How We Analyze Berkshire Hathaway

How we analyze Berkshire Hathaway: the business first

How we analyze Berkshire Hathaway is the way we analyze any company: business quality first, then the balance sheet, then valuation, then the long-term record. The order is deliberate, and none of the four steps is optional. Berkshire is not a normal operating company; it is a holding company that owns dozens of businesses outright, from the BNSF railroad and a large electric utility to insurers, a candy maker and a paint company, and it holds a stock portfolio worth hundreds of billions on top. The first thing to understand is how those pieces fit together. You can follow along in the live Berkshire report on Tenet, which pulls the same figures used below.

Analyzing Berkshire means analyzing a portfolio, not a single product line. Its value is the sum of the businesses it owns, the stocks it holds, and the cash it sits on, all managed by a tiny central team that allocates capital among them. Judging it well requires looking through the conglomerate structure to the parts underneath, and resisting the urge to reduce a company this varied to a single tidy number. The pieces do not all behave the same way, and treating them as one blurs the picture rather than sharpening it.

Insurance float and the wholly-owned businesses

Berkshire's distinctive engine is insurance float, and it is the key to how the whole machine is funded. An insurer collects premiums today and pays claims later, sometimes years later. The money it holds in between is called float, and Berkshire gets to invest that float for its own account. As long as its insurance operations at least break even on the policies they write, that float is effectively free money to invest, and Berkshire carries tens of billions of it. Ordinary companies must raise capital by borrowing or issuing stock; Berkshire is handed low-cost capital by its own insurers.

That float has funded the second layer: the wholly-owned businesses. Over decades Berkshire used its investable capital to buy entire companies, which now generate the bulk of its operating profit. These range from capital-heavy operations like the railroad and utility to asset-light brands, and together they throw off steady cash that the central team redeploys. The skill that ties it all together is capital allocation, deciding where each dollar of profit and float goes next, which is the subject of capital allocation explained and arguably Berkshire's single greatest advantage.

Float is not free of risk, and it is worth understanding the catch. If Berkshire's insurers write bad policies and pay out more in claims than they collect in premiums, the float carries a real cost, and a single catastrophe year can produce large underwriting losses. Berkshire's edge is a long culture of underwriting discipline, a willingness to shrink when prices are poor rather than chase volume. That discipline is what has kept the float cheap over decades, but it depends on managers who are willing to say no to growth, which is not guaranteed forever.

The honest concern with Berkshire is size and succession. The company is now so large that few acquisitions can move the needle, which caps its growth, and its returns are unlikely to match the extraordinary rates of past decades. When a business is worth close to a trillion dollars, even a brilliant billion-dollar purchase barely registers. It has also depended heavily on the judgment of its long-time leadership, and the transition to a new generation of managers is a real variable that every buyer has to weigh. A wide moat can coexist with a slowing growth rate, and any fair analysis says so plainly rather than assuming the past repeats.

Why book value now understates the business

For most of Berkshire's history, book value per share was the yardstick, and management pointed investors to it directly. That guide has become misleading, and understanding why is central to valuing the company today. Book value records the wholly-owned businesses at what Berkshire paid for them long ago, plus retained profits, not at what they are worth now. A railroad or a group of consumer brands bought decades ago and grown ever since can be worth far more than its carrying value on the balance sheet.

As Berkshire shifted from holding marketable stocks, which are carried at current market prices, toward owning whole businesses, which are carried at cost, the gap between book value and true worth widened. The result is that book value has become a floor rather than a fair estimate: Berkshire is worth at least its book value and, for the operating businesses, quite a bit more. This is exactly the kind of case where the price-to-book ratio must be read with care, a caution explained in price-to-book (P/B) ratio. At the end of fiscal 2025 Berkshire reported shareholders' equity of about $717 billion, and the shares traded at roughly 1.5 times that book value, a multiple that looks higher until you remember book value understates the assets behind it.

GAAP earnings noise and the fortress balance sheet

Berkshire's reported earnings are among the noisiest of any large company, and reading them wrong is the most common analytical mistake. Accounting rules require Berkshire to run the unrealized gains and losses on its enormous stock portfolio through net income every period. Because that portfolio is worth hundreds of billions, a normal market swing can add or subtract tens of billions from reported profit, none of which reflects how the underlying businesses performed.

The five-year record makes the point vividly. Berkshire reported net income of about $42.5 billion in 2020, a huge $89.9 billion in 2021, a $22.8 billion loss in 2022 as stock markets fell, then $96.2 billion in 2023 and $89.0 billion in 2024 as they rose, before about $67.0 billion in 2025. The businesses did not lurch that way; the stock portfolio did. This is why Berkshire itself directs investors to operating earnings, which exclude those paper swings, and why the revenue-and-cash view matters more than headline net income here.

Figures are from Berkshire's fiscal 2025 10-K and Tenet data as of July 2026. Past results do not predict future returns; the record shows what the business has done, not what it will do. The balance sheet behind those numbers is a genuine fortress: at the end of fiscal 2025 Berkshire held roughly $373 billion in cash and short-term investments against modest borrowings, a war chest that gives it the power to buy whole companies or step in during a crisis when others cannot. That optionality is part of the value.

Valuation discipline: valuing a mosaic

Business quality is only half of an investment decision. The other half is price, and Berkshire is unusually hard to price because it is a mosaic. In July 2026 the Class B shares traded around $507, giving the company a large market value and a trailing price-to-earnings multiple near 16, though that ratio is nearly meaningless given how noisy the earnings are.

The honest way to value Berkshire is to add the parts: the operating businesses at a fair multiple of their earnings, the stock portfolio at market, and the cash at face value, then compare that sum to the market price. That is harder than reading a single P/E, but it is the only method that respects the structure. A related shortcut some investors use is to watch what Berkshire itself does: the company buys back its own shares only when management judges them to be trading below a conservative estimate of intrinsic value, so heavy buyback activity is a quiet signal that the people with the best information think the stock is cheap. That is a useful cross-check, though never a substitute for doing the sum yourself. The discipline is to admit the imprecision rather than hide behind a clean-looking multiple, the same honesty required for the complex holding company examined in how we analyze Brookfield Corporation. One large piece of Berkshire's portfolio, its stake in Apple, is a business we look at in how we analyze Apple.

Where to go from here

Berkshire Hathaway is a clear example of the Tenet lens meeting an unusual structure: a conglomerate funded by low-cost insurance float, whose book value understates its worth and whose reported earnings are pure noise from quarter to quarter, backed by a fortress balance sheet. To see the same figures laid out as a 10-K, read reading Berkshire Hathaway's annual report, then open the live Berkshire report on Tenet and look through the headline earnings to the businesses underneath for yourself. The habit of reading past a noisy number to the value beneath it is the whole point of the Tenet lens.