Wednesday, September 06, 2006

Strategies for Maximizing Dividend Returns Differ

But weighting stocks on the basis of their total dividend payments is the most basic approach

Investment Executive, Carlyle Dunbar, 30 August 2006

Getting the best results from dividend-paying stocks has become a more urgent matter as lower stock market returns become more probable in the next decade because the economy and the market will be reacting to the super-gains of the past 20 years.

The rise of income trusts illustrates the demand for income and better yields. And the stock market reaction since 1999 has made no serious change in average dividend yields, which are stuck in a historically low range.

This obscures the fact that, over time, dividends have provided the greater part of investment returns. Mind you, many such measurements depend on reinvested dividends. But, apart from stocks offering dividend reinvestment plans, there is no easy way for the average investor to reinvest small amounts of dividend income.

Strategies to maximize dividend returns without sacrificing price strength take different forms. A popular one is to pick stocks that have had frequent dividend increases, or are expected to provide them in the future. Buying stocks with high yields is risky because unusually high yield generally equates to high risk.

Another approach is to pick stocks with better dividend yields than their peers.

But the most successful conservative dividend strategy has been employed for years by the U.S. advisory service Investment Quality Trends, which rates stocks as “buys” or “sells” based on each stock’s individual historical range of yields. The historical period can range for two or three decades — something not possible with new stocks or stocks that have paid a dividend for a shorter period.

Now, indexing is entering the game. This follows the argument that most investors should not bother to seek active portfolio management, but rather buy into a fund or portfolio that replicates a market index.

The dividend searchers say this is an incomplete answer because most indices are capitalization-weighted. Why not, they say, use an index weighted by dividends instead of market value?

But what dividends? Some in-dexing strategies are based on dividends per share, dividend growth or yields. The most basic approach, however, may be weighting stocks on the basis of their total dividend payments. That is the idea explored here and outlined in the accompanying table.

Trading on the Toronto Stock Exchange as of mid-year were 232 dividend-paying stocks. This excludes income trusts, of course, but also closed-end funds, split-share income trusts, foreign companies and exchangeable shares.

By my calculation, the 232 stocks — which includes two classes of stock for some companies — are paying a total of about $22.55 billion in dividends. The largest dividend payer is a major chartered bank, the Royal Bank of Canada. It is paying shareholders at the rate of $1.85 billion per year, or 8.2% of the total of all dividends paid in Canada.

The accompanying table summarizes the 30 largest dividend-paying TSX stocks. Altogether, they pay 78% of all common dividends.

As well as the dividend weighting is each stock’s mid-year weighting in the capitalization-based S&P/TSX composite index. Comparisons show, for example, that banks rank much higher in a dividend-weighted index. Royal Bank’s capitalization weighting is 4.51%, a little more than half its dividend-weighted proportion. But firms such as Barrick Gold Corp. and Falconbridge Ltd. have much lower weights in a dividend-weighted index. The 30 stocks in the table account for only 48% of the capitalization-weighted TSX composite index.

The table overweights the Power Corp. group of companies because no allowance has been made for intercorporate dividend transfers. But there is no doubt they would rank high in any dividend-weighted index.

One researcher of dividend-weighted portfolios is Jeremy Siegel, author of Stocks for the Long Run. He is involved in a U.S. firm developing dividend-weighted indices.

This is what he says about a study of using total dividends paid: “A backtest of the hypothetical, historical performance from 1964 to 2005 of a dividend-weighted index consisting of U.S. companies that paid regular cash dividends showed the annualized total return of this index exceeded the annualized total return of the S&P 500 index for the same period by 138 basis points a year, and did so with lower volatility.

“Moreover, my analysis of the hypothetical backtested data for this period supports the proposi-tion that dividend-weighted indices generally outperformed comparable cap-weighted indices during bear markets.”