Berkshire Hathaway 2012 — key metrics
Selected figures from the 2012 Annual Report
Buffett 2012 — "We do better when the wind is in our face"
From Warren Buffett's 2012 Berkshire Hathaway shareholder letter: the asymmetry behind trailing the S&P and why it was by design; insurance float hitting $73 billion for a tenth consecutive underwriting-profit year; the Heinz deal introducing a preferred-plus-equity partnership structure with 3G Capital; a $344 million thesis on community newspaper moats and "survival of the fattest"; look-through earnings and why $2.8 billion of unreported income is "every bit as valuable"; and the sell-off vs. dividend mathematics with a warning on motivated reasoning.
Published: March 1, 2013 — Warren Buffett, Berkshire Hathaway 2012 Annual Report
The 2012 letter is the one where the streak almost ended. After 43 consecutive five-year periods of outperforming the S&P 500, Berkshire trailed the index in calendar 2012 — book value up 14.4% against the S&P's 16.0% total return — and Buffett said plainly in the first page that another up year in 2013 would snap the run. 1
What makes the letter worth studying is not the underperformance itself, which was minor and accompanied by a $24.1 billion absolute gain in net worth. 1 What makes it worth studying is how Buffett handles it: no revised yardstick, no softened comparison, and a statistical observation that cuts against the instinct to apologize. In eight of the nine years when Berkshire's book value gain fell below the S&P's, the index itself had gained 15% or more. The underperformance is not random — it clusters in the years when the market is the most ebullient. That is by design.
正在加载统计卡片…
"We do better when the wind is in our face"
The performance section opens with a long-range table comparing Berkshire's per-share book value compounded annually since 1965 (19.7%) against the S&P 500 (9.4%), producing an overall gain of 586,817% against 7,433%. 1 Buffett then immediately turns to the current year's shortfall and explains the asymmetry:
"Charlie and I believe the gain in Berkshire's intrinsic value will over time likely surpass the S&P returns by a small margin... Our relative performance, however, is almost certain to be better when the market is down or flat. In years when the market is particularly strong, expect us to fall short." 1
The logic behind this is structural, not psychological. Berkshire's insurance float, its railroad, its manufacturing businesses — none of them respond to P/E multiple expansion the way an index fund does in a pure bull market. What they offer is durability: the ability to earn reliably when the index is collapsing. That durability has a cost in exuberant years, and Buffett tells shareholders to expect it.
His response to the end of the five-year streak was characteristically brief: "One thing of which you can be certain: Whatever Berkshire's results, my partner Charlie Munger, the company's Vice Chairman, and I will not change yardsticks." 1 That sentence is both accountability and commitment: he is keeping the measure that makes him look bad in good years, and he is refusing to replace it with one that would.
For the investor: the willingness to report underperformance using an unchanged benchmark — in a letter that goes to hundreds of thousands of shareholders — is rarer than it looks. Most management teams quietly retire the benchmark when it becomes unflattering. Buffett's refusal to do so is itself a signal about how the business is run.
Float hits $73 billion: ten years of cake and eating it too
By yearend 2012, Berkshire's insurance float reached $73.125 billion — up $2.5 billion from 2011 — and the insurance operations collectively earned a $1.625 billion underwriting profit, marking the tenth consecutive year of profitable underwriting. Cumulative pre-tax underwriting profit over the decade: $18.6 billion. 1
正在加载图表…
Buffett had written in the 2011 letter that float had likely reached its peak. His CEOs proved him wrong, adding $2.5 billion in a single year, and he acknowledged the error without ceremony. 1 On the accounting treatment, he offered the clearest statement yet of why the liability on Berkshire's balance sheet overstates the true economic obligation: if the float is both costless and long-enduring, "the true value of this liability is dramatically less than the accounting liability." 1 This gap — between what GAAP records and what the business is actually worth — is one of the two running explanations for why Berkshire's intrinsic value substantially exceeds its book value.
Buffett summarized the 2012 result with characteristic directness: "Our insurance operations shot the lights out last year. While giving Berkshire $73 billion of free money to invest, they also delivered a $1.6 billion underwriting gain, the tenth consecutive year of profitable underwriting. This is truly having your cake and eating it too." 1
He also flagged a coming headwind: the legacy bond portfolios that once earned high yields on Berkshire's float are rolling into lower-yielding securities. "Today's bond portfolios are, in effect, wasting assets." 1 The warning applies to every insurer with a large fixed-income portfolio accumulated during higher-rate periods.
