In the comments on this article "Six" posted a comment that raised some interesting issues about dividend investing versus growth investing that I thought would be worthy of some analysis.
The first paragraph of Six's comment said (typos corrected):
"$100k invested into a dividend paying stock paying a dividend of 3% per year yields $3000/year for reinvestment and allows the opportunity to compound the interest. Over 10 years if you can compound growth at 3% you create over $34K of additional value. With no compounding it creates $30K and no additional risk was necessary."
First, funny that Six called it "interest" because he set this example up like an annually compounding 3%, ten year CD which, pretax at least, will get you to the same place. Just to be clear, the compounding creates about $4,392 in additional value over and above the $30,000 in pure uncompounded interest, not $34,000....I think that's what Six meant. What dividend growth investors want even more though is for the "interest" (really the dividend) itself to grow at least annually. The "compounding" is a separate issue. I believe Six tries to clear this up in the next sentence
"The compounding is the secret of DG stocks. When you compound compounding through reinvestment in a DG stock the advantage is tremendous. "
I presume what Six means in saying "compound the compounding" is that you get a double benefit from reinvesting dividends plus the growth of the dividend itself. Nonetheless, I couldn't help but be reminded of the "Fidelity Fiduciary Bank" scene from Many Poppins when I read this.
"When you compare this with other investment strategies it becomes clear that the DG strategy has a key advantage. If you purchase a growth stock that has no dividend and it increases in value too, what do you have? A statement that says you have more money."
I part ways with Six a bit here. I am presuming Six is continuing the example of reinvesting dividends to "compound the compounding" since this is the next sentence, but if you do that, you don't have any money in your pocket because you reinvested the dividends. All you really have is the same account statement that says you have more money...right?
But getting to more of the substance of Six's statement, I decided to test whether an investment in a dividend stock with reinvested dividends is superior to a growth stock, apples-to-apples. To do this, one must assume that over time, the dividend yield plus price appreciation (so called "total return") will converge with earnings growth so that, for example, if earnings growth averages 6% per annum, then over an extended time period, the combination of the dividend yield plus the appreciation should equal roughly the same 6% per annum on average. This manifests itself in the analysis by calling two stocks that have the sum of the annual dividend yield plus the annual price appreciation equal to 6% "apples to apples".
Following on this, as seen in spreadsheet 1 below, Account A is a taxable account, in which $10,000 is invested in 1,000 shares of a $10 per share dividend growth stock that has a 3% dividend yield and 3% price appreciation per annum, a total 6% return per annum (this effectively implies the dividend is growing at 3% per annum). Dividends are reinvested annually net of taxes. Account B is a taxable account as well, in which there is also $10,000 that is invested in 1,000 shares of a $10 per share growth stock that has no dividend, but 6% price appreciation per annum.
We advance the time clock and open the account statements after ten years... Account A has $17,162.57 while Account B has $17,908.48. The growth stock account has a higher balance? Why is this? Because in Account A, you had to pay taxes on the dividends every year and the compounding effect on that money is lost (I back-checked the model by taking taxes out as if it were in an IRA and the results become identical between the two accounts so the entire difference is attributable to the effect of taxes). Now granted, if you sell the growth stock in Account B at the end, you will owe higher capital gains tax than if you sell the dividend stock in Account A, and this makes the final result much closer, but this still does not "even it out". The growth stock, apples-to-apples, is the winner in a taxable account.
"You must sell the stock to create a real profit. If you sell you risk selling what you may believe is a good company. This is the catch 22 of growth investing, you have to sell good companies to realize ANY profit. And if the price of the stock falls you can't do anything but wait for it to appreciate or accept losses."
Here the commenter seems to be referring to a non-reinvestment case appearing to suggest that you can take the dividend as a "profit" and go buy some widgets or whatever. But with a growth stock, you have to sell it before you can do this. But can't you just sell a small portion of your growth stock to achieve the same end? Some seem to have the perception that selling a stock is an all-or-nothing proposition. With transaction costs reduced to almost nothing today, though, it is easy to sell a portion of a stock to mimic a dividend if you wish. Selling is your choice, not the company's choice as it is with a dividend.
As such, I decided to test this as well. I used the same assumptions as above for Account A and Account B, but instead, in Account A, dividends are not reinvested, and, in Account B, a portion of stock, equal to what you receive as a dividend in Account A, is sold each year creating a "synthetic dividend". As above, in Account A, tax is paid on the full amount of the dividend received while, in Account B, tax is paid on the gain on the amount of stock you sold. So what was the result? Well...after ten years you have the same exact amount, $13,439.16, left in Account A and Account B (although a vastly different number of shares and share price) but while you've had a total cash flow over the same period from Account A of $2,923.29, the cash flow from Account B is $3,297.64 (see Spreadsheet 2 below). Why is the cash flow in Account B higher? In Account B, you are only paying taxes on the portion of the synthetic dividend sale that is a capital gain whereas in Account A you have to pay taxes on the full amount of the dividend. Again, the growth stock is the slightly better choice after accounting for taxes.
"A DG investor has a different set of options. While their stock appreciates it pays dividends that allow them to profit or reinvest while they STILL OWN THE STOCK. If the price of the stock falls and they still consider it a good investment then they can take advantage of the lower cost or profit while the stock is falling."
