Article by Steve Alexander
James O’Shaughnessy is a fan of mechanical investing. Mechanical investing refers to selecting stocks for investment directly from a mechanical screen that ranks stocks by a particular statistic or set of statistics. The first edition of his book What Works on Wall Street was released in the late 1990′s, just when computers began to play a big role in stock analysis, and historical databases were becoming available. Thus, it was one of the first times a wide range of mechanical investing strategies were able to be back-tested and compared to each other. This article will highlight some of the more interesting findings from the book, and comment on their relevance to those following the Magic Formula Investing (MFI) strategy. For more detail, I strongly recommend picking up the book as it is very insightful and detailed.
First, the book itself. What Works on Wall Street reads very much like a research report. The book is littered with sorted tables, graphs, and charts, with some sparse commentary inserted by O’Shaughnessy. The data really speaks for itself. The chapter organization focuses on the individual strategies tested. O’Shaughnessy starts by simply using mechanical screens focused on a single statistic, and creates portfolios of the top 50 stocks as ranked by that statistic, with the portfolio rebalanced annually (very similar to the MFI strategy). Some example statistics are: price to earnings, price to sales, price to cash flow, price to book, relative strength (defined here as 12 month stock performance), etc. As the book progresses, O’Shaughnessy moves towards multiple factor screens, which combine two or more of the above factors. He then uses the data from these simple screens to create two united strategies that appear to provide the best risk-adjusted performance, called the “cornerstone growth” and “cornerstone value” strategies. Last, and really all most investors will want to know, all of the strategies analyzed in the book are put together into one big table, sorted by performance and compared against the S&P 500′s performance.
The analysis O’Shaughnessy performs yields some very interesting results. The five most interesting facts MagicDiligence took from this book were:
1. Value Based Strategies Greatly Outperform Growth Based Strategies.
This should be no surprise. Strategies that were based wholly or in part on value statistics such as low price-to-earnings, price-to-sales, price-to-book, or high dividend yield dominated the top 20, all of them returning 14% or more annually. The bottom of the table was almost fully comprised of single-statistic growth measures, such as high P/E, P/S, P/B. Overvalued stocks underperform in the long run.
2. 12-Month Relative Strength Adds Significant Returns to Value Strategies.
O’Shaughnessy defines 12-month relative strength as the stock return over a single year. Stocks that have 12-month momentum and value criteria such as those detailed above far outperformed stocks that were just cheap. For example, combining a low price-to-book ratio with high relative strength returned 17.3% per year, while focusing just on low price-to-book ratio returned just 14.4%. While this may seem like a small difference, the compounding difference of 3% per year over long periods of time is extremely significant.
3. Some Growth Criteria Do Outperform.
Namely, a history of rising earnings per share (at least 5 years) and the previously mentioned relative strength. In fact, just a portfolio of stocks with these two characteristics ranked in the top #5 overall performers, the only growth-only strategy to perform relatively well. However, it is important to note that this strategy also was much more volatile than value based strategies. Can you stick with a strategy that underperforms the market for a year or more?
4. The Quality of a Business Does Matter – But Not as Much as Price.
It is interesting to quantify this, as it’s the second part of the Magic Formula strategy. O’Shaughnessy uses a single statistic for business efficiency – return on equity, which can be misleading for companies with large debt components. He does not combine ROE and any value measures, so there is no strategy analogous to the Magic Formula. But a strategy of high ROE and good relative strength placed in the top 5 performers, with nearly 17% annual returns.
5. Limiting Your Investment Universe to Large-Cap Stocks is a Bad Idea.
The top 15 strategies all considered both small and large cap stocks. Not until #15 did a strategy using just large cap stocks rank on the list. This drives home a point made here: You Must Own Small-Cap Stocks!
Another rather interesting finding is that price-to-sales ratio, not price-to-earnings, was the best performing value statistic. This point has been made before, most visibly in Ken Fisher’s Super Stocks. It’s not intuitive that this would be the case. Most low P/S stocks are low margin businesses such as retailing. However, I suppose it makes sense as it is easier (and cheaper, usually) for a company to improve margins than to grow revenues.
It would have been interesting to see a strategy that was a rough analog to the Magic Formula be included in the study. Greenblatt’s 17-year trial period is much shorter than O’Shaughnessy’s 40 year one, and the MFI study was conducted during an unprecedented bull market. MFI’s stated 31% annual return would likely be more modest over this book’s time period, but would it outperform these strategies?
The findings in this book are part of my toolbox when digging up Top Buys from the Magic Formula screen. By using these characteristics, as well as by looking for competitive moats, MagicDiligence weeds out the losers for you and finds the most likely winners.
