The Buffett Puzzle
Warren Buffett has described the ideal business with perfect clarity. "The best business," he once said, "is one that is a royalty on the growth of others." He has also observed that "you'll never buy companies as cheap as stocks sometimes get."
Together, these two statements imply a fantastic investment strategy: steadily deploy capital into businesses that are either royalties or have "royalty-like" economics, dollar-cost averaging into minority stakes using dividends and cash flows from existing investments. A five or ten percent stake in Marriott International might go on sale when investors are worried about a recession or a slowdown in travel, even when the entire company would not be obtainable at a cheap price. And when the cash flows in and no royalty business looks attractive, simply repurchase shares of one's own holding company.
Buffett identified this ideal strategy early, articulated it repeatedly, and has spent the past several decades largely ignoring it. That is what we consider the Buffett Puzzle.
The evidence of the missed opportunity is not subtle. Over the past couple of decades, businesses with exactly the characteristics Buffett described have compounded quietly and relentlessly. Companies like CME Group or Visa Inc. require little reinvestment, tend to generate cash in excess of reported earnings rather than consuming it, sit atop irreplaceable assets, and benefit from network effects or structural monopolies. They are either monopolies or oligopolies and they carry very high free cash flow margins on revenue. They are, in Buffett's own language, royalties on the growth of others. Yet he did not buy them, or bought them too little and too late.
A comparison of capital-light and capital-intensive businesses over the past thirty-five years shows an enormous performance gap, with the capital-light group pulling ahead so consistently that the divergence is not a matter of picking the right year to measure. It is a structural feature of modern capitalism. Businesses that can grow without consuming capital are simply worth more, in ways that compound over long periods. Buffett, who understands compounding better than almost anyone alive, appears to have understood this principle intellectually while failing to act on it in practice.
Consider two businesses that have the attributes that Buffett spent his career describing as ideal.
Google is a royalty-like interest on the broader economy. Its core business connects customers with merchants of every kind, both online and brick-and-mortar, and it earns a small toll on an enormous and steadily growing share of global commerce. The capital requirements relative to the earnings power are modest. The network effects are extraordinary. The business has gone on sale repeatedly, but Berkshire never owned it until 2025. Google traded for less than 20 times earnings in 2008, again in 2011 and 2012, in 2022, and for parts of 2025. In June 2012 specifically, the stock traded at 17 times earnings. Earnings per share have risen roughly fifteen-fold since then. Anyone who thought 17 times earnings was expensive in 2012 was not merely wrong but spectacularly wrong, and the magnitude of that error is a useful calibration point for thinking about what cheapness even means in a business with this kind of compounding power.
Marriott is a royalty specifically on travel lodging. It does not own most of the hotels that carry its brands; it manages and franchises them, collecting fees on the revenue and profits that flow through its system. The capital intensity sits with the property owners, while the brand, the loyalty program, the distribution system, and the global reservation network sit with Marriott. Travel demand grows roughly with global wealth, and Marriott captures a steady percentage of an expanding pie without having to fund the buildings. The business has gone on sale regularly. Marriott traded for less than 20 times earnings for most of 2008, in 2011, in 2013, for much of 2016, in 2020, and again at several points in 2023. Each of those windows offered a chance to build a position in a "royalty-like" business that is growing without using incremental capital.
Perhaps Buffett does not think 20 times earnings is sufficiently cheap. We will return to that question later in the essay. But the dollar-cost-averaging strategy I described at the outset would have generated meaningful positions in both companies many times over, at prices that look in retrospect like obvious bargains.
What he bought instead were extremely capital-intensive businesses that have certainly under-performed the "royalty-like" businesses and in many cases have been outright disappointments. Berkshire's railroad acquisition BNSF earns a large headline profit. But in his 2023 annual letter, Buffett finally acknowledged what the net income number had been obscuring for years. Since Berkshire's purchase of BNSF fourteen years earlier, capital expenditures had exceeded depreciation charges by more than $22 billion, amounting to over $1.5 billion annually. "Ouch," he wrote. The depreciation charge was actually insufficient to maintain the road, and so the dividends BNSF paid to Berkshire consistently fell short of stated earnings. "Berkshire is receiving an acceptable return on its purchase price," he admitted, "though less than it might appear, and also a pittance on the replacement value of the property."
