- Income investors generally have to choose between high dividend yield or high dividend growth.
- A simple model reveals no difference in total return in stocks with the same "Chowder number".
- Paradoxically, those focusing solely on a growing income stream may actually be better served by investing in high dividend yield rather than high dividend growth.
The repression of yields by the Fed's zero-interest rate policy has increased the popularity of a number of income securities, such as dividend growth stocks, REITs, MLPs, BDCs, royalty trusts, and high-yield bonds. As income investors navigate through the distinctive characteristics of each of those asset classes, however, a basal dynamic that inevitably emerges is the issue of dividend yield versus dividend growth.
Is it better to invest in high-yielding, slow-growing dividend stocks, or low-yield, fast-growing dividend stocks? At one end of the spectrum, we have the fast-growing MasterCard (NYSE:MA), which has a 5-year annual DGR of 30%, but only a 0.5% current yield (see recent Tim McAleenan Jr. "rocket ship" piece here), while the income favorite Realty Income (NYSE:O) has a current yield of 5.0%, but a much lower 5-year annual DGR of 5.1% (see recent Brad Thomas "Noah's ark" piece here).
The "Chowder Rule" is a popular approach devised by Seeking Alpha contributor Chowder that can been used to identify stocks that show a satisfactory combination of dividend yield and growth. To calculate the required Chowder number, which corresponds to a total return % target, you simply add the dividend yield to the historical dividend growth rate. Chowder himself proposed a minimum Chowder number of 8%, but with 12% as a criterion for opening new positions to provide a "buffer or dividend moat against [a] total return objective."
Modeling dividend yield vs. dividend growth
To study the issue of dividend yield vs. growth, a simple model was used. Let's consider four imaginary stocks, A, B, C and D with different dividend yield and dividend growth characteristics, but with the same Chowder number of 12%. Representative stocks with similar metrics are provided as comparative examples (thank you to David Fish's CCC list for the data).
- Stock A ("12/0") has a yield of 12% and a DGR of 0%. This would actually be similar to a high-yield bond or bond fund such as PIMCO High Income Fund (NYSE:PHK), which has a current yield of 10.3% and a DGR of 0% (Chowder = 10.3%).
- Stock B ("8/4") has a yield of 8% and a DGR of 4%. AmeriGas Partners LP (NYSE:APU) has a current yield of 7.4% and a 10-year DGR of 4.2% (Chowder = 11.6%).
- Stock C ("4/8") has a yield of 4% and a DGR of 8%. BAE Systems plc (OTCPK:BAESY) has a current yield of 4.3% and a 10-year DGR of 7.8% (Chowder = 12.1%).
- Stock D ("0/12") has a yield of 0% and an EPS growth rate of 12%. Gilead Sciences (NASDAQ:GILD) does not pay a dividend, but has a 5-year EPS growth rate of 11.9% (Chowder = 11.9%).
This model was conducted using the following rules and assumptions:
- An investor started with 100 shares of each stock valued at $100.
- The stock price of the stock rises at the same rate as the dividend growth rate. While this is a big assumption, it does make sense from a fundamental point of view as growing dividends are usually a result of growing earnings, and one could expect the share price to rise accordingly to maintain the same PE and payout ratios.
- Dividends are paid out at the end of each year, and are entirely reinvested into the same stock at year-end price.
Data on the stock price, number of shares, annual dividends paid and total return over a 30-year period as generated by the model are given below.
The above graph shows the share price of the stock over a 30-year period. Unsurprisingly, the bond-like 12/0 stock shows no price appreciation while the 0/12 growth stock has the highest per-share gain.
The second graph shows the number of shares obtained over a 30-year period. The bond-like 12/0 stock will have the greatest increase in the number of shares because its high yield is being reinvested into purchasing new shares. On the other hand, the no-dividend stock will not see its share count grow at all.
This is where it starts to get interesting. The model shows that a low-yield, high-DGR stock (e.g. 4/8) will never catch up to the bond-like 12/0 stock in dividends paid per year! This is in contrast with previous articles (e.g. by Cash King here, although stocks with different Chowder numbers to each other were used) that show a low-yield high-DGR stock eventually overtaking a high-yield low-DGR stock in dividends paid per year.
