Treaty reinsurance underwriting software that reads, rebuilds and ranks every renewal submission
A first view of every program in hours, not days.
See one case, screen by screen ↓Software that takes a reinsurer’s treaty submissions from the broker’s email to a priced, referred decision. The Treaty Underwriting Workbench reads every submission as it arrives, rebuilds the cedant’s loss ratio from the data in it, compares exposure with last year and ranks the season by expected result and fit.
How to choose AI solutions for insurance →
Updated
At each renewal date a treaty team gets hundreds of broker submissions — a slip, a narrative PDF, an Excel loss workbook with a dozen tabs, a large-loss listing, an exposure schedule — each in the broker’s own layout. Analysts rebuild the loss triangles and exposure tables by hand, and in peak weeks the queue runs past a week. So many programs are priced on the cedant’s own summary.
That summary is honest but chosen: old development factors, large losses capped as shock losses, rate changes taken as filed rather than achieved. Each choice is defensible; together they can move a loss ratio ten points or more. In a falling market, a reinsurer that cannot tell good programs from flattering ones writes the flattering ones.
Estimated split, in days, for one casualty quota share rebuilt by an analyst by hand and with the solution.
Six agents read, rebuild and rank every submission; the underwriter decides the line and the terms.
Kestrel Point Re’s US casualty team has 214 submissions for 1 January 2027. Daniel Okafor has 22 programs and 81 days to inception. Here is one of them — Llano Ridge Mutual’s Commercial Auto Quota Share — screen by screen, in the working solution.
Daniel’s 22 programs, ranked by expected result on our line less $4k for each point of fit below 100. Each row has a loss-ratio strip: a hollow dot where the cedant puts it, a filled dot where the rebuild does. As stated, the book expects +$4.70M and no program loses money. Llano Ridge sits third.
“214 submissions for 1 January across the team · Rebuilt and ranked 168 · Being read 31 · Waiting on the broker for data 15”
He switches from “Rank as the cedant states” to “Rank on the rebuild”. The rows re-sort. Seven programs move five points or more; five now lose money on our line. Llano Ridge drops 15 places, from #3 to #18, at −$251k and 88% rate adequacy.
Ashby Greene Re’s email carried the slip, a 64-page narrative, a 9-tab loss workbook with 16,412 claim rows, large-loss narratives, two years of fleet schedules, the limits profile, and Kestrel Point’s own 2026 quote and signed line. The first view was on the board 23 minutes after the email arrived.
Never guesses a missing figure — it lists it as a question.
The program opens on the rebuild: the cedant’s 58.0% from narrative p. 14, beside 71.0% rebuilt from 16,412 claims and 9 workbook tabs. A bridge shows the three corrections. The run log on the right shows each agent’s step and time — the triangle rebuilt and tied to the cedant’s Inc_Tri_Capped tab within $0.01M in all 28 cells.
“Rebuilt from its own data · 16,412 claims · 9 workbook tabs · 3 corrections, each cited”
The summary counts every loss at no more than $1M. The quota share follows original policies with limits up to $5M and has no per-loss cap (slip, clause 4). The listing has 16 losses above $1M, $59.68M in all — $43.68M above the cap. Three are at the $5M policy limit. That is +3.3 points; the filed 11.5% rate taken for 2025, against 7.7% achieved on the rate monitor, adds +2.1.
Never removes a loss the treaty covers.
The incurred triangle, capped at $1M per loss as supplied, with age-to-age factors on three rows: the cedant’s (2016–2019), the rebuild’s latest 3 years volume-weighted, and the Kestrel Point commercial auto 2025 benchmark. The cedant’s factors give 63.2%, the rebuild 71.0%, the benchmark 70.6%.
Outside data is used only for the tail, and it is labelled.
The fleet schedules of 30 June 2026 and 2025, side by side: 4,812 power units against 4,690, tractors up 22.3%, long haul over 200 miles from 844 to 1,299 units — 18% to 27% of the fleet — and $5M limit policies from 141 to 212. Each change note links to its source and can go straight into a question.
At the cedant’s 58.0% the slide pays 32.3% commission and our line expects +$299k. At 71.0% it hits its 27% floor and loses −$251k. The recommended counter keeps the slide’s width and moves it: last year’s 10% line, 22.5% provisional, sliding 18.5% at 71% to 30% at 58% — +$287k expected, 100% rate adequacy.
Never quotes, binds or sends — the ranking and the counter are advice.
Five questions to Ashby Greene Re — development factors, the 16 capped large losses, the 2025 rate, the move to long haul, and the Fulton County verdict of $11.4M against a $5M limit — each tied to a tab or page. Beside them, the referral note to the chief underwriting officer: why it is referred, what was found, the recommendation.
“Drafted only · Daniel Okafor sends”
Daniel sends the questions to Owen Pryce and signs the referral. Owen has already answered question 3: 7.7% is the achieved 2025 rate. Helen approves the counter — 10% line at 22.5% provisional — and it goes to the broker. Every step is on the activity trail with who and when.
The same program, seen by the four people at the reinsurer who carry it — what their season looked like, and what it looks like now.
Six agents read, rebuild and rank every submission; the underwriter decides the line and the terms.
Sorts and reads every document in a broker’s submission — slip, narrative, workbook tabs, schedules — maps each tab to what it holds, extracts the slip terms and links last year’s submission and quote.
Rebuilds paid and incurred triangles from the cedant’s claim records, ties them to the cedant’s own triangles, and selects development with a labelled benchmark tail.