This was an extraordinary statement. The most celebrated capital allocator in history had spent fourteen years and tens of billions of dollars on a business that was, by his own accounting, delivering a pittance on replacement value.
Newsletter writer Porter Stansberry, in a recent open letter to Berkshire's board of directors, calculated that on a capital-weighted basis, the gap between what Berkshire earned from its whole-company acquisitions and what it would have earned from simply holding the superior publicly traded businesses in the same industries amounts to nearly one trillion dollars of lost value.
Embarrassingly, a comparison of each major Berkshire acquisition against the obvious public market alternative, available at the time of purchase, shows that Berkshire consistently chose the inferior business: See's Candies instead of Hershey, Nebraska Furniture Mart instead of Home Depot, Dairy Queen instead of McDonald's, BNSF instead of a basket of the publicly traded railroads. (See's Candies deserves a partial exemption from this list. Buffett has said it transformed his understanding of earning power divorced from capital requirements. The mistake was not buying See's; it was failing to apply the lesson at scale in the decades that followed.)
The Conventional Defense, and Why It Falls Short
The standard explanation for Buffett's capital allocation choices is size. Apologists say that Berkshire became so large that only a handful of investments could move the needle, and the universe of businesses large enough to matter grew small. But it does not fully explain the pattern. It might explain why Buffett could not buy small royalty businesses. It does not explain why he didn't accumulate large minority stakes in the best capital-light businesses available, since they were and are large enough to absorb serious capital. Google and Marriott are not small companies. The size constraint has become relevant over time, but it can only excuse so much.
Alex Rubalcava, a venture capitalist in Los Angeles, asked a question that goes to the heart of this tension at the 2003 Berkshire annual meeting. He noted that in his writings, Buffett expressed admiration for businesses that could employ large amounts of capital at high returns, while in other contexts Buffett and Charlie Munger both praised businesses that required very little capital. He asked whether these statements were contradictory.
Buffett's answer invoked Berkshire's scale: the sheer amount of capital it needed to deploy required businesses that could absorb it productively. At the time, this probably sounded reasonable to most of the audience. But the answer was actually an admission dressed up as an explanation. The real problem it revealed was that Buffett was unwilling to do what the situation logically required: return capital to shareholders.
A manager who genuinely could not find enough wonderful businesses at fair prices to buy could have simply paid dividends or repurchased stock. Buffett resisted this for decades, preferring to keep the capital inside Berkshire and working. And since he needed to deploy capital but was too cheap to pay fair prices for great businesses, he ended up buying a great many mediocre ones at prices that felt comfortable. BNSF. Berkshire Hathaway Energy. A collection of capital-intensive industrial businesses that satisfied the value investor's need for a familiar-looking multiple while quietly requiring capital investment in excess of what the depreciation charges showed. The scale argument was not an explanation for why Berkshire couldn't find better investments. It was a rationalization for why Buffett didn't have to confront the alternative.
The Long Decline in Returns on Capital
If the size argument is really a rationalization for not returning capital, the question becomes why Buffett insisted on deploying capital into the businesses he chose. The answer begins with cheapness. Buffett's deepest instinct is that of a value investor, and value investors are cheap. Cheapness is usually a virtue. It is the bedrock of margin of safety, the discipline that keeps an investor from paying for optimism that never arrives. But cheapness has a corresponding weakness: it makes it very difficult to pay what a genuinely great business is actually worth. The royalty businesses Buffett should have been buying rarely looked cheap on conventional metrics. Companies like Google, Marriott, CME Group, or Texas Pacific Land have almost never traded at the kind of headline P/E multiples that would have satisfied a classically trained value investor. Buying them required accepting that the world had changed, and that returns on capital have been falling across the entire economy as the world has grown wealthier.
It is worth taking this point seriously, because it is the central macroeconomic fact of the past two centuries and the one Buffett's framework has had the hardest time absorbing.
There was a time when a person could live off the yield on government bonds. If you read a Jane Austen novel, the characters are getting a certain number of pounds per year, which was the interest on their consols, the perpetual government debt of the British empire. (See: 1, 2.) The yield was probably around four percent in sound money, which was a significantly higher real yield than the 2.7% that thirty-year TIPS offer today. (Not to mention the important difference in income tax rates.) A modest fortune invested in consols funded a country house, servants, horses, and a respectable place in society. Today, the real yield on government debt, particularly after income tax, varies between negligible and negative. The same nominal capital that supported a Regency-era family in idleness would have trouble supporting a retired couple in a nice place to live today.
