Traditional active management has been receiving increasingly tough press thanks to its high costs and apparent failure to outperform index tracking funds over the long run. But one specific brand of active investing has been gaining more attention for its ability to do exactly the opposite – provide long-term outperformance at a low cost. So called ‘factor investing’, often marketed under the guise of ‘smart beta’, ‘strategic beta’, or ‘fundamental indexing’, has been heavily researched by academics for decades, and certain specific factors have been discovered which have outperformed market-cap weighed indices. Thanks to its growing popularity, many financial institutions have jumped on the factor investing bandwagon and launched products designed to capture factor exposure.
At its most basic, factor investing is selecting stocks which share a certain characteristic. For example, buying only stocks with low price relative to the value of a company would be a form of factor investing, and is the basic idea behind ‘value’ investing. Buying stocks that have risen the most in price over the last 12 months would also be a form of factor investing, and is the basis of ‘momentum’ investing. They are, at their heart, quantitative strategies that don’t rely on anyone’s qualitative assessment of a company. There’s no portfolio manager picking stocks based on their growth prospects, their barriers to entry, or management quality. Instead, portfolios are created to include only companies which share a fundamental characteristic. A portfolio of companies with low P/B ratios would be a simple value factor portfolio. As a result of needing less human intervention, factor investing is usually cheaper than traditional active investing.
Factor investing has fascinated academic researchers since the 1980s, and much research has been dedicated to finding the factors that provide the best chance of outperforming a market-cap benchmark. Whilst there’s still a huge amount of research being conducted on an almost infinite number of possible factors, a small number of factors have been shown repeatedly in academia to outperform a market-cap weighted benchmark over a long period of time, across asset classes, and across geographies.
I’ve provided an introduction below the most popular of today’s factors, which tend to be those that have withstood the most rigorous academic testing, showing long term-outperformance among many different asset classes, geographies, and time periods, as well as possessing explanations for why the factors are likely to persist going forwards. At the end of the article you’ll find a selection of evidence supporting each factor. It’s by no means supposed to be an exhaustive list of all the research available, but merely a starting point based on papers/articles I’ve personally found the most useful when navigating my way through the factor zoo (full disclosure: I haven’t read every single one cover-to-cover).
One of the first factors to be discovered was the ‘value’ factor. Value investing involves buying stocks that are cheap relative to the underlying business – i.e. the manager’s opinion of what the business is worth is more than what the market currently thinks. At its simplest, value in investing is bargain hunting. Quantifying how much of a bargain a stock looks to be is usually done using ratios. For example, stocks that are considered ‘value’ stocks tend to have low price-to-something ratios (e.g. P/E, P/B, P/S, P/FCF) and/or high dividend yields. The idea was that that the stocks would be bought when they were cheap (the market for whatever reason hasn’t fully recognised the true value of the company) and they would be sold when they became fairly valued. It’s a simple and intuitive method of stock selection which has been around forever, and which is still incredibly popular today. It’s been subject to intense scrutiny ever since its discovery, and has managed to withstand the pressure, remaining one of the major factors that investors try to harvest today.
Alongside value and market beta, size was the other factor that made it into the original Fama French 3 factor model. The size premium was associated with buying smaller, riskier companies, as the increased risk of buying smaller, less liquid, companies was thought to provide extra returns over the long run. Following research subsequent to the publication of the original Fama French papers, the size factor was found to be less robust than originally thought, and many have questioned its very existence. It has, however, undergone a revival more recently with the qualifier that the size factor may be effective only once a ‘quality’ screen has been applied to the small-cap stocks – the small cap effect only works once the ‘junk’ stocks have been filtered out. In the words of the researchers at AQR, ‘Size Matters, If You Control Your Junk’.
Momentum investing involves buying stocks that have gone up the most in price over a defined period. Momentum can be either time series (aka ‘absolute’ momentum: buying stocks that have gone up the most over the last X months), or cross sectional (aka ‘relative’ momentum: buying stocks/asset classes that have gone up the most versus other stocks/asset classes). Academics are still debating the reasons behind why stocks that have risen in the past tend to continue to rise, but most agree that there is at least some behavioural element. Investors have been shown to hold onto losing stocks for too long, sell winners early, prefer gains disproportionately more than disliking equivalent losses, and underreact to market news. There is also some evidence that momentum investors are being rewarded for taking on extra risk, in the form of infrequent but large crashes. Whatever the reasons behind the momentum premium, it’s one the most researched factors and has been found to be pervasive over many different geographies, asset classes, and factors, with backtested data going back over 200 years.
