Hartford: A Genuine Moat Priced at Its Own Estimate

HIG · Hartford (The) · Published Aug 9, 2026

The Question Worth Asking

The central question in any deep moat review is a simple one: if the business is this good, why is it not yet owned at this price?

For The Hartford, the answer starts with a flattering number. On an owner-earnings basis derived from the screening’s primary calculation — using the August 7 Friday close — the return on earnings stands at 9.81%, the highest among the 36 names under observation. That figure leads the board. It is also, on inspection, a numerator story.

The $3.854B numerator is built from 2025 earnings — core EPS up 30%, ROE of 19.4% — the crest of a hard market. Normalize for softening rates and the fading of favorable reserve development, and the numerator falls to $3.3–3.5B, bringing the yield down to roughly 8.7%. Apply a 12x multiple with modest growth to that normalized figure and intrinsic value lands near $40.8B — against a current market cap of $39.28B. That is a discount of 3.7%. The price of an excellent company, in other words. Not a price that absorbs error.

And error tolerance is precisely what the current moment demands.

Underwriting Discipline: Five Years of Combined Ratios

The evidence of pricing power sits in the underlying combined ratios of the Business Insurance (formerly Commercial Lines) segment:

A combined ratio pinned in the high 80s across the cycle — while premiums grow — is the quantitative signature of pricing power: the company kept its rates without surrendering volume. The five-year average ROE of 16.3% with a standard deviation of just 3.3% is the shadow of the same fact: unusual consistency for an insurer.

Two caveats temper the picture. Management itself has acknowledged softening in property insurance and intensifying competition in the small-business space during 2026. And favorable prior-year development shrank from $146M to $52M quarter over quarter — a signal, in our reading, that the cycle has passed its peak.

What the Moat Actually Is

The moat’s source is a scale-plus-data cost advantage in small-business commercial insurance. Hartford holds roughly 11% market share, ranking among the top three, with more than 1.5 million policyholders and around $5 billion in annual premium. Small-business policies carry small premiums per contract, so unit costs of underwriting and claims settlement decide the game — and decades of loss data plus a digital underwriting platform have built that cost structure. Secondary advantages include a two-century brand, the exclusive AARP channel in personal lines, and a dense agency network.

The honest comparison, though, is what the moat is not. Hartford lacks Progressive’s consumer brand pull in direct personal auto, and has nothing of Chubb’s high-net-worth or global specialty standing. This is a “deep advantage in a narrow market” moat — the American small-business market. Outside that lane, in mid-to-large commercial and personal lines, management’s own words concede intensifying competition: a competent, ordinary business.

On the AI question, the current call is mildly positive. Small-business underwriting is a prime candidate for data-driven automated underwriting, and Hartford — with proprietary loss data and distribution already in hand — fits the “AI as a feature benefiting the strong” pattern. The counterargument has substance: AI-native upstarts have demonstrated quotes in minutes in small-business workers’ comp, and if third-party data access erodes the value of accumulated data, the relative edge narrows. But the predictability of demand itself is undisturbed — small businesses will still be required to carry workers’ comp a decade from now — so the moat’s direction is graded “maintained to modestly widening,” one notch more conservative than before.

The Reserve Question: One-Off or Discipline Breach?

The second quarter of 2026 brought adverse development of $46M in general liability (rising large-loss frequency in excess and umbrella, accident years 2017–19 and 2022–23) and $26M in commercial auto (severity upgrades for 2023–24). Workers’ comp favorable development of $110M in the first half offset these, leaving net prior-year development favorable at $52M.

The verdict: one-off in magnitude, a warning in pattern. The dollar amounts are small against written premiums, and net development remains favorable. But a tripwire set out in mid-July — two consecutive quarters of adverse development in general liability and commercial auto — has effectively been tripped (a $70M legacy liability charge in Q1, then $46M in GL in Q2). Social inflation — jury verdicts, rising attorney involvement — is an industry-wide pressure rather than proof of company-specific indiscipline. Still, the threshold matters: if general liability runs adverse for a third quarter, normalized owner earnings get cut to $3.1–3.3B and the buy line gets recalculated.

Whether consistent ROE is genuine underwriting profit or partly reserve release is the single most dangerous argument against this thesis. Right now, it cannot be distinguished — and when something cannot be distinguished, the only acceptable price is one that assumes it will.

Capital Allocation

The record here is reassuring. A $4.2B buyback authorization was approved in Q2, running through December 2028 — up 27% from the prior $3.3B — with quarterly repurchases raised to $475M; total Q2 capital return was $615M ($450M of buybacks plus $165M in dividends). The sale of Hartford Funds to Wellington, expected to close in Q1 2027, is a divestiture of non-core asset management, with part of the proceeds funding the larger repurchase — empire-building run in reverse, which is the right direction. Dividends have been raised for more than 14 consecutive years, and the 2021 refusal of Chubb’s $65-per-share offer looks well judged against today’s $143 price. Repurchases have been executed at roughly 10x owner earnings, well below the 12x alarm threshold.

Valuation

At the August 7 close, $143.28 across 274.13M shares gives a market cap of $39.28B. That is 9.81% on the screening’s primary owner-earnings figure, or 8.40–8.91% (center 8.66%) normalized. Against a 10-year Treasury yield of 4.660%, even the normalized yield clears the 6% hurdle by about 4 percentage points. Intrinsic value ranges from $34B at a no-growth 10x to $40.8B at a low-growth 12x; the current price sits at a 3.7% discount to the center.

The buy line, set at 30% below the 12x normalized valuation, is roughly $104 per share — a market cap of $28.6B. The current price sits 37.5% above it.

The Gate Summary and the Verdict

The verdict remains watch. This is a genuinely well-built business wearing a price that merely agrees with itself. Three triggers would move the call: the stock approaching $104; the direction of general liability prior-year development in Q3; or simultaneous deterioration in small-business growth and underlying combined ratios.

Until one of those arrives, patience is the position. The moat is real. So is the bill for certainty.

This analysis is AI-generated, educational, and not investment advice. Figures may contain errors or be delayed. Disclaimer