The same thing has happened to real estate investments within living memory. During our lifetimes, there were times and places when you could buy residential real estate (like a single family house) and the rent would cover the mortgage. The cap rate on the property was perhaps a hundred basis points higher than the cost of the mortgage, so the property paid you to own it from day one. That has never recurred, even during the housing crash and the various housing corrections that followed. Cap rates on residential real estate and on similar streams of safe cash flow have compressed steadily toward the TIPS yield. The cap rate on a house in a nice part of the U.S. will generally sit a couple hundred basis points below the cost of a mortgage. The economics of being a small landlord have gone from positive carry to substantial negative carry, and the change is structural rather than cyclical.
It is possible, and I think likely, that equities are going through a similar transition. I am not saying this is certain. It is a hypothesis. But, as Kris Abdelmessih has written, the mass adoption of passive investing may be the invisible hand wringing the equity index risk premium out of the market. The mechanism is straightforward. As markets have built and popularized increasingly cost-effective ways to diversify away idiosyncratic risk, the price discounts that risky investments need to offer in order to attract capital have come down. A century ago, holding a diversified equity portfolio was expensive and difficult, and the equity risk premium that compensated investors reflected that friction. The friction is now nearly zero. There is no obvious reason for the premium to stop compressing.
Future investors may eventually look back at twentieth-century equity returns the way we now look at twentieth-century housing affordability or government bond yields. Can you believe people earned 8-10% real in stocks and thought that should last forever?
There is a second change in the investment landscape that compounds the first. As we pointed out in our review of the American Express corporate history, large company management teams have become meaningfully better at capital allocation over the past three decades, and they have purified their business models to deliver higher returns on capital. The conglomerate era of the 1960s and 1970s produced sprawling businesses with undisciplined balance sheets, divisions that subsidized losers with the cash flow from winners, and managements that confused empire building with value creation. The shareholder revolution of the 1980s and 1990s, however uncomfortable and uneven, forced companies to think harder about what they actually did well. The businesses that survived and thrived tended to be those that identified their core advantage and stopped subsidizing everything else. They divested non-core operations, returned capital aggressively, focused on the highest-return uses of incremental dollars, and managed the balance sheet with discipline.
The result is that the best public companies today are structurally better businesses than their predecessors of a generation ago. They are not merely benefiting from accommodative monetary policy or temporary advantages. They are run by people who have absorbed forty years of accumulated lessons about how capital should be deployed and which businesses deserve to be inside a corporate structure at all. A company like CME Group is not expensive because the market has lost its mind; it is priced to reflect a business model that has been refined to near-perfection and sits atop a franchise that is genuinely difficult to displace. These businesses deserve to trade at higher multiples than what a value investor trained in the 1950s would consider normal, both because returns on alternative capital have collapsed and because the businesses themselves have improved.
The Insurance Mind
We can think of an additional explanation for why cheapness is so ingrained in Buffett's behavior, despite the two factors that we mention above (falling returns on capital and rising business quality) and it has to do with where Buffett's temperament was formed.
Berkshire's insurance operations are genuinely extraordinary. Buffett is one of the great insurance minds of the twentieth century, and the discipline he built at GEICO, General Re, and the reinsurance operations reflects a profound understanding of how that business works. In insurance underwriting, patience and stubbornness are not just desirable attributes, they are the whole kahuna. In a hard market, pricing is attractive, terms are disciplined, and underwriters earn good returns. In a soft market, competitors cut prices to buy volume, and the disciplined underwriter who refuses to write business at inadequate rates will appear to be leaving money on the table right up until the cycle turns and the undisciplined writers get crushed. The market cycles from hard to soft and back again with regularity. The curmudgeon who said no all through the soft market is vindicated. Stubbornness, in the insurance business, is the strategy.
