You’ve probably heard talk about the key role that reinvesting dividends play in total returns.
This CNBC article from 2010 says “dividends make up 90% of total return”.
Since 1970, more than 80 percent of European equity returns came from dividend yield and growth, it stated, citing data from GMO.
And this 2019 article from The Balance quotes two different sources on the importance of reinvested dividends. Eagle Asset Management says reinvested dividends account for 97% of total return since 1871. John Bogle says they account for 95% of total return since 1926.
Articles in this vein are often accompanied by a chart that looks like this:
It looks like dividends make up 50% of the total return since 1990!
The first problem is that you can’t decompose total return like that. Total return is a geometric, not arithmetic, result. It is the result of multiplication, not addition. What if we take the same chart but show “the power of capital gains”?
Somehow price increases now look like they are responsible for over 80% of the return!
This is exacerbated because of the compounding nature of returns. The longer the timeframe we look at, the bigger the differences become. In our first chart, it looked like dividends made up 50% of the total return since 1990. But John Bogle said they made up 95% of the total return since 1926.
We can see this effect via multiple charts, each with increasing time horizon.
We can also see it by charting the “relative contribution of price gains” over different lengths of time.
We can see that if measure from 1871-present (the far left of the chart) it is minuscule. Even from 1940-present it is under 10%. From 1980-present it is ~35%.
When we say that “dividends make up 95% of total return since 1926” we aren’t actually saying anything other than “compounding things compound”. We haven’t said anything insightful about the nature of dividends or capital gains or anything else.
To see this even more clearly, we can look at a side-by-side comparison of this “relative contribution” for both dividends and capital gains.
Notice that they don’t “sum up to 1”? They can both be below 10%. Or they can both be above 50%. The entire notion of talking about “dividend reinvestments make up X% of total return” is misleading to the point of being useless.
While the entire notion of “relative contribution” is mostly broken, there is one thing we can use it for: comparing different time periods against one another. We can compare the relative contribution of capital gains over rolling 10-year periods and see the increased dominance of capital gains since the early 1940s. Even at the worst part of the 1970s, capital gains were still more important than any time in the period before the 1950s.
The trend is even more obvious when we switch to 30-year rolling periods (30-years is a reasonable approximation of one’s investing life).
Dividends have become less and less important to total return since the 1910s.
