Dividend Dilettante

Dividend Dilettante
Contributor since: 2013
I just wanted to provide another angle to consider. Taking everything into consideration, I think MSFT is the clear winner as far as dividends are concerned, but it's worth nothing, I think, that during the time frame I considered, MSFT had 3 dividend payments below the conservative estimate, while all of DIS's increases were well above it. That being said, DIS did of course freeze their dividend while MSFT kept on increasing it.
Thanks for the article.
When you looked at dividend growth, you looked at 'absolute' growth, i.e. simply how much the dividend grew each year. An alternative way of understanding dividend growth is by looking at the growth relative to how much leftover cash the company had available to spend on dividends in any given year.
I've made a blog post on SA with two graphs comparing DIS's and MSFT's dividends in this way from 2004-2012. They are here:
http://bit.ly/IoGFnz
Information about the "dividend drill model" I used is here:
http://bit.ly/1chhxKt
AgAuMoney, thanks for commenting.
I think that's a great idea. I've cleaned up the code, added comments and uploaded it to:
http://bit.ly/1bC1ubx
As a software engineer, you'll probably find my code messy and inefficient - I wrote it mostly for my own use, and I am, by no stretch of the imagination, an expert programmer. Let me know if you have any questions about anything.
Also, there were two minor typos in the article that I should point out:
1) MO was not included in the analysis.
2) The start date for the back tests was actually September 19, 2000, not July 17, 2000.
Richard Berger,
Thanks for your accolades.
I agree that analyzing how long the collected dividends sit around uninvested in the manual reinvestment case is important. As suggested above, this could explain why the DRIP is usually the out-performer.
David Fish,
Thanks for commenting (the dividend Champion/Contender/Cha... spreadsheet has been a starting point for a lot of my research and is a fantastic resource). My intuition is also that while the cash is not invested between buy signals the manual reinvestment strategy looses ground on the DRIP. If this were the case, however, one would expect that as the portfolio size increases the manual reinvestment strategy performance would improve since presumably there would be buy signals more often, i.e. the cash would be sitting around uninvested for a shorter amount of time. The first bar chart demonstrates, however, that the manual reinvestment strategy actually performs progressively more poorly as the portfolio size increases. I'll dig a little bit deeper and see how long the time is between buy signals in the manual reinvestment strategy.
Readjusting the buy criteria to what you suggest in parentheses would be an interesting experiment.
"Who on earth plans a retirement assuming a 12% total return for 38 years ?????? "
Where total return = DGR + initial yield, and total return = 11%, the answer is, obviously: I do.
I think we mean different things by "total return"
Thanks for the links, SDS
Great point Gary!
I should have taken this into consideration
Thanks for commenting!
Thanks stink, I appreciate it!
Thanks for the fact checking!
SDS,
Thanks for your comment.
It does look like a log-normal distribution - is there a reason you're mentioning this?
The figure you're talking about is attempting to communicate the following: If I'm considering purchasing a stock yielding, say, 2.5%, I'm going to require a total return (where total return = DGR + current yield) of about 13.75%. This is calculated by substituting 2.5% for x in the equation. This means an implied DGR of ~11.25%.
Does that make sense?
Thanks for commenting.
"First, finding "dividend growth stocks" that have a 5% yield is going to present a slight problem."
That's not what I'm saying. What I'm saying is: A 2.5% yield is fine, but I'm going to need a total return of about 13.75% for that initial yield to reach my goal, meaning an implied dividend growth rate of ~11.25% - for example.
"Second, building a basket of stocks with a yield of 1-8% means that you will have to have the basket more heavily represented by the 5% and above yielding stocks."
Sorry, I don't think that's correct - the important thing for me is DGR + initial yield. So I can have an average yield of, say, a mere 2.5%, but as long as the DGRs are high enough the result should be (based on the simulations I describe) about the same as if I were holding stocks yielding more but growing less. That's what my 3rd figure is attempting to communicate.
"Fourth, your assumption about you need 4.5 million dollars of stock to throw off your income target is incorrect. As Robert points out, if you had 4.5 million dollars today and invested it at 3.5% then you'd achieve your income target."
