Cincinnati Financial Q3 2026: EPS fell 58% on missing stock gains; storms lifted combined ratio to 90.9%

Cincinnati Financial's third quarter of 2026 looks like a collapse on the headline: net income fell 59% to $459 million ($2.96 per diluted share) from $1.122 billion ($7.11). Almost all of that drop is a comparison effect. In Q3 2025 the stocks the company owns rose sharply in value and $853 million of pretax investment gains flowed into net income; this quarter that figure was $20 million. The company's own profit measure that strips those gains out, non-GAAP operating income, was $443 million ($2.86 per share), against $449 million ($2.85) a year ago. Inside that flat number, catastrophe losses roughly tripled, and two other items made up the difference (all figures from the 10-Q filed October 26, 2026, and the same-day earnings release).

Why net income fell while the business held steady

Cincinnati Financial owns an unusually large stock portfolio for an insurer: $13.1 billion of equities, 39.7% of total investments at September 30. Under US accounting rules, changes in the market value of stocks it still holds go straight into net income, even if nothing is sold. In Q3 2025 those unrealized gains were $855 million pretax; in Q3 2026 they were $1 million. After tax, investment gains fell from $673 million ($4.26 per share) to $16 million ($0.10).

The release breaks down the $663 million fall in net income: −$657 million from lower after-tax investment gains, −$41 million from lower after-tax property-casualty underwriting profit, and +$24 million from higher after-tax investment income. Total revenues fell 18% to $3.047 billion for the same reason: the investment-gain line is counted as revenue. The recurring revenue lines grew: earned premiums rose 5% to $2.690 billion and pretax investment income rose 10% to $325 million.

Key metrics

Net written premiums are the premiums the company sold in the quarter after passing part of the risk to reinsurers; they show growth before it shows up in earned premiums. The loss ratio here is claims plus the cost of handling them, as a share of earned premiums.

Underwriting: storms came back, the underlying book improved

The combined ratio is the share of each premium dollar spent on claims plus operating expenses; under 100% means the insurer made an underwriting profit. At 90.9%, property-casualty operations earned a $240 million underwriting profit, down from $293 million.

The 10-Q is direct about the cause: the $53 million decline "included an unfavorable increase of $185 million in losses from catastrophes, mostly caused by severe weather." The largest events were storms on August 9–17 ($82 million), July 24–30 ($50 million) and June 28–July 8 ($36 million), plus $106 million from smaller 2026 events. The CEO described catastrophe losses as "moving closer to historical averages for the third quarter," which is a fair read: Q3 2025 at 3.7 points was unusually light.

Two things offset most of the storm damage:

• Reserve releases. Claims from earlier years cost less than the company had set aside, so it released $94 million of reserves (3.6 points of the combined ratio), up from $22 million (0.9 points). In personal lines the swing was largest: $38 million favorable this year vs $14 million unfavorable last year.
• A better underlying book. The ratio of current-year claims before catastrophes fell to 54.0% from 55.4%. That is the cleanest gauge of whether pricing is keeping up with claim costs, and it improved in every segment.

Personal lines (home and auto) took the brunt of the storms: catastrophes were 19.2 points of premium. Excluding them, the segment's current-year combined ratio fell to 74.7% from 78.5%, helped by renewal rate increases near the low end of the high-single digits. The cost is growth: new business premiums fell 34% to $76 million, and net written premiums grew only 1%.

Commercial lines barely grew (+$1 million of net written premiums), with renewal price increases in the low-single digits and new business down 2% in what the company calls "a highly competitive market." Excess & surplus lines (risks standard insurers decline) again grew fastest at 5%. The underwriting expense ratio rose 0.4 points to 29.7%, which the 10-Q attributes to "timing of recognition of certain expenses."

Investments

Pretax investment income rose 10% to $325 million: bond interest rose 11% to $252 million as the company bought $1.107 billion more bonds in the quarter, while stock dividends fell 4% to $66 million. After tax, investment income was $268 million, more than half of operating income. The pretax bond yield slipped to 5.04% from 5.10%.

Rising bond yields worked the other way on the balance sheet: the change in unrealized bond values was −$1.162 billion pretax in the quarter, which does not touch net income but does reduce book value. That is why book value per share fell $3.22 in the quarter even though the company earned $2.89 per share before investment gains.

Takeaway: The 58% EPS drop is a stock-market comparison, not an insurance problem. Underneath, a quarter with three times last year's storm losses still produced a 90.9% combined ratio, because underlying claims ran lower and reserves from earlier years proved more than adequate. The $94 million reserve release did a large share of that work, so it is the number to watch.

What the headline numbers hide

• The prior-year gain makes the comparison harsh. Q3 2025 EPS of $7.11 included $4.26 of after-tax investment gains; Q3 2026 includes $0.10. On operating EPS the two quarters are essentially equal ($2.86 vs $2.85).
• Reserve releases propped up operating income. Favorable prior-year development rose by $72 million pretax, roughly $57 million or about $0.37 per share after tax at a 21% rate. Without that increase, operating EPS would have fallen by roughly 13%. The company is still adding to reserves overall (net loss reserves up $1.260 billion or 11% since year-end, $997 million of it for claims not yet reported), so the releases do not look like a thinning cushion, but they can swing from quarter to quarter.
• Buybacks helped a little; there were almost none this quarter. Diluted shares fell 1.7% to 155.1 million, which is why operating income fell 1% while operating EPS was flat. Repurchases in the quarter were only about $1 million (7,600 shares) against $60 million a year ago; the $396 million for the first nine months was spent in the first half. Dividends paid were $142 million in the quarter.
• A lower tax rate. The effective tax rate fell to 17.0% from 20.6%, which the 10-Q attributes to the mix of underwriting income, investment gains and investment income; tax-advantaged bond and dividend income weighs more when gains are small.
• Cash conversion is fine. Nine-month operating cash flow was $2.147 billion, versus $997 million of operating income and $1.988 billion of net income that includes $1.242 billion of non-cash investment gains.

On balance the quarter is understated: reported EPS fell 58% because a $853 million prior-year investment gain did not repeat, while operating EPS was flat even after $185 million more in storm losses, with a larger reserve release as the main offset.

Did last time's read hold up?

Our Q2 2026 report said three things mattered for the second half: commercial-lines claims rising faster than its low-single-digit price increases, personal-lines growth slowing as rate increases drove customers away, and the stock market moving net income far more than insurance. The second and third played out exactly: personal-lines new business fell 34% and retention stayed in the upper-80% range, and the absence of stock gains cut net income by 59%. The first did not, at least this quarter: commercial lines' current-year loss ratio before catastrophes improved to 58.0% from 59.2%, although for nine months it is still 1.0 point worse than 2025. We also flagged hurricane season; catastrophe losses did rise to $280 million, but mostly from inland summer storms rather than hurricanes.

Outlook

The company gives no earnings guidance. Its stated long-term target is a combined ratio of 92% to 98%; at 95.8% for nine months, it is within that range with one quarter left. Management says it expects price increases and "pricing precision" to support underwriting results and that it intends to "remain prudent in reserving." The board added a 15 million-share repurchase authorization on August 21, leaving 16.8 million shares available, but used almost none of it in Q3.

In our view, the underwriting book is in better shape than a year ago (current-year combined ratio before catastrophes 83.7% vs 84.7%), but premium growth is slowing: net written premiums grew 2% in the quarter versus 4% for nine months, and new business fell 12% overall. For Q4, the figures to watch are whether favorable reserve development stays near this quarter's 3.6 points, whether personal-lines new business stabilises, and the stock market, which will again move reported EPS more than anything the insurance business does.