TCS Daily

Dogs of the Dow

By James K. Glassman - January 14, 2003 12:00 AM

When I heard that President Bush was proposing to eliminate the double taxation of dividends, I immediately exclaimed, "Arf!"

I was thinking of the Dogs of the Dow, those high-yielding blue chips that gained notoriety nearly a decade ago. If it becomes law, the White House plan should significantly enhance the value of companies that pass a large chunk of their profits through to shareholders. And that's what the Dogs do.

The Dogs of the Dow system was invented a quarter-century ago by a money manager named Michael O'Higgins, who discovered a simple formula that was consistently beating the market as a whole. It works like this: At the start of each year, invest equal amounts in the 10 stocks with the highest dividend yields among the 30 stocks of the Dow Jones industrial average. At the start of each subsequent year, sell those 10 and repeat the process.

In his book "How to Retire Rich," analyst James O'Shaughnessy tracked the system from 1952 to 1996 and found that an initial investment of $1,000 in the Dogs would have grown to $660,000, while the same amount in the benchmark Standard & Poor's 500-stock index grew to only $170,000.

In the mid-1990s, financial writers like me introduced the system to readers, and it enjoyed a vogue. Wall Street firms packaged unit investment trusts (often with absurdly high fees) based on the Dogs, and online services including the Motley Fool promoted user-friendly variations such as the Foolish Four, the Dow Five and the Small Dogs.

But suddenly, just as investors were embracing the system, it seemed to stop working. In 1997, the S&P beat the Dogs by 15 percentage points; in 1998, by 18 points; and in 1999, by 17. The next year, the Dogs staged a comeback, but by then fickle investors had stopped paying attention.

They should snap to. The Dogs are back, with a vengeance. In each of the past three years, the Dogs have whipped both the S&P and the Dow as a whole. Overall, an investment of $1,000 in the Dogs at the start of 2000 was worth $927 while $1,000 in the S&P dwindled to $623 and $1,000 in Fidelity Magellan, the largest managed fund in the world, declined to $621.

Better yet, the 2003 edition of the Dogs has the highest average dividend yield in years: 4.3 percent. That compares with 3.5 percent in 2002 and 3 percent in both 2000 and 2001. On Wednesday, the five-year Treasury note was yielding 3 percent and the 10-year note 4 percent -- and they're taxable.

Assume that the Bush tax plan passes and that the Dogs make no changes in their dividend payouts this year. If you're in a 35 percent federal tax bracket, an investment of $1,000 in each of the 10 stocks will put $430 in dividends into your pocket (before any state or local taxes), but put $10,000 into a five-year T-note and you'll earn just $195 in after-tax interest. Yes, the Treasury securities are safer, but over five years the stocks are likely both to rise in price and to offer higher dividends.

This year's list of Dogs is headed by Philip Morris (MO), maker of Marlboro cigarettes, Miller beer and Kraft cheese. At year-end 2002, Philip Morris had an "indicated" annual dividend of $2.56 (that is, its current quarterly dividend multiplied by four) and was trading at $40.53. Divide the dividend by the price, and you get a yield of 6.3 percent.

The nine others this year, according to the Dogs of the Dow Web site, the system's unofficial scorekeeper, are J.P. Morgan Chase (JPM), yielding 5.7 percent; General Motors (GM), 5.4 percent; Eastman Kodak (EK), 5.1 percent; SBC Communications (SBC), 4 percent; DuPont (DD), 3.3 percent; Honeywell (HON), 3.1 percent; General Electric (GE), 3.1 percent; Caterpillar (CAT), 3.1 percent; and AT&T (T), 2.9 percent. All of the stocks except the last three were also on the list last year. They replaced International Paper (IP), Merck (MRK) and Exxon Mobil (XOM).

It is a decently diversified group, with consumer, telecom, transportation, chemical, financial and manufacturing sectors represented. Still, an energy company and a pure high-tech firm, among others, are lacking, and you should not consider the Dogs alone as a core holding, but a solid value-oriented piece of your total stock portfolio.

Why the Dogs? I am skeptical of systems in general. Some appear to work simply because the laws of chance prescribe that at least a few of the billions of formulas for picking stocks or horses or lottery tickets pay off when they are "back-tested," or judged against history. The future is quite another matter. But Dogs of the Dow has logic on its side: Dogs stocks are often undervalued since, by definition, a high yield indicates either an abnormally high payout or an abnormally low price.

So, often the Dow's highest yielders are its cheapest stocks by such valuation measures as price-to-earnings (P/E) ratios -- hence, Dogs. Philip Morris and DuPont, for instance, both have P/Es of 8.

A high yield can also indicate that a company is in serious trouble and may not make its next dividend payment. In fact, O'Shaughnessy's research found that a high yield alone is not a recommendation to own a stock. The Value Line Investment Survey runs a list of high-yielding stocks each week, and it's not a savory bunch. Of the five conventional stocks (that is, companies that are not real estate investment trusts and others with special tax treatment) that have the highest yields, three have rock-bottom ratings of "5" and the others are rated "4" (below average).

