Arch Capital Group Ltd.

Sources: FY2025 10-K XBRL data (via EDGAR-derived tables), FY2024 and FY2025 Annual Reports, and Q4 2025 supplement. Derived figures are labelled.

Income statement

Notes and gaps:

• FY2021 GPW: the company "growth story" page shows about $12.5B (rounded). It is not entered in the table because it is not exact.

• Q4 2025 was $434M (+6.4% QoQ). Derived from total revenues, NII plus equity-method income plus other income was about $2.18B.

• FY2022–FY2023 EPS and BVPS figures are from prior company releases (analyst knowledge, high confidence). FY2024–FY2025 figures are verified in the 2025 Annual Report.

• Gross profit, operating income in the industrial sense, and EBITDA are not applicable to an insurer. Underwriting income and after-tax operating income are the analogues used here.

Margins and growth (derived)

• The derived combined ratio is 1 − underwriting income ÷ (NPE + other underwriting income).

Revenue CAGR FY2021–FY2025: 21.2% (from $9,248M to $19,929M). NPE CAGR: 20.5%.

Commentary.

2021–2022: strong underwriting was offset by 2022 mark-to-market investment losses ($663M realized losses) as rates rose. Net income fell to $1.44B and BVPS fell.

2023: an inflection year. Combined ratio was about 79%, investment income doubled, and a large one-off deferred tax benefit arose from the Bermuda CIT enactment. These drove net income to $4.4B and a derived net margin of 32%.

2024: a record $3.5B of operating income despite heavy catastrophe losses from Hurricanes Helene and Milton. The MCE Acquisition boosted premium.

2025: another record operating year ($3.7B) despite the January 2025 California wildfires, which added 9.5 points of catastrophe losses to Q1 Insurance. Net premium growth decelerated to about 5%, and Q4 2025 NPW fell 4.5% YoY.

H1 2026: net income was about $2.0B ($2.88 plus $3.00 per diluted share). Operating ROE was about 15.3–15.4%, below 2025's 17.1% because of lower earned premium and Iran-conflict losses.

Balance sheet

Debt (holding-company senior notes, face value; FY2025 10-K and Q2 2026 10-Q):

Total senior debt at FY2025: about $2.75B face, all long-term. Short-term debt is nil apart from the 2026 notes, which fell due within 12 months at year-end 2025.

Pro forma after the 2026 refinancing: about $3.83B, excluding the $500M 2026 notes to be retired.

Debt to total capital: 10.1% at Dec 31, 2025 versus 11.6% at Dec 31, 2024 (company-reported via Q4 2025 summary).

Net debt, derived: about $1.76B at FY2025 (senior notes less consolidated cash). Net debt is of limited relevance for a regulated insurer whose cash and investments support policyholder liabilities.

Working capital: not applicable. Insurers present unclassified balance sheets.

Cash flow

Capital returned:

• 2025 share repurchases: about $1.9B, including $798M in Q4 2025.

• 2024 special dividend: about $1.9B, at $5.00 per share.

• H1 2026 repurchases: 20.7M shares for about $1.9B ($783M in Q1 and $1.2B in Q2).

• Since inception, Arch has repurchased 475.7M shares for $9.7B (Q2 2026 10-Q).

• Preferred dividends run at about $40M per year ($10M per quarter in Q2 2026).

Note on the "cash flow" line: the 4.5% headline figure sometimes cited for 2025 refers to Q4 OCF ($1,404M, −10.7% YoY). Full-year 2025 OCF was $6,172M, down 7.5% (Q4 2025 supplement).

Ratios

Basis and gaps:

• FY2022–FY2023 ROE is derived on average common equity (total equity less $830M preferred). FY2024–FY2025 figures are company-reported.

ROIC: not reported. Derived as after-tax operating income ÷ (common equity + senior debt), it is about 14.5% for FY2025 ($3,700M ÷ about $25.5B average).

Interest coverage: FY2022 about 12.4x (derived; pre-tax $1,487M plus interest $131M, divided by $131M). Recent quarterly interest expense was $44M in Q2 2026 versus $38M in Q2 2025. Coverage on 2025 earnings is well above 20x (derived, approximate).

• Current ratio, cash conversion cycle, asset turnover, net debt/EBITDA: not applicable or not meaningful for an insurer.

Trend commentary.

• Arch's capital ratios are conservative and improving. Equity-to-assets is about 30%, debt-to-capital fell to 10.1%, and underwriting leverage declined to about 0.68x NPW-to-equity. This gave the company room for the 2024 special dividend and the 2025–26 buyback acceleration.

