Factor investing – an introduction
Academic research and many years of practice have shown that factor-based strategies can help to significantly improve the return-risk profile of a portfolio.
Introduction
Any list of the hottest topics in investment management circles over the last few years is almost certain to contain ‘Factor Investing’.
Sitting between highly active, alpha seeking investing and low-cost, index hugging investments, factor investing is effectively a ‘third pillar of investing’, combining the transparent, rules-based and low-cost nature of passive investing with the outperformance opportunities of active investing.
As such, factor investing strategies are likely to be of great relevance – and value – to financial advisers and their clients.
Factor investing is underpinned by the premise that the performance of a portfolio is often largely attributable to the presence of one or more observable ‘factors’ or characteristics, rather than the skill of the individual fund manager. Examples of these factors – known as factor premiums– include ‘low volatility’, ‘value’, ‘quality’, and ‘momentum’.
Although not without its detractors[1], the quantitative evidence in support of factor investing is overwhelming, and as a result, factor investing continues to grow in popularity, especially with investors for whom cost and transparency are important.
Indeed, a 2018 study[2] found that over 70% of institutional investors were using factor strategies, and more than 60% were planning to increase their use of them in the following years. Furthermore, in a report[3] published in October 2017, Morgan Stanley estimated that almost USD 1.5 trillion were invested in smart beta, quant and factor-based strategies and that assets under management have been growing 17% per year on average since 2010.
The benefits of factor investing are accessible to individual investors too, via a variety of widely available products designed to meet different objectives and which can therefore be tailored to the unique needs of your clients.
But whilst interest in, use of, and discussion about, this third pillar of investing has never been higher, factor investing is no overnight sensation, and the genesis of the factor approach can be traced back more than four decades, to academic studies from the 1970s.
In this article, we will discuss the origins of factor-based strategies and examine in more detail the evolution of specific factor premiums. We will explain the terminology of factor investing as well as exploring the effectiveness of factors in driving performance and risk management outcomes within portfolios.
What is factor investing?
At the heart of factor investing is the identification of discrete and common characteristics which explain differences in returns between equities (and bonds), and which can be harnessed via a rules-based approach to deliver higher returns, greater diversification, and lower risk, over time.
Selecting stocks and building portfolios which exhibit these characteristics – called factor premiums – can add significant value to investors over the longer term.
Whilst the last few decades have seen researchers propose literally hundreds of factors, most of these have not proven particularly robust, and only a relatively small number have achieved universal acceptance and attention amongst experts. Figure 1 below defines the four most widely accepted factors, whilst Figure 2 illustrates the extent to which these factors have been proven to drive out-performance over a five-decade period in the US.
A fifth factor subscribed to by some, but not all, fund managers is ‘size’, based on the
tendency of bonds issued by companies with little debt outstanding, and small-capitalization stocks, to outperform the market.
Although most discussions – including this article – tend to focus on the five ‘style factors’ described above, experts also acknowledge those well-known and intuitive ‘macro factors’ which can influence securities pricing, including economic growth, inflation, and interest and exchange rates.
The academic origins of factor investing
The existence of factor premiums – and their proven ability to explain higher than market returns – was first documented in 1970s and early 1980s, when academics begun to uncover anomalies in the Capital Asset Pricing Model (CAPM) of risk and return.
The CAPM model – the dominant market theorem for a time – posited that the relationship between risk and return was linear. Simplistically it said those securities which were inherently riskier needed to reward investors in the form of higher returns. The model was developed during the 1960s – a time characterised by a relative lack of data and low computing power – and as such was more theoretical than evidence based, assuming, amongst other things, the existence of rational, perfectly informed investors.
By the 1970s the ability to collect and analyse data was increasing, and several studies were released which showed the CAPM risk/return relationship was much weaker than had been previously accepted, and that other ‘factors’ were in fact responsible for driving investment performance differences.
A brief chronology of these first studies – and the factors they effectively introduced to the world – is shown below.
1972 – low volatility
A Robert Haugen and A. James Heins study[4] showed that less volatile stocks had consistently outperformed more volatile ones over the 1929-1971 period. (Note: whilst this study was based solely on US stocks, Robeco’s David Blitz and Pim van Vliet showed in their award-winning paper[5] from 2007 that this also held true across Europe and Japan).
1977 – value
Sanjoy Basu observed and documented[6] the value factor for the first time in 1977. After ranking stocks according to their price-earnings ratio, Basu found an inverse relationship between the price-earnings ratio of a stock and its return. In other words, stocks featuring a lower valuation tended to achieve higher returns than the CAPM would suggest.