But a high-yielding member of the Dow Jones industrial average is another matter. Dow stocks tend to be solid citizens with good balance sheets, and the chances of their cutting or missing a dividend are low -- though certainly not nonexistent. GM, for example, slashed its annual dividend three years in a row, from $3 in 1990 to 80 cents in 1993. (Today, GM pays $2 a share.) In short, the Dogs system is a bargain play. It's a simple way to own blue-chip value stocks with an exceptional record of both high returns and low volatility. The Dogs have beaten the S&P in 31 out of the past 52 years -- a winning ratio of better than 60 percent. In the past 10 years the system has returned an annual average of 12.8 percent, compared with 11.2 percent for the S&P 500.

If the Bush proposal becomes law, the Dogs will be more attractive than ever. Instead of pocketing only 36 cents of every dollar in before-tax corporate profits, the typical shareholder will be able to pocket 60 cents. That's a two-thirds increase -- a big number. It's almost certain that investors will demand higher payouts from companies, which have been slicing their dividend ratios dramatically over the past decade, partly using double taxation as a justification.

Of course, not every company will raise its dividend. But managers will have to convince investors that the firm can earn more by investing profits internally than their shareholders can on their own. For many firms, such as Cisco Systems (CSCO) or Dell Computer (DELL), that case won't be hard to make. An excellent article in the Dec. 20 issue of Grant's Interest Rate Observer points to EchoStar Communications (DISH), the satellite TV company, as an example of a company that does not -- and should not -- pay dividends because it produces such high returns all by itself. It costs EchoStar about $450 to secure a customer, but in the first year thecompany realizes $340 in revenue, and similar amounts keep flowing in for an average of six years -- with almost no additional costs.

Grant's cites the great investment guru Philip Fisher, who wrote in his book "Common Stocks and Uncommon Profits," first published in 1958 and reissued recently by Wylie: "Over a span of five to 10 years, the best dividend results will come not from a high-yield stock but from those with a relatively low yield. So profitable are the results of the ventures opened up by exceptional managements that, while they still continue the policy of paying out a relatively low proportion of current earnings, the actual number of dollars paid out progressively exceeds what could have been obtained from high-yield shares. Why shouldn't this logical and natural trend continue in the future?"

Look at Johnson & Johnson (JNJ). In 1987, the stock paid a dividend of 10 cents a share, for a yield of 1.9 percent, well below the average for the time. But over the next 15 years, the dividend rose to 80 cents. In other words, an investor who put $1,000 into J&J in 1987 would today be receiving an annual dividend payment of $152 -- for an annual return of 15 percent on her original investment, and rising.

These Fisher-style companies generate lots of free cash, put some of the money to work in capital investments (like a new plant or a new computer system) that will generate high returns, and send the rest to shareholders as dividends. Another good example is Gillette Co. (G), which earns an incredible return of 50 percent on its equity. Gillette had an average yield of only 1.3 percent during the 1990s, but its actual dividend payouts rose during that time from 14 cents a year to 65 cents. Gillette's price has fallen lately, so its yield today is a lovely 2.1 percent.

During the Roaring Nineties, a dividend was a sign of stodginess. Fast-growing companies, especially in technology, paid no dividend -- and were proud of it. But that attitude is changing, and it's not hard to find go-go firms with terrific payout records.

Look at Paychex (PAYX), which provides payroll and tax-filing services to small businesses and has increased its earnings at an incredible average annual rate of 35 percent over the past 10 years. Paychex started paying dividends in 1990 at a rate of just a penny a share (adjusted for splits). Today, the dividend is 42 cents. Growth is slowing (Value Line projects earnings growth will average a mere 19 percent for the next five years), and the stock has fallen by more than half since 2001. As a result, Paychex, now sports a yield of 1.5 percent, about twice its average for the past decade.

Still, if the dividend tax is eliminated for investors, it is high-yielding stocks that should gain the most. Take Genuine Parts (GPC), the giant distributor of auto parts that has increased its dividend for each of the past 46 years. Genuine currently pays a dividend of $1.16, and the stock on Friday was trading at $31.47, for a yield of 3.7 percent. Buy 100 shares for about $3,150, and you will be paid $116 tax-free in 2003 (actually, it will almost certainly be a few dollars more when the company raises its payout later this year).

What if you hold dividend-paying stocks in an IRA or other tax-deferred account? You may want to move them, but understand that eliminating the tax should boost the prices of dividend-paying stocks (they will be worth more because they put more money in your pocket), so you will benefit anyway.

Because of SEC rules requiring broad diversification, there is no mutual fund that strictly follows the Dogs of the Dow strategy. Sun America's Focused Dividend Strategy Portfolio (SDWAX), formerly called Dogs of Wall Street Fund, owns 30 Dogs-type stocks.

A fund I mentioned in a previous column, Hennessy Total Return (HDOGX), comes closer to the actual strategy. The 10 Dogs stocks account for 75 percent of its assets, and the rest (to satisfy regulatory requirements) is Treasury bills. But the fund carries a lofty expense ratio of 2.3 percent, according to Morningstar. The Sun America fund, in its "A" version, has a front load of 5.75 percent and an expense ratio of 0.95 percent -- not particularly attractive either.

I expect the dividend tax reduction to pass, but even if it fails, consider the Dogs of the Dow, in whole or part, for your portfolio.

TCS Daily Archives