Following the discovery of the CAPM model, which argues that increasing portfolio returns can only be achieved through increasing the risk of the portfolio, researchers began to discover that this might not actually be the case. Increasing risk didn’t appear to correlate with increasing returns, and in fact it was the opposite that appeared to be true – stocks with low betas showed higher returns than those with higher betas. Again, the reasons for the factor’s existence are still debated, but reasons so far include 1) the fact that many investors are unwilling/unable to use leverage, so high-beta stocks become overweighted in higher-risk portfolios, reducing their future returns, 2) compensation structures and internal stock selection processes at asset management firms lead institutional investors on average to hold more volatile stocks, and 3) high beta stocks are overpriced, as investors favour those stocks with lottery-type characteristics. In any case, there is considerable evidence for the factor’s existence.
‘Quality’ investing is the approach of buying stocks that are profitable, growing, well managed, and stable. Quality has long been established as an investment approach, but it is less well researched as a factor, especially when compared with value, size, yield, momentum and low volatility. Its reasons for existing are less well understood – finding a risk-based explanation for why quality stocks are underpriced and junk stocks are overpriced has proved nearly impossible to date. Because of this lack of risk-based explanation, some researchers have suggested that quality be used more as a factor ‘enhancer’ than a standalone factor. For example, introducing a quality screen has been found to enhance both the value factor and the size factor. Whether it’s used as an enhancer or a standalone factor, its existence is well documented in academic literature, and remains one of the major factors that both investors and fund providers are championing.
One of the newer and another one of the less well-researched factors is the carry, or ‘yield’, factor. That is, stocks with high dividend yields have a tendency to outperform those with lower yields. Whilst there is evidence of the strategy working across asset classes, some researchers have questioned whether the factor is just an amalgamation of the value and low beta factors.
A special mention ought to be given to what should probably be considered the original factor. It’s the most basic factor that underpins the CAPM model, and has persevered in almost all pricing theories since the Fama French 3 factor. All students of finance learn the basics of CAPM in their studies, and yet the theory on which it’s based is often overlooked as a factor in its own right. The market factor (or beta), is a bet on the equity risk premium – i.e. it’s a bet that the equity market will continue to outperform government bonds over the long run. From a risk-based perspective it makes sense – equities will always be riskier than bonds, so investors will demand a higher rate of return. Interestingly, this makes all investors, including indexers, factor investors. Passive investors tracking a market index will be betting on the market factor, whether they realise it or not.
“Market cap weighting is the best form of index construction, except for all the others” – Winston Churchill, ish.
Research has found that indices weighted by any other attribute other than their market cap have consistently outperformed regular market cap indices over the long run. As soon as you weight an index by something other than its market cap, it’s likely to outperform. Even indices constructed according to the Scrabble scores of the tickers of the underlying stocks outperformed the equivalent market cap index. And the reason? Factors. As soon as you increase the weighting of smaller companies in an index versus a traditional market cap weighted index, you benefit from factor exposure (notably the size and value factors).
There are a huge number of possible factors in the factor zoo, accompanied by an almost unlimited amount of research. The major factors have all shown signs of historic outperformance, with rationale underpinning the reasons for why the outperformance should continue. Factor investing, however, does come with a whole host of problems, and is far from a panacea to cure all investing woes. I’ve dedicated a series of posts to the problems with factor investing here, and these problems should be fully understood before allocating to a factor strategy. However, for those understanding the risks involved, and who continue to keep costs low, factors remain one of the best chances an investor has to earn long term outperformance. For those interested in introducing factors into a real portfolio, please also refer to this post on multi-factor investing.
A few of the reasons why factors outperform have been mentioned above, but I take a deeper dive into possible reasons why these factors outperform in the next article (click here).
Appendix: Factors – the evidence