Excess casualty tower analysis: every underlying policy read into one tower, with its gaps shown before you price your layer
Limits, attachments and follow-form gaps shown before you price your layer.
See one case, screen by screen ↓Reading every policy beneath an excess layer — primary, lead umbrella and each excess layer — to confirm limits, attachment points, aggregate erosion and whether each layer truly follows form. Excess Casualty Tower Analysis does that reading with six agents, draws the tower to scale and marks every gap with the page it came from, before the underwriter prices the layer.
How to choose AI solutions for insurance →
Updated
You are asked for $10M xs $20M on a tower other carriers built. Before the price means anything you need to know what sits underneath: every limit and attachment point, how much of each underlying aggregate is already eroded, and whether each layer really follows form to the lead umbrella or quietly carves something out in an endorsement. That means reading a dozen policies, the schedule of underlying insurance and five years of loss runs from more than one carrier.
Every carrier says it follows form. The gaps hide in the endorsements — an auto exclusion on page 14, defence counted inside a limit, a lead umbrella that ends 17 days before the rest of the tower — and the broker’s schedule says “follow form, all lines” over the top of it. Miss one and it surfaces at claim time, as a coverage dispute between layers or as your layer sitting over a hole.
Estimated split for one excess submission of about a dozen documents, by hand and with the solution, in working days. With the solution the agents read the tower in under an hour; the time left is the underwriter checking the cracks and finishing the note.
Six agents read every policy into one tower and mark its cracks; the underwriter prices and signs.
Cinderford Wholesale Partners asks Harrowgate Specialty to quote $10M xs $20M for Pellham Ridge Building Supply, a Fort Worth building-materials distributor with 186 power units. Here is what happens next, screen by screen, in the working solution.
Theo opens his excess submissions: nine in flight, each with its tower read before anyone prices it. Pellham Ridge — XS-26-04417, received Tue 6 Oct, quote due Fri 16 Oct — shows “Tower ready” with 4 cracks, 1 critical. Kessling Field Services is waiting on the broker; Bluebank Foods is still being read, 9 of 14 documents.
The policy reader took 41 limits and sublimits, 5 attachments and 67 endorsements out of 12 documents, each tied to its page. The schedule matcher checked the broker’s schedule and ACORD 131/212 against 9 policies and found 3 mismatches. The follow-form comparer ran 12 provisions across 5 layers; the loss run reader read 182 claims from 2 carriers.
“Note writer — draft for $10M xs $20M: $468,000 subject to 3 conditions · referral to Ingrid Solberg.”
$45M in 5 layers from 5 carriers: Bellwether Casualty primary, Ironbridge Specialty’s $8M xs $2M lead umbrella, Calder Bay’s $10M xs $10M, Theo’s requested $10M xs $20M and Westmoor’s $15M xs $30M on top. Tower premium with his layer is $4,628,000; his layer is indicated at $468,000, a rate on line of 4.68%. The four cracks are drawn where they sit.
Theo picks $28M and Auto · tractor-trailer — a multi-vehicle crash, one occurrence — and drops it. Bellwether pays $2M and exhausts. Ironbridge pays $8M and exhausts. Then the loss stops at layer 3: Calder Bay pays $0, and Pellham Ridge carries $10M uninsured, inside the tower.
“Stops at layer 3 · auto excluded — Calder Bay CB-XS 117, page 14 · insured carries $10M.”
Critical: auto excluded in layer 3. High: the EL disease limit is short of the schedule — Ironbridge sits over $1M disease each employee; Bellwether gives $500K. Medium: the lead umbrella ends 17 days early. Medium: defence counted inside Calder Bay’s limit, so layer 3 wears down faster. Each one can be shown on the tower, asked of the broker, or accepted.
Crack 1 opens on the wording of endorsement CB-XS 117: no cover for any liability arising out of any “auto”, “regardless of whether such liability is covered by the controlling underlying insurance”. The broker’s schedule lists Calder Bay as “follow form, all lines”. A $10M hole for any auto loss between $10M and $20M, directly beneath his layer.
