This article covers these topics:
- Why dividend investing is not necessarily the most efficient way to compound capital.
- Dividends are paid out of book value, but usually reinvested above book value. Why accept to receive a dividend at book value when you can for example sell shares at two times book value?
- Where is the best place for your capital? In the hands of the company or in your hands?
- By looking at the reinvestment math, retained earnings need a lower marginal rate of return to beat a reinvested dividend’s marginal rate of return.
- I believe great managers and investors have a true understanding of marginal rate of return and how to allocate capital in the most efficient way.
Before you read further you might want to read my two previous articles on dividend investing:
- Don’t Be Fooled By Your Dividend Bias: Sell Shares To Get Income
- The Foolishness Of Dividend Investing
Of course, a shareholder in our dividend-paying scenario could turn around and use his dividends to purchase more shares. But he would take a beating in doing so: He would both incur taxes and also pay a 25% premium to get his dividend reinvested. (Keep remembering, open-market purchases of the stock take place at 125% of book value.)
The quote is taken from Berkshire Hathaway’s shareholder letter of 2012, in which Warren Buffett has a discussion about dividend reinvestments. I strongly recommend dividend investors to read that letter in full.
To create wealth you need to both compound and withdraw your capital in an efficient way. One very important question to ask is this:
Where does your capital compound best; in your hands or in the hands of the company? To me it seems the great majority of dividend investors believe that capital compounds best at their own hands. However, we need to look at ways to estimate the future marginal rate of return to figure out where capital compounds best. The marginal rate of return is unknown, but looking at math and numbers help to understand where your capital most likely is of better use.
Before we start I want to direct you to a very good article about marginal rate of return, a piece was written as long ago as 2011 on Seeking Alpha by a contributor nicknamed Briar. This is an excellent article! I highly recommend the article as it explains marginal returns very well. Briar has been an investor in Berkshire for four decades and has studied both Buffett and Munger extensively (I know Briar). I believe there is a lot to learn from Berkshire shareholders because learning and adapting compounds really well over time. My article is built on much the same arguments as Briar’s, but I believe his article is much better than the one you are reading now.
I also want to emphasize that my article is not a case against dividends; I own several dividend stocks. Nevertheless, there are a lot of negatives and opportunity costs with returning capital as dividends.
The importance of compounding:
Wealth creation is dependent on two factors: return and time. The higher the return, the more wealth in the future, which is pretty obvious. What is perhaps not so obvious is the importance of time: It’s far better to start saving early than saving more later. The reason is the compounding effect, the most important factor in investing. By compounding you get an exponential return on your investment. It snowballs. I believe Warren Buffett has made around 99% of his wealth after he turned fifty. This shows the value of time: the longer you wait, the more you get, of course given a positive return. However, time is not the only reason why Buffett has managed to become very wealthy (together with his shareholders): He understands where he can get the best marginal rate of return. This is an important concept because marginal rate of return has a huge impact on how your capital compounds. Even “small” differences in returns have a huge impact on future wealth. For example, 100 000 invested today is worth 1.08 million after 25 years with a 10% return, while this amount is reduced to 862 000 and 684 000 if the return drops to 9% or 8%. In other words, your wealth in 25 years time drops 37% if it compounds 2% less.
What is the marginal rate of return?
In this article I have defined marginal rate of return as the the return you get on reinvested earnings, for example capital retained in the business or distributed as dividends. Let’s assume you have an account worth 100 000 and your portfolio returns 3 000 in dividends and 3 000 in retained earnings, in total 6 000. The return you get on those 6 000 will in the long run decide your compounding rate more than the original investment, ie the marginal rate of return is the most important factor!
In the bond market the most important factor is the interest you make on the interest, the reinvestment rate (the marginal rate of return – incremental return). In this video Howard Marks explains this very well starting from 13 minutes. Marks emphasizes the fact that over long periods of time most of the return comes from the interest you make on the interest, not from the original coupon rate. If you invest 100 in a bond that yields 5%, you have 105 at year end. That is of course a 5% return on capital. If you chose not to reinvest, you have 110 at the end of the second year. In the second year your return is only 4.76%. Why? Because you have a marginal rate of return of zero on the coupon (which is not invested). If you reinvest at a lower rate than the coupon, logically your total return will gravitate downwards.
Under a compounding scenario you are reinvesting the interest as it is earned, and thus making interest on both the original principal and the later accumulated/reinvested interest/coupon. The difference between compounding and not compounding is enormous over long time periods: at 10% a 100 000 investment grows to 400 000 after 30 years with no reinvestment (when marginal rate of return is zero), while it grows to 1.75 million with reinvestment at 10%. Presumably Albert Einstein called this the eighth wonder!
