The Dogs of the Dow is an investment strategy popularized by Michael B. O'Higgins, in 1991 which proposes that an investor annually select for investment the ten Dow Jones Industrial Average stocks whose dividend is the highest fraction of their price.
Proponents of the Dogs of the Dow strategy argue that blue chip companies do not alter their dividend to reflect trading conditions and, therefore, the dividend is a measure of the average worth of the company; the stock price, in contrast, fluctuates through the business cycle. This should mean that companies with a high yield, with high dividend relative to price, are near the bottom of their business cycle and are likely to see their stock price increase faster than low yield companies. Under this model, an investor annually reinvesting in high-yield companies should out-perform the overall market. The logic behind this is that a high dividend yield suggests both that the stock is oversold and that management believes in its company's prospects and is willing to back that up by paying out a relatively high dividend. Investors are thereby hoping to benefit from both above average stock price gains as well as a relatively high quarterly dividend. Of course, several assumptions are made in this argument. The first assumption is that the dividend price reflects the company size rather than the company business model. The second is that companies have a natural, repeating cycle in which good performances are predicted by bad ones.
Contents |
O'Higgins and others back-tested the strategy as far back as the 1920s and found that investing in the Dogs consistently outperformed the Market as a whole. Once these findings were announced in the early 1990s, a wave of mutual funds and institutional investors adopted the strategy, only to find that the Dogs lagged the general market by 2 or 3% annually for the next decade. The current thinking on the Dogs of the Dow findings were either a result of data mining or that a once reasonable strategy has been exploited to the extent that it no longer confers an advantage. By "data mining" analysts mean that in a large enough sample of data patterns will undoubtedly emerge, but this does not necessary require that those patterns are meaningful or bound to continue. For example, we might find that during the 2000s companies whose names started with M, ended with T and had CEOs whose names ended in a sibilant outperformed the market by 5%, but that does not imply that this is a good basis for future investments. A similar strategy, the Foolish Four (see below), found that one could outdo the dogs by looking at square roots of dividends and discarding the very lowest, an inexplicable finding that was, by its creators' own admission, a case of blind data mining. The other possibility is that the Dogs of the Dow were indeed a significant market-beating strategy but upon its discovery was soon exploited by investors so that it lost all market advantage (see: Efficient Market Hypothesis). Either way, today the Dogs method has been largely discredited and is no longer followed by major funds or institutional investors.
If you look at Dow stocks on January 1 then the high yielding stocks did significantly better than average before 1991 than they did after 1991. The Motley Fool created the Foolish Four which used the square root of the price and dividend in order to create back tested results better than the Dogs of the Dow. All of these methods were dependent on the fact that the stocks picked on January 1 performed better than stocks picked at other times during the period of back testing.