This claim came from Kevin O’Leary when promoting his new dividend-centered ETF. Mr O’Leary stars in a TV show where they put him in a tank full of sharks or something like that. While the statistic might be inspirational in starting a dividend investing strategy, I couldn’t resist taking a closer look and doing some fact checking.
I originally came across this article over at the Bogleheads Forum which has much more detailed commentary on the article as might be expected.
One school of investing thought argues that dividends are irrelevant in terms of total return. Saying they’re irrelevant doesn’t mean that dividends are bad or to be avoided. It’s like someone saying “I measured the average speed of 2,000 cars at random and on average, red cars seem to go faster than blue cars. My next car will be red so that I can get around town faster.” Was the paint color really the most important factor? Or was it other factors, like the driver, the engine performance, more cars of one color being located in more congested areas or the age of the car etc.? Either way, red cars aren’t bad but a personal preference.
Anyway, I digress.
There’s also nothing fundamentally wrong with OUSA, Mr. O’Leary’s ETF which is being promoted in the Forbes interview. However, if you do watch his video explanation, at least keep an open mind. He also argues that non-dividend paying stocks no have monetary value. In the real world, ownership of a profitable non-dividend paying company is actually worth something. Just ask Mr. Buffet for some free BRK.A shares.
OUSA aims to hold at least 20 stocks over at least 5 sectors (it currently holds ~153 stocks) and it favors ‘lower volatility’ companies. It aims to reduce volatility via a low allocation to the Energy and Financial sectors and a high allocation to Technology and Consumer Defensive sectors. It pays an average 2.3% yield.
I am personally not interested in it because of the 0.48% fees but that’s just me.
However, I am interested in the math. How did he arrive at that number since there’s no citation mentioned? And can I reproduce it? It’s time for some fact checking.
Calculating total return vs price return
The following chart shows the effect of re-investing dividends vs spending dividends in VFINX, which I’m using as a real-world equivalent of the virtual S&P 500 index. I could only get dividend data from Yahoo starting in 1980 so the calculations below are based from the 1 Jan 1980 to 31 Dec 2016.
If you had invested $10,000 back at the start of 1980, you’d have over $403,000 at the end of December 2016 if you’d re-invested all the dividends. Had you not re-invested them, your portfolio would be worth only $146,000.
A simple calculation shows:
Total Value with dividends re-invested: $403,188
Total Value without dividends: $146,155
“Dividend-only” amount of Total Value: $403,188 – $146,155 = $257,032
Percentage of Total Return due to Dividends: $257,032 / $403,188 = 64%
So, not quite 71%. My data is only from 1980 which is 36 years and not the forty-years mentioned. It’s fairly close and I’ll give Mr. O’Leary the benefit of the doubt!
In reality, the above calculation doesn’t show “dividend-only” contributions at all. When dividends are re-invested, the new share purchases are also subject to capital gains (or losses). So a fairly large portion of the “dividend only” amount above includes capital gains from new shares purchased by the re-invested dividends. It’s a little disingenuous to include capital gains as part of a “dividend only” metric so let’s exclude that part.
In the above example, the total payments of dividends that were re-invested was $94,189. This number includes the compounding effect from a larger number of shares due to re-invested dividends. It’s a pretty awesome result from a $10,000 investment just on its own!
Let’s try the same math again:
Total Return with dividends re-invested: $403,188
Dividend only amount of Total Return: $94,189
Percentage of Total Return due to Dividends: $94,189 / $403,188 = 23%
So not such a great result, compared to the earlier 64%.
OK, I know the arguments. If the dividends weren’t re-invested, there wouldn’t have been the associated capital gains. This is all true, but it is still a fact that a significant amount of the return of those extra shares from re-invested dividends were due to capital gains.
One more time…
Given that the average dividend yield over that time period was about 2.8% and that the total return was about 10.8% it should be fairly intuitive that 70% of the returns couldn’t come from dividends alone. Dividend yields steadily decreased from around 5% in 1980 to around 2% in 2016.
Here’s one last example using the values above.
$10,000 at 10.8% CAGR grows to $401,265 in 36 years.
With 2.8% Dividends only, $10,000 at 2.8% CAGR grows to $27,024 in 36 years.
With 8.0% Capital Gains only, $10,000 at 8.0% CAGR grows to $159,681 in 36 years.
$27,024 + $159,681 = $186,705 and not the full $401,265.
Or in mathematical terms:
1.028^36 + 1.08^36 does not equal (1.028+1.08)^36
So any comparison where either dividends or capital gains are “removed” gives a misleading answer. Since we’re in the realm of shaky math, we could also simply divide 2.8 as a percentage of 10.8 and reach a result of 26% from dividends.
Are you getting a sense that none of these results are meaningful yet?
So which value is correct?
Well, if you’re trying to promote a dividend-focused ETF, then the first answer (64%) is better for obvious reasons!
However, neither answer matters because the percentage values are meaningless as part of an investing decision.
The results only demonstrate that:
- Some companies in the S&P chose to pay dividends.
- Dividend payments are a part of Total Return.
The more interesting question, “was the S&P total return higher because of those dividends?” can’t be answered by the total return calculations above.
Total Return is a combination of Capital Gains plus Dividends and it’s not so simple just to separate one from the other. If you plan to live off dividends then your “Safe Withdrawal Rate” is simply the same as the dividend yield. If this meets your income objectives in retirement then you’ve succeeded.
Consider the source
More importantly however, when reading reports, investing strategies or even blog posts, consider the source of the article. Data can be presented in many different ways, some more misleading than others. Question the data and consider what the motivation behind the numbers might be. There are two sides to every story after all.
In this case, the purpose was to encourage investing in a dividend-based ETF by the fund owner who receives a percentage of the value of the assets being managed by the fund. By suggesting that dividends contribute a high percentage of total returns, you’re led towards concluding that the total return of the S&P was higher because of those dividends.
Be wary too of confirmation bias which is where you favor data to support your beliefs.
One more thing
Just for fun, here’s the performance of OUSA since its inception just over a year ago. I’ve compared it to a Total Stock Market fund (VTSAX) as well as to the Vanguard High Dividend Yield Index ETF (VYM) and the Vanguard S&P 500 fund (VFINX) as a benchmark.
Don’t read too much into the above chart since it’s only a short timeframe and past performance says nothing about the future. OUSA got off to a good start but is currently the lowest of the four investments. All three Vanguard funds have expenses below 0.10% giving them an automatic +0.38% yearly advantage over OUSA. Fees are important since they’re a continual drag on returns.
Quote of the Day
Just taking risks for risk’s sake, that doesn’t do it for me. I’m willing to take risks that I think are worth it, and I’ve worked so hard to make sure that I survive.