Is marginal rate of return relevant for dividend investors? It’s highly relevant. I believe most investors, including me until recently, don’t understand that dividend reinvestment is mostly done at a lower marginal rate of return than the original rate. This is because of taxes and premiums to book value:
Taxes on dividends:
The first and most obvious headwind is taxes. Dividends are taxed unless you have a tax-deferred account. For me, a non-US resident, it means I only keep 85% of the dividend as 15% is withheld at source. Because a dividend is distribution of retained earnings, the same capital is taxed twice: first as earnings in the company and later as dividends.
We all know that dividend reinvestment has been a very important factor in total returns, but this is gross returns. All research is done ignoring taxes, which of course overrates the real return.
How much do taxes diminish returns? Let’s look at Vanguard Real Estate ETF (VNQ):
The blue line is dividend reinvestment done with no taxes, and the pink line is deducting 20% for taxes. This table summarizes the results:
|Investment done in December 2004:||End result September 2019:||CAGR|
|56 750 (no taxes)||181 775||8.07%|
|56 750 (20% taxes)||159 300||7.12%|
With 25% taxes the CAGR drops to 6.9%.
Johnson & Johnson (JNJ) is one of the most popular dividend growth stocks. This graph shows the difference between no tax and 20% tax:
|Investment done in February 2000:||End result September 2019:||CAGR|
|39 000 (no taxes)||212 700||8.85%|
|39 000 (20% taxes)||192 400||8.3%|
The difference is less for JNJ than for VNQ, simply because less is paid out as dividends. Clearly, the graphs show that the impact of taxes gets bigger and bigger as time goes on, indicating the importance of taxes for long-term savers. Even small drops in CAGR have huge effects on future wealth!
Reinvestment above book value:
And, remember, every dollar of net worth attributable to each of us can be sold for $1.25.
– Berkshire’s letter of 2012.
Almost all stocks currently trade at a premium to book value. Book value is the assets minus the liabilities – also called shareholder’s equity. The equity is simply the assets that the shareholders receive if the company were to be liquidated: liabilities are subtracted from the assets, what is left is sold, and capital is returned back to the owners – the shareholders.
A dividend is paid out from the book value/shareholder’s equity, thus reducing the value of the company. Capital is simply transferred from the company to the shareholders’ accounts (less taxes). Because of this the stock drops the same amount as the dividend when it goes ex. dividend. Thus, a dividend is in practice a partial liquidation of the company.
Now ask yourself the following: Why accept to receive a dividend at book value when you can sell shares at a premium to book value? This is exactly what Buffett meant with the quote that started this paragraph: Why should shareholders of Berkshire want a dividend when they can sell shares at a 25% premium to book value? As Buffett said: You take a beating in doing so.
Let me illustrate: assume you bought a stock at 100 per share ten years ago, at the time having 3% dividend yield. Ten years later the company has managed to grow both earnings and the dividend at 10%, giving a share price of 259 and a dividend of 7.78%, giving a yield on cost price of 7.78%. What happens when you receive a dividend? In effect you give up the cost yield of 7.78% in exchange for a reinvestment yield of only 3% (the current yield is still 3%) when you reinvest in the tenth year.
Here is another example: Let’s say you own shares in company ABC where historical return on equity is 10%, book value is 100 per share and market price is 200, meaning it’s trading at two times book value. If you receive one dollar per share in dividend naturally book value declines to 99 at ex. dividend date. Dividends are always valued at one for one compared to the book value. You chose to reinvest the dividend in company ABC. At what marginal rate of return do you reinvest?
Let’s assume you are withheld 15% of the dividend at source for taxes. This leaves you with only 85 cents per share to reinvest. But what most investors fail to realize, is that they reinvest at a much higher cost price than they received the dividend: You reinvest at two times book value, which means your share of the book value is only 42.5 cents (85 cents divided by two). This equals a marginal rate of return on the equity of only 4.25%!
This is perhaps better illustrated with a fixed income example. Let’s say you invested in a 10% coupon bond several years ago, but now the prevailing interest rate is 5%. This means your bond has appreciated in value (selling above par). If you had the choice, what would you rather do: receive a coupon (10% of cost price) in cash or have the coupon reinvested in the 10% yielding bond? The obvious answer is that a rational investor would chose to reinvest (getting additional bonds). If the investor needed cash, he would sell part of the bonds later at market because the bond trades above par, ie. above 100 cents on the dollar. The same thinking applies just as well to dividends. Why reinvest at prices above your original shareholder’s equity?
If the stock trades at a discount to book value, then it obviously makes sense to both pay and reinvest a dividend. Gazprom (OGZPY) currently trades at about 0.5 times the equity which means that after reinvesting you own double the equity than before (you would be receiving a dollar in book value but reinvesting at just half the dollar, thus receiving 2x of the equity). This additionally means that every dollar (or rouble) retained in Gazprom is only transformed to about 50 cents of market value. In the letter of 1984 Buffett called the latter “gold-into-lead-process”, and meant that most earnings should be distributed back to shareholders in such a discount to book scenario. Opposite, when a company trades at a premium to book value, retained earnings is automatically converted into x times the book value.
