Jason is a favorite columnist of mine in part because of his affinity to Ben Graham and value investing, and in part because he is a great guy. His Saturday column, The Intelligent Investor, named after the Graham book that Buffett says changed his professional life, is must reading. But, Jason is a bigger fan of indexing than I am, so it is interesting that his latest column moves him a little closer to his Graham roots.
In his latest column (3/2/13), Jason refers to a paper about to be published by the University of Rochester’s Robert Novy-Marx (RNV). The column:
RNV has written often on the value premium–the item that my investment firm has been trying (with some success) to capture. He has been trying to understand why and where it exists, such as in the following paper that I read a few years ago that relates the value premium to operating leverage (Note: the draft of the operating leverage paper that I read was dated May 18, 2007):
In RNV’s latest paper, which was the impetus for Jason’s column, RNV introduces a quality formula to improve the value premium. It seems to be generating a lot of buzz because Jason wrote that DFA’s founder David Booth and AQR’s Cliff Asness are planning to create funds based on RNV’s quality paper. Booth even called it one of those investing ideas that only come along once every twenty years or so.
But, RNV compares his measures with several tools that my firm has used since day one, including those with a quality component, specifically value investing tools based on the work of Fama, French, Lakonishok, Shleifer, Vishny (see F&F and LSV tab above), Haugen, Piotroski, and Greenblatt.
Fama and French (F&F) were not the first to point out that high book-to-market (BtM) stocks (value) trounced low BtM stocks (glamour) in generating returns, but their 1992 paper brought the issue to the forefront because they are staunch defenders of the Efficient Market Hypothesis (EMH), which their paper seemed to discredit. F&F swiped that cognitive dissonance aside by claiming (“hoping” actually) that the extra returns were compensation for risk (that they did not quantify).
LSV and Haugen later showed that value’s better performance was earned with less risk than the market leaving F&F with only hope. In 2000, Piotroski showed that the performance of the F&F model could be further improved with a nine-point measure of quality based on nine financial statement metrics. Piotroski called it an F score. And Greenblatt used a two factor joint measure of quality and price.
When executing its strategy, my firm has leaned on F&F’s BtM research and on Piotroski’s F score in addition to the research from the others mentioned above. I have always thought of our process as a joint value/quality approach, so I find it interesting that many feel as if this is something new. It may be new in that the research is presented in a new way with data through 2011, but this approach has been practiced by many value investors for a while and RNV’s results are not very different from the results of other Value/Quality practitioners. In addition, I have called my process a sorting process and not a screening process, because we sort rather than screen for the best investment ideas. RNV used a similar sorting process.
In the end, however, most of a practitioner’s ability to capture the value premium is going to be determined by whether they have the stomach to enter the order and buy some temporarily ugly looking businesses at the height of their grotesqueness (when their prices are dropping) and to sell those businesses when those prices bounce back to intrinsic value. As cliff Asness once said, a model never loses its nerve. I would add that a model never gets greedy. The ability to manage fear and greed are paramount.
Here is a draft of RNV’s paper to which Jason refers in his column:
It’s good stuff if only to help explain why so many value investors have beaten the market for so long.