Gwern's Reviews > Berkshire Hathaway Letters to Shareholders
(online letters) The famous annual letters of Buffett laying out the progress of Berkshire Hathaway and his & Munger’s investment beliefs. Apropos of an investigation into Long Bets as a charitable giving opportunity, I read through them. Perhaps the most interesting part about them is how skewed Buffett’s reputation is, given how often he is cited as a counter-example to EMH, despite him clearly explaining his strategy many times. What is most striking about Buffett’s returns is how, despite talking about how irrational “Mr. Market” is, Buffett and Berkshire Hathaway stay away from the general stock market. Reading through, I am stuck by the critical roles played by captive insurance companies and by buying private companies which are not on the stock markets; the methods are radically different from those of hedge funds like Thorp or RenTech, which focus on mispricings in the stock market and whose long-term successes might indeed show substantial weaknesses to the weak EMH. To summarize, the overall arc of Buffett/Hathaway appears to have been:
small-scale stock market trading in the 1940s, picking up minor inefficiencies in dead-end companies or warrants or whatnot. (All those opportunities have of course long since vanished.)
parlaying that into a series of purchases of private, off-market purchases of small cap businesses with steady cashflow, by cultivating a folksy mystique & offering very fast deals in which the owners do not shop around for alternatives, then folding their future earnings into the conglomerate but otherwise not investing in them & leaving them entirely alone with the same CEOs etc (1950s-1980s)
expanding holdings in insurers in order to benefit from their float (1970s-1990s)
using float plus cashflow to maintain a large capital which can be deployed for brief once-in-a-decade opportunities like stock market crashes or companies in sudden distress or (in one particularly interesting example) resolving the Lloyd’s of London reinsurance spiral, reaping extraordinary returns (1960s-present). As Buffett puts it:
Occasionally, though, either because of company-specific problems or a worldwide shortage of credit, maturities must actually be met by payment. For that, only cash will do the job. Borrowers then learn that credit is like oxygen. When either is abundant, its presence goes unnoticed. When either is missing, that’s all that is noticed. Even a short absence of credit can bring a company to its knees. [We, at Berkshire] have pledged that we will hold at least $10 billion of cash. We customarily keep at least $20 billion. By being so cautious in respect to leverage, we penalize our returns by a minor amount. Having loads of liquidity, though, lets us sleep well. Moreover, during the episodes of financial chaos that occasionally erupt in our economy, we will be equipped both financially and emotionally to play offense while others scramble for survival. That’s what allowed us to invest $15.6 billion in 25 days of panic following the Lehman bankruptcy in 2008.
as capital expands & returns fall, with small private companies no longer worth the time it takes to investigate, move to purchasing/investing in mega-corporations with vast capital investments required (2000s-present)
Step #2 is interesting. Buffett praises the businesses he bought in step #2 like See’s Candy as all being solid businesses that an idiot could run, with little competition, long streams of earnings, requiring little or no capital investment and benefiting from no synergies with other BH companies, and which the existing managers (often family) could be left in place to run as before they sold out (all being highly capable managers Buffett praises in the most extravagant terms), and notes that cumulatively they earned billions of dollars which fueled his purchases of GEICO and General Re etc. Given all these facts, one has to ask: why did any of these companies’ owners ever sell to Buffett in the first place? And the answer is… I don’t know. Even by Buffett’s presumably favorable recounting, many of the reasons were downright idiotic, like the furniture store owner who sold to ‘stop her kids from fighting over it’ (which is one of the worst forms of estate planning I’ve ever heard of and doesn’t actually resolve the issue since presumably then they would simply fight over the estate’s cash). A better reason is that they don’t want to go public and corrupt their vision, so they sell to Buffett instead, but that’s not a great reason either as they forfeit a tremendous cashflow. Given his famously fast offers and handshake deals, I can only guess that either sometimes even highly experienced entrepreneurs go temporarily insane & decide to sell on the spur of the moment, or they develop bizarrely high discount rates and decide to sell lucrative income streams for a pittance up front. After finishing, I went looking for commentary and found Matt Levine and an investment banker noting the same thing (albeit in much more cynical terms), and I can’t really disagree: reading Buffett’s own descriptions, he drove extremely hard bargains and almost all of the business owners (except Dexter Shoes, who correctly insisted on BH stock instead of cash) would’ve made a lot more money selling to anyone but the kind avuncular Sage of Omaha. To paraphrase Groucho Marx, you should never sell any business to Buffett that Buffett would buy; and given the observed results and how he preys on sexagenarian & older, 'attempting to sell to BH' could justifiably be added to the DSM-5's diagnostic criteria for senile dementia... I don’t blame Buffett for doing this, as it is the owners’ faults for being so easily out-negotiated, and certainly Buffett is making far better use of his money in philanthropy than any of them would have, but still, this is not a model which can be emulated and such a method can hardly be considered a strike against the EMH.