The next logical question would be:
Where is the best place for my capital? In the company or in my pocket? The company’s retained earnings can have a lower marginal rate of return and still make more sense than paying a dividend for reinvestment.
Given a proper management and high marginal returns on retained earnings, it’s of course better to let it compound in the company instead of distributing it. The truly great compounder frees the investor to sit on his or her ass:
Sit on your ass investing:
What this little tale tells us is that tax-paying investors will realize a far, far greater sum from a single investment that compounds internally at a given rate than from a succession of investments compounding at the same rate. But I suspect many Berkshire shareholders figured that out long ago.
– Berkshire letter of 1993
Charlie Munger said the perfect stock is the one that “lets you sit on your ass”. Why? Because investing requires work. You spend time deciding and finding an attractive stock, and later you need to spend time deciding when to sell. The same thing goes for dividends: When you receive a dividend you need to spend time to reinvest it. Because this requires many decisions over long periods of time, mistakes compound over time. Charlie Munger prefers to find good stocks and just “sit on it”. This requires only one decision. In addition, you have the friction costs mentioned in this article: taxes and reinvestment above book value.
Reinvestment opportunities could be just as important as cash flow:
Ideally you would want to invest in a business that throws off a lot of cash relative to the invested capital (ROIC). But the reinvestment rate is often just as important because of the difficulties of getting the same marginal rate of return on the distributed capital. Ideally you want a business that does not send you a dividend check, but has great opportunities for reinvestment of the earnings.
Let’s again look at some math:
Two identical companies are making a 12% return on their capital (ROIC), but have different reinvestment opportunities. Company A distributes 50% of earnings as dividends and reinvests the rest, while Company B retains and reinvests all earnings. For the sake of simplicity, we assume that both companies trade in the market today at 150 dollars, 15 times earnings. Which company compounds faster? After 10 years Company A is trading at 268 (EPS of 17.91 times 15 multiple) and has paid out about 75 dollars in dividends, while company B is trading at 466 (EPS of 31.06 times 15 multiple). The difference is due to the dividend distributions: Company A compounds only at 6% (because 50% is paid out in dividends). This means Company A has appreciated from 150 to 343 (268 plus 75 in dividends) compared to 466 for Company B. CAGR is 8.62% vs 12%.
Now, you might argue that this is unfair because the dividends from Company A have been laying idle in your account. But what is the end result if you invest the dividend at 12% returns? This means Company A is still priced at 268, but your dividends have increased from 75 to 131, giving a total appreciation to 399 (268 + 131). That is still below Company B’s price of 466. Because Company B is able to redeploy earnings at the same marginal rate of return, it simply snowballs the EPS much faster than Company A. To make up for this difference, you need to compound the dividends at 20%. If we include a 20% haircut for taxes on the dividends, you need to manage a CAGR of 24% on the dividends.
Even if Company B only managed to compound retained earnings at 10%, and not 12%, you would end up with the same appreciation as if you compounded the dividends at 12% in Company A.
The math shows that reinvestment opportunities matter. The more a company can reinvest retained earnings, the better (at acceptable rates, of course). I guess the shareholders of Berkshire knew this long ago, as Buffett says.
Great investors understand marginal rate of return
I believe great managers and investors both have a true understanding of marginal rate of return and how to allocate capital in the most efficient way. As the above examples show, it’s very hard to reinvest at the original rate. We often use return on capital employed, IRR, ROIC and ROE as if they were marginal rate of returns (but they are not). It’s easy to calculate those mentioned numbers, but they are most of the time (sadly) incorrect when evaluating marginal rate of return. And, as I have explained in this article, the marginal rate of return is the main factor of your future wealth.
What is your marginal rate of return?
Given that nearly all companies trade at a premium to book value, you can have a look at your dividend paying portfolio and estimate how much you compound before and after payouts: simply use the payout ratio and multiple to book value (and perhaps less dividend taxes).
All you get with Berkshire stock is that you can stick it in your safe deposit box, and every year you take it out and fondle it.
– Warren Buffett
The best place to compound is in a stock that has the ability to grow and expand without any distributions to shareholders (given an acceptable return on assets). Even if the marginal rate of return on the retained earnings is slightly lower than your return on reinvested dividends, you are better off having the capital reinvested in the company. If you need “income”, sell some shares. Additionally, this has the added perks of deferring taxes and letting you pursue other interests than following the markets.
I end the article by listing what I believe to be relevant points to ask or elaborate before any investment, based on my arguments in this article:
- What is historical ROIC?
- How much of earnings can be reinvested back into the business at historical ROIC?
- How long can the company reinvest its earnings before size limits opportunities?
- Is it reasonable to expect a marginal rate of return similar to the historical rates?
- How does management and board manage excess capital if it can’t reinvest all earnings into existing businesses (value enhancing dividends, buybacks, M&A etc)?
Suggested further reading:
http://www.berkshirehathaway.com/letters/2012ltr.pdf (pages 19-21)