1981 – size
A study by Rolf Banz[7] identified that shares in smaller companies tended to outperform those of larger companies.
1993 – momentum
Jegadeesh and Titman ‘discover’ the momentum factor[8], based on the premise that the outperformers of the recent past are seen as the outperformers of the future. Momentum can be in share price and earnings.
1993 – Fama and French three factor model
In essence a synthesis of the various earlier studies already described, Nobel Prize winning economist Eugene Fama[9] and his researcher Kenneth French proposed the ‘three factor model’, within which asset pricing was shown to be a result of the interplay between three factors; size, value, and market risk.
Unhappy Norwegians – the real-world catalyst for factor investing
As strong as the academic underpinnings of factor investing had been, they remained largely ignored by most investors for several years.
The breakthrough for factor investing came after the 2009 publication of a research report[10] analysing the performance of one of the world’s largest sovereign wealth funds, NBIM, which invests Norwegian oil revenues. Although famed for being one of the happiest nations on earth[11], the GFC had seen the NBIM lose 23% in value during 2008, leaving the fund’s managers distinctly unhappy!
To understand exactly why the fund had performed the way it had, NBIM management commissioned a study by high profile business academics Andrew Ang, William Goetzman and Stephen Schaefer. Their study showed that approximately 70% of all active returns (alpha) since NBIM’s inception in 1998 could be explained by implicit exposures to factor premiums and therefore did not reflect true investment management skill. The analysis also highlighted that these factor exposures were merely a by-product of the bottom-up security selection by the active managers NBIM had hired and not a deliberate investment decision.
The authors recommended NBIM to begin using a top-down approach to intentionally obtain strategic factor exposures and to examine how the individual factor premiums could be harvested in the most efficient manner. After this research was published, strategic allocation to factor premiums was dubbed by some as ‘the Norway model’.
Escaping the zoo – what makes a factor?
In recent years, the combination of rising computing power and greater data availability has led to a dramatic rise in the number of market anomalies reported by academics. Purported factors have become so numerous that a number of experts have characterised their ubiquity as a factor ‘zoo’[12].
However, many of these factors subsequently prove to lack robustness, either because they are nothing more than different ways of measuring the same phenomenon, or because they only work over short periods of time or in limited market segments.
Most experts agree that that it is possible to bring the number of anomalies included in the zoo down to a handful of relevant factors, which consistently perform over multiple time periods and across markets.
According to Robeco, one of the pioneers of factor investing, a factor should meet the following criteria to be regarded as relevant:
- Performing: show strong premium with superior risk-adjusted returns;
- Proven: surmounted attempts for falsification (within academia and in-house research);
- Persistent: observable in different markets, stable over time, robust to different definitions;
- Explainable: have an economic rationale with strong academic underpinnings;
- Executable: implementable in practice; e.g. survive after trading costs and other market frictions.
Investing via factor-based strategies
In general, a portfolio diversified along factor lines can reduce risk and enhance return potential over the longer term when compared to the broader market. The rules-based approach to generating superior performance is generally achieved at a lower cost than traditional active management.
Investors can access factor investment strategies – which can be active, or index based – in many different ways and can use them to accomplish a range of objectives, spanning high level objectives such as return enhancement, cost and risk reduction, and diversity, as well as more specific objectives in areas such as income generation, ESG integration and regulatory obligations.
Improving returns
As already illustrated in Figure 2, above, there is a substantial body of evidence proving that stocks exhibiting value, momentum and quality factor characteristics achieve higher returns over the longer term. This has also proven true for corporate bonds with attractive value, momentum and size factor characteristics, as shown in Figure 3 below.
Lowering risk
The low volatility factor is grounded in empirical evidence that securities generating stable returns relative to the broader market have achieved higher risk-adjusted returns than riskier ones over the longer term[13]. Various studies have confirmed this effect holds true in equity markets across the world and also in other asset classes, in particular the corporate bond market.
Increasing diversity
Diversification is one of the most fundamental risk management principles, and asset owners have long applied this principle by dividing their holdings across different asset classes and regions. But the dramatic increase in correlations between asset class returns during the market turmoil of the 2000s cast doubt on the benefits of traditional diversification frameworks, and many investors have turned to factor investing in the quest for more robust diversification techniques.
Various empirical studies have demonstrated the superior diversification benefits of factor investing, compared to classic diversification.
