Join our community of smart investors

Beta Blocks

James Norrington takes a look at the factors you can harness to build smarter portfolio
June 12, 2015

The financial services industry is rife with acronyms and buzz words, and with the increasing prominence of passive strategies ‘Smart Beta’ is another term that has become part of investors’ vocabularies. An exact definition isn’t easy to pin down but essentially smart beta has been used to describe indices which use alternative weighting methodologies with the aim of achieving better risk-adjusted returns and specific objectives such as reduced volatility or diversification. The phrase has caught on and is also being used by the ETF industry to market new products which track indices with strategic tilts towards return factors such as value, momentum, quality, income or size.

Before looking at what smart beta means for investors, it is worth recapping how financial indices are constructed and used. Indices are useful to track the fluctuations in price and returns of different asset classes over time and how these compare with broader economic data such as GDP and inflation. As well as providing benchmarks to judge the performance of investment portfolios, financial indices are frequently replicated by providers of mutual funds and exchange traded products to give investors a whole of market play on the returns of the underlying assets. The advantage of buying an underlying market index, or ‘investing to the benchmark’, is that investors can eliminate the idiosyncratic risk of individual securities, so that all returns are linked to beta – the market risk that cannot be diversified away.

For asset classes such as bonds and equities, markets are made up of many investments with their own beta, ie sensitivity to systematic market risk. The way an index is constructed impacts the level of exposure to different securities and hence an investor’s returns. In the case of the FTSE 100 for example, the index is weighted by market capitalisation which means an investment tracking it is skewed towards the biggest companies. Alternatively, an equal-weighted version of the FTSE 100 would give the same exposure to each company in the index regardless of size.

Different weighting methodologies are now being used to target diversification benefits, liquidity or volatility reduction. In addition, index selection criteria can be set to give investors a strategic tilt towards specific return factors. For example selecting companies on the basis of price-to-book is one factor used to achieve a value tilt. Index providers dislike the term smart beta and are keen to distinguish ‘factor indices’ which, although often alternatively-weighted, provide a sharper focus on factor returns. While many investors are probably not concerned with semantics, of more interest will be new opportunities to profit from factors that in some cases have hitherto been inaccessible for reasons of expense or practicality.

 

Factors that beat the market

The index provider MSCI has identified six risk premium factors with proven results: value, low size, low volatility, high dividend yield, quality and momentum. Integrating Barra’s risk-factor model (see box below) with MSCI’s granular global securities data enabled 40-year back-testing from 1975 to 2015. In this period all of MSCI’s factor targeting indices achieved a premium performance compared to their parent index (the MSCI World index). The fluctuations in the chart below show that there have also been periods of considerable underperformance by different factors. However, there are negative correlations that investors can exploit to diversify risk.

 

Two of the best attested return factors are value and momentum. Both have outperformed the stock market over time and historically have been negatively correlated, so they provide a fine illustration for the potential of factor-optimised investing. As the MSCI graph shows, value has been one of the least volatile factors overall so contrasts nicely with momentum, which has experienced some nasty troughs. The inverse relationship between value and momentum is the subject of a study by Asness, Moskowitz and Pedersen (Asness et.al. 2013) which, from a survey conducted between January 1972 and July 2011, found -0.53 correlation between the two factors in the US and -0.52 globally.

Investors Chronicle readers will of course be familiar with the behavioural concepts that underpin value and momentum. In the case of value, the tendency to extrapolate past disappointing news can lead to the market undervaluing shares with solid fundamentals. For momentum the cognitive bias of under-reaction to good news about companies has often led to continuing strong gains as the market catches on. There exists a significant body of empirical research in support of both factors.

The groundbreaking three factor equity return model of Eugene Fama and Kenneth French (1992) used comprehensive US stock market data from July 1963 to December 1990 and found strong evidence for the predictive importance of the Price to Book ratio. Examining more recent data from 1990 to 2011, the same authors demonstrated value premiums based on Book to Market Value ratios in four developed market regions: North America, Europe, Japan and Asia-Pacific. Other studies have also proven the predictive powers of value factors such as price to cash flow.

In the case of momentum, the work of Jegadeesh and Titman (2001) updating their initial 1993 study, demonstrated that between 1965 and 1998 the best performing US shares over six month periods consistently outperformed the worst performing shares over the next six months. The momentum phenomenon is not restricted to just the US as the UK momentum portfolio of Investors Chronicle’s Chris Dillow has shown over the past decade. Recently, in the 2012 study cited above, Fama and French added momentum as a fourth equity return factor and showed it to work in all major developed regions apart from Japan.

