WRB — Q3 2026 Financial Report Analysis

W. R. Berkley, one of the largest US commercial insurers (it covers businesses rather than households), earned $446.0 million, or $1.14 per diluted share, in the third quarter of 2026. Net income fell 12.7% from $511.0 million ($1.28) a year ago, but the drop came almost entirely from the investment portfolio's market swings: a $18.3 million pre-tax loss on investments this quarter versus a $79.0 million gain a year ago. Strip those out, and management's operating income rose 3.7% to $456.5 million, or $1.17 per share, up from $1.10. The insurance business itself did slightly better: the combined ratio was 90.8% versus 90.9%, and steady growth in bond income carried the rest. This analysis is based on the October 19, 2026 earnings release; the full 10-Q, with reserve and segment commentary, has not been filed yet.

Key figures

A few terms. Net premiums written is the insurance Berkley sold in the quarter after passing part of the risk to its own reinsurers. The combined ratio is claims plus expenses as a share of premiums earned; under 100% means the insurance itself makes a profit before any investment income. It is the sum of the loss ratio (claims) and the expense ratio (commissions and overhead). The accident-year combined ratio excluding catastrophes strips out storms and any revisions to estimates for claims from earlier years, which makes it the cleanest measure of current pricing against current claims.

Where the earnings came from

Pre-tax income fell $73.8 million to $574.3 million. The moving parts:

• Investment gains and losses: down $97.3 million, from a $79.0 million gain to an $18.3 million loss. Realized gains on sales fell to $27.8 million from $69.2 million. The value of Berkley's stock holdings fell $45.4 million, against a $9.6 million rise a year ago. These swings reflect market prices and have little to do with running the business, which is why management excludes them from operating income.
• Net investment income: up $32.3 million (+9.2%). Income from the "core portfolio" (bonds, stocks, cash, real estate and loans) rose 12.0% to $369.1 million. CEO W. Robert Berkley, Jr. said income from fixed-maturity securities rose 10.3%, and that reinvestment yields "continue to meaningfully exceed our portfolio's book yield": maturing bonds are being replaced with higher-yielding ones. The two volatile pieces fell: investment funds earned $2.1 million versus $5.4 million, and the merger-arbitrage trading account earned $12.3 million versus $16.3 million.
• Underwriting: roughly flat to slightly better. Premiums earned rose 1.8% and the combined ratio improved 0.1 point. In the Insurance segment, pre-tax underwriting income rose 10.6% to $237.7 million.
• Corporate "other expenses" rose to $32.4 million from $17.6 million, and currency gains shrank to $5.1 million from $12.0 million. The release doesn't explain the expense increase.

The two segments

Insurance (89% of net premiums written) grew net premiums written 3.2% to $2,899.0 million and gross premiums written 4.9%. Professional liability grew fastest (+5.0% to $322.4 million), followed by auto (+5.2% to $438.7 million). Other liability, the largest line, grew 3.7%. Workers' compensation was flat (+0.7%). The segment's combined ratio improved to 91.7% from 92.3% even though its catastrophe losses rose to $82.9 million from $69.8 million. By our arithmetic, catastrophes added about 2.9 points to the loss ratio (2.5 a year ago). That puts the Insurance loss ratio excluding catastrophes at about 60.4%, versus about 61.4%. This rough figure still includes any changes to old-year reserves, which the release doesn't break out by segment.

Reinsurance & Monoline Excess keeps shrinking. Net premiums written fell 11.6% to $368.6 million, led by a 25.2% drop in casualty reinsurance (to $144.2 million). Property fell 2.2% and monoline excess rose 2.5%. The segment's combined ratio rose to 83.4% from 81.1%, still far better than the Insurance segment. But it went up even though catastrophe losses fell to $3.0 million from $8.7 million. Excluding catastrophes, our arithmetic puts the segment's loss ratio at about 52.3%, up from about 49.0%. Pre-tax income fell 7.5% to $133.3 million.

Did last time's read hold up?

