"Dogs of the Dow" and the "Foolish Four"

The "Dogs of the Dow" is a stock picking strategy that has been the subject of a great deal of attention in the last few years. Proponents of the strategy cite the fact that it has outperformed the Dow Jones Industrial Average and other indexes by significant margins. Many cite figures for the last 25 years with others going back even farther. As a result, thousands of investors have proceeded to invest in various "Dogs of the Dow" strategies, either individually, or through packaged funds. Unfortunately, while the track record is undeniably impressive, there are a number of reasons for those considering the strategy to waiver and for those who have already taken the bait, to rethink.

Strategies of investing in unpopular stocks in the DJIA are nothing new. In fact, an astute reader of Benjamin Graham's all-time classic The Intelligent Investor will find a reference to a study by H. G. Schneider published in the June 1951 issue of the Journal of Finance that documents a strategy of investing in unpopular DJIA issues from 1917-1950. A second study noted in the book covers the years 1933-1969. The studies looked at strategies of buying either the six or ten issues in the DJIA selling at the lowest earnings multiples and rebalancing at holding periods ranging from one to five years. The strategy proved unprofitable from 1917-1933 but from 1937-1969 a strategy of investing in the low multiple ten soundly and consistently beat the high multiple ten and the DJIA.

The recent excitement has been focused on high dividend yield stocks in the DJIA. Outperformance of high dividend Dow stocks was apparently discovered by John Slatter in the late 1980's. The strategy began to increase in popularity in the early nineties following Michael O'Higgins book Beating The Dow. The Dow-10 strategy consists of buying the ten highest yielding dow stocks and rebalancing annually. Some proponents "tweak" the Dow-10 strategy by ranking those ten according to price and selecting the lowest priced of the five. The Motley Fool has endorsed their modification (The Foolish Four) in The Motley Fool Investment Guide. "The Foolish Four" consists of investing 40% of a portfolio in the second lowest priced of the ten and 20% each in the third, fourth, and fifth lowest priced. The rational for "The Foolish Four" strategy is back-tested results showing the strategy to have yielded 25.5% annually over a twenty year period.

The 11th chapter of The Motley Fool Investment Guide addresses arguments against the strategy, and its there that counter arguments begin to appear. The first issue is volatility and diversification. Both the Foolish Four and the Dow-10 are likely to be more volatile and riskier than investing in the DJIA or a broader index (more on this later). The second issue is the relevance of the twenty year period. On that page, the Gardners disclose that a staff member had indeed run the numbers back to 1961. Say goodbye to the 25% returns unless you happened to start investing in the strategy precisely at the right time. The longer term results brought the annual rate down to 18.35% - still a healthy lead over the market's 10.02% but not nearly as impressive (there are apparently no investors claiming to have actually used the strategy for either period).

The third issue discussed in The Motley Fool Investment Guide is "Overpopularity." They argue that "sales of our book are not going to wreak havoc" with the market. Ironically, they probably underestimated their own future success, having gone on to become extremely popular on many fronts. But the issue is much larger. The Motley Fool was, after all, only elaborating on O'Higgin's book and they were only one of many organizations hyping the strategy. Barron's (which coined the phrase "Dogs of the Dow") and many others had also documented the strategy and numerous mutual funds and trusts have sprouted up that are attempting to cash in. The Select-10, a unit trust (actually made up of a number of funds) that invests in the Dow-10 and is offered by Merrill Lynch and other full-service brokers has grown to over $14 billion and several commentators have estimated that over $20 billion is investing in Dow dividend strategies. With countless individuals investing on their own, the issue is without a doubt significant. According to Andrew Bary's December 28, 1998 column in Barron's (titled "Bound for the Pound?"), the Select 10 trust has become the largest holders of Dow Dogs J.P. Morgan, International Paper, and Kodak.