For instance, a 2012 paper[14] by Antti Ilmanen and Jared Kizer analysing data on several asset classes dating back to 1927 reported that diversification into and across factors has been much more effective in reducing portfolio volatility and market directionality than traditional asset class-based approaches
Smart beta
One popular way to accessing the proven benefits of factor investing is via index-based factor strategies, known as ‘smart beta’. Smart beta strategies explicitly target factor premiums and represent an alternative to traditional market capitalization-weighted indices (beta).
Motivations to use smart beta and factor strategies
A 2019 survey[15] of European investment professionals, conducted by EDHEC Risk Institute, found the most important motivation behind the adoption of smart beta and factor investing strategies is to improve performance. On a scale from 0 (no motivation) to 5 (strong motivation), respondents gave an average score of 3.76 to ‘Improve performance’. ‘Manage risk’, which is in second position among key motivations (score of 3.25), is also an important element of choice when it comes to smart beta and factor investing strategies (see Figure 4).
Single factor v multi factor
Another key decision for factor investors to make is whether to invest in a specific factor (a single factor ‘tilt’) or across multiple factors. As always, the answer is very much dependent on the investment objectives being pursued.
Value strategies are a good example of how single factor strategies are neither unusual nor new. For decades, prominent investors have advocated buying securities trading below their intrinsic value and many active managers have been offering so-called value strategies[16]. With the advent of factor investing, many investors have turned towards this kind of approach, as a systematic and cost-efficient way to achieve the kind of exposures they were previously seeking with ‘fundamental’ strategies.
Notwithstanding the legitimacy of single factor strategies however, there is a strong body of evidence suggesting that multi-factor strategies are particularly effective.
A research paper[17] by Joop Huij and Eduard van Gelderen analysed the returns of US equity mutual funds over the 1990-2010 period and found large differences between the funds with significant exposure to one or more proven factors and those without factor exposures. Only 20% of the funds with no exposure to factors yielded outperformance (relative to the market) in the long run. For funds that did have significant exposure to proven factors, this figure was substantially more favourable, ranging from 51% for single factor funds to 68% for two-factor funds, and 78% for three-factor funds.
Assisting, not replacing, active managers
Rather than undermine the importance of active management, factor investing should be thought of as adding to the investors’ armoury and would typically be an approach used alongside other strategies.
In this regard, the aftermath of the previously discussed NBIM Fund report is instructive.
Remembering that the initial report found that 70% of fund performance could be attributed to factor premiums, the impact of active investment management was therefore still recognised as substantial.
Subsequent to this report, the NBIM Fund adopted a formal factor-based investing approach based on size, value and growth factors. A follow-up report[18] noted that the Fund’s factor approach became just one facet of a multi-pronged approach, which also included security selection by way of external active managers and internal programmes.
Put another way, the cost and effort savings accrued from applying a factor-based strategy can be re-directed towards seeking truly complementary active management to harvest returns from specific investment insights and idiosyncratic risk, thus helping achieve even higher performance.
The future
Whilst the empirical evidence supporting the efficacy of factors is substantial, our understanding of factors continues to evolve, in line with an ever-growing data set, and increasingly sophisticated- AI based – data mining techniques.
Research continues apace, and whilst the factor zoo continues to welcome – then farewell – many exotic new species, most studies subsequent to those seminal works discussed earlier have strengthened the basis for the core factor premiums, reinforcing their robustness.
Notwithstanding this, not all factors are created equal, and some have proven more powerful than others over specific time periods and in different regions of the world.
Real world, once in a lifetime, events such as the Covid 19 pandemic will continue to challenge and inform our understanding of investment markets. But whilst the exact impact of Covid 19 is yet to be fully understood, the intellectual basis for factor premiums is built on a global data set spanning the Great Depression, two world wars, the Spanish Flu, the GFC and multiple tech-bubbles, and as such, investors should be confident in their continued value and relevance in an increasingly complex world.
Conclusion
Academic research and many years of practice have shown that factor-based strategies can help to significantly improve the return-risk profile of a portfolio, for example by reducing downside risk or enhancing long-term returns. As a result, the popularity of factor investing continues to grow.
Factor-based strategies can be active or passive and are easily accessible to individuals via a variety of products which can be tailored to meet each investor’s unique objectives.
Financial Advisers who understand the academic basis of factor investing, and the options available to access a factor-based strategy, either as a standalone approach or to complement other strategies, will be much better equipped to help their clients achieve their investment and lifestyle goals.
Read Part two: CPD: Factor investing in action, across regions, cycles and asset classes
———–
You must be logged in to post or view comments.