Stephen Donaghy told analysts on Universal Insurance Holdings' second-quarter call that the company generated a 33.2% annualized adjusted return on common equity, alongside a 91.6% combined ratio and a net loss ratio that improved 7.5 points to 64.8% (Universal Insurance Holdings, July 2026). The largest Florida-focused homeowners writer now prints margins the wider P&C industry reserves for its best hard-market years.

Universal reported after the close on July 23, with the call the following morning, and the numbers land as more than one carrier's good quarter. Because Universal writes the state's largest concentrated book of Florida homeowners business, its loss ratios, its reinsurance spend, and its reserve language together form the most current measurement available of where the post-reform Florida market sits in its cycle. Read that way, the release describes a market past the repair phase and entering the phase where returns this wide start attracting the forces that narrow them.

Where the 33.2% Comes From

The release arithmetic splits the return into three verifiable parts. An 8.4-point underwriting margin on $377.3 million of net premiums earned implies roughly $32 million of quarterly underwriting gain, the largest single contributor. Net investment income added $20.2 million, up from $17.3 million a year earlier on higher reinvestment yields and a larger invested asset base, and commission revenue plus policy fees contributed another $19.8 million (Universal Insurance Holdings, July 2026). The GAAP return ran higher still at 38.8% annualized, with the gap to the 33.2% adjusted figure sitting in investment gains the adjusted measure strips out: diluted GAAP earnings of $2.04 per share against $1.84 adjusted. The return is underwriting-led.

The denominator makes the numerator more demanding, not less. Book value per share ended the quarter at $22.89, up 39.7% in twelve months, so the 33.2% was earned against an equity base two-fifths larger than the one that carried last summer's returns (Universal Insurance Holdings, July 2026). A capital-depleted carrier can flatter its return ratio with a hollowed denominator. Universal is compounding a rebuilt one, and it spent only $4.5 million on buybacks during the quarter while doing it.

The premium side carries the cycle signal. Direct premiums written rose 4.1% to $621.3 million, but the mix is lopsided: Florida grew 0.8% to $453.2 million while other-states business grew 14.4% to $168.1 million (Universal Insurance Holdings, July 2026). Policies in force reached 934,371, up 7.1%, and the spread between 4.1% premium growth and 7.1% unit growth means average written premium per policy fell roughly 3% year over year on the release's own figures. Universal is earning its 33.2% while giving back rate, not while taking it.

MetricQ2 2026Q2 2025Change
Combined ratio91.6%97.8%Improved 6.2 pts
Net loss ratio64.8%72.3%Improved 7.5 pts
Net expense ratio26.8%25.5%Up 1.3 pts
Ceded premium ratio30.8%31.2%Down 0.4 pts
Direct premiums written$621.3M+4.1%
Florida / other states DPW growth+0.8% / +14.4%
Policies in force934,371+7.1%
Net investment income$20.2M$17.3M+16.8%
Book value per share$22.89+39.7%

Source: Universal Insurance Holdings Q2 2026 earnings release, July 2026. Prior-year ratios per the release's stated year-over-year changes.

The Loss Ratio Against the State's Litigation Data

The quarter's engine is the 7.5-point loss ratio improvement, and the release attributes it to better current accident year results. Donaghy drew the causal line to the legislature. In his framing, the favorable claims and litigation trends are a direct product of Florida's reforms: the December 2022 special session bill that ended one-way attorney fees and assignments of benefits for property claims, plus the follow-on 2023 bad-faith changes. He went further on the stock of open disputes: "Our litigation inventory is back down to levels that preceded Florida's litigation crisis, and the impact of pre-reform claims practices is behind us" (Universal Insurance Holdings, July 2026).

The state's own measurement system supports him. The OIR's January 2026 Property Insurance Stability Report counts personal residential legal service of process filings down 23% in 2024, then down a further 26% through the first eleven months of 2025 against the same period a year earlier (Florida OIR, January 2026). The declines run consecutively back to the 2021 peak. Defense cost and containment expense per claim fell from $992.89 in 2022 to $817.64 in 2024, on $107.4 million of total domestic homeowners defense spend in 2024 (Florida OIR, January 2026). Set against national benchmarks the shift is starker. Florida generated more than 79% of US homeowners claims lawsuits in 2020 on about 9% of the country's homeowners claims; by 2025 the lawsuit share had fallen to 41% on roughly 5% of claims, per OIR data reported by Insurance Journal (Insurance Journal, July 2026). Guy Carpenter puts the cumulative litigation decline at roughly 66% from its peak (Guy Carpenter, June 2026).