In the past, investors faced difficulty getting exposure to different return factors due to the cost of frequent trading and the need for access to high quality data. These barriers can now be overcome with the next generation of low cost ETFs that replicate indices with specific factor tilts.

Last year, BlackRock iShares launched a group of London-listed ETFs tracking the MSCI Factor indices and investors can expect to see more providers offer more smart-beta products. Going a step further, there are also products emerging that combine factor strategies such as value and momentum. The First Trust United Kingdom AlphaDEX UCITS ETF is one of the first to use a quantitative stock selection strategy in this way.

The drawback of using products with a very short period of time behind them is the lack of a performance overview that covers periods of market stress and reflects how the average annual return has measured up to established benchmarks over time. That said, with the weight of research covering a number of decades, there is certainly no lack of support for the factors that the products are targeting. The performance of UK-listed, factor-targeting ETFs goes back only months and there is barely more data for US products. However, it is worth noting that a 50:50 split investment in iShares’ US Value and Momentum ETFs would have made 14 per cent compared to 12 per cent from the S&P 500, between April 2014 and April 2015.

 

Constructing a Smart Beta portfolio

There are now indices with tilts towards value, momentum, quality and size giving ever-widening scope for ETF providers to launch innovative products. As opportunities arise it is important to consider the underlying reasons to believe any given factor will continue to deliver risk-adjusted returns; the volatilities and correlations among different factors; and the levels and periods of drawdown* that factors have experienced historically. With any portfolio strategy, investors need to consider their personal objectives, tolerance for risk and ability to sustain any losses over the short-term. Negatively-correlated factors, as in the value and momentum example, can diversify away risk to an extent but investors should be conscious that some factors are cyclical and have delivered periods of significant underperformance. It is also worth noting that in a balanced portfolio, factor-led strategies should not necessarily be to the exclusion of market-cap index investing which has the advantage of reflecting more fully the available opportunity set of equity investments as well as the aggregate holdings of all investors.

ETF investors also need to be mindful of liquidity. Some more esoteric smart beta strategies are likely to be more thinly traded and therefore introduce the cost of wider spreads. Index methodologies such as equal weighting also by definition increase exposure to smaller and less liquid underlying stocks. In addition, alternatively-weighted indices require more maintenance in terms of rebalancing trades which contributes to generally higher charges for smart beta products.

While rules-based strategies are mainly described as passive, with important risk diversification and product specific considerations, there remains plenty of scope for active management especially at the portfolio level for smart beta investors. Managers can add value by monitoring risk-adjusted returns against benchmarks and timing cyclical factor exposures. However, the returns that can be achieved with largely passive strategies does bring the issue of charges sharply into focus. A report for iShares by Ronald Kahn and Michael Lemmon estimated that 35% of performance by international equity managers between April 2011 and March 2014 was attributable to smart beta factor exposure. Greater transparency on the source of returns should advance the cause that where gains are due to smart beta factors, then, in terms of fees, investors should be paying lower smart beta prices.

*Please note drawdown in this context refers to peak-to-trough price falls, not pension income drawdown.

 

Well-known systematic return factors

Systematic factorsHistorical risk premium Historical factors to target
ValueCaptures excess returns to stocks that have low prices relative to their fundamental valuePrice to book, earnings to price book value, sales, earnings, cash earnings, net profit, dividends, cash flow
Low size (small cap)Captures excess returns of smaller firms (by market cap) relative to larger counterpartsMarket capitalisation (full or free float)
Momentum Reflects excess returns to stocks with stronger past performanceRelative returns (3-month, 6-month, 12-month, sometimes with last 1 month excluded), historical alpha
Low volatilityCaptures excess returns to stocks with lower than average volatility, beta and/or idiosyncratic riskStandard deviations (1-yr, 2-yrs, 3-yrs), downside standard deviation, standard deviation of idiosyncratic returns, Beta
Dividend yieldCaptures excess returns to stocks that have higher-than-average dividend yieldsDividend yield
Quality Captures excess returns to stocks that are characterised by low debt, stable earnings growth, and other "quality" metricsReturn on equity, earnings stability, dividend growth stability, strength of balance sheet, financial leverage, accounting policies, strength of management, accruals, cash flows