“The policy, not the schedule, decides.” Found by the follow-form comparer — flagged for you, not decided.
Twelve provisions — auto, defence costs, employers liability, policy period, exhaustion, PFAS, abuse or molestation, punitive damages, products aggregate, contractual liability, wildfire, territory — across five layers, compared against Ironbridge IB-UMB 2024. Sixty cells, each citing its page. Westmoor’s own exhaustion wording and its wildfire exclusion are marked too; both sit above Theo’s layer.
Granite Prairie Mutual 2022–2023 and Bellwether Casualty 2024–2026 read together: auto, 41 claims and $9.84M incurred, valued 30 Sep 2026. One loss in five years reached an umbrella — the I-35 jackknife of 14 Mar 2023, $3.4M. The products-completed operations aggregate is 41% used by two open engineered-joist claims, over the 33% warning.
“Aggregates reset on 1 Jan 2027; the joist claims may not.”
The risk, the tower, the four cracks, the losses and the scenario, each cited. Option A quotes $468,000 with auto following form, subject to closing the cracks; option B quotes $186,000 with auto excluded (HS-AU 04) if Calder Bay keeps CB-XS 117; option C declines. The build-up loads +2.0% for defence inside Calder Bay’s limit and credits −4.3% for telematics and driver scoring since 2023.
You choose the option, edit and sign; nothing reaches the broker until you send it.
186 power units is over 150, and there is a critical crack: guideline G-9 makes it a referral. Theo refers it to Ingrid with the tower, the cracks and the note; she co-signs, and only then can he sign option A. The subjectivities go to Dana Whitlock at Cinderford as a separate step — no premium is sent until the note is signed.
Ingrid Solberg: “Agree option A. Crack 1 must be closed before binding.”
The same submission, seen by the four people who carry it — what their week looked like, and what it looks like now.
Six agents read every policy into one tower and mark its cracks; the underwriter prices and signs.
Reads every policy, declarations page and endorsement into limits, aggregates, attachments, periods, exclusions and definitions — each with its page.
Checks the broker’s schedule of underlying insurance and ACORD 131/212 against the policies actually supplied: carriers, lines, limits and periods.
Compares every layer, provision by provision, with the lead umbrella — follows, narrower or own wording — and marks the cracks.
Reads loss runs from every carrier into one view by line and year, and shows aggregate erosion and the losses that reached a layer.
Pours a hypothetical loss down the tower from the terms as read: who pays, who exhausts, where payment stops and what the insured carries.
Drafts the underwriting note, the pricing build-up and the subjectivities, every statement cited, with options for the underwriter to choose from.
Prices and signs; head of casualty co-signs referrals. The agents propose; a named person decides.
Ask in plain words, or have it drop a loss on the tower for you. Every answer points to the page it rests on.
Why does a $28M truck loss stop at layer 3?
Because Calder Bay’s layer follows the lead umbrella except where its own endorsements say otherwise, and endorsement CB-XS 117 excludes any liability arising out of the use of an auto. The broker’s schedule lists Calder Bay as “follow form, all lines” — but the policy decides, not the schedule. Bellwether pays $2M, Ironbridge $8M, Calder Bay $0; Pellham Ridge carries $10M uninsured, and Harrowgate’s $8M is reached only if Pellham Ridge pays that $10M.
What does Calder Bay’s defence wording cost our layer?
$1.2M on a defended $28M loss. Calder Bay counts defence inside its $10M; Ironbridge pays defence on top of its limit. With $1.2M of defence, Calder Bay’s damages run out at $8.8M, so we pay $9.2M instead of $8M. The note loads +2.0% (+$9,591) for it, within guideline G-5.
Does the broker’s schedule match the policies?
No — three mismatches; the other 38 limits and all 5 attachments agree. Calder Bay: the schedule says all lines, the policy excludes auto. Ironbridge’s period: the schedule says 1 Jan 2027, the policy says 15 Dec 2026. EL disease each employee: the schedule says $1M, the policy says $500K. All three are subjectivities in option A.
What if Calder Bay won’t delete the auto exclusion?
