Topic hub

Treaty reinsurance: quota share, surplus, and excess of loss operations

Treaty is automatic capacity for a class. The ops work is keeping bordereaux, wordings, and technical price inputs on the same year and the same definitions. Pricing essays without a cited attachment are not ops.

Treaty reinsurance is automatic capacity for a defined class. The reinsurer agrees in advance to accept risks that meet the treaty. The cedent does not offer each policy for a separate yes. That sentence is the legal idea. The operations idea is harder: keep the wording, the bordereaux, the pricing inputs, and the accounts on the same year and the same definitions.

This hub treats quota share, surplus, and excess of loss as ops problems, not as textbook diagrams. If you need the one-paragraph definition, use what is treaty reinsurance? and the treaty reinsurance glossary entry. If you need facultative — optional, risk by risk — use the facultative hub. Mixing the two in one inbox is how a facultative limit gets posted to a treaty year.

Pricing essays without a cited attachment are not ops. The treaty pricing guide is the pricing-prep sibling. This page is the operating system around that prep: wordings, bordereaux, inputs, accounting.

Quota share as an operations problem

Quota share is proportional. A fixed percentage of premium and of loss moves to the reinsurer, usually against a ceding commission. In a 30 percent quota share, 30 percent of the subject book is meant to move, on the definitions in the wording, not on the definitions in last quarter's spreadsheet header.

The ops failures are specific.

The bordereau includes a class the treaty excludes. The file is booked anyway because the total looked close to last quarter. The ceding commission in the accounts uses a rate from the placing email, not the rate in the wording, and nobody can show the clause. Profit commission is calculated on a loss ratio that used a different earned premium basis than the contract. Currency is mixed without a rate as-at. A special acceptance sits in a mailbox and never reaches the bordereau filter.

Quota share does not fail because someone forgot the meaning of "30 percent." It fails because 30 percent of the wrong rows is a different treaty. Extraction has to pull the share, the class, the territory, the commission basis, and the exclusions from the wording with spans, then validate inbound rows against those spans. The Treaty Agent is the wording side of that loop. The Bordereaux automation guide is the file side. What is quota share reinsurance? is the definition you can cite without this page.

Special acceptances are quota-share ops, not folklore. A risk that sits outside class or occupancy is either declined, or accepted in writing, or it leaks into the bordereau and becomes a silent extra. The acceptance is a document. It needs a span, a period, and a join to the policy reference. A mailbox yes without a file is how you discover the extra occupancy at the loss.

Year of account is the other quiet failure. Premium written in December and reported in January is either this treaty or next, according to the wording, not according to when the spreadsheet was emailed. If the inbound file mixes years without a column you can cite, stop. Do not split rows by instinct.

Surplus as an operations problem

Surplus is still proportional, but participation varies by risk size. The cedent's line is a retention. Capacity above that line cedes, in lines, up to the surplus limit. Small risks cede little. Large risks cede more. The legal idea is flexible capacity. The ops idea is that every risk now needs a cited sum insured and a cited line, or the cession is a guess.

A surplus bordereau that omits the underlying limit cannot be checked against surplus capacity. Flag it. Do not impute a line from the average of the file. Do not assume this quarter's "SI" column is last quarter's "TIV" column. If the cedent reports a risk above the surplus limit, that is an exception: facultative leftover, error, or a special acceptance. It is not a rounding item.

Line definition has to come from the wording. Some surplus treaties define the line as a currency amount. Some define it as a percentage of a maximum. Some change the line by occupancy. If the extracted line and the bordereau implied line disagree, you have a conflict. Show both. Accountants should not invent a third line to make the cession footprint look smooth.

Surplus also collides with facultative when a risk is thrown out of the treaty and offered as a pack. That is two files. The treaty exception list should cite the surplus wording. The facultative pack should not be asked to explain the treaty. See the facultative hub for the pack. Keep surplus administration here.

Excess of loss as an operations problem

Excess of loss is non-proportional. The reinsurer responds when a loss, or an event, or an aggregate, exceeds an attachment, up to a limit. Per-risk XoL, cat XoL, and aggregate XoL are different attachments and different event definitions. They are not interchangeable just because the slide said XoL.

The ops failures cluster around four clauses people paraphrase and then cannot find.

Attachment: where cover starts. If the wording says USD 10,000,000 any one risk and the claims bordereau attaches a loss at USD 8,000,000 because someone used a different deductible, the recovery is wrong. What is an attachment point in reinsurance? is the definition. The field on the pack must cite the page, not the placing slip's shorthand if the signed wording differs.

Limit: where cover stops. Exhaustion is an accounting event. If paid plus outstanding on an event exceeds the cited limit, the next question is reinstatement, not a silent extra recovery.