Buffett's insurance success caused him to internalize this model deeply. He then applied it to equity markets, where it also worked for a long time. The logic transferred reasonably well: be patient, wait for the "hard" market in stocks (a crash), and then deploy capital aggressively when prices reach levels that others find unbearable. The approach was spectacularly rewarded in 1974, in 1987, in 2002, and again in 2008. Each of those dislocations produced the kind of prices that justified decades of waiting.
But the hard markets in equities are getting structurally shallower, for two related reasons. The first is that the world is wealthier, and more capital is now competing for distressed assets. When prices fall, the buyers appear faster and in greater numbers than they did in previous generations. The dislocations do not go as deep or last as long. The second reason is fiat currency and the modern toolkit of central bank intervention. The authorities now have both the instruments and the political will to short-circuit financial crises before they reach the depths that create Buffett-style generational buying opportunities. The Federal Reserve's response to the 2020 pandemic is a perfect example. What looked briefly like a 2008-style dislocation was resolved in a matter of weeks. Would-be buyers of distressed assets, including Buffett, barely had time to act before the floor was put back under markets. It happened again in 2023 where a Federal Reserve bailout (Bank Term Funding Program) preempted any distressed investment by Berkshire in the regional banks.
So the cash piles up at Berkshire. He is waiting, with the discipline and the temperament that made him one of the greatest investors in history, for a hard market that the modern financial system is increasingly engineered to prevent. The insurance mind, which was his greatest professional asset in the business that built Berkshire, has become a constraint in a world where the cycles that once justified it are being deliberately suppressed. He is not merely cheap in the abstract. He is cheap in a specific, structured way, expecting the world to present him with the prices it presented in 1974. That world might not be coming back, and Berkshire's cash pile is the visible evidence of a strategy waiting for conditions that may no longer reliably arrive.
The Seduction of Scale
Cheapness, as reinforced by experience in insurance underwriting, may not fully explain the pattern. There may be another, less flattering force at work.
There is a meaningful parallel to James Ling, who built LTV Corporation into one of the great conglomerate empires of the 1960s through relentless acquisition. Ling was a genuine financial genius, a student of deal structure and capital markets who understood things about reorganizing corporate balance sheets that few of his contemporaries grasped. He was also, ultimately, a man for whom size had become a goal in its own right, a psychological reward that gradually overwhelmed his judgment about what actually created value. The behavioral fingerprint is this: a highly intelligent person who articulates sound principles about capital allocation and then consistently makes decisions that build scale rather than per-share value. LTV eventually collapsed under the weight of its ambitions, and Ling was left with almost nothing.
Buffett acolytes would reject the comparison, and the differences are real. He has not used leverage recklessly. He has not chased acquisitions for the sake of novelty or prestige in any obvious way. But the behavioral pattern is recognizable. Scale brings attention. It brings adulation. It brings Becky Quick flying to Omaha. It brings heads of state returning calls. It transforms a successful investor into an American institution, a monument, a figure whose annual letter is studied by millions as a kind of secular scripture.
A portfolio of quietly compounding royalty interests in land companies, pipeline partnerships, and hotel franchisors would not generate that kind of gravity. It is harder to narrate as a legacy. You cannot easily explain to a television audience why you own four percent of Marriott. You can explain, in terms that feel weighty and serious, why you own the entire BNSF railway.
There is also a subtler dynamic. Businesses that generate substantial free cash flow without obvious opportunities for reinvestment create pressure to return capital. If CME Group is throwing off cash that Berkshire cannot redeploy at scale, it must be dividended out or used to buy back shares. The empire, in that sense, stops growing, or at least grows differently, returning capital to owners rather than accumulating assets. That is the correct outcome for shareholders. It is a different story than the one Buffett has been telling for sixty years, and it is a story that does not lend itself to the particular kind of immortality he has spent his career constructing.
There is a sympathetic counterargument worth acknowledging. Buffett's model of permanent, patient ownership built something with genuine value beyond the returns: a reputation that attracted exceptional businesses on favorable terms, a culture of decentralized autonomy that good managers found appealing, a guarantee that the businesses Berkshire bought would not be flipped or dismembered. The argument is that this reputational capital created deal flow that a minority-stake approach could not have replicated. The counterargument is real, but it does not survive contact with the BNSF numbers. A model that generates deal flow into capital-intensive businesses delivering a pittance on replacement value is not a competitive advantage; it is a more polite way of destroying shareholder value.