Again, in the second sentence, Six seems to view the sale of a stock as an all-or-nothing proposition. However, it is not, as I discussed above, but the last sentence is the most interesting in the entire comment and is actually quite subtle and complex. In my mind, it brings out the importance of the psychology involved in determining which investment style one prefers. It is occasionally written in these forums that a falling stock price is "good" because then you can buy the stock "on sale". But this is true for a growth stock too, isn't it? Somehow, it seems different for a dividend stock, particularly one that churns out consistent earnings and dividends even when its price is down due to macro market factors.
I know that I find comfort when a seemingly benevolent entity, a share of which I own, says "Yes, our stock is down but, we're OK and here's a little money back to prove it. You can do something else with it if you want, but you can also buy more of our stock because it's cheaper now". Other people, as we know, are perfectly comfortable riding the roller coaster of growth stocks, buying and selling at any point in the market cycle while experiencing little cognitive dissonance (the ultimate example being day traders). As a buy-and-hold person though, that's not me. I fret over the decision almost every time, so having a little capital "forced" back to me in the form of a dividend is probably not a bad thing.
Numerically analyzing a more real world case where stock prices don't move in a linear fashion (as my above models use) is a little complicated but, I was curious. Might this tip the math scale in favor of dividend stocks because you will be able, on occasion, to reinvest dividends when the stock price has actually fallen? I expanded my model to include a random function that allows the annual return to vary year-to-year with the same expected 6% return being the median point. I tied Account A and Account B together so that they were both experiencing the same "random market" effect from year to year. For the Account A model, I maintained the dividend the same as in the steady return model to simulate the real world steadiness of a dividend "stalwart" independent of general market influences on the stock price.
I provided a maximum variance up and down, per annum, for the dividend stock price of 20% meaning, at most, the stock could lose 17% or gain 23% in any given year. For the growth stock in Account B, I allowed a maximum variance of 25% per annum in the stock price meaning at most the stock could lose 19% or gain 31% in any given year (*see note below).
By recalculating the spreadsheet (F9 for you Excel mavens), I could run a virtually limitless number of tests. Approximately 90% of the time, the value in the tenth year for the growth stock account exceeded the dividend stock account but, the opposite result occurred particularly when "the market" was weak in the early years which allowed reinvested dividends to purchase more shares. But even when the growth stock "won the race", there were often points during the ten-year period when, if you had to liquidate the accounts, the value of the dividend stock account would have exceeded the growth stock account. In fact, if you plug in the returns on the S&P 500 for 2000 through 2009 and subtract 3% per annum from price appreciation of the dividend stock to account for the 3% dividend (see Spreadsheet 3 below), it still outperformed the growth stock for this period. Looks like Six has a point here and it might explain, in part, why dividend investing has waxed in popularity over the past several years.
"Of course NOTHING in life is guaranteed... be it dividends or growth. All equity is (should) be based on buying undervalued companies. The rewards of finding an undervalued DG stock is getting paid while your investment increases in value. Starting from a list of companies that have a history of increasing dividends offers an investor the opportunity to find values that will pay them to wait for share price appreciation. Then they can compound their compounding dividends. "
Well first, of course a popular version of the adage says "nothing is guaranteed in life, except death and taxes" (oddly two subjects I find myself writing about quite a bit). As for "buying undervalued companies", that opens up a new subject that strays a bit from the main point of this article, but an interesting one nonetheless that I plan to address in a future article. Finally, as to getting paid while a stock appreciates in value, again, the math indicates that, apples-to-apples, in a taxable account, because of taxes, getting paid while the value increases is generally not as good as not getting paid and having the value additionally increase by the amount you would have gotten paid (sounds like the line at the end of the movie Tootsie "I was a better man with you as a woman than I ever was with a woman as a man"). Somehow though, my heart still tells me something different.
"All equity investing is about "2 in the bush" and different investors approach it different ways. I love cash flow..."
I think that is a great line! Six is right on the money (pun intended). Investing is always about trading a bird in the hand to get two in the bush whatever your style is. There is uncertainty in any method (including holding on to the bird in the hand). I particularly liked it though when Six says "I love cash flow" as I think it was a raw emoting of why Six really likes dividend stocks. Perhaps for some, maybe even most dividend growth stock investors, if they step back and really think about it, cash flow is the yin that satisfies the conservative side that is saying "I want to feel like I am in a relatively safe investment that has a steady income" while the wilder yang side is satisfied by "I still have a good chance for dividend growth and price appreciation".
That probably explains why I overweight dividend growth stocks in my portfolio even though the math indicates I should perhaps not, particularly as a tax sensitive investor. I am comforted by the fact that dividends tend to make stock prices less volatile meaning that if I do have to sell when the market has taken down stocks, I am less likely to suffer a large loss. I guess, in a sense, I view the taxes paid on dividends as sort of an insurance premium that allows me to own a less volatile stock and gives me (well, really forces upon me) the option of taking some money off the table without having to experience the "cognitive dissonance" of selling a stock. See, even a spreadsheet wonk like me can be subject to human emotions.
*The modest difference in variance is meant to account for the real life difference in volatility for growth versus dividend stocks. When I changed the variance, it did affect the outcome to differing degrees. I relied on my best judgment as opposed to hard data to settle on the final variances used. Further, because the random function is only used for 10 "answers" per run (one per year), there are not enough samples to maintain the total within an acceptable range that would allow the median return to be achieved closely enough an acceptable number of times. To counter this, I employed an "adjustment factor" to drive the result back to the median. This, infrequently caused the return in a particular year to slightly exceed the maximum or minimum thresholds. I do not believe this prejudices the analysis.