Starbucks' chief executive officer Howard Schultz recently announced the company would pay its first-ever dividend. Some analysts, believing Starbucks had run out of opportunities for growth, recommended selling the stock. Others argued that the decision to return profits to owners would be good for its shareholders.
Although Starbucks' stock subsequently rose, the ensuing debate was not really about Starbucks. It turned on a much more crucial question - whether investors are better off when a company elects to pay out profits to shareholders through a dividend rather than spend the money on expanding the business. On this issue, the evidence is clear. Companies with higher dividend yields have, on average, provided far better total returns to investors.
Dividends matter far more than most investors realize. The conventional wisdom is that common shares provide gains first and income second. However, over time, it is the dividends - particularly reinvested dividends - that have made up most of the total return from stocks.
In a 2002 article in the Financial Analysts Journal, entitled Dividends and the Three Dwarfs, Robert Arnott analyzed returns on U.S. stocks from 1802 to 2002. The total return during this period was 7.8 per cent a year. Of this, 5 per cent - 64 per cent of the total - represented the return from dividends. Mr. Arnott's calculation, which assumes dividends are reinvested, demonstrates the powerful effect of compounding over long periods of time.
If we look at real returns (after inflation), the result is even more dramatic. Jeremy Siegel, in his book The Future for Investors, has calculated that from 1871 through 2003, 97 per cent of the total after-inflation return from stocks came from reinvesting dividends. Only 3 per cent came from capital gains.
Given the importance of dividends, one would expect that higher yielding stocks would produce a greater total return than lower yielding stocks. A number of independent studies have proven just this. For example, Mr. Siegel examined the record of firms in the S&P 500 index from its beginning in 1957 up to 2002. At the end of each year, he sorted the companies into five quintiles by dividend yield (dollar amount of dividend per share divided by share price) and then calculated the return over the next calendar year. The portfolios with higher dividend yields offered investors, in strictly increasing order, higher total returns.
In Canada, the strategy of investing for the next 12 months in the highest yielding bank stock at the beginning of each year has been shown to far outperform the best performing bank stock over a 10-year period.
Perhaps more important than providing higher total returns over time, reinvesting dividends can protect investors in falling or range-bound markets.
A classic example of this is the Great Crash. The 1929 bull market peaked on Sept. 3, 1929. The Dow Jones Industrial Average remained below that bull market peak until Nov. 24, 1954 - more than a quarter of a century. For many who had borrowed on margin, the stock market crash was a disaster. But the long-term stock investor who did not buy on margin and stayed the course for those 25 years had a much better result than you would expect. Stockholders who reinvested their dividends realized an annual rate of return over this disastrous period of more than 6 per cent a year. The capital gain was zero. But the return solely from receiving and reinvesting dividends was almost twice the return from bonds and four times that from short-term Treasury bills.
North American stocks provided a total return during the 20th century of just more than 9 per cent a year. But there were several periods, lasting on average about 15 years, when the broader index (which measures price gains but not dividends) was essentially flat. In these range-bound markets the total return to investors was about 5.5 per cent a year - and all this return came from dividends.
If you believe that the greatest bull market in history, lasting from 1982 to 2000, is likely to be followed by an extended range-bound market, you may want to ensure that the companies you invest in pay you dividends while you wait for higher stock prices.
There is, of course, more to investing than picking high dividend-paying companies. But if you start by researching companies that pay a higher than average dividend, you will increase your baseline probability of beating the market. And, crucially, the likelihood of higher dividend-paying stocks outperforming is greatest precisely when that outperformance is most needed - in a falling or flat stock market.