The Theory of the Puzzle
The theory of the Buffett Puzzle, then, has three interlocking parts. The first is that Buffett's temperament was formed in insurance, where stubborn patience waiting for the hard market is the correct strategy. He transferred that model to equities, where it also worked for several decades. But fiat currency, central bank activism, and the general increase in global wealth have made the hard markets shallower and less frequent. The insurance mind is now waiting for conditions that the modern financial system is built to prevent.
The second is that cheapness, already a deep instinct, prevented him from recognizing that the world had changed. Returns on safe capital have been falling gradually for two centuries and falling more sharply for several decades, in a long-running compression that has touched bonds, real estate, and now arguably equities. At the same time, the best public companies have become structurally better businesses than their predecessors. Both forces argue that royalty businesses with durable franchises deserve to trade at multiples that look expensive by historical standards. Buffett could not bring himself to pay those multiples, and so he passed on the businesses he had described as ideal.
The third is that scale provided an emotional reward for the alternative. Deploying large capital into large, visible, controllable businesses built a monument. Accumulating quiet minority stakes in the best businesses available and returning excess cash to shareholders would have been better for owners but harder to narrate as a life's work.
These three forces reinforced each other. The insurance mind supplied the patience. The cheapness supplied the principled-sounding reason to pass on royalty businesses priced to reflect a permanently changed investment landscape. And the appetite for scale supplied the emotional reward for deploying that capital into something more visible instead. The result was BNSF, Berkshire Hathaway Energy, and a collection of capital-intensive industrial businesses that look impressive on an organizational chart and have consumed, by Buffett's own accounting, tens of billions of dollars more capital than their reported earnings suggest.
The Question for Greg Abel
Warren Buffett is now stepping back, and Greg Abel is taking over as CEO of Berkshire Hathaway. Abel inherits a company shaped by the forces this essay has described. As he acknowledged in his inaugural letter to shareholders, there is a collection of capital-intensive businesses acquired at prices that felt comfortable, generating less free cash flow than their reported earnings suggest, sitting alongside a cash pile so large it has become its own problem.
Abel also inherits an opportunity that is, by any measure, extraordinary. He could choose to become one of the great capital allocators in the history of American business! The path to doing so would require something that would strike most Berkshire observers as bizarre, even heretical. He should shrink the company.
William Thorndike's book The Outsiders (see review) profiles eight CEOs whose returns greatly surpassed the market over long periods. The common thread was not operational brilliance, though several were excellent operators. It was capital allocation discipline, and specifically the willingness to do whatever the situation actually required rather than what felt comfortable or looked impressive. Henry Singleton of Teledyne bought back nearly ninety percent of his company's shares when his stock was cheap. Bill Anders of General Dynamics sold divisions and returned the cash when he could not find acquisitions that made sense. Tom Murphy of Capital Cities avoided the temptation to build for its own sake and focused relentlessly on per-share value. All of them were willing to make the company smaller when making it smaller was the right answer. None of them were building monuments.
Ironically, Buffett himself appears in the book, and was once exactly this kind of CEO. The early Buffett, running the partnership and the early years of Berkshire, was as disciplined about per-share value and capital return as anyone Thorndike profiles. The later Buffett, presiding over an accumulation of capital-intensive businesses too large to sell and too mediocre to celebrate, is a different story.
Abel's most consequential early decisions may well be subtractions rather than additions. Prune or spinoff underperforming operating businesses. Begin buying minority stakes in the best capital-light businesses available rather than acquiring whole companies at negotiated prices. And if Berkshire's own stock is the best thing available, buy that. What would be genuinely bold, and genuinely correct, is to stop managing Berkshire for scale and start managing it for per-share value, even if that means the company gets smaller, quieter, and harder to narrate as a legacy.
Warren Buffett is one of the greatest investors who has ever lived. That is not in question. He described the ideal destination, saw it clearly, and then drove somewhere else. A combination of forces got him there: a temperament built for insurance cycles that the modern financial system keeps suppressing, a value investor's cheapness anchored to a world where returns on capital were higher and businesses were worse than they are today, and the seductive pull of scale and the attention it brings. None of these forces were obviously bad in isolation. Together, they compounded into a decades-long detour.
The question for Greg Abel is whether he has the nerve to turn the car around.
Posts
