Speculation as to whether Berkshire Hathaway (BRK.B, BRK.A) will begin dividend payments to shareholders has risen again, but this time with fresh, convincing arguments in favor. Barron's issued a report on Berkshire here suggesting that the company may pay a dividend as early as this year.
Analysts at Stifel Nikolaus & Co. reiterated their “sell” rating in response, noting reserve requirements in Berkshire's insurance business were favorable through the September period. While Stifel analysts continue to miss the “big picture” on Berkshire's undervaluation, namely its earnings power and ultimately profitable short puts positions we wrote about here, the insurance business is the key reason why Buffett will refrain from issuing a dividend.
The spark behind the speculation of a dividend stems from Berkshire's under-performance relative to the legacy Berkshire of the last few decades and its relatively poor showing versus the S&P 500 (NYSEARCA:SPY). Berkshire crushed the market for decades by returning an annual growth rate of 22% and 20.3% growth in book value according to the company's latest annual report.
Berkshire shares are cheap in their own right and should trade around the $145,000 mark ($96 per B share) based on Berkshire's historic book value median of 1.6x. Most analysts continue to model Berkshire primarily as an insurance company which is meritable, given that Berkshire's insurance businesses we value around $85,000 per A share. However, the Burlington (BNI) acquisition has transformed Berkshire's earnings power to suggest a book value analysis consistent with that of a railroad for its representative part of earnings. Most railroads carry a book value between 2.0x and 2.5x which suggests Berkshire at 1.7x or 1.8x with all else equal. A second flaw in most analyst models interprets market prices as the fair value of Berkshire holdings-- to this end, large holdings like American Express (NYSE:AXP) and smaller holdings like US Gypsum (NYSE:USG) remain significantly undervalued, and that percolates through the models to an inaccurate, lower Berkshire valuation.
Misunderstanding valuation and a stagnant price need not militate a dividend policy, however. Buffett views the entirety of Berkshire as an insurance company. Consistent with this view he has entered into short put positions on various market indices (effectively amounting to an insurance policy). Buffett also takes his position as a financial steward, not only for shareholders but also all Americans, very seriously. In the 2009 Annual report from last spring, Buffett wrote:
We will never become dependent on the kindness of strangers. Too-big-to-fail is not a fallback position at Berkshire. Instead, we will always arrange our affairs so that any requirements for cash we may conceivably have will be dwarfed by our own liquidity. Moreover, that liquidity will be constantly refreshed by a gusher of earnings from our many and diverse businesses. When the financial system went into cardiac arrest in September 2008, Berkshire was a supplier of liquidity and capital to the system, not a supplicant. At the very peak of the crisis, we poured $15.5 billion into a business world that could otherwise look only to the federal government for help. Of that, $9 billion went to bolster capital at three highly-regarded and previously-secure American businesses that needed – without delay – our tangible vote of confidence.
Buffett's approach with Berkshire departs from other insurers and conglomerates, including majority-owned Wesco (NYSEMKT:WSC), in that his aim is to never directly risk permanent impairment to liquidity. He even emphasized that Berkshire was a liquidity supplier. Thus, while maintaining a dividend policy at Wesco is acceptable, a dividend constrains Buffett's Berkshire from being nimble with its cash at the time that it is most fruitful to deploy it.
It is on this note that Stifel Nikolaus analysts have it right but they could take it a step further. Initiating a dividend is not likely because of the backstop a hoard of cash provides to Berkshire's insurance businesses. It is furthermore unlikely because Buffett views the entire entity as insurer, and not just the part we value in the $80,000 (per A share) range.
Disclosure: I am long BRK.B. Author is also short puts on BRK.B.