# Understanding Compounding: Berkshire's Not-So-Hidden Dividend Contrarian Secret

Alice Schroeder’s biography of Warren Buffett, The Snowball, takes its title from a quote of his, “Life is like a snowball: the trick is finding wet snow and a long hill.”

Warren Buffett has long been lauded as the best investor of our day. And Berkshire Hathaway (BRK.A, BRK.B) has been envied for its superior long-term performance. Many reasons have been given for this record. One that has received perhaps less attention is the efficiency with which Berkshire has been able to compound.

To try to understand better how compounding manifests itself in investing, let’s start with a simple model: You have an account which holds a single position of \$10,000 all of it in a bond bought at par paying 10% and maturing in 20 years.

The first question is, what is the return on the bond at maturity? Yield to Maturity (YTM) says 10%. But, now, assuming that nothing is withdrawn, what will be the return on the account at the end of 20 years? We don’t know. While YTM assumes the coupon was reinvested at a 10% return, we don’t know what actually happened. If the coupons simply accumulated in the account for 20 years the IRR would decline to about 5%. If, on the other hand, you withdrew the coupon each year, the IRR makes the same assumption as YTM, that what you withdrew compounded at the coupon rate. Hence, it would tell you that the account had compounded at 10%, even though there will be only \$10,000 in the account at the end of 20 years, when Cinderella turns into a pumpkin.

This example helps us to get at the realities of compounding in the real world. How well the account performs depends crucially on the marginal, actual return on the coupons. In the case of the bond the marginal return is zero and so the compounding of the account will asymptotically approach zero.

As analysts we focus on return on capital, but this focus can deceive. In the beginning the ROC was 10%, but with the marginal return on capital zero, the ROC gradually declines to zero, reaching about 3% after 20 years. Companies don’t publish marginal returns, because, like analysts, they don’t know what it is: it is nearly impossible to calculate. But just because it is hard to know doesn’t mean it is not important. ROC is a poor substitute to RMC. What really determines how well you eat is RMC.

Withdrawing money from an account has the effect of producing a higher IRR. It also reduces actual compounding, in spite of what IRR says. Investors who think they can have growth and income will deceive themselves, unless they can reinvest at a rate high enough to overcome taxes. You can’t have your cake and eat it too when investing; it is always growth or income.

Our example allows another important distinction. The nature of the bond (bought at par) is that the burden of compounding rests outside of it: to compound effectively the investor must act. The account as a whole is comparable to an equity in that the compounding resides either within the account or outside, depending on payout policy. So, in the case of a company, the compounding is determined by the company’s ability 1) to produce free cash flow; 2) to achieve a high rate of return on retained CF; and 3) its dividend policy. The higher the payout, the greater the transfer of the burden of compounding to the shareholder. When Charlie Munger observed that the truly great company was one that allowed you to sit on your ass, he meant that the truly great company had a high rate of compounding - that is, a high marginal rate of return and that it retained the full burden of compounding.

How does the foregoing relate to Berkshire Hathaway? Let’s look at an example. Some 35 years ago, Berkshire Hathaway bought See’s Candy. A couple of years ago, Warren Buffett wrote in his annual letter that See’s had produced in the prior year \$80 million of pre-tax operating earnings on \$40 million invested capital. If See’s had been a standalone public company it would have had one whale of a job reinvesting its free cash flow, or it would have paid a large dividend and in doing so transferred the burden of finding growth to shareholders after paying taxes. See’s has little opportunity to grow: Notice that total earnings haven’t grown that much over 35 years. It needs very little cash to support operations, but it is a prodigious cash machine.

As a standalone it would have limited opportunity and value, even though the market would erroneously give it a high multiple based on its high return on capital. But within Berkshire and in the hands of a master capital allocator, See’s is a gem of a company. The real story of See’s is what Buffett has done with all the cash that See’s has sent him over the past 35 years; that is also the story of Berkshire Hathaway. It is an efficient compounding machine.

Some dividend lovers have tried to claim that Buffett is really a closet compatriot, because many of Berkshire’s marketable holdings pay substantial dividends. This misses a crucial point. See’s is able to upstream at no cost its free cash flow to its parent where it can be productively redeployed. This ease in up streaming is one important reason why Berkshire prefers to purchase whole companies.