• The June 2026 refinancing extends duration with long-dated 2056 paper. It modestly raises gross leverage and interest expense, which rose about $6M per quarter year-on-year in Q2 2026.

• Returns have normalized from the 2023 peak, which was flattered by the tax benefit. They are trending toward the mid-teens on an operating basis as reinsurance pricing softens.

Management guidance

Arch provides no numerical EPS or premium guidance. Qualitative guidance from the 2026 calls and 10-Qs:

• continued discipline and shrinkage in property cat and short-tail reinsurance where returns are below target;

• about $250M of 2026 NPW headwind from MCE program non-renewals;

• selective growth in E&S casualty and specialty insurance, and in post-Iran political violence and war pricing;

• mortgage underwriting income expected to remain attractive given in-force credit quality and embedded equity;

• buybacks continuing, with $2.2B of capacity at June 30, 2026.

Consensus expectations

Q3 2026 EPS: about $1.89–1.96 (Zacks; OpenCapital), versus about $2.77 a year earlier as implied by the 29% decline cited. This consensus was assembled before any Q3 catastrophe disclosure.

Forward P/E: 9.93x.

Revenue: one aggregator (Simply Wall St) projects about 6% annual revenue decline over three years, consistent with the soft cycle.

Bull case

1. Underwriting edge persists through the soft market. A derived combined ratio of about 83% in 2025, against 87–99% for peers, gives a margin cushion before profitability is impaired.

2. Capital return compounds per-share value. About $1.9B of buybacks in H1 2026 alone (over 5% of market cap annualized) at about 1.5x book still compounds BVPS at mid-teens rates.

3. Optionality from excess capital and low leverage. Debt-to-capital of about 10% (about 13–14% pro forma for the June 2026 issuance), plus $2.2B of remaining authorization and the A+/AA- franchise, position Arch to deploy aggressively into any post-catastrophe hard market. It did this in 2001–02 and 2019–23.

Bear case

1. Top-line erosion: Q2 2026 NPE fell 8.1% and reinsurance NPW fell 10.4%. Earnings power declines mechanically as earned premium shrinks.

2. Insurance segment underperformance: a 98.5% Q2 2026 combined ratio and a rising expense ratio (36.3% in Q1 2026) raise questions about MCE economics.

3. Tail risks are concentrated in current events: Iran-conflict political violence and war losses, inflation-driven reserve risk flagged in the 10-Q, and housing-credit risk in mortgage. Any could reverse years of favorable development.

Catalysts and monitorables (next 12 months)

Concluding analyst verdict

Arch Capital enters late 2026 as one of the highest-quality balance sheets and underwriting franchises in global specialty (re)insurance. It now faces the phase of the cycle that tests its discipline most.

The evidence is strong. Arch posted a record $3.7B of after-tax operating income in 2025 and grew book value per share 22.6%. Its derived combined ratio of about 83% compares with 87–99% for its closest listed peers. It also delivered a fourth consecutive year of at least $1B of underwriting income from a mortgage franchise that no P&C competitor can replicate.

Leverage is modest, and the June 2026 refinancing pushed maturities out to 2034–2056. Capital return is running at about $3.8B annualized, which absorbs much of the earnings at roughly 1.5x book.

The near-term picture is less flattering. Earned premium is contracting, with Q2 2026 NPE down 8.1%, as management deliberately walks away from soft property-cat pricing. The Insurance segment remains the weakest link: Iran-conflict losses and MCE transition costs pushed its Q2 combined ratio to 98.5%. Consensus expects Q3 2026 EPS to fall about 29% year on year, and the October wave of target cuts reflects a sector-wide soft-market narrative.

Our synthesis is that Arch is executing exactly what its cycle-management philosophy prescribes. A shrinking top line is a feature of that strategy, not a defect, provided underwriting margins hold in the low-to-mid 80s, mortgage stays well above $800M of underwriting income, and buybacks continue near book-accretive levels.

At about 7.9x trailing earnings and about 1.5x book, against a mid-teens operating ROE, the valuation already discounts much of the softening. The key risks to this view are:

• a large catastrophe or man-made event that coincides with soft pricing;

• adverse reserve development in long-tail casualty;

• a U.S. credit downturn affecting mortgage.

Verdict: a high-quality compounder in a cyclical trough. Investors should expect slower but still double-digit BVPS growth. Upside optionality materializes if a market dislocation lets Arch redeploy its excess capital at hard-market returns.