I use a simple model when it comes to investments. Your returns are a function of:
1. Dividends
2. Growth in earnings per share
3. Change in the valuation multiple
As long-term investor, I care about the first two items, which are the fundamental returns. If I select a good quality company at a decent entry price, the goal is then sit back and hopefully let it keep growing earnings, dividends and intrinsic value. I do not try to bet on multiples shrinking or expanding. In general, if you look at long-term equity returns, the change in multiple matters the least in the long-run. It does matter in the short-run, like the next 5 years or so. But the longer one holds a business that keeps working, the less the impact of changes in the valuation multiple. This of course works on average, since multiples were generally around 15 - 20 times earnings. There were times where multiples were higher, but those were offset by times where multiples were lower. If we get starting multiples of 50 or 100 times earnings, then things would be different. But this has never been the case in the US, so we'd worry about it if/when it happens (though it did happen in Japan in 1989, but that's the story of another time).
For the US stock market in general, we have often heard that it has delivered a total annual return of 10%/year since either 1802, 1871 or 1926, whichever study is being referenced. This has been achieved through 6% annualized growth in earnings per share coupled with a 4% average dividend yield. Dividend growth has been close to 6%/year as well. As we all know rising earnings are the source behind future rising dividends.
6% + 4% = 10%.
This exercise really helps me think about the trade-off between dividend yield and dividend growth.
If equities never grow earnings per share, but are available at a dividend yield of 10%, then future total returns would be 10%/year, on average.
If equities never paid dividends, but grew earnings per share by 10%/year, then future total returns would be 10%/year on average as well (provided we do not
This model is very helpful, when it comes to evaluating individual companies as well. Let's look at two companies, Archer Daniels Midland (ADM) and NextEra Energy (NEE).
As far as the companies are concerned, they are both members of the dividend aristocrats list. In order to gain entry to this list, a company needs to be a member of S&P 500 and increase annual dividends to shareholders for at least 25 consecutive years. This last item does not happen by accident. It is usually a sign of a strong competitive advantage that allows a company to not only grow earnings per share for decades, but also to generate rising torrents of excess cashflows for decades as well. I do believe that a rising stream of annual dividends is a good initial indicator of quality.
However, this merely puts a stock on my list for further research. The next steps involved reviewing the trends in earnings per share, dividends per share, dividend payout ratios, and checking on valuations, in order to determine if the company is still fundamentally sound and also available at a good price.
Archer Daniels Midland (ADM) sells for 11.50 times forward earnings and yields 3.15%. It has a 5 year annualized dividend growth of 6.76%. If I add 3.15 and 6.75 I end up with 9.90, which is very close to an estimated total return of 10%/year. Plus, if that annual dividend of $2/share managed to grow at 7%/year over the next decade, we may see an annual dividend of $4/share in 2034/2025, for an yield on cost of 6.30%. That's future yield on cost without taking into account future dividend reinvestments. Check my review of ADM for more information about the company.
NextEra Energy (NEE) sells for 21 times forward earnings and yields 2.85%. It has a 5 year annualized dividend growth of 10.75%. If I add 2.85 and 10.75 I end up with 13.60, which is higher than the estimated total return of 10%/year. However, we also have some margin of safety in growth expectations, especially if they come in 3% - 4% lower than the past. If that annual dividend of $2/share managed to grow at 7%/year over the next decade, we may see an annual dividend of $4/share in 2034/2025, for an yield on cost of 5.70%. If it grew at 10%/year, it would double in 7 years. That's future yield on cost without taking into account future dividend reinvestments. Check my review of NEE for more information about the company.
This model is also good for making sense of company returns over certain time periods. It's a helpful tool to review past returns, but also to stress-test various possible future scenarios.
As we saw above, dividends and earnings per share growth are the fundamental returns. But the changes in valuations are speculative returns.
For example, let's look at the best performing Dividend Aristocrat over the past decade. That is Cintas (CTAS).
Today, the stock sells for 45.67 times forward earnings and yields 0.82%. The current price is $761.59 as of July 26th, 2024. It is expected to earn $16.67/share in 2025 and pays $6.24/share in annual dividends. The company earned $15.40/share in 2024. This was a very high growth from the $2.75/share that the company earned in 2014.
The company's share price was 62.12 at the end of May 2014. This translates into a P/E of 22.58 in 2014. The yield was 1.25%, as annual dividend was 77 cents/share for 2013. It was raise to 85 cents/share in 2014.
The company successfully managed to grow earnings per share from $2.75/share in 2014 to $15.40/share in 2024. And have a forward EPS of $16.67/share for 2025. This 506% increase in earnings per share fueled a large portion of returns. The small, but growing dividend, reinvested over time, also helped fuel returns. But if the P/E ratio had stayed at 22.58, the stock price would have only been at $376.41 today. So you can see that the doubling of P/E ratio definitely helped fuel future returns. As the company did well fundamentally, investors felt comfortable paying a higher multiple for its earnings stream. The problem can come if their animal spirits cool off, and the stock valuation returns to a more reasonable 20 - 25 times earnings. As a result, investors who buy the stock today for its growth potential may not realize a return on their investment for as long as a decade, even if earnings per share double from here. It's unlikely that earnings per share would grow by 500% through 2035 however. The company is expected to earn $33/share in 2033, which is the farthest Wall Street estimate I could find.
This is where I always try to evaluate fundamentals in terms of earnings per share growth and dividend growth. I try to evaluate the likelihood of those fundamental changes continuing. But I also try to evaluate valuations, and see if I have any margin of safety there.
Today we learned about the sources of returns. We also learned how to use this simple technique on a backwards, current and foward measure. We looked at a few examples, and hopefully learned something that we could use in the future.