In the case of partial ownership of publicly traded companies, however, there is a cost of redeploying cash: Berkshire must pay a tax on the dividend received, but corporations receive an 80% exclusion and thus pay only about 7% in taxes, instead of the statutory 35%, which it would pay on a realized gain. The individual shareholder pays a 15% tax plus state tax and likely doesn’t possess the compounding skills of a Buffett. So, given a choice between a lower-taxed dividend and a higher-taxed capital gain, Berkshire will opt for the former, but would prefer to own the entire business. In their eagerness to claim a compatriot, the dividend lovers fail to explain why Berkshire itself does not pay a dividend.

The Buffett view of company management is that its job is to allocate capital to compound at an optimal, efficient rate. The job of the shareholder is to add or withdraw from capital through capital transactions. Managements often mouth that paying a dividend is doing something for shareholders. It is precisely the opposite. Instead they are saying, “Sorry, we can’t do anything with it; see what you can do. Good luck!”

Many high quality companies produce more cash than they can productively redeploy at a high rate of return. This reality leads to the question of the best way to return money to shareholders: To pay a dividend or to buy back shares. Particularly since the financial crisis, the former has become enormously popular. Popularity, however, does not imply sound investment advice.

Few managements, sadly, are skilled in capital allocation in the manner of a Buffett and few are competent to know when buying back their stock is a good investment decision. Hence, shareholders are happy to have a dividend check, fearing that management will get into things they know little about. Thus the reference to Aesop’s Fables, a bird in the hand is worth two in the bush—even though both birds were initially in the investor’s hand.

But paying a dividend is definitely inferior to good capital allocation by a well-schooled management. Here is why:

1. Dividends force all shareholders to take cash whether they need it or not;
2. Dividends are fully taxable. This is comparable to selling a zero cost-basis position;
3. It is actually worse, since dividends are paid in cash and therefore directly reduce book value;
4. while a sale of shares is at market, often a multiple of book.

If a stock is selling at, say, twice book, why should one accept a dollar when one could sell the same amount of book for twice the amount? And the investor pays tax only on the realized gain; the rest is a return of capital.

One rejoinder given by dividend lovers is that they would rather have the security of a regular check. But does this hold water? First, if one could get twice book selling at market and the stock dropped 25% then one would still be better off by selling at 1½ times book. Second, this insecurity can be handled by maintaining a rainy day or buffer account of liquid cash. Third, a diversified account usually provides flexibility.

Dividends easily deceive: The noise in the market generally masks the ex-dividend drop in price, and the check either shows up in the mail, or appears in the account, as if from nowhere. Note that when a stock goes ex-dividend, the price drops by the amount of the dividend. The market has priced the dividend at book value, which is what it is, even though it might price the rest of the company at a multiple to book.

What can be said about dividend reinvestment plans? Don’t fool yourself! What do you gain? You have been forced to realize a tax liability, if the stock is in a taxable account, and the round trip is a joke. You receive a dollar of book value and have it reinvested by buying shares at market, a multiple of book, all the while having to pony up the taxes, which you paid out of another pocket. Since the dividend is reinvested pre-tax, the tax portion amounts to a required additional investment.

A stock buyback, on the other hand, reduces outstanding shares, thus increasing the remaining shareholders’ compounding. It does so without requiring a tax payment or an additional investment. Dividend reinvestment, on the other hand, increases outstanding shares at a higher price and dilutes compounding. The investor who thinks that dividend reinvestment plans solves his problem of having to get off his butt is fooling himself. At a twice-book ratio he has reduced his compounding by half—more if you net taxes.

The logic against dividends as an efficient means of compounding or of withdrawing capital for living should be persuasive. The Baby Boomers, however, have reached retirement and are now positively in love with dividends, after the bursting of the dot.com bubble and the downside of the real estate frenzy. Rather than clamor for dividends they would do better by themselves and their heirs if they understood better the hurdles to compounding in the real world. To return to the quote at the beginning, dividends are not the wet snow nor is a long record of paying a dividend the long hill Buffett had in mind.

I have owned Berkshire Hathaway since 1979.

Disclosure: I am long BRK.A.