Our Q2 2026 report said the question was underwriting. Renewal rate increases had slowed to 3.8% (excluding workers' comp), and we expected the Insurance segment's underlying loss ratio to keep creeping up. It also said catastrophes would be the main swing factor in Q3. That read held up only partly. Rate increases did not slow further: they were about 4.0%, a little above Q2, though still about half the 7.6% of Q3 2025. The company-wide accident-year combined ratio excluding catastrophes improved to 88.1% from 88.4%, not worse, and the rough Insurance loss ratio excluding catastrophes fell. Catastrophes were moderate ($85.9 million, 2.7 points), so they were not the swing factor this time. The investment portfolio was, through market losses and a rate-driven hit to book value.

Takeaway: The 12.7% drop in net income is mostly market noise. Operating earnings grew, and the main insurance business priced slightly ahead of claims even with rate increases at about 4%. The number that moved the wrong way is book value: rising bond yields widened after-tax unrealized losses on the portfolio from $317 million at June 30 to $928 million. That cut book value per share 2.4% in a quarter in which Berkley earned $446 million. Those losses reverse as the bonds approach maturity, and the same higher yields are what is growing investment income 12%.

What the headline numbers hide

• Reserves were not used to flatter the result. Insurers hold reserves for claims already incurred but not yet paid. When they cut those estimates, profit rises ("favorable development"). The accident-year ex-catastrophe combined ratio (88.1%) plus catastrophe losses (2.7 points) equals the reported 90.8%, which implies roughly zero net prior-year reserve development. The same arithmetic gives about zero for Q3 2025 (88.4% + 2.5 = 90.9%). The 10-Q will give the exact figure and say whether adverse development in other liability and auto, flagged in the first half, continued.
• Cash conversion weakened but remains solid. Operating cash flow was $731.3 million, about 1.6 times net income, but down 35.8% from $1,139.9 million a year ago. Year to date it is $2,199.2 million, down 15.0%. Loss reserves grew $1.4 billion this year to $23.6 billion, so premiums are still arriving well ahead of claim payments. The release gives no reason for the lower cash flow. Higher claim payments or timing are both possible; the 10-Q's cash flow statement should show which.
• GAAP vs operating earnings. Operating income excludes after-tax investment gains and losses ($18.6 million pre-tax loss net of related expenses) and currency gains ($5.1 million). This quarter that puts operating EPS 3 cents above GAAP EPS. A year ago it was 18 cents below, because investment gains inflated GAAP results. So the comparison runs: GAAP EPS -10.9%, operating EPS +6.4%.
• Part of the operating EPS growth is buybacks. Diluted shares fell 2.2% to 391.3 million, so operating income growth of 3.7% became 6.4% per share. Buybacks slowed sharply in Q3: 309,508 shares for $21.1 million, versus $111.5 million in Q2 and about $302 million in Q1. Year to date Berkley returned $728.6 million, including $435.1 million in buybacks and about $293.5 million in dividends.
• The tax rate was higher, not lower. The effective rate rose to 21.9% from 21.0%, a small drag rather than a boost.
• Return on equity looks lower partly because of last year's investment gains. Annualized ROE was 18.4% versus 24.3%; on an operating basis it was 18.8% versus 21.0%.

Outlook

Management gives no numeric guidance. The CEO pointed to two drivers: "the strong prevailing interest rate tailwind" for investment income, and "greater utilization of technology" in underwriting and claims. The bond portfolio has an average duration (roughly, how long until the money is reinvested) of 3.3 years and an average AA- rating. Bonds that mature get reinvested at today's higher yields, so investment income should keep growing as long as rates stay up. That also means unrealized losses on the existing bonds stay elevated, so book value per share will lag earnings in the meantime.

Underwriting looks steadier than our Q2 read feared. Rate increases have settled around 4% instead of falling further, and the underlying combined ratio is holding near 88%. The risks are in the mix. Reinsurance, the more profitable segment, is shrinking quickly (casualty reinsurance -25%) and its own underlying loss ratio rose this quarter. Workers' compensation and the wider liability lines are where pricing is softest. With Q3's hurricane-season peak passed and only $85.9 million of catastrophe losses, the nine-month combined ratio of 90.5% (91.1% a year ago) is a reasonable base for the full year unless Q4 brings a large event.

What to watch in the 10-Q (expected around early November) and Q4:

• Prior-year reserve development by segment, especially other liability and auto, where the first half showed adverse development.
• Why operating cash flow fell 36% year on year.
• Whether buybacks resume after the near-pause in Q3, and whether special dividends continue.
• The Insurance segment's underlying loss ratio as rate increases stay near 4%.