in all nine original strategies but would switch any of the nine investment parts to a cash equivalent holding if that strategy was not exhibiting sufficient momentum.

While both strategies displayed superior risk-adjusted return profiles when compared to any of the original smart beta strategies, the second method scored significantly higher. This was due in part to a significantly reduced maximum drawdown statistic of only 13.7%, compared with 46.3% for the first momentum strategy and 50.9% for the market cap-weighted approach. By protecting investor wealth during significant market downturns, the second strategy was able to significantly improve the risk-adjusted return over singular smart beta strategies.

Paper 4: Monitoring smart beta investment strategies.

In the fourth paper produced by Cass the due diligence requirements of smart beta investing relative to an active fund management approach are investigated. The paper highlights that, in many ways, the due diligence of smart beta is less daunting than that required when investing with an active asset manager. The investment process is generally very transparent and based upon a set of unchanging rules. Secondly, the majority of these strategies have been tested over a diverse range of markets and varying environments, adding a level of familiarity and reliability to its processes. Lastly, as the ‘fund manager’ in terms of a smart beta investment strategy is the strategy itself, investors need not worry about the human element which relates to active managers: such as behavioural biases, changes in life circumstances, human errors, and even relocating to another company.

To further ensure the robustness of their investment strategy, investors may wish to qualify the administrator of their benchmarks. In 2013, the International Organisation of Securities Commissions (IOSCO) published a report entitled Principles for Financial Benchmarks. The report covered key areas such as governance, benchmark quality, methodology, and accountability. For further security it is recommended that investors ascertain whether their administrator has committed to a code of standards similar to that produced by the IOSCO. Lastly, although there is no need to worry about managerial biases, regular scheduled reviews of the chosen strategy may be appropriate.

Opportunities abound, but education is key

In conclusion, Cass appears to have validated the statistical significance of the superior return generated through smart beta investment strategies when compared to conventional market cap-weighted approaches.

Additionally, the insight into factor combining points to a potential rich new seam of product innovation and development if fund providers move beyond suites of single-factor exposures and launch multi-factor strategies, particularly if factor-timing models are introduced.

All in all, the outlook for smart beta appears strong, but, the oft-used mantra of ‘education, education and education’ will prove to be key if smart beta is to fulfil its potential.

Tags: , , , , , , ,

Comments are closed.