I believe that the reason for this comes down to dividend reinvestment. For the 4/8 stock, the dividend/share rises from $4 to $37 after 30 years, which is a near 10-fold increase in dividends paid per year (yield on cost = 37%!). This seems highly impressive compared to constant dividend/share amount of $12 for the 12/0 stock, until you consider that the number of shares for the 12/0 stock ballooned from 100 to 2675 over 30 years. By comparison, the share count of the 4/8 stock increased from 100 to only 287 over the same period. This model therefore suggests that with the same Chowder numbers, a low-yield high-growth stock will never produce more income than a high-yield low-growth stock, all other things being equal.
Finally, the total return data is shown. All four lines are identical, showing that with the same Chowder number (which remember corresponds to a total return % target), the theoretical portfolio value at the end of the experiment is the same.
This can also be demonstrated algebraically. Let the yield (%) of a stock be x. Let the annual dividend growth rate (%) be y. Let the Chowder number (%) be N. Hence N = x + y.
If at the start of the year, an investor owns 100 shares valued at $100, his portfolio value will be $10,000. At the end of one year, his dividends paid out will be $10,000 * x = 10,000x. At the same time, his shares have appreciated by $10,000 * y = 10,000y. Therefore his portfolio value at the end of one year will be $10,000 + 10,000x + 10,000y = $10,000 * (1 + x + y) = $10,000 * (1 + N). In other words, the total return after one year (or after any number of years, by extension), is simply given by the total return number, i.e. the Chowder number, and is irrespective of the distribution of dividend yield and growth.
These results corroborate with an article published by Seeking Alpha contributor Eric Landis last year, showing that stocks with the same Chowder number have very similar total returns. The present study provides graphical data to help readers understand this relationship further, and also demonstrates that with the given set of assumptions used here, the total return is actually exactly the same.
When I first started to construct this model, I had not yet read Mr Landis' article, and hence had not considered what result to expect. In fact, I had actually predicted that although the bond-like stock A (12/0) would provide a higher current income in the early years, it would eventually be overtaken by stocks B (8/4) and C (4/8) due to their significantly higher DGRs. Combined with the higher share price appreciation that was conferred to B and C (due to their increasing earnings), I had expected that the total returns of B and C would leave the return of A in the dust after a certain period of time.
Hence, the result that the total return of all four stocks was identical was surprising to me. Upon deeper inspection of the model, the reason for this became clear. The fact that the high-yield bond-like stock A exhibited zero dividend growth is actually an advantage in that the 12% yield of the stock can be continually invested at the same initial price. On the other hand, the share price appreciation of stocks B and C means that the even though the dividends/share is increasing, relatively fewer shares can be purchased each year (compared to stock A).
In summary, the model shows that stocks with equal Chowder numbers (yield + DGR) would be expected to have the same total return profiles.
A small subset of dividend growth investors profess to not ever care about the current price of their stocks, but to only focus on growing an income stream. They point to dividend champions like Coca-Cola (NYSE:KO) and Colgate-Palmolive (NYSE:CL), that although sport relatively low dividend yields of only 2-3%, have dutifully increased their dividends by 10-11% year after year (Chowder = ca. 13%).
However, if a growing income stream is one's only goal, the results of this study suggests, paradoxically, that buying and reinvesting in a high-yield, low-DGR stock actually produces a faster-growing income stream compared to a low-yield, high-DGR stock (assuming they have the same Chowder numbers).
To address this from an angle, consider that a purely income-focused investor will not "benefit" from the significant price appreciation conferred by a high DGR stock, as he is only concerned with the income stream and not with the price of the underlying security. In fact, the higher the DGR and the lower the current yield, the greater the proportion of the total return will come from price appreciation of existing shares, rather than from the reinvestment of dividend income to purchase additional shares (assuming the same Chowder numbers).
This result may seem surprising to "income investors" who invest in these stocks purely for income and are completely agnostic to share price. Do they have something in common with "total return investors" after all?
In the subsequent article, the dividend growth vs. dividend yield issue will be further analyzed. Various aspects will be discussed, including the pros and cons of dividend yield vs. dividend growth, the applicability and limitations of the model, psychological aspects, tax considerations, and historical data. Readers are invited to share their thoughts on these issues in the comments section below.
Editor's Note: This article discusses one or more securities that do not trade on a major exchange. Please be aware of the risks associated with these stocks.