ALLSTATE CORP (ALL) stock intrinsic value fair value estimate & owner earnings analysis

Allstate is one of the largest personal auto and homeowners insurers in the United States, with a smaller Canadian business. It makes money two ways: pricing policies so that premiums exceed claims and expenses, and investing the roughly $83 billion of reserves and capital it holds until claims are paid. Earned premiums were $61.4 billion in 2025, up from $33.6 billion in 2016. How we measure earnings. We start from operating cash flow and remove the swings caused by reserves and unearned premiums building or releasing, stock-based pay, and the capital spending needed to keep the business running. We also deduct the roughly $117 million a year of preferred dividends, because the $2.05 billion of preferred stock ranks ahead of the common. Because borrowing, reserves and float are the raw material of an insurer, we value these earnings after interest and do not separately subtract the company's $7.5 billion of debt. On that basis Allstate earned about $33.08 a share in 2025 and $18.43 in 2024, but only $0.22 in 2022 and $1.41 in 2023, when repair and medical costs outran prices across the industry. The ten-year record works out to a median of about 9.7 cents of owner earnings per dollar of earned premium, and we use that full cycle, crisis years included, rather than the recent peak, as the measure of normal earning power. First-half 2026 results were stronger again (net income $5.7 billion versus $2.7 billion a year earlier), which supports a strong near-term year but does not change the long-run level on its own. Growth path. Owner earnings rose about 79% in 2025. Our base case assumes the first year is lower, not higher: about $8.0 billion, or roughly $31 a share, as competitors' catch-up pricing narrows the current margin. Earned premiums have risen every year since 2016, including through the 2022-2023 crisis, at about 7% a year; much of that was catch-up rate increases rather than more customers, so we credit only 2% a year, roughly the pace at which a renewing book's premiums keep up with repair and replacement costs. We give no credit for gaining market share, new products or permanently higher margins. Base case. Premiums grow 2% a year; the margin steps down from about 13% next year to 11.5% and then settles at 9.5% of premiums, slightly below the ten-year median. Annual earnings to common shareholders run from about $8.0 billion to $7.0 billion by year ten, and auto and home policies renew continuously, so we carry the year-ten level forward without further growth. That values the business at about $115.0 billion, or $443.91 a share. Bear case. Within two years the industry repeats a 2022-style episode, where claims costs rise faster than approved rates, and it coincides with a heavy catastrophe year. For scale, Allstate's catastrophe losses were $5.6 billion in 2023 and $5.0 billion in each of 2024 and 2025; in 2025, the year of the California wildfires, gross catastrophe losses were $6.2 billion before reinsurance. In this case the company loses about $4.1 billion in year two, roughly breaks even in year three, and recovers to only 8% of premiums, below its long-run median, with premiums growing 1% a year thereafter. That values the business at about $89.6 billion, or $345.83 a share. Tangible equity is only about $108 a share, so this case rests on the renewal franchise surviving a bad stretch: through 2022-2023 the diluted share count fell rather than rose, and at year-end 2025 debt was about 20% of debt plus equity against a 37.5% debt-to-capital limit in its credit facility. Bull case. Margins hold closer to today's levels, about 14% next year and 11.3% of premiums over the long run, which is roughly the 2024 level, with premiums growing 3% a year. That values the business at about $146.7 billion, or $566.20 a share. What could make this wrong. The largest uncertainty is the normal margin. If the long-run margin is 7.5% rather than 9.5%, the base case falls to about $360 a share; with no premium growth as well, to about $310. Insurance also carries risks no case models well: a major earthquake in California or a hurricane striking a large eastern or gulf coast city, regulators in states such as California restricting rate increases or cancellations, and the possibility that self-driving and safety technology shrinks the auto insurance pool over the long run. Margin-of-safety price. We are reasonably confident in the long-run earning power because it is drawn from a full cycle that includes the 2022-2023 loss years, but the gap between a good year ($33 a share) and a bad one (near zero) is wide and can arrive with little warning. The structure would survive a severe year: debt is modest, reserves are backed by a largely fixed-income portfolio, and the business is spread across individual households rather than a few large customers. The dividend is currently $1.08 a quarter, about $4.32 a year, paid while waiting. We set the margin-of-safety price at $190. Assuming a five-year holding period, from that price the base case implies roughly 20% a year including dividends, the bear case roughly 18%, and the bull case roughly 27%; if the bear case's bad years arrive early and the shares recover to that case's value in three years, it implies about 29% a year. If the long-run margin proves to be 7.5%, $190 implies roughly 15% a year. The previous estimates were built on a different basis and are not comparable to these figures.