Incidentally, at the start of 1998, the Motley Fool "modified" The Foolish Four formula. The second variation invests equally in the four lowest priced of the Dow-10 over 18 month periods, however if the lowest priced of the ten is also the highest yielding stock, it is thrown out and the second through fifth lowest priced are used. The new version was described in You Have More Than You Think. The Foolish Four was modified a third time some time later. See the Foolish Four Evolves, Explained, and History. As for O'Higgin's, he no longer invests in the Dow Dogs at all according to an article in the 12/8/97 issue of Time. The article (The Dow Dogs Won't Hunt) quotes O'Higgins as noting that the Dow Dogs "have become too popular and the market has become too high" for the gambit to keep working. His latest book is titled Beating the Dow With Bonds (see also The Bond Bard by Scott Burns from Worth - 6/99).

A rigorous analysis of the Dow strategies topic was published in the July/August 1997 issue of the Financial Analysts Journal. In "Does the 'Dow-10 Investment Strategy' Beat the Dow Statistically and Economically?" Grant McQueen, Kay Shields and Steven R. Thorley thoroughly analyzed the strategy from 1946 to 1995 and deal with the issues of risk, taxes, transactions costs, and the potential problems of "investor learning" and "data mining" (Abstract). The authors found that the Dow-10 did in fact produce significant excess returns over the 50 year period. The average annual return (arithmetic mean) for the Dow-10 was 16.77% versus 13.71% for the Dow-30. Higher risk as measured by standard deviation (19.10% versus 16.64%) accompanied the higher returns.

The authors point out that more of the Dow-10 returns come from dividends (as you would expect) which can not be deferred and are taxed at a higher rate. Further the Dow-10 strategy requires annual rebalancing exposing taxable accounts to taxes on gains. The Dow-10 authors state that a formal analysis of the tax differences is not possible because the tax payments will depend on each individual's tax rate and other considerations. However, transactions costs and risk explain most of the Dow-10 excess return, and they believe that most if not all of the remaining excess return would have gone to the IRS. The authors also looked at subperiods and found that during some extended periods the strategy outperformed, but during other long stretches (decades) the authors suggest that economically, an investor would have been better off (after adjusting for risk, transactions costs, and taxes) in the Dow-30.

The authors then discuss the impact of "investor learning" (see "Overpopularity" above) or the tendency for investors following the strategy to drive up the price of the stocks thereby reducing or eliminating excess returns. The Dow-10 authors also discuss an issue referred to as "data mining" or the "file drawer problem." This is the possibility that certain correlations between variables will occur randomly in financial and other data purely by chance. If one searches long enough these chance or random correlations will be found. However, "the true significance of successful investment strategies can be assessed only after quantifying the number of unreported or unpublished failures gathering dust in the file drawers of stock market analysts, traders, and researchers."

Professors McQueen and Thorley followed up on the Dow Dogs article by examining the Motley Fool's Foolish Four. "Mining Fool's Gold" appeared in the March/April 1999 issue of the Financial Analysts Journal and the Professors have posted a "lighthearted" version of the paper on the BYU server. The data used in the study can be downloaded here. See the Data Mining page for a discussion of the paper.

The June 16, 1997 Issue of Business Week included a commentary by Peter Coy titled "He who mines data may strike fool's gold." The article discussed data mining and the fact that patterns will occur in data by pure chance, particularly if you consider many factors. Many cases of data mining are immune to statistical verification or rebuttal. In describing the pitfalls of data mining, Coy cited an example from David J. Leinweber, Ph.D. who "sifted through a United Nations CD-ROM and discovered that historically, the single best predictor of the Standard & Poor's 500-stock index was butter production in Bangladesh." The lesson to learn according to Coy is a "formula that happens to fit the data of the past won't necessarily have any predictive value." See also What's the Stock Market Got to Do with the Production of Butter in Bangladesh? from Money (March 1998).

James O'Shaughnessy (author of What Works on Wall Street) is another proponent of the Dow Dogs and has researched the strategy back to 1929. His latest book (How to Retire Rich) includes a table of annual Dogs of the Dow returns from 1929 through 1996. He found that "The Dogs of the Dow compounded at 12.7 percent a year, while the S&P grew at 9.89% a year." O'Shaughnessy Funds, Inc. offers a "Dogs of the Market Fund" that invests half of the portfolio in the Dogs of the Dow and the other half in high-yield, large-cap stocks. See also returns of the Dogs of Dow strategy from 1929 to 1997.

