The Progressive Corporation

Progressive sells insurance, overwhelmingly for cars. A driver pays a premium for six or twelve months of cover; Progressive pays for the crashes, thefts and hailstorms that follow, plus the cost of finding, selling to and serving that driver. If the premiums exceed the claims and costs, it makes an underwriting profit. Meanwhile it holds the money between receiving premiums and paying claims — about $99bn of investments at the end of September 2026 — and earns interest on it. A ten-year-old's version: lots of people each pay a little so the few who crash get paid a lot, and Progressive is very good at guessing who will crash.

Its edge is that guess. Progressive prices each driver on more variables, and more accurately, than almost any rival, and since 2008 it has collected billions of miles of real driving data through its Snapshot programme. That lets it charge good drivers less and bad drivers more, so it wins the profitable customers and leaves the unprofitable ones to competitors. The one number that summarises how well this works is the combined ratio: claims plus costs as a share of premiums earned. Progressive's goal, unchanged for decades, is to grow "as fast as we can" while keeping it at or below 96 — a 4% underwriting profit. In 2025 it was 87.4, and for the first nine months of 2026, 87.5.

That is a margin of about 12 cents in every premium dollar, in an industry that over five years made roughly nothing on its underwriting. It produced earnings of $19.23 a share in 2025 and an $8bn dividend in January 2026. At $229.25 the shares are where they were at the end of 2024, 21% below the March 2025 record, because the market has noticed two things: car-insurance prices are now falling, and Progressive's growth has halved.

Two lessons for an owner. The first is that Progressive's advantage is cultural and has compounded for sixty years: it was built by a man who insisted on pricing every risk and reporting the results every month, and the company still publishes its combined ratio monthly — something almost no other listed insurer does. Buffett said of Lewis in 2022 that "he knew insurance backwards and forwards". The second is that the business is cyclical even when the company is not: earnings per share were $9.66 in 2020, $1.18 in 2022 and $19.23 in 2025. Progressive never lost money on underwriting in those years, but the profits of a car insurer breathe with the price cycle, and an owner has to value the breath, not the peak.

The underwriting profit. In 2025 Progressive earned $81.7bn of premiums and paid out about 67.6 cents of each dollar in claims for that year's accidents, plus adjustments to earlier years' estimates, plus about 21.5 cents of expenses — including a record $5.1bn of advertising and $1.2bn returned to Florida drivers under a state law that caps insurers' profits there. The result was a 12.6% underwriting margin and $10.2bn of pre-tax underwriting profit, against $3.0bn in 2023.

The float. Because premiums arrive before claims are paid, Progressive holds money that belongs, in effect, to its policyholders: about $47bn of loss reserves and $28bn of unearned premiums in September 2026 — some $51bn net of receivables, by our arithmetic. When the underwriting is profitable, that money costs nothing; Progressive is paid to hold it. Invested mostly in short-dated, high-quality bonds (duration 3.6 years, average rating AA−), it produced $3.6bn of investment income in 2025 at a 4.1% book yield. This is the same economics Buffett has described for decades at GEICO and Berkshire's other insurers.

The limit. Insurance regulators require capital in proportion to premiums; Progressive aims for about $1 of statutory surplus for every $3 of premiums. Each year of growth therefore needs extra capital, which comes from retained earnings. That is why the cash flow statement flatters an insurer: operating cash flow of $17.5bn in 2025 includes the growth of the float, which is not the owners' to spend, and the owners' true share — owner earnings — is net income less the capital the business must keep to grow (Part XI).

Buffett's description of insurance is not flattering: a commodity product, sold on price, in an industry that habitually under-prices. A durable advantage in such a business can come from only two places — lower costs or better information — and GEICO, which Buffett first bought in 1951, has always competed on the first. Progressive has built the second. Its edge is in the loss ratio: it knows which drivers are profitable, prices them correctly and lets the rest go. Ajit Jain, Berkshire's insurance chief, put the gap at about five points in 2019. Five years of underwriting margins of 8% against an industry at zero are the measured result.

The weakness of the moat is the customer. Nobody is locked in; drivers shop every renewal, and when rivals cut prices — as State Farm, GEICO and Allstate are doing in 2026 — Progressive must choose between matching them and accepting slower growth. It has chosen growth at a profit, and growth has slowed. The data advantage is also being copied: GEICO is rebuilding its telematics and pricing, and its combined ratio beat Progressive's in 2024–25, though by raising prices and shedding customers. We score the moat 7: a genuine and measurable information advantage, in a business with no switching costs.

How the cycle works. Car insurance moves in waves. When repair and medical costs jump faster than premiums — as in 2021–22, when used-car prices soared after the pandemic — insurers lose money, raise prices sharply and stop chasing customers. Once prices have caught up, profits are excellent, insurers compete to grow again, prices fall, and margins erode until the next shock. Progressive's 2025 results were the top of such a wave: an 87.4 combined ratio, 12% premium growth and earnings of $19.23 a share.