For the investor: when evaluating an insurance holding, the question is not "how large is the float?" but "how much does the float cost?" An underwriting profit means the cost is negative — the company is paid to hold and invest the capital. A combined ratio above 100 means the float is expensive leverage. Ten consecutive years of profitable underwriting across the range of catastrophes that 2012 included (Hurricane Sandy caused GEICO's largest single loss in history — 46,906 vehicles, more than triple Katrina's losses — yet the operation still earned a profit) is not luck. 1 It is underwriting discipline systematically maintained.
The Heinz deal: Berkshire meets 3G
The biggest piece of news in the 2012 letter was not from 2012. On February 14, 2013, Berkshire and 3G Capital — the Brazilian-led private equity group co-founded by Jorge Paulo Lemann — agreed to jointly acquire H. J. Heinz. 1
The structure is worth examining. Berkshire and 3G each contributed approximately $4 billion for common equity in a new holding company. On top of its equity share, Berkshire also invested $8 billion in preferred shares paying a 9% dividend, redeemable at a significant premium, with warrants to acquire 5% of the common stock for a nominal sum. Total Berkshire commitment: roughly $12 billion — "much of what Berkshire earned last year," in Buffett's words. 1
This preferred-plus-equity structure recurs across Berkshire's large private deployments: Goldman Sachs in 2008, Bank of America in 2011. The mechanics are similar each time — equity-like upside through warrants, creditor-like priority through the preferred dividend, plus a redemption premium that adds further return if the instrument is called. It is a structure that extracts multiple layers of return from a single capital commitment.
Buffett framed the partnership's operational logic in a single sentence: "We couldn't be in better company. Jorge Paulo is a long-time friend of mine and an extraordinary manager." 1 The implicit point is that 3G brings operational intensity — the systematic zero-based budgeting and management discipline it applies to consumer brands — while Berkshire provides permanent capital with no exit mandate. Neither side supplies what the other brings.
He also signaled that this structure could repeat: "Our luck, however, changed early this year." 1 For an investor trying to understand what Berkshire was becoming at scale — too large for most deals to move the needle, but capable of deploying $12 billion in a single transaction — the Heinz structure was a preview of Berkshire's acquisition architecture at the elephant level.
For the investor: the preferred-plus-equity structure is a repeatable template for large capital deployments where the investor wants debt-like downside protection but equity-like participation in the upside. The key variables are the dividend rate (9% here), the redemption premium (which increases effective yield on the preferred), and the warrant strike price (nominal in this case, making them essentially free equity optionality). Any time this structure appears in a Berkshire announcement, the investor should evaluate all three terms together, not the headline equity stake alone.
Newspapers: "survival of the fattest"
One of the most unexpected sections in the 2012 letter is the 1,500-word thesis on local newspaper economics. Over the prior 15 months, Berkshire acquired 28 daily newspapers for a total of $344 million — an average of roughly $12 million per paper. 1
Buffett acknowledged the objection immediately: he had long predicted that newspaper circulation, advertising, and profits would decline. These papers also fall far below Berkshire's usual acquisition size threshold. So why buy?
His thesis rested on a narrow claim about community moats:
"Wherever there is a pervasive sense of community, a paper that serves the special informational needs of that community will remain indispensable to a significant portion of its residents." 1
The mechanism was not advertising nostalgia but a specific kind of local monopoly. In two-paper markets, whichever paper captured more readers attracted more advertisers, which funded more content, which retained more readers — what Buffett called a "survival of the fattest" dynamic. 1 That process eliminated most competitors decades ago. What survived was usually a single paper with a near-monopoly on local civic information: court records, school sports, local elections, zoning decisions.
The operating requirements he prescribed were specific: comprehensive local news coverage, a sensible internet pay strategy, daily publication frequency, and independent editorial judgment. A less-than-daily schedule, he wrote, "seems certain to diminish the papers' relevance over time." 1 The six small dailies owned throughout 2012 had unchanged revenues; the two large papers (Buffalo News and Omaha World-Herald) held revenue loss to 3%. 1
His summary of news itself — "News, to put it simply, is what people don't know that they want to know" 1 — is as clean a statement of the editorial value proposition as exists in the investing literature.
He was not sentimental about the economics: cash earnings "will almost certainly trend downward over time." 1 The investment case depended entirely on buying at a low enough multiple of current earnings that even a declining stream would exceed Berkshire's acquisition hurdle.
For the investor: the "survival of the fattest" dynamic — the virtuous circle where scale in distribution leads to more advertiser revenue, which funds more content, which retains more subscribers — describes the structural advantage in any platform or network-effect business, not just newspapers. Buffett's investment discipline here is also worth noting: he entered a secularly declining industry at prices low enough to justify the investment purely on the existing earnings base, without requiring the decline to reverse. The discount to intrinsic value does the analytical work that growth assumptions cannot.