Louis Rukeyser, the popular host of Wall $treet Week, discussed the Dogs of the Dow on his 2/5/99 show. His comments included the following:

Let's recall the wisdom of late financial genius G. M. Loeb who once told me "Lou, whenever you think you've found the key to the market, some SOB changes the lock." Lately a lot of new comers to finance have been trumpeting an alleged sure-fire way to beat the averages by buying the so-called Dogs of the Dow. The problem is, as a little research reveals, that in more years than not of late, they really were dogs . . . By the way, it used to be that the real Dogs of the Dow were simply the indexes ten worst performers of the prior year, but that theory didn't work as promised. So supporters apparently figured that this way would give them a better shot . . . Vast amounts of phony commentary have been paraded on this subject and mutual funds and Unit Investment Trusts have been formed to exploit the market. But the reality is, as with so many so-called sure-fire theories, that just about the time you hear about it, somehow it stops working.
According to Rukeyser, the Dogs of the Dow have underperformed the DJIA more often than not in the past ten years and its compounded return over the period has lagged the DJIA.

An interesting discussion of the Dow Dogs theory is also included in Peter Tanous' 1997 book Investment Gurus. In it, Tanous and Nobel Laureate William Sharpe discuss the strategy in the context of value investing. Sharpe makes the point that on one hand, if you search hard enough with a large number of random data sets, you will eventually "find some strategy that would have made you a fortune." On the other hand, the results from the strategy could be the value stock effect. An argument can be made in favor of the Dow Dividend strategy given the scores of studies that have documented outperformance of "value" stocks (See Fundamental Anomalies).

Andrew Tobias also addressed the Dogs of the Dow and Foolish Four in his daily comments (Playing the Fool 1/23/97 and Motley Fool Dog Track - Revisited 1/30/97) on the Ceres Securities site (now Ameritrade). Tobias focused on the tax issue and argued that it seems aggressive and optimistic to assume that the strategies will outperform an index fund by significant margins (5% or more) in the future. While acknowledging that the strategy is not a crazy speculation, Tobias argued that "backtested systems rarely are the winners going forward that they were when 'discovered' by looking back" and because "the Dogs of the Dow strategy entails considerable annual turnover, it must appreciably outperform the index funds (unless you're investing through tax-deferred accounts) just to keep up."

Recent changes in the tax law are another issue that Dow Dividend strategy investors should consider in evaluating these strategies. Capital gains tax rates are lower for longer holding periods and lengthening the holding period of any strategy in order to take advantage of the new rate is certainly a worthwhile consideration.

Perhaps the greatest cause for concern with the Dow Dividend strategies is the central assumption inherent in these strategies. That is, the assumption that the future will be like the past. Unfortunately, past experiences with disappearing anomalies, the impact of the $ billions already invested according to the strategies, and the inconsistent performance of the Dow strategies over significant subperiods (see Morningstar link below) imply that this may be dangerous assumption. The irony is that the strategy is based on the idea that unpopular stocks outperform. Yet, given the publicity of the Dow strategies, the Dow Dogs now seem to be just the opposite (popular). Keep in mind, that the phrase "past performance is no guarantee of future performance" is nothing more than a legal disclaimer. The phrase "past performance may be no indication of future performance" is a more appropriate description in some cases.

While its clearly possible that the Dow-10 (and derivative strategies) will outperform in the future, its apparent that many investors have overly optimistic expectations (based on rear-view mirror, back-tested returns for specific time periods) and haven't accounted for the added risk, transactions, and tax costs inherent in the strategy. The strategy's consistency with other value approaches is certainly appealing, but current and future funds committed to the strategies may end up reducing or eliminating any future outperformance. The Dow-10 authors concluded their article by asking whether Dow-10 investors will "reap any adjusted (for risk, transaction costs, and taxes) premiums in the future?" Their response - "We wouldn't bet on it."