Review of “What Works on Wall Street”

A friend asked me to review the book What Works on Wall Street, by James P. O’Shaughnessy. The book was published this year (1997) by McGraw-Hill and is subtitled A Guide to the Best-Performing Investment Strategies of All Time. These are my comments:

1. It is a useful book-chock full of data.
2. The book is limited in scope, focusing on the “Wall Street” in the title instead of the “Investment Strategies” in the subtitle.
3. The book is all hindsight. It should be titled, “What Would Have Worked on Wall Street.”
4. The author briefly mentions the most interesting aspects of his data, but then ignores them.

I’ll expand on each comment in turn.
The data is plentiful. The author uses Standard & Poor’s Compustat Active and Research Database from 1950 through 1994. This gives 44 years of annual returns. He takes into account various biases (such as survivorship bias and inflation) that detract from the validity of most studies. Equally important, he is specific about how the data is used. He then analyzes this data by capitalization and by a number of popular investment strategies. The resulting tables alone are easily worth the price of the book. Chapter 4, which lists annual returns by market capitalization, is a must-read for anyone investing in small stocks because of their reported long-term outperformance.
O’Shaughnessy finds that the source of the small-stock outperformance was in stocks of less than $25 million capitalization. Even “small cap” funds invest in stocks averaging over 10 times $25 million.

The book is limited in scope.  It looks only at stocks. Granted that is the charter in the title, but at various times during the last 44 years, there have been periods when the best-performing investment strategies included real estate, gold, bonds, certificates of deposit, Chinese ceramics, oil and gas, and others. I mention this more as a reminder than as a criticism. If the data is available, which I doubt, a similar treatment of these vehicles would make a useful book.

The book is all hindsight. I believe Warren Buffett has said, “If the future were a repeat of the past, all librarians would be rich.” The book should be titled, What Would Have Worked on Wall Street-If We’d Had a Crystal Ball Forty-Four Years Ago. But even a crystal ball is insufficient. As a practical matter, a profitable strategy is only useful if it was identifiable at the time, and if the strategy had a strong enough rationale to cause investors to believe in it with enough conviction to put money on the line at the time. Otherwise, saying low price-to-sales is a useful strategy is no more profound than saying, “I wish I’d have bought Coca-Cola.”

O’Shaughnessy ignores prior attempts at determining “the best performing investment strategies.” These attempts include Risk and Returns from Common Stocks and Security Prices in a Competitive Market by Richard Brealy, which were published by MIT Press in 1969 and 1971, respectively. Stock Market Logic by Norman Fosback, published in 1976, is a rational, comprehensive review of stock market indicators, econometrics, and stock selection theories, culminating in “A Total Financial Management System.” If O’Shaughnessy’s unstated premise (what has worked in the past will work in the future) is true, then Stock Market Logic, which described what works well on Wall Street, should have provided the key to wealth. Fosback now publishes Mutual Funds Magazine. It would be interesting to read his review of the current book.

However, my real criticism is: the book doesn’t make good use of the data it presents. O’Shaughnessy’s conclusions are based on the results of the entire 44-year period. Yet on page 144, O’Shaughnessy prints tables showing the Compound Annual Rates of Return by Decade for the strategies of High & Low Price to Earnings, High & Low Price to Book, High & Low Price to Cash Flow, High & Low Price to Sales, and High Yield.

A quick look shows that none of the strategies was optimal for two successive decades. In the large stock universe, the optimal strategy in the 1950s was low price to cash flow. In the 1960s, it was high price to cash flow. In the 1970s, it returned to low price to cash flow. Note that each decade’s optimal strategy is a reverse of the prior decade’s optimal strategy. Someone using a strategy derived from using O’Shaughnessy’s methods, based on his data, over a decade would have found their strategy to be backwards in the following decade. In chapter 14, O’Shaughnessy reviews Returns on Shareholder Equity. In his Implications (summary) on pages 182 and 187, he points out that high ROE was a very successful strategy from 1952 through 1967-successful enough to make someone using an O’Shaughnessy method a true believer, only to have the strategy not work in the 1970s.

Presumably, this tendency for optimal strategies to change over time is the reason he uses the entire 44-year period. Unfortunately this averages out, and therefore hides, the most interesting aspect of the data. More to the point, it assumes that investors have such conviction in the conclusions of his studies that they are willing to stay with it through a decade or more of underperformance.

O’Shaughnessy never entertains the idea that shifts in markets and optimal strategies might be normal. An idea that such shifts are normal would make them expected, and possibly observable. We expect such shifts. We expect them to occur when there is a change in the investment climate (our phrase). We think that the shifts in the optimal investment strategies likely occurred in the periods 1965-68, 1979-82, and 1990-93, when we believe the investment climate changed. It is our intent to do such a study, based on O’Shaughnessy’s annual data, but we haven’t done it yet. Stay tuned.

Editor’s Note
No matter how good the data is, the analysis of the data determines if you are successful. O’Shaughnessy’s data is good, but like many analysts, he misses a critical point. He fails to recognize that investment climate changes. Therefore, the success of an investment strategy depends upon the climate in which it is used. A successful investor must recognize the climate (and changes in climate) and choose the appropriate strategy for that climate.