The Big Four and look-through earnings
Berkshire's four largest equity positions — American Express, Coca-Cola, IBM, and Wells Fargo — generated $3.9 billion in total earnings attributable to Berkshire's ownership stakes in 2012. Of that, only $1.1 billion appeared on Berkshire's income statement as dividends. The remaining $2.8 billion, retained and reinvested by the underlying companies, was not recorded in Berkshire's reported income. 1
"The $2.8 billion of earnings we do not report is every bit as valuable to us as what we record." 1
This is the "look-through earnings" concept. GAAP income from a minority equity stake captures only the cash distributed; the retained earnings compound at the underlying company's return on equity and accrue to Berkshire's intrinsic value without appearing in the income statement. At yearend 2012, unrealized capital gains on the Big Four totaled $26.7 billion. 1
Berkshire increased its ownership in all four through the year: Wells Fargo grew from 7.6% to 8.7%, IBM from 5.5% to 6.0%, with Coca-Cola and American Express increased through buybacks at those companies. Buffett quoted Mae West on the direction of travel: "Berkshire's ownership interest in all four companies is likely to increase in the future. Mae West had it right: 'Too much of a good thing can be wonderful.'" 1
For the investor: look-through earnings provide a more accurate picture of a holding company's earnings power than reported net income alone. Whenever a company holds large minority stakes in profitable businesses, the investor should calculate the look-through earnings by multiplying each ownership percentage by the underlying company's total earnings, then compare that figure to what actually flows through to GAAP income. The gap is often larger than it appears on the face of the income statement — and it is systematically understated by the accounting rules that govern minority stakes.
The dividend vs. sell-off mathematics — and the motivated-reasoning warning
The capital allocation section of the 2012 letter includes two arguments that reward careful reading.
The first is a detailed mathematical demonstration of dividend policy. Buffett constructed a hypothetical company earning 12% on net worth and selling at 125% of book value — conditions he described as reasonable for Berkshire, though not assured. Under a 3.2% annual sell-off (the shareholder liquidates 3.2% of holdings each year), the investor ends up after ten years with roughly 4% more capital value and 4% more annual cash than under a dividend policy paying out the equivalent cash amount. 1 The difference arises because retained earnings compound at the company's return on equity rather than being redeployed by the shareholder at potentially lower rates.
He capped the dividend discussion with Phil Fisher's restaurant analogy: "You can successfully run a restaurant that serves hamburgers or, alternatively, one that features Chinese food. But you can't switch capriciously between the two and retain the fans of either." 1 Consistency of capital allocation policy matters not because any single year's decision is large, but because the ownership base a company attracts depends on knowing what it will do with earnings.
The second argument is on motivated reasoning in corporate decision-making. Buffett described the characteristic failure mode of executives who destroy value through acquisitions or excessive share issuance: "The usual cause of failure is that they start with the answer they want and then work backwards to find a supporting rationale. Of course, the process is subconscious; that's what makes it so dangerous." 1
The subconscious element is the key qualifier. An executive who knows they are rationalizing can stop. The executive who is unaware of it cannot. Buffett's prescription is structural: design incentives and governance so that the answers executives are "supposed" to reach are actually correct ones.
For the investor: the sell-off vs. dividend math is a direct analytical tool. Given a company's return on equity and its market-to-book ratio, the calculation shows whether a dividend policy is value-creative or merely a preference concession to income-seeking shareholders. At high ROEs and above-book market prices, retained earnings compound faster than distributed cash can be redeployed — the Berkshire sell-off model wins. At low ROEs or below-book prices, the calculus may reverse. Run the numbers before attributing moral value to a company's dividend policy.
正在加载内容卡片…
The 2012 letter closes with a sentence about capital expenditures that lands differently from the surrounding prose: "We will keep our foot to the floor and will almost certainly set still another record for capital expenditures in 2013. Opportunities abound in America." 1 BNSF, which had just announced $4 billion in 2013 capex (roughly double its depreciation), and MidAmerican Energy, with $13 billion committed to renewables, were the primary vehicles. The railroad was transporting 500,000 barrels of oil per day — 10% of lower-48 U.S. production — a revenue stream that simply did not exist when Buffett bought the company in 2009. 1
The year Berkshire trailed the S&P, it was also deploying capital at the highest rate in its history, building infrastructure for decades of future earnings. That juxtaposition — a subpar annual number alongside record long-term investment — is probably the most compressed illustration in any of these letters of what long-term ownership actually means.
Cover image: AI-generated illustration
围绕这条内容继续补充观点或上下文。