Monday, January 29, 2018

Your Complete Guide to Factor-Based Investing

If you’ve ever wondered whether you should be taking advantage of the historically above average returns of small stocks, value stocks, momentum stocks, and other types of market anomalies, Your Complete Guide to Factor-Based Investing, by Andrew Berkin and Larry Swedroe, is the book for you. It’s based entirely on “evidence from peer-reviewed academic journals,” and it helped me focus my thoughts on the degree to which I want to pursue factors.

The authors begin with a treatment of the seven factors they consider “worthy of investment”: market beta, size, value, momentum, profitability and quality, term, and carry. For each of these factors they discuss persistence, pervasiveness, robustness, whether they are investable given real world concerns such as trading costs, and whether there are logical explanations for the existence of above-average returns.

In the case of size and value factors, “While small-cap stocks as a whole have provided higher returns (the size premium), small-cap growth stocks have produced below-market returns.” This is the reason why I invest in the exchange-traded fund VBR; it contains only small value stocks.

Financial institutions “are momentum traders, while private households are instead contrarians taking the other side.” If this is representative of who makes money from momentum and who loses, the ever-shrinking role of private households as direct stock investors would seem to point to a reduced momentum premium in the future. But I’m not making any predictions.

“While momentum has offered investors the highest risk-adjusted return of all the factors we have discussed, it also has a ‘dark side’ – it has experienced the worst crashes.” That’s enough to scare me off momentum investing, particularly now that I’m retired.

The carry factor consists of assets that yield a high return if the asset’s price remains the same. One example is choosing to hold the currency of a country with high interest rates. “Carry can be like picking up nickels in front of a steam roller. It has been profitable of the long term, but one must be sure they can handle being run over every so often.” That leaves me out.

After going through the investable factors, the authors address what has happened after factors are published. “You need to be aware that the publication of research on anomalies does lead to increased cash flows from investors seeking to gain exposure to their premiums, which can then lead to lower future realized returns.”

Throughout the book, each factor always gave impressive-looking returns, but the section on implementing a diversified portfolio of factors gave a single example where using factors boosted market returns by 0.3% annually, and standard deviation dropped by 1.8%. This shows that those readers who don’t understand the math behind the analysis can be fooled by seemingly big factor returns.

The book contains ten appendices, including one on smart beta: “while there really is in fact such a thing as smart beta, much of it is nothing more than a marketing gimmick – a repackaging of well-known factors.” Other appendices explain why dividends are not a factor, and why other possible factors were rejected based on the authors’ criteria.

In an appendix discussing the default factor for corporate bonds, the authors point out that “Like many risky assets, high-yield debt does not have a normal return distribution.” This criticism extends to virtually all of the academic analysis this book is based on. Stock returns don’t follow the normal distribution, making discussions of volatility, Sharpe ratios, and the entire capital-asset pricing model (CAPM) suspect. It’s a lot to ask, but I’d like to see factor analysis using a pricing model built on a stable distribution that better follows the fait tails we actually see in stocks returns. No model is perfect, but the normal distribution underestimates the probability of extreme events by such a huge margin that it treats extreme events essentially as though they can’t happen at all, which is dangerous.

Some important information for factor investors didn’t appear until the seventh appendix. “If a portfolio already has exposure to market beta, size, and value, adding exposure to momentum cannot contribute that much more in the way of incremental returns.” So, each factor looks impressive on its own, but combined they are less than the sum of their parts. “When adding exposure to additional factors, you may also increase the turnover of the portfolio, raising trading costs and reducing tax efficiency.”

The final appendix lists mutual funds and exchange-traded funds that give access to the investable factors. Some of the funds have quite high expense ratios. This represents costs that you pay whether the factors deliver higher returns or not. If we add other fund costs such as trading costs to the expense ratios, and we eliminate funds whose costs are above, say, 0.25% per year, some factors no longer seem investable.

Overall, I found this book very helpful in understanding factor-based investing. However, it’s not for the casual reader. A decent grasp of CAPM is needed to follow the discussion well. I don’t know if I’ve chosen the right level of factor tilts for the stock portion of my portfolio (80% market beta and 20% small value), but I’ve made my peace with my choices.


  1. "I don’t know if I’ve chosen the right level of factor tilts for the stock portion of my portfolio".

    I'm guessing it doesn't really matter that much if your tilts are off a wee bit.

    1. @Larry C: Agreed. I think it's likely a bigger mistake to go all in with factors than it is to ignore them altogether.

  2. More giggles. Swedroe has little credibility.

    It's funny when some ETF funds are larger than the underlying market that they're supposed to represent. Do people really understand that they're now using options on a small market?

    1. @Anonymous: Once again I disagree. Credibility is a rare thing in the financial industry, but I find Swedroe gives an honest assessment of the research and the potential pitfalls of projecting the research results forward. He might draw different conclusions than I do about the best path forward, but I find his analysis useful.