Berkshire Hathaway is a holding company, which means it owns a large collection of completely different businesses rather than running just one. The biggest pieces are insurance (GEICO, General Re, and others), the BNSF freight railroad, and Berkshire Hathaway Energy, which runs power utilities and natural gas pipelines. It also owns manufacturing, retail, and service companies, plus a giant portfolio of stocks in companies like Apple and Coca-Cola. The insurance businesses are the engine: they collect premiums from customers upfront, then invest that money while waiting to pay out claims. That pool of investable money, called float, stood at roughly $176 billion at the end of 2025. The diagram below traces where the money flows through this machine.

The float has grown steadily, from roughly $138 billion at the end of 2020 to $176 billion at the end of 2025. That expansion matters because a larger float means more money available to invest, and the insurance businesses generated pre-tax underwriting profits in each of the past three years. In other words, Berkshire was paid to hold other people's money.

Five years of financial data tell a story of a business that generates enormous cash but whose reported profits swing wildly. Operating cash flow ranged from $30.6 billion in 2024 to $49.2 billion in 2023. Reported revenue held roughly flat, between $23.2 billion and $25.8 billion across the period, because the XBRL revenue figure captures only a slice of the total business. Net earnings attributed to shareholders dropped from $96.2 billion in 2023 to $67.0 billion in 2025, but most of that swing came from unrealized gains and losses on stocks, not from the operating businesses themselves. The company's own management says those investment gains and losses are essentially meaningless for understanding how the underlying business is actually doing.

The operating businesses underneath all that noise have been more stable. GEICO earned $6.8 billion before tax in 2025, down from $7.8 billion in 2024 but up sharply from $3.6 billion in 2023, when it was struggling with rising claims costs. BNSF earned $5.5 billion after tax in 2025, up 8.8% from 2024, mostly by cutting costs rather than growing revenue, since railroad revenues were almost flat across all three years. Berkshire Hathaway Energy recovered from wildfire-related losses in prior years and earned $4.0 billion after tax in 2025. The balance sheet shows a net cash position, meaning the company holds more cash than it owes in debt, and that cushion has grown to $51.9 billion at the end of 2025.

The risk picture at Berkshire is dominated by the sheer scale of what could go wrong in insurance. The company carries $151.8 billion in estimated unpaid losses across its insurance businesses. Even a small percentage error in those estimates can move earnings significantly. Berkshire intentionally takes on more catastrophe risk from single events than any other insurer in the world. It tries to cap losses from one catastrophe at $15 billion, but the 10-K acknowledges that unexpected events could exceed that limit. The 2025 Southern California wildfires cost the insurance businesses roughly $850 million after tax, a real but manageable hit. A larger event, or a series of events in the same year, would be a different story.

BNSF faces a specific structural threat from the coal business. Coal volumes at BNSF dropped 17.9% in 2024 when natural gas prices were low, then recovered slightly in 2025 when natural gas became more expensive again. That volatility shows how dependent coal revenue is on conditions BNSF cannot control. Separately, the utility business at Berkshire Hathaway Energy is exposed to regulators who set the rates customers pay. If those regulators do not allow BHE to recover rising costs, including costs tied to new environmental rules, profits at BHE decline. PacifiCorp already dealt with large wildfire loss charges in recent years.

The leadership transition adds a layer of uncertainty that is genuinely new. For decades, capital allocation at Berkshire meant one mind making the biggest decisions. The 10-K now identifies Greg Abel and two other executives as the critical people, and it states plainly that losing any of them could materially harm operations. Abel's early deals, including the Alphabet commitment and the Taylor Morrison acquisition, suggest a willingness to deploy capital in new directions. Whether his judgment over time matches the standard that built this institution is simply unknown.