Allstate Looks Cheap on the Surface. Strip Out Catastrophe Luck and the Story Changes

The headline number for Allstate in the second quarter of 2026 was a recorded combined ratio of 86.6 — an improvement of 4.5 points from the prior year. Total revenues hit $18.6 billion, up 11.8%. Adjusted EPS came in at $8.99, well above the consensus estimate of $6.06. The stock has climbed roughly 29% year-to-date and trades at a trailing P/E of 5.1, the lowest of any major US property-casualty carrier.

The number that matters more is the underlying combined ratio: 79.4, virtually flat versus 79.5 a year ago. The 4.5-point improvement in the headline ratio came almost entirely from two things that don't belong in the operating story — lower catastrophe losses and favorable prior-year reserve development. Strip those out, and the business is not materially better than it was twelve months ago.

That distinction changes how you read the valuation.

The catastrophe tailwind

Allstate's catastrophe losses in Q2 2026 totaled $1.7 billion, down 13.5% from $1.99 billion a year earlier. The first quarter of 2026 saw catastrophe losses of roughly $1.2 billion, a 43% drop. Pre-tax catastrophe losses for April and May combined were $1.16 billion, well below the heavy loss months that defined 2024 and 2025. May 2026 alone was just $289 million — a 63% decline from May 2025.

Catastrophe frequency is cyclical. The last two years were brutal; this one is quiet so far. That's a real tailwind for the recorded combined ratio, but it's not operating improvement. When you're evaluating whether earnings power is durable, you need to know which part of the result was earned and which part was luck.

The underlying trend is not improving

Auto insurance — Allstate's core book — delivered an underlying combined ratio of 87.6% in Q2 2026, down only 0.2 points from 87.8% a year ago. In Q1 2026, the underlying auto ratio excluding reserve benefits and catastrophes was 89.5%, up 1.7 points year-over-year. The 5-year underlying average for auto sits at 95.2%, so Allstate is still running below its own long-term norm. But the trajectory isn't getting better.

Homeowners is where the underlying trend is moving in the wrong direction. The underlying combined ratio rose 2.9 points to 61.5%, reflecting higher loss costs. Average gross written premiums for Allstate-brand homeowners increased 5.8%, driven by rate hikes and rising replacement costs — and even that pricing pressure wasn't enough to offset the higher loss severity. The segment swung from a $76 million underwriting loss in Q2 2025 to a $226 million profit in Q2 2026, but that turnaround was built on lower catastrophe losses, not a fundamentally improving loss experience.

Vehicle repair and maintenance costs rose 6.1% year-over-year in May 2026. Labor shortages and parts scarcity continue to push claim costs upward. The Consumer Price Index rose 4.2% in May from a year earlier, marking the highest inflation reading in three years. Allstate has largely completed its recent repricing cycle and is now rebuilding policy counts after prioritizing profitability. The next test is whether pricing stays ahead of claim costs as growth comes back on the table.

The P/E of 5.1 is an accounting illusion

Allstate's trailing P/E of 5.14 looks absurdly cheap next to Progressive at 10.66, Travelers at 9.68, Chubb at 12.08, and Hartford at 8.93. The reason it's so low is that Q2 2025 posted a net loss of $1.6 billion on record catastrophe losses, dragging down the trailing twelve-month earnings denominator. The TTM EPS isn't representative of the current quarter, which posted $12.51 of diluted EPS including investment gains.

The forward P/E of 12.8x strips out the distortion — and even at 12.8x, Allstate is the cheapest in the group. But 12.8x is not a margin-of-safety number. It's what the market would pay if this year's quietly low catastrophe environment and current return on equity (44.2% adjusted ROE on a trailing basis) are treated as the new normal.

Investment income is real, but not a moat

Q2 2026 net investment income rose $255 million year-over-year to $1.0 billion, helped by an asset portfolio of $87.8 billion benefiting from higher fixed-income yields. Net gains on investments swung from a $144 million loss to a $1.1 billion gain, driven by mark-to-market increases in public equities. That's a genuine contributor to bottom-line earnings.

But investment income is a function of asset balances and interest rates, not underwriting skill. When rates normalize — which they eventually will — that contribution shrinks. The market can't price Allstate as if a 2.6% quarterly investment return is a permanent structural advantage.

What the factor stack says

Valuation (Grade: B+). Forward P/E of 12.8x is the cheapest among the five major carriers, and the book value multiple of 2.0x is below Travelers at 2.4x and Progressive at 3.6x. But the trailing P/E is artificially compressed, and the forward multiple assumes catastrophe luck holds. Relative value is real; it's just not as extreme as the headline multiple suggests.

Growth (Grade: B). Auto earned premiums grew 1.2%, homeowners grew 11.4%, and total revenue is up 11.8% year-over-year. Policy counts are expanding — auto up 2.8%, homeowners up 2.9%, with new business surging 8.8% in auto and 16.4% in homeowners. Volume growth is coming back after the company pulled back to protect margins. The growth story is credible but not exceptional.

Profitability (Grade: C+). Adjusted ROE of 44.2% is outstanding, but it's buoyed by a quiet catastrophe period and strong investment returns. The underlying combined ratio is flat in auto and deteriorating in homeowners. The profitability number is high today, but the trend line underneath isn't improving.

Safety (Grade: A). Net debt of $7.5 billion against $33.7 billion in equity, with a debt-to-equity ratio of 22.3%. Operating cash flow of $12.5 billion on a trailing basis. Capital returns of $3.5 billion to shareholders over the last year, with buybacks increased to $1.0 billion in Q2 alone. The balance sheet is fortress-grade — this is the one factor that doesn't have a caveat.

Momentum (Grade: A). The stock is up 29% year-to-date, up 29.3% over the last 120 days, and trading just 3.2% below its 52-week high of $277.22. Earnings have beaten consensus by a wide margin in both Q1 and Q2. The market has voted with its feet — but the price action has moved ahead of the underlying operating evidence.

AInvest's aggregate signal labels Allstate a Hold

That aligns with what the factor stack is telling you: solid balance sheet and relative value, but the operating momentum isn't carrying the premium that the price action implies.

What to watch

The sustainability question isn't whether Allstate can keep posting attractive quarterly numbers. In a low-catastrophe environment with rising interest income, it can. The question is whether the underlying combined ratios — flat in auto, worsening in homeowners — can improve when catastrophe losses return to a more normal level. If Q3 and Q4 bring heavier cat activity (September through November is the peak hurricane window), the recorded ratio will test whether the underlying book is strong enough to absorb it.

Allstate is not a broken business. It's a well-capitalized carrier with real pricing power and a disciplined management team that pulled the right levers over the past two years. But the operating improvement the market appears to be pricing in doesn't exist in the underlying ratios. What exists is a one-time tailwind from catastrophe luck and an artificially compressed trailing P/E. When both normalize — and they will — the stock will need actual underwriting momentum to justify where it trades.

The portfolio role for Allstate right now is defensive value: a dividend-paying (1.6% yield, 24 consecutive years) name that sits below peers on a forward basis and can be held for mean reversion on the multiple. It doesn't belong in the growth sleeve. The trigger to upgrade it there is a decline in the underlying combined ratio, not another quiet catastrophe quarter.

Volatility is usually a signal that the market can't figure out what comes next. In Allstate's case, the signal is already there. The factor stack is telling you which parts of the story are durable and which are weather.