In this article, I created a model to allow sensitivity analyses to be performed on a dividend stock portfolio being targeted towards producing a specific level of income, as opposed to a value.
There are eight variables that comprise the model, but I would contend five of these are within the investor's control. Beginning Value and Going-in Dividend Yield, are based on a current state of facts that any devoted dividend growth investor has ready access to. Values for Additional Contributions, Income Target and Investment Horizon are chosen by the investor and can be altered in the investor's discretion at any future time.
Actual changes in the three remaining variables, Inflation, Dividend Growth Rate and Future Share Prices, will be beyond the investor's control. Nonetheless, in the context of this model, these variables have some interesting characteristics:
Inflation - Its effect in this model is limited to inflating the Income Target to the end of the Investment Horizon.
Dividend Growth Rate - Due to the observed "stickiness" in payment patterns, particularly for reasonable-yielding blue chip companies with a clear dividend history and stable operating metrics, an investor should be able to project a reasonable range for this variable for short to intermediate Investment Horizons with a greater degree of confidence than for price changes. Of course, the higher the current yield, the more risk there is in predicting dividend growth or sustainability (the so-called "yield trap" effect).
Future Share Prices - This is arguably the single most difficult variable for an investor to estimate due to the added element of "Mr. Market". However, in a model designed to target a level of income as opposed to value, the effect of this variable is significantly limited as price is only needed to determine in the number of future shares purchased through additional investments and reinvesting dividends.
This model is clearly different from an asset allocation model which, as it is targeting portfolio value and not income, must fully account for asset price changes. Asset allocation models also must consider additional variables such as interest rates and percentage of allocations. This does not mean that I believe there is less risk in dividend growth investing! It just means that it is easier mathematically to relate cause and effect.
With this (boring stuff) as background, let's do some additional sensitivity analysis. As a refresher, Larry, a 55 year old, saved well for retirement having an IRA that has a $400,000 current value. It is comprised of roughly equal amounts of four stocks: Procter & Gamble (PG), Johnson & Johnson (JNJ), McDonald's (MCD) and Exxon (XOM). Combined, these stocks have an average current yield of about 3% and experienced a dividend growth rate of about 10% per annum from 2007 to 2010. I converted these four stocks into "Larry's ETF" and assigned a hypothetical number of shares at 10,000 and current share price of $40 per share. I fixed the annual anticipated stock price appreciation on the portfolio at 3%, but this only serves to determine the number of shares purchased when dividends and new contributions are invested.
Let's mix things up a bit from my first article and assume for this one that Larry has done some cash budgeting and he has determined that he will be able to contribute $10,000 per year to his IRA for each of the next ten years. What combinations of Going-in Dividend Yield and Dividend Growth rates will allow him to achieve his goal of $50,000 of dividend income per annum when he turns 65?
As seen in the spreadsheet below, with a 3% Going-in Dividend Yield and a $10,000 per annum contribution, Larry would need a dividend growth rate of 9.25% per annum to achieve his goal. This is 0.75% less than the average annual dividend growth rate for this portfolio from 2007 to 2010. That's a fairly tight margin of error.
Click to enlarge:
Let's say now that we replace McDonald's and Exxon with Consolidated Edison (ED) and Waste Management (WM) in Larry's portfolio. The Going-in Dividend Yield for this grouping is about 4% and the average dividend growth rate was about 6.75% per annum from 2007 to 2010. What dividend growth rate going forward would allow Larry to achieve his Income Target of $50,000 in this case? As seen in the spreadsheet below, a dividend growth rate of 5.68% per annum would be required. As a result, if this portfolio has a dividend growth rate of not more than 1.07% less than the dividend growth rate achieved from 2007 to 2010, then Larry will achieve his goal.
Finally, let's say we now replace Procter & Gamble and Johnson & Johnson with AT&T (T) and Eli Lilly, (LLY) (so the portfolio is now AT&T, Eli Lilly, Consolidated Edison and Waste Management) bringing the Going-in Dividend Yield to about 5%. The average dividend growth rate for this grouping was about 5% per annum from 2007 to 2010. As seen in the spreadsheet below, a Dividend Growth Rate of 2.93% per annum would be required to achieve a $50,000 Income Target. As a result, if this portfolio continues to have a dividend growth rate of not more than 2.07% less than the dividend growth rate achieved from 2007 to 2010, then Larry will achieve his goal.
Conclusion: A higher Going-in Yield significantly reduces the Dividend Growth Rate that is mathematically required to achieve an Income Target for a dividend growth portfolio with a ten year investment horizon. However, higher current yields are sometimes associated with higher dividend vulnerability. As a result, investors should weigh risks carefully before choosing specific stocks to increase their current yield.
Also, the above combinations of going-in yields and dividend growth rates are specific to the selected value for the model variables, namely: The $400,000 Beginning Value, Additional Contributions of $10,000 per annum, a $50,000 Income Target, a ten-year Investment Horizon and 3% per annum Future Share Price increases. While altering any of these variables in a reasonable manner will not change the inverse relationship between the going-in yield and the corresponding required dividend growth rate, it could materially change the specific combinations that are necessary to achieve an income target. Any investor trying to replicate this approach should carefully choose values for the above variables that match their personal situation.
In the next article, I will demonstrate the effect of changes in the Investment Horizon on the above relationships.