Commentary

September 30, 2019

The Importance of Dividends

 
I have long believed that financial instruments, in a sense, compete with each other based on investors’ risk and return preferences.  I started my career as a bond trader and I thought then, as I do now, that bond yields affect the valuation of common stocks.  The premise is that the higher the level of interest rates on bonds, the lower the valuation multiples on common stocks.  I believe this view is currently the conventional wisdom.
 
However, at the present time we are seeing $17 trillion of sovereign debt (bonds issued by governments around the world) carrying a negative yield to maturity.  Therefore, it becomes difficult to understand how these yields are economically justified.  Mathematically, as bond yields approach zero, equity valuations should approach infinity, and once bond yields become negative, it requires imaginary numbers to connect these levels to equity valuations.  Since equity valuations have remained relatively “normal,” it is clear that this long term relationship between stocks and bonds has now been broken.
 
Consequently, it seems to me that more attention should be paid to dividends, dividend yields and corporate dividend policy in valuing common stocks.  At present, the S&P 500 index produces a dividend yield of roughly 1.8% per year.  Put into historical context, dividends have generally made up 40% of the return investors received from an equity investment.  If that relationship were to hold true today and going forward, the outlook for equity market returns would be 4.5% per year over the long term.  However, the appreciation of the S&P 500 over a similarly long term has been 6.5% per year.  Adding a 1.8% dividend yield to that produces a total return to investors of 8.3% per year over the long term.  That is both more pleasing to investors and closer to long run stock market returns.
 
I believe that corporations that pay dividends should have a clearly enunciated dividend policy so that investors can judge the prospective future value of the company’s stock.  Home Depot has announced a policy of paying out roughly 40% of trailing earnings, and as the company has prospered, so have the shareholders.  Other companies, such as AES, have announced its intention to pay higher dividends over a set number of years.  In the case of AES, the company proposed in 2014 paying 10% more each year for 4 years. As a result, the stock almost doubled during that time period as investors saw the company deliver dividends as promised.
 
Unfortunately, in my view, some companies have included borrowing funds to buy back their own stock as part of their dividend payment policy.  If a company, such as 3M, is able to borrow at less than 3%, and buy back shares with a dividend yield of 2.5%, it can claim that the borrowing pays for itself.  Since interest costs are tax deductible and dividend payments are not, the effective cost of this “financial engineering” is slightly negative on an after tax basis.  This has been 3M’s plan for the last several years and it shows up as a $1 billion increase in debt each year.  The cost of carrying this debt is currently negative, but that may not always be true.  Also, it is not possible to continue such an approach indefinitely as the debt on the corporate balance sheet has limits, based on credit rating service criteria. 
 
I would like to see dividends become tax deductible to corporations to discourage the practice of borrowing to buy back stock.  I would also support taxing dividends to individuals at current income tax rates, to keep taxes in balance.
 
In conclusion, I believe that corporations that pay dividends and have a clearly enunciated dividend policy, would provide a far superior method for investors to judge the prospective future value of the company’s stock, as opposed to linking valuations to bond yield levels.


Raymond A. Beplat, CFA
Chief Investment Officer







Archived Commentary