Step #3 is interesting since float isn’t necessarily that profitable, but aside from some discussions of “supercat insurance” (a very interesting area of reinsurance I always enjoy reading about), Buffett doesn’t explain exactly why it worked so well; as far as I can tell, none of the supercat events wound up severely impacting BH and so BH got very lucky in those contracts. If he went into why BH insurance was so efficient and well-priced, I might have been more impressed by his skills.
Step #4 is something of an exception that proves the rule. If Buffett has to wait a decade for major vulture buying opportunities, that implies that such mispricings are actually quite rare, since he is not out routinely making such deals. Here he clearly benefits from the unique access to insurance floats to have lots of large but very cheap capital to throw around. One has to suspect that any returns from that are not that impressive after accounting for that, a pool of money simply not available to most investors.
And in step #5, returns have fallen drastically and are no longer anything much to explain.
So the critical steps are either forever vanished (there will never be inefficiencies in stocks and warrants as enormous as they were in the 1940s-1950s, and there will perhaps never again be an economic boom to compare with the US post-WWII return on all American equities) or irreproducible (there are only so many insurers), and Buffett appears to have benefited from a large helping of luck: luck that nuclear war didn’t break out, luck that he has lived so long and in such health that his persona could pay off in off-market purchases, luck that BH avoided all the supercatastrophes in the 1990s, luck that he cottoned onto GEICO early on, luck that BH never fell prey to legal action like some hedge fund competitors such as Thorp did etc. Looking at how he did it, I feel certain that if Buffett were reborn today and handed a copy of Graham, he would find it thoroughly useless and not die a billionaire.
And one wonders how much return Buffett’s mistakes cost him. For example, the initial foray into textiles he often mentions, but he repeated it with Dexter Shoes - I read the letter announcing the purchase and thought to myself, “buying another textile/clothing company in New England? Why on earth? It doesn’t sound like it has any monopoly or regulatory capture or other moat at all. This is not going to end well” and indeed it did not - and in early letters he focuses heavily on inflation long after it had been tamed (at what cost to investment decisions, one wonders), and then in the 2000s forecast doom for the US via the trade deficit (which thus far has not eventuated and he quietly dropped the topic altogether). I can’t say I feel like I learned all that much from reading his analyses.
So to sum up based on his letters: Buffett made his money largely off the efficient stock markets in irreproducible ways exploiting individual irrationality while benefiting from historical and personal luck and is a poor example for anyone trying to argue that the weak EMH is sufficiently false as to make stockpicking a good idea. His investment advice is not particularly impressive or actionable (does anyone need to be told to invest in indexes now?) while the more interesting technical areas like his actual securities trading and insurance pricing methods are deftly concealed under rustic bonhomie, and his writings, while clear, are increasingly repetitive and recycling of jokes toward the end (hopefully as a result of needing to reiterate basics to the growing legions of BH shareholders and not because of senility). All in all, I came out less impressed with Buffett’s investment acumen than I started. There are probably better materials to read on stock markets and investment (certainly, Fortune’s Formula was much more interesting).