BERKSHIRE HATHAWAY INC (BRK-A) stock intrinsic value fair value estimate & owner earnings analysis

All figures in this analysis are per Class A share. A Class B share is worth 1/1,500th of these amounts. A correction first: our previous analysis expressed every per-share figure on a Class B basis, which made our value estimate look about 1/1,500th of what it should have been next to the Class A price. This analysis fixes that and also revisits several assumptions, explained below. Berkshire Hathaway is a holding company with two halves. The first is a set of wholly owned operating businesses: property-casualty insurance and reinsurance (GEICO, the Berkshire Hathaway Primary Group and the Berkshire Hathaway Reinsurance Group), the BNSF railroad, Berkshire Hathaway Energy (regulated utilities including PacifiCorp, MidAmerican and NV Energy, plus natural gas pipelines), and a large manufacturing, service and retailing group that includes Precision Castparts, Lubrizol, Clayton Homes, Pilot travel centers, See's Candies and, since 2026, OxyChem and the homebuilder Taylor Morrison. The second is an investment portfolio worth about $727 billion at June 30, 2026: roughly $365 billion of cash and U.S. Treasury Bills, $324 billion of stocks (the largest holdings are Alphabet, American Express, Apple, Bank of America and Coca-Cola), $17 billion of bonds, and about $21 billion at market value in Kraft Heinz and Occidental Petroleum. How we valued it. A single cash-flow valuation of the whole company would only see the dividends and interest the portfolio throws off and would badly understate what it is worth. So we valued Berkshire in pieces: the portfolio at market value, plus a discounted cash-flow value of the operating businesses' own earnings (leaving out investment income so nothing is counted twice), minus what the company owes. The portfolio. We counted it at June 30 market values, holding back $30 billion of cash, the level below which Berkshire says it will not let its cash fall. About $177.5 billion of the portfolio is funded by insurance float, meaning premiums collected before claims are paid. We treated ongoing float as a permanent, cost-free source of funds rather than as debt: it has grown almost without interruption for decades, and the insurance businesses have earned underwriting profits in most years, so policyholders have effectively paid Berkshire to hold their money. The exception is about $22 billion of net liabilities from old retroactive reinsurance contracts, which are winding down and will be paid out; we deducted that in every case. Berkshire also owes tax on about $217 billion of unrealized gains in its stock portfolio, roughly $46 billion at the 21% corporate rate. That tax is only paid when shares are sold, and Berkshire has been selling (taxable gains of $9.5 billion in the first half of 2026 alone), so we deducted the full amount in the bear case, half in the base case, and a quarter in the bull case, reflecting that its oldest holdings such as Coca-Cola and American Express have been held for decades without sale. Note that because we value the stocks at market prices, a fall in the stock market would lower our estimate directly. The railroad, energy and manufacturing/service/retail businesses. Measured before interest and after deducting all capital spending, these businesses generated between $20.7 billion and $24.0 billion a year for their owner from 2021 to 2025; we used the $22.3 billion median as our starting point. That figure has drifted lower because capital spending at the railroad and especially the utilities has grown faster than depreciation: BNSF and Berkshire Hathaway Energy spent $14.4 billion in 2025 and expect about $15 billion in 2026. We counted all of that spending as a cost of staying in business, and in turn credit the modest earnings growth it produces. This measure fell about 3% in 2025, while our base case assumes about 2% growth in the first year. The reason is the first half of 2026, which was markedly stronger: compared with the first half of 2025, BNSF earnings rose 9.5%, Berkshire Hathaway Energy's rose 11.5%, and manufacturing, service and retail rose 15%, partly from OxyChem. Figures for 2021 and 2022 exclude Pilot, which was then only partly owned, so the starting point is slightly conservative. Insurance underwriting is treated separately because it swings so much. After-tax underwriting profit was $7.3 billion in 2025 and $9.0 billion in 2024, but management itself describes the last three years as exceptional compared with longer periods and warns that competition and rising claim costs may reduce earnings. From 2016 through 2022 the figure ranged from a $2.2 billion loss to a $1.6 billion profit. We assumed $3.0 billion a year in the base case, $0.5 billion in the bear case and $5.5 billion in the bull case, with