I don't understand your point here. I don't think that's what Robert was pointing out, unless I misunderstood his comment. I should mention that $ 4.5M is simply an estimate I use, based on a projected average future yield and costs of living.
Perhaps I didn't explain it in the right way in my article, but my plan is to slowly save money over 37 years, investing in DG companies along the way using dividend reinvestment, then collect the dividends as income during retirement.
leanne, thanks for commenting. I would, however, suggest that you consider my article as a model with important limitations and assumptions to be considered while interpreting the results. Also, it's good to do your own home work and not consider my results as 'the right way of thinking' about the topic -- but thanks! :)
Chief, good point. However, different investors have different commission costs/brokerage policies, so it depends on one's individual situation. For most, probably, (like me), commission costs are something to keep in mind when investing dividends without a DRIP.
tomlos, I'm glad liked it! I enjoyed doing the research
DVK, thanks for your comment! I would have rather written something that better reflects how people actually invest (a portfolio with more than one stock, for example), but I think that my analysis is still interesting to think about.
maybenot, thanks for the comment. I agree with you
LH, thanks for commenting. Another factor (as someone mentioned below) is that commission costs could build up without a DRIP
Roady, thanks for your comment
Hawmps, very true
Be Here, interesting idea, thanks!
I adjusted the DRIP model to reinvest 23 days after the ex-date (after a cursory look at CL's ex-date/pay-date relationship, this is about the average distance I observed). Here are the current values (04/15/13) from a starting $1,000 investment made on 01/01/02:
DRIP CL, invest $500 at same time as div. reinvestment: $ 47,537.70
Collect divs, buy on dips, invest $500 every quarter: $49,406.02
It shows whole, but behind the scenes it must be using fractional
Pendragon, for your first point: I see what you mean. However: If you go to Colgate Palmolive's total return calculator (http://bit.ly/10Bm9BV) and enter January 2, 2002 as the purchase date of $10,000 worth of stock and reinvest dividends, the current value they give is $ 25,804.34. Presumably their calculator uses the actual pay-dates to reinvest dividends. If I adjust my model to a DRIP with no additional contributions (this should be the same thing as what CL's calculator is doing) I get a current value that is higher than this, at $26,135.55
shootpar, Thanks for your criticism. The problem is that 20% drops don't come along very often, and while waiting for it to happen the cash that's being held earns nothing. I don't fully understand how "the current market run-up works to my advantage in this simple example"
Could you say more?
Pendragon, thanks for commenting.
I agree that my technical buy signal isn't perfect, but neither is the ability of people to identify the point at which the stock's price has bottomed. I'm not sure I understand your point when you say, "even the tiny differnce you found is wiped out by the dip you missed." Could you elaborate on that?
As for your second point, I ran another test just now with Consolidated Edison Inc. (ED) investing $1,000.00 on Aug 1, 1989. Here is the result:

Dividend Reinvestment investment value on 4/12/13: $187,708.70
Collect and buy on dips investment value on 4/12/13: $184,030.32
xinge, thanks for your comment. The window was April 4, 2010 to April 4, 2013
Jeep5ter, I'm glad you like it! I enjoyed doing the research!
Part-time,
Thanks for your comment. The PAAY systems looks good. I actually read about it a few months ago in one of your articles and made a comment about perhaps allocating to each stock an amount of money proportional to the PAAY rather than an equal amount to the highest 10. I think my results here support the rationale behind your approach generally, but I also think my results support my 'proportional allocation' modification to your PAAY system. I'll think about how I might go about using python to back test and compare an 'equal allocation' PAAY system with a 'proportional allocation' PAAY system. It's an interesting question.
whisper, You are correct in that that is a limitation of my analysis, however I disagree that my overall point of there being a close relationship between initial yield and total value is affected. While you are correct that the absolute value of the total values I report would be different than the actual values, the relative analysis is still valid. For example, if you go to Colgate Palmolive's investment calculator on their website (http://bit.ly/10Bm9BV) and type in the dates for the example I gave above, the percent difference between the total returns for those two days is the same as the percent difference I report.
maybenot, thanks for commenting, I appreciate it!
Thanks chasingalpha, glad you liked it