Norman Tweed's  Instablog

Myron J. Gordon wrote the classic paper Dividends, Earnings, and Stock Prices back in May 1959.

http://www.wiso.uni-hamburg.de/fileadmin/sozialoekonomie/bwl/bassen/Lehre/International_Finance_I/Assignments/1959_Gordon.pdf

In his paper he argued that the investor buys the dividend when he acquires a share of stock because the dividend is literally the payment stream that he expects to receive. In implementing the hypothesis it must be recognized that the stockholder is interested in the entire sequence of dividend payments that he may expect and not merely the current value. “For the purpose of arriving at an operational model we may represent this infinite sequence by two quantities, one the current dividend and the other a measure of the expected growth in the dividend.”

This model can be simplified and written as P=D+G. Where P is the price of the stock today, K is the required return, G is the dividend growth rate (Gordon speaks of 5-yr average dividend growth rate in the paper, but mathematically derives it for perpetuity), and D is next year's dividend.

This formula has been simplified and rewritten as: P= ΣD*(1+G)t/(1+k)t, summing the infinite series we get, P=D*(1+g)/(K-G) but it is only meaningful in this form if K>G. I further modify it into K=D/P +G in which D/P is current %yield and G is dividend growth %. What this tells you is that constant future dividend growth is additional yield. Gordon speaks about earnings growth also in the paper. However, this is a highly conservative usage, leaving out pure growth stocks and concentrating on yield only. It is most applicable for utilities and slow growth rate stocks.

With these bond-substitute stocks, p/e is determined as 1/yield. Thus if the yield is 5%, then the p/e is 20. If the yield is 10% then the p/e is 10. Gordon talks about companies having earnings per share growth rates of 4-10%. Thus in this small range, the yield + dividend growth rate provides a fair value of p/e.

• My first example is Dominion Resources (D): The yield is 4%. The 5-yr dividend growth rate is 6.4%. Thus the Gordon yield would be 10.4%. The fair p/e would be 1/ Gordon yield=9.61. The higher the dividend growth rate, the more the stock is improperly modeled—due to its deviation from bond status. In practice, I simply add the yield + dividend growth rate to get a fair p/e of 10.4. At the time of this writing, the actual p/e is 15.4 (First Call).

• My second example is Southern Company (SO): The yield is 4.62%. The 5-yr dividend growth rate is 4.1%. The fair p/e would be 1/ Gordon yield=11.46. The actual p/e at time of writing is 17.3 (First Call). My simplified formula would provide a fair p/e of 8.72.

• My final example is Frontier Communications (FTR): The yield is 10.3%. The 5-yr dividend growth rate is 0. Thus the Gordon yield would be 10.3%. The fair p/e would be 1/Gordon yield=9.7. The actual p/e at time of writing is 28.8 (First Call). My simplified formula would provided a fair p/e of 10.3.

What should we get from this rule of thumb? Well, it is a highly applicable first screen for dividend stocks, especially utilities and telecommunications stocks. It should not be used on growth stocks. It is only a first cut tool, one must study the company to see if the dividends are covered by cash flow and be sure that the earnings are growing to provide for dividend growth. In times of frothy markets, like today, it is important to get the income stream, regardless of price swings, especially if you are retired. I personally require a minimum 4% yield before I calculate any growth factors for dividends or price appreciation in the stock.