Many investors goal is "Maximum real total returns after tax" as it was defined by Sir John Templeton many years ago. My goal is slightly different and it is "As big as possible positive real total returns after tax during intermediate (~ 3-5 years) and long time periods"
Often financial advisers just split total stocks returns into 2 parts as shown in Credit Suisse (NYSE:CS) picture below.
From this picture it is clear that capital appreciation part dominates during 1926-2006. Therefore many investors nowadays just ignore dividend contribution. In order to capture the capital appreciation a "regular" investor (without short positions) must sell stocks and I'll show elsewhere cons and pros of such decision. But if I go 1 step beyond this picture and try to understand causes of the capital appreciation I conclude that most probably the following factors are important in the long run:
a) growth of earnings (or another metric that investors in masse use to buy stocks, for example organic grow, revenue, etc..) due to better productivity and partially due to inflation;
b) growth of valuation due to increase of money inflow into stock market (in average life conditions on our planet improve with time and more and more people have some extra money they can invest. BTW, USA stock market still has good reputation and it is important for money inflow into it).
On long run earnings and divdends are correlated: both grow with time although at different spead in various time periods as shown below.
Although for many companies payout ratio fluctuates with time (mostly because of earnings fluctuations) for a big set of companies (e,g, all 500 stocks in S&P500) during some resonable long average period these fluctuations are zero out as shown above. Prof. R. Shiller demonstrated the same for inflation-adjusted numbers in his excellent book "Irrational Exuberance":
If we ignore fluctuations we can see that both stock prices (i.e. , capital appreciation) and dividends increase with time, so dividends might play a second role in total return.
Therefore total capital appreciation can be splited into growth of dividends (or earnings) and change of valuation as Societe Generale researches did in the following pictures (please note change in color codes).
In a long run (~10 years) about 94% of my goal can be achieved with Dividend Growth Investing (NYSE:DGI) according to graphs below I took from a good James Montier book "Value Investing: Tools and Techniques for Intelligent Investment".
Prof. R. Shiller, Credit Suisse (CS) and Société Générale (SG) quants used all stocks traded in USA to build these graphs. A dividend investor by definition does NOT invest in stocks without dividends. So subset of abovementioned dataset should be used. I do not have such information and look on a simple task - to figure out how dividends and stock prices are correlated for seasoned dividend growth companies. I took 125 companies which increase dividends at least 20 years from today's David Fish list (dripinvesting.org/Tools/U.S.DividendCham...) and "verified" their dividends at Yahoo.com (please note that David Fish information is more precise but I wanted to do quick and dirty test). I trimmed out companies that have not enought Yahoo data or which dividends increased in more than 50 times (probably because of recent huge special dividends). As the result I got list of 100 companies. I compared dividends and prices grow numbers for these 100 companies for last 20 years and got the following graph:
Both factors are purely correlated. For most of companies dividends increase is below 20 /median ~ 4.6 /and for ALL companies price increase is below 5.5 /median ~ 1.88/. Nevertheless up trend is visible at the graph. Hence for last 20 years SG results are valid for companies with growing dividends. See also 25 March 2012 addition.
I mentioned that payout ratio (dividends/earnings) fluctuate due to significant earning noise. Indeed the following CS picture for dividends and earnings variations confirms this observation:
As probably any investor I dislike variations so smaller variations in dividends seems me atractive (well, as a nanotechnology researcher I do see a lot of noises and quantum fluctuations at work and even use them for banks dividend analysis in seekingalpha.com/instablog/725729-sds-se...). But I'd like to stress the big (in my opinion) difference: dividends can be only positive or zero while earnings can be positive, negative or zero and any "regular" investor hate negatives (behavior scientists found that a human hate losses about 3 times stronge than love gains). I calculated Geometric Average Dividend Payments Growth coefficient from R. Shiller data for 1871-2010 and it is 3.27%, so a dividend growth investor (DGI) can pressume such increase of dividends and hence real variations DGI faces even smaller than numbers calculated by CS quants. Also I calculated Geometric Average Earnings Growth coefficient from R. Shiller data for 1871-2010 and it is 3.86%. Therefore Dividend Payments Growth is equal to 84% of Earnings Growth now I know how SG quants got the number.
Do you still remember that Sir John Templeton and I wrote "returns after tax", so what is about taxes? In short I hate them (esp. in Silicon Valley, California there at the top of federal and huge state taxes we have ~ 10% sales taxes) but pay them. But I demand from all politicians that "represent" me at least to keep low taxes on dividends and I beleive any dividend zealot must do the same. And the max I think that double taxation of dividends in US should be stopped.
