This book promises that a magic formula will help anyone to beat the market. It delivers this message in breezy, informal prose that assumes no prior knowledge about the stock market. While peddling a magic formula raises the specter of a snake-oil salesman, Greenblatt explains the underlying logic of the formula in plain English. And the underlying logic is based on Benjamin Graham's investment philosophy which focuses on the fundamentals of firms' situations. Truly remarkable past results are presented for a strategy that picks about thirty stocks, and Greenblatt provides reasons why the strategy is likely to work in the future.
The underlying idea is simple. As in a fundamental analysis, pick stocks that would be candidates using initial screens: in this case, high return on capital and earnings yield. Then, if further analysis will not be done, pick a portfolio of these stocks based on their combined rank on these two screens. If the screens are effective, the stocks will on average be better than many others and, while returns will not be as good as they would be with further analysis, returns will be better than returns from picking stocks at random or holding all stocks. To avoid buying and holding klunkers and to benefit from tax laws in the U.S., sell the stocks after one year and pick new ones. That's it.
Results that would have been obtained using this rule in the past are used to justify the magic formula's usefulness. The past returns are for 1988 through 2004.
Greenblatt presents results for the top 30 stocks for 1988 through 2004 among the largest 3500 and 1000 firms, mentioning results for the largest 2500. The results are great: 30.8 percent on the portfolio based on the largest 3500 and 22.9 percent for the largest 1000 firms, compared to returns of 12.3, 12.4 and 11.7 percent for an equally weighted portfolio of the same set of firms.
Greenblatt then looks at returns of portfolios for deciles based on ranking. He finds that the average returns for 1988 through 2004 are highest for the highest decile and fall with the rank order of the decile. The return for the highest decile is 17.9 percent and the return for the lowest decile is 2.5 percent.
Good news: Returns are not always better with the magic formula. If returns always were spectacularly better, why would someone else not have noticed before, bought the stocks and in the end raised the prices to eliminate the high expected returns? Even if investors had not noticed these returns before, there would be every reason to think that the magic formula would not work after this book's publication.
In fact, the magic formula generated worse returns than market averages for one quarter of the 12-month periods from 1988 through 2004. Given the tendency of mutual fund investors to chase performance, this means that the magic formula is not likely to be a guaranteed success for a mutual fund. It also means that it is liable to be hard for most of the book's readers to follow the strategy! Still, for about 95 percent of the three-year periods from 1988 through 2004, the magic formula generated better returns than market averages and never generated negative returns. (This is based on the largest 1,000 firms.)
What are the transactions costs of this strategy? After all, the underlying idea is to average a lot of choices over time, and every one of those choices generates brokerage fees. Suppose that each stock trade costs $10. Buying and selling 30 stocks a year generates transactions costs of $600 per year. Suppose that the difference in actual return is 15 percent per year. This entire difference in return will be eaten up by transactions charges if the initial portfolio is $4000. In any case, it is hard to imagine owning $133 in stock in 30 firms.
Suppose that you wanted to invest $1000 in each of the companies and could. This is a $30,000 portfolio. Then the 15 percent difference in return on a $30,000 portfolio is $4500 and the net difference in return is 13 percent. This is not too bad, if one can get transactions costs this low. This calculation is based on the average return, of course, and a few negative returns initially could make it difficult to acquire stakes in 30 firms later.