no growth in any case. OxyChem, bought for $9.4 billion in January 2026, is not in our historical earnings, so we added it at what Berkshire paid. Taylor Morrison, bought for $6.8 billion in July 2026, swapped cash for a business, which we treated as an even exchange. Debt. Berkshire and its subsidiaries owe about $129 billion, mostly long-dated fixed-rate bonds at BNSF and Berkshire Hathaway Energy plus Berkshire's own yen- and euro-denominated notes, at an average interest cost of about 3.9%. Because this debt costs well below the 6% rate we use to discount future cash, its economic burden is less than its face value; we carried it at about $103 billion. Outside owners of partly owned subsidiaries account for another $2 billion. The operating businesses cover their interest roughly four times over even in the bear case, and no maturity or covenant materially affects the outcome. Bear case: underwriting falls back to its 2016 to 2022 norm; the non-insurance businesses start from their 2025 level and shrink 1% a year for a decade before leveling off, reflecting rising capital needs, continuing wildfire litigation at PacifiCorp, and the retail weakness seen in 2025; the full deferred tax is paid; and the OxyChem and Taylor Morrison purchases prove worth 25% less than Berkshire paid, as happened when it wrote down roughly $10 billion of its Precision Castparts acquisition in 2020. The operating businesses are worth about $328 billion and owe about $103 billion, leaving $225 billion; adding the portfolio and the other items above gives about $598,000 per share. Base case: underwriting earns $3.0 billion a year, and the non-insurance businesses grow 2% a year for ten years and 1.5% a year for the following twenty before leveling off. That growth rests on two mechanisms each demonstrated over decades: the railroad's pricing power and the regulated utilities' ability to earn allowed returns on the capital they invest. The operating businesses are worth about $535 billion and owe about $103 billion, leaving $432 billion; in total about $762,000 per share. Bull case: underwriting earns $5.5 billion a year, below the first half of 2026's annualized pace of about $6.9 billion; the non-insurance businesses grow 3% a year for ten years as the $15 billion-a-year utility and rail investment programs earn their returns and the railroad's 2025 to 2026 efficiency gains continue, then 2% a year for twenty years. The operating businesses are worth about $640 billion and owe about $103 billion, leaving $537 billion; in total about $843,000 per share. What we left out. We gave no value to redeploying the cash pile into new acquisitions or buybacks at attractive prices. Berkshire has a long record of doing so, but the timing and terms cannot be known, so the cash is counted at face value only. Greg Abel became CEO at the start of 2026; we made no assumption about whether his capital allocation will be better or worse than Warren Buffett's. Margin-of-safety price. More than half of what a share is worth is cash, Treasury Bills and widely traded stocks counted at market value, so the range of outcomes is unusually narrow, with the bull case only about 40% above the bear. Berkshire pays no dividend, so a return has to come from growth in per-share value, which we estimate at roughly 5% to 6% a year, plus the share price closing the gap to our estimate. The margin-of-safety price is $560,000, about 1.07 times book value of roughly $522,000 per share and about 6% below even the bear case; if the price returns to our base-case value within about two years, with value growing in the meantime, the base case implies roughly 20% a year. What would make this wrong: a large catastrophe or reserve problems in the insurance businesses, a sharp decline in the stock market, wildfire liabilities at PacifiCorp beyond what has been recorded, or a major acquisition made at a poor price. Accounting notes. Berkshire's reported net income swings enormously because accounting rules require unrealized stock gains and losses to run through earnings; management tells investors to look past this, and none of our figures are built from reported net income. We also did not rely on total operating cash flow for the operating businesses, because Berkshire reports most of its Treasury Bill interest (about $12 billion before tax in 2025) as an investing item while reporting the taxes on it in operating cash flow, which makes operating cash flow understate what the company earns. Instead we built the operating figure from each business's after-tax earnings plus depreciation minus capital spending. We could not cleanly separate short-term working-capital swings from ordinary growth in insurance float, so we treated that item as zero; float growth has historically been a durable source of funds rather than short-term noise.