Two ways to quote, one to walk away: A, ask for a $10M auto buffer xs $10M beneath us, at $468,000; B, exclude auto in our layer too (HS-AU 04), at $186,000, the minimum premium; or C, decline. Guideline G-3 says a line excluded beneath our attachment is a subjectivity, not a reason to decline. Option B leaves Pellham Ridge with no auto cover from $10M to $45M — Dana Whitlock will want to know that in writing.
13 screens from the working solution, on its sample data. Pick one to see it large.
Authority, what counts as a crack and how losses are modelled are settings — versioned, on the audit trail, and re-checked on every tower in flight when they change.
| Setting | Default | Choose from |
|---|---|---|
| Largest limit an underwriter may write alone | $10M | $5M to $25M |
| Refer fleets above | 150 power units | 100 · 150 · 250 power units |
| Co-signer | Ingrid Solberg | Ingrid Solberg · Ingrid Solberg or Marcus Ode |
| Any line excluded below our attachment | Always critical | Fixed — auto, EL or products: a hole beneath our layer |
| Underlying limit below the schedule | On | On · Off |
| Policy period mismatch tolerance | 0 days | 0 · 15 · 30 days |
| Aggregate erosion warning | 33% used | 25% · 33% · 50% used |
| Defence assumed on a defended loss | $1.2M | $0.8M · $1.2M · $2.0M |
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 submission · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published figures (Insurance Thought Leadership — an E&S underwriter reviews 16–20 submissions a day). Context for reading each excess policy’s own text first: in Ohio Casualty v. Patterson-UTI Energy (Tex. 2024) the Texas Supreme Court held that “follow form” takes in the underlying policy only as far as the excess policy’s own words say. People, companies, carriers and policies named on this page are fictional — characters and sample data in the working solution.
Reading every policy beneath an excess layer — primary, lead umbrella and each excess layer — to confirm limits, attachment points, aggregate erosion and whether each layer truly follows form. Excess Casualty Tower Analysis does that reading with six agents, draws the tower to scale and marks every gap with the page it came from, before the underwriter prices the layer.
The follow-form comparer compares every layer, provision by provision, with the lead umbrella the layers above say they follow — auto, defence costs, employers liability, policy period, exhaustion, exclusions, aggregates and more. Any provision that leaves a loss uncovered between attachments, or wears a layer down faster, is marked as a crack and ranked by severity.
No. The schedule matcher checks the schedule of underlying and ACORD 131/212 against the policies actually supplied, and where they differ the policy decides. In the working solution it found three mismatches, including a layer the schedule called “follow form, all lines” that excludes auto by endorsement.
Yes. Pick a loss size and a cause — auto, premises or products — and drop it on the tower. The tower loss modeller applies each layer’s own exclusions, defence and exhaustion wording and shows who pays, who exhausts, where payment stops and what the insured carries. It is a scenario, never a price or a reserve.
No. The agents read, compare and draft; the underwriter chooses the option, edits and signs the note. Referrals above authority need the head of excess casualty to co-sign, and nothing reaches the broker until the underwriter sends it.
Yes. Your excess casualty guidelines — minimum attachments, required underlying limits, exclusions you always carry, defence and exhaustion rules, referral triggers — are what the agents check every tower against. Authority, the referral threshold for fleets, period tolerance and the aggregate erosion warning are settings, and every change is versioned.
Primary and underlying excess policies with their endorsements, the schedule of underlying insurance, loss runs from each carrier, and the broker submission with its ACORD forms, large-loss narratives and requested layer. A document added later is read and the tower, the cracks and the note are re-checked.
The Agentic Solution Engine builds and deploys it from your requirements — your excess guidelines, rating plan and a sample of past submissions — and it goes live once every quality gate has passed. We will walk you through it on one of your own towers first.
Control gaps found before the quote, each linked to the answer or scan.
1-min demoCyber Accumulation ManagementExposed insureds and limit listed the day a vendor event breaks.
1-min demoPolicy Wording ComparisonEvery deviation from your form found in minutes, none bound without an approver.
1-min demoWe’ll run Excess Casualty Tower Analysis on one of your own excess submissions.