The turn. The signs of a soft market arrived in 2026. The consumer price index for car insurance turned negative in May and was 5.0% lower than a year earlier in September. Progressive cut its own rates by less than 1% in 2025 and again in the second quarter; State Farm, GEICO, Allstate and USAA all cut rates in Florida. Rivals advertised more and kept their customers longer. Progressive's premium growth slowed from 13% to 5%, and its policies in force from +12% to +6%; in September the number of policies fell from August for the first time. The underlying cost of claims for the current year rose about two points, to 69.6% of premiums, partly from higher severity and tariffs on car parts. On the June results the stock fell 9.4% in a day.

Our reading: a soft market, handled from strength — but earnings will fall. Progressive enters this phase with the widest margin in the industry: it can cut prices, keep advertising and still earn far more than its 96 target. That is why its management talks about growth, not profit. But the arithmetic is unforgiving. If the combined ratio drifts back to the low 90s — roughly its five-year average — on premiums growing 5–6% a year, earnings per share fall from about $20 to about $15–16, which is what analysts' 2027 estimates already suggest. The 2026 results are also helped by $1.5bn of releases from earlier years' reserves, which will not recur at that pace. The stock at $229 is 11.5 times today's earnings and about 15 times a normal year's. The question for an owner is not whether Progressive will remain the best car insurer in America — we think it will — but how much of the next two years' decline in profits is already in the price. Our answer: most of it, not all.

Progressive is one of the very few listed companies that lets its owners check the thesis every month, usually in the third week. The variable dividend is usually declared in December and paid in January. The full third-quarter 10-Q is due on 2 November 2026.

The duel with GEICO. For Berkshire shareholders, Progressive is the yardstick. At Berkshire's 2019 meeting Buffett said he would "bet significant money that GEICO increases its market share in the next five years", and wondered aloud "which one of us passes State Farm first" (as transcribed by Insurance Journal). GEICO's share fell from about 13.8% in 2022 to about 11.6% in 2024, according to Berkshire's own filings, and it is Progressive that has drawn level with State Farm. GEICO answered in 2023–25 by raising prices, cutting costs and shedding customers, and its margins were spectacular — Buffett called its 2024 improvement "spectacular" in his last letter — but Greg Abel warned in February 2026 that "competitors' rate reductions may extend that pressure into 2026". In the first half of 2026 GEICO's combined ratio (89.3) was worse than Progressive's (86.9) and its premiums barely grew. The two have different weapons — GEICO the lowest costs, Progressive the best pricing — and in a soft market the second has proved the better one so far.

Integrity: high. Progressive reports monthly, explains its reserve changes, discloses its litigation in detail and has met its stated profit rule every year we checked. Capital allocation: excellent. It keeps the capital the business needs, pays out the rest in a lump when it has it, and buys back shares only when it judges them cheap — Buffett's 2011 test, "what is smart at one price is dumb at another", applied in practice. Owner mentality: good, with a caveat. The Gainshare bonus aligns nearly every employee with the owners' goals, but executives' wealth comes from stock grants, and none bought shares with their own money this year. We score management and capital 8.

Reading an insurer's accounts. Three things make Progressive's numbers easy to misread. First, the earnings include changes in the estimates of past years' claims: in 2023 those added $1.1bn of costs; in the first nine months of 2026 they removed $1.5bn. Underlying profitability is better judged on the current year's losses, which rose two points in 2026. Second, net income includes gains and losses on the investment portfolio. Third — and most important for valuation — the cash flow statement is not a measure of owners' cash. Progressive's operating cash flow was $17.5bn in 2025, half as much again as its $11.3bn of net income, because a growing insurer collects premiums before paying claims. That money is owed to policyholders; it is the float, and it must stay invested in safe bonds. It is valuable — Buffett built Berkshire on it — but it is not distributable.

The verdict on the dividend: generous in good years, absent in bad ones — and that is the point. Progressive is not an income stock in the usual sense: anyone buying it for a steady, growing payout will be disappointed. What it offers instead is a dividend that returns the capital the business cannot use, as Buffett argues a company should. In a soft market, with growth slowing, that may be a lot: the company will need less new capital to support premiums growing 5% than it did at 12%. Our expectation is a smaller variable dividend for 2026 than the record $13.50, but a meaningful one, plus continued buybacks below fair value. Compare Chubb, whose regular dividend grows every year but takes a small share of earnings.

What is it worth? We discounted owner earnings of about $12 a share at 10% for ten years. If they grow 7% a year for five years and 6% for the next five — premiums growing mid-single digits, the share count falling slightly — and the business is then valued at 16 times owner earnings, Progressive is worth about $240. If the soft market is long and deep — owner earnings of $9, growing only 3–4% a year, ending at 12 times — about $125. If Progressive keeps taking share at a low-90s combined ratio and owner earnings of $14.50 grow 9% and then 7%, ending at 18 times, about $350.

At $229.25 the price sits about 5% below our central value. That is fair, not cheap. Our own Buffett screen put Progressive on its Buy list this morning with a value of $449; that figure treats the growth of the float as cash the owners could take out, which no insurer can. Our playbook does not demand a deep discount for an exceptional business at a fair price, and Progressive is the best operator in its industry — but its earnings are at a cyclical peak and set to fall, and we score the moat 7, not 8 or more. For that we want a margin of safety. Our advice: hold if you own it; buy below about $195 — about 13 times normal earnings, close to the May 2026 low of $190. A soft market that frightens investors will probably offer it.