Untangling Warren Buffett’s unique firm
And its mediocre performance
Feb 27th 2020
ON FEBRUARY 22nd Warren Buffett reported that the conglomerate he runs, Berkshire Hathaway, earned net income of $81.4bn in 2019. That makes Berkshire, America’s biggest non-tech firm by market value, more profitable than any other company anywhere bar Saudi Aramco, an oil giant. Yet after years of mostly level-pegging or outperforming the broader market, Berkshire’s shares did only one-third as well as the soaring S&P 500 index last year (see chart). What is going on?
Assessing the conglomerate’s true success is a complicated business, because the business of Berkshire is complicated. Worse, a change in accounting principles two years ago forced Berkshire to start booking changes in the value of its $248bn equity portfolio as earnings. Last year that resulted in $53.7bn of unrealised capital gains filtering through to the bottom line—and a return on equity of 19%. The year before hefty unrealised losses meant a return on equity of just 1%.
The surge in unrealised gains was driven by the performance of Berkshire’s holdings in giant public companies such as Apple and Bank of America, which Mr Buffett and his colleagues pick like any old asset manager. Last year these stakes did a bit better than the S&P 500 as a whole—chiefly thanks to an epic big-tech bull run, which supercharged the returns from Berkshire’s 5.7% stake in the iPhone-maker.
Mr Buffett prefers the “real world” to “accounting-land”, as he put it in his annual letter, referring to the new standards on treatment of unrealised gains and losses. But in recent years he has struggled to articulate a consistent way of measuring the firm. At points he has endorsed tracking book value and at other times “operating earnings”, a proxy for the cash generated by the businesses Berkshire owns outright (plus the dividends from minority stakes).
One way of getting your head round Berkshire is to split it into two parts: a volatile financial arm, which includes its equity portfolio and insurance activities, and a steadier industrial group, composed of unlisted firms such as BNSF, a railway, and Precision Castparts, a manufacturer. The second category are the sort of company the no-nonsense Nebraskan professes to understand, and has spent a decade buying up. Their weight in Berkshire’s portfolio has grown: the industrial arm now makes up roughly half of total assets, up from a third a decade ago. But Mr Buffett paid up to acquire these firms, leaving the industrial arm’s return on equity at about 8%. Not terrible—but nothing to write home about.?
Here is the link to Buffett's latest shareholder letter:
https://www.berkshirehathaway.com/letters/2019ltr.pdf
The section on retained earnings is worth reading.
For those new to investing, It's easy to think that profits retained by a company are not important to shareholders -- after all, they only get the dividends that are paid (if any). What good is owning an asset if you can't keep all of the money it generates?
Nothing could be further from the truth.
I like to think of a company like a circuit board. An electric current (cash) enters the circuit board (a business), is directed in various ways to perform a range of tasks (capital and operational expenditure), and then generates an output (profit or loss).
But unlike a circuit board, that must operate under the second law of thermodynamics, a company can produce a lot more cash than what goes into it. Obviously, from an investor's standpoint, the higher the better.
The best companies are self-sustaining perpetual motion machines that spin out ever increasing amounts of cash. (The worst are those that forever burn up their supply of precious cash, and survive only because owners continiue to sustain them with fresh capital.)
Now, as a CEO, the main responsibility is that of capital allocation. (They are the architects of the circuit board). Given the cash they have available (from contributed equity and retained profits), they can choose to:
The 'north star' when allocating capital -- when designing the circuit board -- should be to focus on what activity has the best risk-adjusted return potential.
CSL (ASX:CSL), for example, generates a ~40% return on equity. And has done so pretty consistently due to some very attractive business characteristics (which we'll go into another time). The point is, when you have a machine that can spit out $1.40 for every dollar employed, you'd be crazy not to feed it as much capital as possible. While CSL could easily more than double its dividends, it would be a huge disservice to shareholders who are unlikely to get a reliable 40% return elsewhere.
Of course, the existing business can only invest so much at these rates of return; the market for its products is not infinite. But where possible, prudent reinvestment can lead to phenomenal compounding within a company. That's why CSL's equity (net assets) has increased over 18-fold in the past 20 years.
If it had paid out most of it's earnings along the way, the company would today be much, much smaller. Forgoing high-return reinvestment potential simply to satisfy shareholders' dividend cravings is not a good thing.
Here you can see a certain irony with stocks that pay out a lot of their profit in dividends. While that's great for those that are focusing on regular income (as opposed to longer term capital gains), it suggests the company has very limited reinvestment potential -- and therefore, often, limited growth potential. That's not a problem, per se, you can still do really well in such companies -- so long as the CEO recognises the limitations of their 'circuit board' and distrubutes profit to the owners rather than employ it in low return opportunities.
The key point here is that how a company allocates the capital at its disposal is the key determinant to long term shareholder returns. Whether it be a new acquisition, new plant & equipment, increased marketing or R&D, geographic expansion or whatever, the expected Return on Investment (ROI) is paramount. Unless it's likely to be better than what shareholders can themselves expect to generate (say ~10%, using the long term market average), companies should distribute as much in the form of dividends (or buybacks) as possible.
Did anyone else have any takeaways from Buffett's letter that they are keen to share?