In what follows, I will describe my personal plan for accumulating wealth for use during retirement. In short, my plan is to invest in high quality companies that have long histories of dividend increases, taking advantage of the power of dividend reinvestment during the accumulation phase and collecting dividend payments for use as income during the distribution phase. Since I plan on investing in a Roth IRA account, I (hopefully) won't have to pay taxes on distributions.
I figure that about $50,000 (maybe more or less depending on where and how one chooses to live, of course), is a safe number to cover expenses throughout the year. Because of inflation, however, the amount of purchasing power of $50,000 in the future will most likely be drastically reduced, so the retirement plan must take this into account. The chart below illustrates this effect.
So, in fifteen years from now, for example, assuming 3.25% annual inflation, about $80,000 will be needed to have the same amount of purchasing power that $50,000 has today. Extend this 37 years into the future, and about $160,000 is needed, so this becomes my target annual income at the beginning of retirement - since I'd like to collect dividends worth this amount every year, I'll need about $4,500,000 of principal. With an average dividend yield of 3.5%, I should receive about $160,000 in dividend payments. Continued dividend increases should provide a protection against further inflation during the mid and late retirement years.
As someone with about 37 years until retirement age, I ask: Among all the factors that are taken into consideration when considering buying a company's stock, how can I be confident that the initial yield and dividend growth rate I'm getting with a dividend growth investment will keep me on track to reach my goal? That's the main question I address below.
Of course there's no way to tell for sure what will happen tomorrow much less 37 years from now, but it can be useful sometimes to run computer simulations to test various possible outcomes to get some idea. Building off of the work I describe here, I have found that a 5% initial dividend yield with an 11% total return yields an average value of about $4,600,000 after 37 years, from 100,000 simulations. The simulation assumes that $1,375 is invested each quarter (based on the $5,500 annual Roth IRA contribution limit) and all dividends are reinvested. The distribution of results is shown in the figure below. For more details about and assumptions used by the algorithm, please read "The Algorithm" section of the article linked above.
So, a 5% initial yield with an 11% total return becomes my base case standard (admittedly, I chose this somewhat arbitrarily). In order to reach my goal, when I now consider buying a stock, it must at least meet or exceed this standard's performance.
To evaluate a candidate stock's initial yield using my criteria, 37 years of the base case standard are simulated and the ending value is stored. With the candidate stock's yield as an input, 37 years are simulated using the same conditions as the base case until the total return % that can exceed the base case standard is found. (Here, Total Return = DGR + Yield). This two step process is iterated 1,000 times, and an average and standard deviation are calculated. This was done for a series of initial yields between 1.5% and 8.0%, and the results are in the figure below.
The line closely follows a power law. Going one step further, this information can be used to calculate, based on a series of dividend growth estimates, an acceptable buy price range. For example, let's say I'm interested in purchasing shares of Microsoft (MSFT), which is paying 0.92 cents per share this year. At the time of writing, it is yielding 2.84% with a stock price of $32.39/share. I use the equation reported on the graph above to figure out that with an initial yield of 2.84%, I would need a total return of at least 13.15% to exceed the base case standard. Subtracting the current yield (2.84%) from the required total return (13.15%) gives a required DGR of 10.31%.
If I then estimate that MSFT's next dividend increase will be between 8% and 12% (just for the sake of demonstration), the table below can be made by finding the price at which the total return is equal (or close to equal) to the equation in the graph above for each possible DGR.
MSFT Estimated DGR | MSFT Buy Price |
8% | $23.58 |
9% | $26.87 |
10% | $30.76 |
11% | $35.35 |
12% | $40.76 |
For another example, let's say I'm interested in possibly purchasing shares of Family Dollar Stores Inc (FDO), which pays a dividend of $1.04 per share this year. At the current stock price of $72.57, it yields 1.43%. Let's say I estimate that FDO's next dividend increase will be between 10% and 14% (again, just for the sake of demonstration).
FDO Estimated DGR | FDO Buy Price |
10% | $34.77 |
11% | $39.96 |
12% | $46.08 |
13% | $53.25 |
14% | $61.59 |
In this case, given my goals, it would be wise to wait for a price pullback to purchase shares of FDO since the current price of $72.57 is well above even the highest DGR estimate buy price.
DG investors use many different criteria to judge whether to make an investment in a particular company. Most of these important measures, like evaluating the safety of the dividend, for example, by looking at dividends paid/free cash flow, percent debt to capital or interest coverage, evaluate the financial health of the company. In addition to these company-specific measures, it is also important, I think, to have clearly defined investor-specific standards by which to evaluate the potential value of an investment, i.e. something to determine whether the investment fits in with the overall plan of the investor.