Of course there are other arguments in favour of dividend investing readers can find them in excellent books:
"The single best investment: creating wealth with dividend growth" by Lowell Miller (actually this 2006 book changed my investment philosophy);
"Dividends don't lie: finding value in blue-chip stocks" by Geraldine Weiss (substitutes are "The Dividend Connection: How Dividends Create Value In The Stock Market" by Geraldine Weiss & Gregory Weiss and "Dividends Still Don't Lie: The Truth About Investing in Blue Chip Stocks and Winning in the Stock Market" byKelley Wright);
"The Strategic Dividend Investor" by Daniel Peris - see seekingalpha.com/instablog/725729-sds-se...;
TOP 40 DIVIDEND STOCKS FOR 2012: How to Create and Maintain a ividend Growth Portfolio by David Van Knapp (DVK published such book each year so previous editions can be used, he also publish good articles at SA - seekingalpha.com/author/david-van-knapp - I sometimes try to challenge).
"The High Dividend Yield Return Advantage" downloadable from www.tweedy.com/resources/library_docs/pa....pdf
I'd recommend to read these books for any stock investor.
Of course there are other and probably even better styles of investment but I prefer to be eclectic dividend investor.
ADDITION 20 Feb. 2012
Graphs like figs. 7.2 and 15.6 of SG shown above for US stocks can be constructed for any country. Indeed famous trio from London Business School / the authors of excellent book "Triumph of the Optimists" (2002)/ analyzed long-term stocks return in 19 countries.
They wrote "Unlike fixed-income investments, equities offer the prospect of dividend growth. Historically, dividends have grown in nominal terms in every country. But what matters is real, inflation- adjusted growth. Figure 4 shows the <analyzed> countries and world index ranked by their annualized real dividend growth over 1900-2010 (the gray bars). Real dividend growth has been lower than is often assumed. Figure 4 shows that 10 out of 19 countries recorded negative real dividend growth since 1900, and only four enjoyed real dividend growth above 1% per year.
Dividends, and probably earnings, have barely outpaced inflation. Dividend growth was lower in the turbulent first half of the last century.... But from 1950 to 2010, real dividends grew everywhere except New Zealand, and the world index enjoyed far healthier real growth of 2.3% per year. Figure 4 also shows how dividend yields have changed over the long run. The red bars show the annualized change in the price/dividend ratio (the reciprocal of the yield) from 1900 to 2010. Over the last 111 years, price/dividend ratios have risen (dividend yields have fallen) in 16 of the 19 countries. The price/dividend ratio of the world index grew by 0.48% per year. Finally, the blue bars in Figure 4 show the mean dividend yield in each country from 1900 to 2010. By definition, the real annualized equity return in each country is equal to the sum of the three bars shown for that country, i.e. the mean dividend yield plus the real growth rate in dividends plus the annualized change in the price/dividend ratio. Dividends have invariably been the largest component of real returns. "
I'd like to compare this recent picture with similar graph published by the "Triumph of the Optimists" authors (Elroy Dimson, Paul Marsh and Mike Staunton) 10 years early:
It seems interesting that USA moved from #4 to # position while Sweden maintained #1 position. I think this difference reflects importance of dividends in "lost decade" of US stock market.
25 March 2012 addition:
Research Affiliates LLC compared stock prices earnings and dividends after inflation based on the same Robert Shiller's data - see picture below
All 3 have almost the same uptrend that again confirms SG results that about 90% of total return is due to dividends and their growth. Let me just remind that dividends come from earnings at least in the long (> 100 years) run.
March 14, 2013
There are few research that shows that dividend indexes outperform S&P500.
S&P500 vs S&P Dividend Aristocrats:
During 1994-2010 returns were 8.0$% vs 10.33% at 5-year volatility 17.82% vs 16.82%. So from standard point Dividend Aristocrats outperformed, although they had higher yield 2.82% vs 1.81%. (S&P data).
It is well know that total return of stocks is higher than just price return (almost at any time scale due to positive dividends). In the long run the difference is about order of magniture (see graph below for US and Canadian stocks markets) mostly due to reinvesting of dividends.
The difference is due to 2 aspects: a) high yield - HY (i.e. opportunity to reinvest dividends more efficiently), and b) positive dividend growth -DG (i.e. opportunity to receive more dividends in a given period of time, and hence reinvest more money). On the level of individual stocks it means that total return is busted due to HY & DG combination. Therefore, I'm HY & DG investor. Many DGis (including me) tilt to value investing - they wait good yield of a high DG stock to purchase a company.
9 Jan 2014
Many stock investors rely on price appreciation. IMO this "next fool" approach is quite risky. Robert Arnott data (www.indexuniverse.com/publications/journ...) shown in the pictures (his figures 2 and 3) below
confirms this opinion. Who can wait 17-44 years in red ink /except communists 8-)/ with non-dividend stocks? Well, most of such "delay" occurs in 1801-1961 when most of companies paid dividends. Such zero real price appreciations periods can be ahead because inflation is more suitable for any goverment issued bonds (like Threasuries) than deflation. During such periods bonds (goverment and corporate) are not a panacea because "In the bond market the borrower issues the bonds and therefore writes the rules. Consequently, bonds are designed to make money for the borrower, not for the lender." (Ronald H. Muhlenkamp). Corporates include redeemable provisions in the bond, goverments print money at speed to keep inflation at about level of interest...
Decomposition of real (after inflation) stocks returns is shown below.
As you can see the result is robust for several developed countries.