Hartford: A Fine Business Waiting for a Fault-Tolerant Price
What the business is
There is a quiet mathematics to insuring a dry cleaner in Ohio or a machine shop in Wisconsin. The premium is small, the claim is rare, and the data needed to price either correctly is the kind of thing you can only accumulate by doing it badly for a few decades first. That is Hartford’s business: it is the largest writer of small-business commercial insurance in America, selling workers’ compensation, commercial packages, and general liability to more than a million small enterprises, alongside personal auto and home policies through its AARP alliance and group life and disability benefits.
The earnings engine is the classic insurance double — underwriting profit from keeping the combined ratio below 100, and investment income from the float that premiums create before claims are paid. In 2025, both major segments ran profitable underwriting: Business Insurance at an 88.8 combined ratio, Personal Insurance at 88.7. The moat is cost advantage built on scale and data — a digital underwriting platform rated the top small-business digital insurer, with decades of loss data — plus intangible assets: a 200-year brand, an exclusive AARP channel, and the number-two position in U.S. commercial multi-peril. The moat is slowly widening, though not without pressure; the CEO has acknowledged intensifying competition in the SME market, and commercial rates have entered a softening phase in 2026. That is cyclical pressure, not a crack in the wall.
What the numbers say
The screening snapshot from late August showed owner earnings of $3.854 billion, but that figure reflects a hard-market peak. The desk’s normalized owner earnings estimate is $3.3–3.5 billion, adjusting for rate softening and the disappearance of favorable prior-year development ($90 million down to $5 million). On a price-adjusted basis, the arithmetic is straightforward:
- Market cap (price-adjusted, 2026-09-04 close): $37.60 billion
- Owner-earnings yield (price-adjusted): 10.25%
- Share price (2026-09-04 close): $138.37
- 10-year Treasury: 4.78%
The 10.25% owner-earnings yield against a 4.78% Treasury is a spread of 5.47 percentage points — superficially attractive. But the intrinsic value range, based on normalized earnings of $3.4 billion at 10–12x, is $34–41 billion. The current market cap sits at a 1,039% premium to the low end of that range. The yield is real, but the numerator is peak earnings; normalize it and the picture changes materially. Management has been disciplined — 14 consecutive years of dividend increases, buybacks averaging roughly 10x owner earnings, and a rejection of Chubb’s $65-per-share offer in 2021 that hindsight has vindicated — but the price today leaves little room for error.
What Wall Street believes vs what I see
The headlines are friendly: “Still A Buy As Revenue Growth Outperforms,” “Strong Cash Flows,” “Impressing With Improving Margins.” The consensus mood is that Hartford is a quality compounder with momentum, and on the qualitative dimensions, I agree. The moat is real, the management is honest, and the five-year average ROE of 16.4% with a standard deviation of 3.3% is unusually consistent for a cyclical industry.
What I see is a yield that depends on a peak-cycle numerator. The 10.25% owner-earnings yield is computed on $3.854 billion of owner earnings; the normalized $3.4 billion center of the range produces something closer to 9% — still above the 4.78% Treasury, but the intrinsic value range of $34–41 billion brackets the current $37.6 billion market cap almost exactly. This is a fair price for an excellent business, not a discount. The market is paying for the cycle to stay favorable; I am watching for the cycle to normalize. The warning light is already flickering: adverse prior-year development in general liability and commercial auto for two consecutive quarters, offset so far by favorable workers’ comp development, is precisely the early pattern that reserve inadequacy would show.
Verdict
Verdict: watch.
The owner-earnings yield of 10.25% exceeds the 4.78% Treasury by a wide margin, which keeps this on the watchlist rather than in the pass bin. But the market cap is more than ten times the low end of the intrinsic value range, and that range already assumes normalized — not peak — earnings. The business clears every qualitative gate: moat, management, honesty. The price does not clear the safety-margin gate. What would change the verdict is a pullback toward the $29 billion market cap level (roughly $100–105 per share), where a 30% discount to normalized intrinsic value would give the reserve clock time to tick without punishing the buyer.
This analysis is AI-generated, educational, and not investment advice. Figures may contain errors or be delayed. Disclaimer