The actuarial mechanics of that collapse run through severity as much as frequency. Under the old statute, a represented water or roof claim carried one-way fee exposure that former deputy insurance commissioner Lisa Miller describes as reaching five or six times the claim amount itself (Insurance Journal, July 2026). Each point of litigated-claim frequency therefore dragged disproportionate loss adjustment expense and settlement leverage into the ratio. Remove the fee asymmetry and the assignment channel, and the claims that still arrive settle earlier, cheaper, and with fewer intermediaries. Universal did not publish an ex-catastrophe loss ratio for the quarter, but the release's attribution of the improvement to the current accident year is the load-bearing disclosure. The 64.8% embeds a live pick on post-reform frequency and severity, not a harvest of old redundancy, and three consecutive years of state data now sit underneath that pick.

One Month of the New Reinsurance Program in the Quarter

Reinsurance is the second margin lever, and the quarter only sampled it. The ceded premium ratio eased to 30.8% from 31.2%, a move the release ties primarily to the new program that incepted June 1, 2026 (Universal Insurance Holdings, July 2026). The timing detail carries the modeling weight: April and May earned under the 2025-2026 treaty and only June under the new one, so the printed 0.4-point improvement blends two months of old pricing with one month of new. The full relief spreads across the treaty year running through May 2027, which makes the ceded line a known tailwind for the next three quarters rather than a lapsed benefit.

The program itself tops out at $2.623 billion of All States first-event coverage above a $45 million retention, up approximately $50 million from the 2025-2026 tower with the retention unchanged (Universal Insurance Holdings, May 2026). Full reinstatement applies on $1.098 billion of non-FHCF first-event limit for a guaranteed second event. The Florida Hurricane Catastrophe Fund provides an estimated $1.39 billion mandatory layer for lead subsidiary UPCIC, more than half the tower, which mutes the program's sensitivity to private-market price swings in either direction. Universal also locked in $352 million of capacity on contractually agreed limits extending into the 2027-2028 treaty year, $277 million of it attaching below the FHCF layer, with Nephila among the lead private participants (Artemis, May 2026).

Those terms were bought into the softest Florida renewal in years. Guy Carpenter marked risk-adjusted Florida property-catastrophe decreases typically in the 15% to 20% range at June 1, on cedent demand up 12% (Guy Carpenter via Carrier Management, June 2026). The 2025 season behind those numbers was the first in a decade without a landfalling tropical storm in the state. Universal's response reads as deliberate cycle management: hold the retention, buy modestly more limit, and convert part of the rate savings into multi-year duration at prices management evidently judges closer to the bottom than the middle. The site mapped the renewal's double-digit drops as they cleared in June 1 renewal coverage; Universal's program is that market-wide repricing showing up inside a single carrier's cost structure.

The Reserving Read at One Year and at Two

Universal's reserve disclosure is qualitative but pointed. The release does not quantify prior-year development; the loss ratio improvement is attributed to the current accident year, and management states its belief that aggregate reserves hold a meaningful margin above expected ultimate losses (Universal Insurance Holdings, July 2026). For a Schedule P reader that pairing is the conservative configuration. If the margin assertion is right, the printed 64.8% overstates what the quarter's business will ultimately cost, and the release is choosing not to monetize the gap through the income statement yet.

The market-level diagonals give the assertion a base rate. When Florida domestic insurers re-measured reserves one year on, 2024's claims came in approximately $169.6 million below original estimates, favorable development the OIR explicitly credits to the reforms (Florida OIR, January 2026). At the two-year look-back, which reaches into pre-reform accident years, development was still adverse by $523 million. That sign flip between diagonals is the signature of a genuine trend break rather than a mix effect: cohorts adjudicated under the new statute are releasing while the tail written under the old one was still deteriorating. The same report notes Florida domestics posted aggregate underwriting gains in 2024 after nine consecutive years of underwriting losses, and Guy Carpenter has the domestic cohort at a 76.8% combined ratio for 2025 (Guy Carpenter via Carrier Management, June 2026).