Reads large-loss listings against the slip’s terms — per-loss caps, policy limits, ECO and XPL — and on-levels premium with the achieved rate, not the filed one.
Compares this year’s exposure schedules with last year’s submission and our quote, and writes year-on-year change notes.
Works out commission on the slide, combined ratio, expected result on our line, rate adequacy and technical commission on both loss ratios; scores fit against appetite; ranks the season.
Drafts the questions to the broker and the referral note to the chief underwriting officer, each line cited.
Decides the line; the chief underwriting officer signs referrals. The agents propose; a named person decides.
The underwriter can ask in plain words — or ask for the questions to be drafted. Each answer points to the tab or page it rests on.
Why is Llano Ridge at 71%, not 58%?
Llano Ridge’s own summary says 58.0%. Three changes, all from data in the same submission, take it to 71.0%: development factors from 2016–2019 when the triangle now develops faster (+7.6), 16 large losses capped at $1M when the treaty has no cap (+3.3), and the filed +11.5% used for 2025 when the achieved rate was +7.7% (+2.1).
Which programs moved most on the rebuild?
Seven programs moved by 5 points or more, all of them upwards — Vantor Specialty, Ironweed Casualty, Llano Ridge Mutual, Mesquite Ridge Casualty, Redfern Mutual, Sable Harbor Specialty and Hatchet Point Insurance. The board switches to the rebuild, filtered to the movers.
What commission makes Llano Ridge work at 71%?
At 71.0% the slide offered pays its 27% floor and our line loses −$251k. To earn the 5% target margin the commission must be 18.5%: 100 − 71.0 − 2.5 brokerage − 3.0 expenses − 5.0 margin. A counter that keeps the slide’s width: 22.5% provisional, sliding 18.5% at 71% to 30% at 58%, at last year’s 10% line — +$287k a year.
Draft the questions to Ashby Greene Re
Five questions are drafted, each tied to the tab or page it refers to: development factors, the 16 capped large losses, the 2025 rate, the move to long haul and the Fulton County verdict above its $5M limit. Owen Pryce’s deadline for answers is Friday 23 October. Nothing is sent until you send it.
13 screens from the working solution, on its sample data. Pick one to see it large.
How submissions are rebuilt and ranked is a setting. Changes are versioned and on the audit trail.
| Setting | Default | Choose from |
|---|---|---|
| Development benchmark | Kestrel Point commercial auto 2025 | Kestrel Point commercial auto 2025 · Kestrel Point commercial auto 2024 · Industry annual statement data 2025 |
| Development factors | Latest 3 years, volume-weighted | Latest 3 years, volume-weighted · Latest 5 years, volume-weighted · All years, simple average |
| Premium on-level | Achieved (rate monitor) | Achieved (rate monitor) · Filed · The lower of the two |
| Put large losses back to the treaty’s own terms | On | On · Off |
| Flag a program when the rebuild differs by | 5 pts | Step up or down |
| Target margin | 5% of premium | 5% · 7.5% · 10% of premium |
| Rank by | Expected result, less $4k per fit point | Expected result, less $4k per fit point · Expected result only · Fit only |
| Refer when rate adequacy is below | 90% | 85% · 90% · 95% |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample renewal data · “target” = the design goal, measured in the live solution · “estimated” = our estimate (2–4 days to rebuild a casualty program by hand, and the split of days) · “typical” = published figures (Guy Carpenter rate-on-line indices via Artemis, 30 Jun 2026 — global property cat rates down about 16% in 2026 so far). People, reinsurers, cedants, brokers and programs named on this page are fictional — characters and sample data in the working solution.
Software that takes a reinsurer’s treaty submissions from the broker’s email to a priced, referred decision. The Treaty Underwriting Workbench reads every submission as it arrives, rebuilds the cedant’s loss ratio from the data in it, compares exposure with last year and ranks the season by expected result and fit.
From the cedant’s own data: it rebuilds paid and incurred triangles from the claim records and ties them to the cedant’s triangles, selects development factors as set in Settings, puts back large losses the treaty covers, and on-levels premium with the achieved rate. Each correction is a step on a bridge, cited to its tab, cell or page.
No. The cedant’s figure is shown next to every rebuilt one. The only outside data in a rebuild is the benchmark development tail, and it is labelled.
No. The agents rank and draft; they never quote, bind or send. The underwriter decides the line and terms and sends questions and counters, and referrals need the chief underwriting officer’s approval.
In the working solution: when rate adequacy is below 90% at the terms offered on the rebuilt loss ratio, when the rebuild differs from the cedant’s by more than 10 points, when our premium is above $5M on one program, for a new cedant, or when declining a program written five years or more.
The broker’s email and everything in it — slips, narratives and other PDFs, scans, spreadsheets with many tabs, exposure schedules, limits profiles — plus last year’s submission and your own quote and signed line.
Yes. The development benchmark, how factors are selected, which rate change to on-level with, the gap that flags a program, the target margin, the expense load, how the board is ranked and the adequacy that refers are all settings, versioned on the audit trail.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your underwriting guidelines, appetite and authority, development benchmarks and a sample of past submissions — and it goes live once every quality gate has passed.
A checked renewal submission in days, not weeks, every figure traced.
1-min demoTreaty Wording ComparisonEvery changed clause found before signing, not when a claim tests it.
1-min demoReinsurance Contract SearchCoverage questions answered with the clause, without waiting on one specialist.
1-min demoWe’ll run the Treaty Underwriting Workbench on a sample of your own past submissions.