The watch item for Universal's coming statutory filings follows directly. Accident years 2023 through 2025 releasing on schedule would validate both the current-year pick and the margin language; renewed creep on 2021 and prior would mean pre-reform claims practices are less fully behind the book than the release asserts. The site's Schedule P work this quarter found prior-year releases doing a growing share of reported combined-ratio improvement across the industry. That sharpens the distinction here: Universal is presenting its improvement as rate and frequency, not as release, and the claim is checkable in next spring's triangles.

Depopulation, Rate Decreases, and the Capacity Aimed at This Margin

The competitive backdrop is moving in one direction. Citizens Property Insurance counted 278,662 policies in early July, an all-time low for the residual carrier, down from the 1.42 million peak of October 2023 (Keys News, July 2026). The 2025 depopulation rounds alone transferred more than 546,000 policies to private carriers, and Guy Carpenter counts more than 1.4 million policies exiting Citizens since 2022 (Citizens Property Insurance, December 2025; Guy Carpenter via Carrier Management, June 2026). Citizens chief executive Tim Cerio says the corporation's probable maximum loss now sits at an all-time low as well (Keys News, July 2026).

Rate is following the policies out the door. Citizens' board recommended a statewide average personal-lines decrease of 2.6% for 2026, with three in five policyholders in line for an average 11.5% premium reduction (Citizens Property Insurance, December 2025). On the supply side, the OIR had approved seventeen new residential property companies since the reforms as of its January report, a count Insurance Journal put at more than twenty by July (Florida OIR, January 2026; Insurance Journal, July 2026). That roster includes the reciprocal-exchange entrants that took large blocks in the May depopulation round the site covered in its 184,000-policy takeout analysis. Universal's own geography shows where that pressure lands first: 0.8% direct premium growth in Florida against 14.4% outside it. Donaghy told analysts that Citizens no longer represents a competitive threat to the company at all (The Markets Daily, July 2026). A specialist whose home-state top line is flat while its policy count grows is a specialist already conceding rate.

The counterweight is the tail. John Rollins, chief executive of Florida carrier Patriot Select, argued in a June viewpoint that a 100-year storm scenario can generate losses of three to four times a year's premium (Insurance Journal, June 2026). Decades of accumulated surplus can be consumed in hours, against regulated profit margins that are modest by design. The same piece notes Guy Carpenter's index of US property-catastrophe reinsurance rates peaked in 2023 at more than double the 2017 trough, a measure of how violently the cost side re-rates after events. A 33.2% adjusted ROCE is simultaneously the payoff for surviving 2017 through 2022 and the standing invitation to the capacity that will compress it. On a one-in-one-hundred view the return is not excessive; on the five-year view the OIR's rate-filing queue will treat it as if it were.

What the Quarter Changes in Pricing, Reserving, and Capital Assumptions

For pricing work on Florida homeowners books, the quarter hardens the case that post-reform frequency is a level shift rather than a cycle dip. Three consecutive years of double-digit declines in suit filings, defense cost per claim down 18% in two years, and a national lawsuit share cut roughly in half now sit behind any trend selection (Florida OIR, January 2026). The discipline problem inverts 2022's. Selections still carrying pre-reform litigation loads will price themselves out of a market where even the residual carrier is filing decreases, while selections that quietly extrapolate a zero-landfall 2025 into the cat load commit the opposite error. Rate-adequacy testing should keep the two effects separated, because litigation reform is in the frequency data and hurricane luck is not a trend.

For reserving, the OIR's diagonal flip is the working template: test reform-era accident years for release capacity separately from the pre-reform years still capable of adverse surprise, and read every Florida carrier's current-accident-year attribution against that split. For capital and planning, the June 1 renewal and Universal's multi-year purchase together say the reinsurance tailwind has at least one more treaty year to run, and the entry wave says the underwriting margin does not. Returns like this quarter's mark the top of a repair cycle. The actuaries pricing 2027 Florida business will be working its down-slope: still adequate, likely no longer this wide, and contested by carriers that never carried the pre-reform scar tissue on their balance sheets.

Further Reading

Sources