Reinstatement: whether capacity comes back, free or paid, automatic or not, once or more. What is a reinstatement clause in reinsurance? is the short answer. The ops answer is a formula with a span. "Two reinstatements" in a summary is not a formula. The premium for a paid reinstatement is in the wording or it is a gap.

Hours and event definition: how 72 hours or 168 hours cuts a weather event, and whether two storms are one event. If the hours clause is missing from the extracted pack, you do not have a cat XoL you can administer. You have a heading.

What is the difference between quota share and excess of loss? is the comparison page. Use it when a new joiner conflates share with attachment. Use this page when the files disagree with the wording.

Wordings

The wording is the system of record for the treaty, not the broker summary, not the pricing memo, not the spreadsheet named Final_v7. The Treaty Agent pulls attachment, limits, exclusions, reinstatement language, territory, class, inception, and claims procedures from the PDF so accountants are not re-typing clauses.

Extraction here follows the same field contract as facultative. Each term is a field with a span. Buried exclusions stay as fields, not as a bullet in a summary titled "key terms." Manuscript amendments and endorsements are documents in the pack. If endorsement 3 changes the hours clause, the live hours field cites endorsement 3, and the original wording span is either superseded with a record of that fact or left as a conflict until someone confirms which text governs.

Renewal comparison is an ops job: last year's cited attachment versus this year's cited attachment, side by side, with both pages. A model that says "broadly similar" has not compared the treaty. A table of fields with spans has. This page will not invent a percentage of clauses that usually change. Operators already know where their wordings drift. Software should show the drift with evidence.

Do not summarise a wording into a chat window and call it a treaty abstract. Summaries drop the clause that bites at the loss. See the provenance hub if you need that argument at length. Here the rule is shorter: no span, no term in the administration system.

Bordereaux

Bordereaux are how the automatic treaty becomes rows. Premium, claims, commission. Monthly or quarterly. Every cedent's layout differs. Column names drift. Totals in the covering email disagree with the sheet. The inbound job is map, validate against wording, exception, then book. The outbound job, on the cedent side, is produce a file that could survive that inbound job.

Reconciliation is not a vibe. It is row-level.

• Does this policy fall inside class, territory, and period as cited in the wording?
• Does the share or the attachment match the treaty, not the previous file's mapping?
• Do claims references exist on the premium bordereau for that year of account?
• Do totals in the sheet match a cited total row, or is the covering email inventing a number?
• Did the cedent restate prior quarters, and is there a bridge?

The bordereaux automation guide is the long guide. This hub only insists on the order: wording spans first, file cells second, booking third. Guessed totals do not close the quarter. They delay the dispute.

A worked pattern, without inventing a cedent. The wording cites a quota share, class property, territory excluding a named flood zone, inception 1 January. The inbound premium sheet has a column the mapper used last quarter as "class" that this quarter holds a branch code. Totals in the covering email match a subtotal row, not the grand total. Two policies sit in the excluded zone. Those two rows are exceptions with spans into the wording and cells into the sheet. The covering-email total is a conflict against the grand total cell. None of that is solved by a paragraph that says the bordereau is broadly in line.

Claims bordereaux add event date versus report date versus period. A loss reported in this quarter for an event in last year's period is either prior-year development or a new claim, according to the wording's basis, not according to the sheet tab name. If event date is empty, attachment to period cannot be shown. That is a gap, not a booking.

Pricing inputs

Treaty pricing consumes documents. It should not consume unsourced summaries of those documents. Subject premium by year. Loss history with large losses identified. Exposure and sums insured. Development for long-tail. CAT model narratives if the class needs them, as documents, not as a pasted paragraph. Current structure: attachment, limit, reinstatement, hours, event definition — cited from the wording you intend to bind, which may not be last year's.

The Pricing Agent assembles those inputs so a human can set technical price. The Actuarial Agent is the report-and-assumption side of the same pile: what the pricing indication actually said, which triangle, which as-at. Neither agent is the pricing committee. The illustration of scaled prep, labelled as an illustration, is the cedent treaty pricing case. The method note is the treaty pricing guide.

Technical price versus quoted price is the same split as on the reinsurer hub. Burning cost that used an attachment the wording does not contain is not burning cost. Exposure rating that used a TIV distribution averaged from a conflicted SOV is not exposure rating. Show the conflict. Let the actuary choose. Do not pre-choose in the extract.

This page will not quote a hours-saved figure for pricing prep. If a number appears on a case-study illustration, it is labelled as such and is not a promise on this hub.

Accounting

Treaty accounting is the booking of premium, commission, recoveries, reinstatement premium, and IBNR movements against the cited contract. The ops failures look like claims leakage under another name: the wrong year of account, the wrong share, a recovery below attachment, a reinstatement premium missed because the clause was in an endorsement nobody extracted.

Close is not "the bordereau total was close enough." Close is: exceptions listed, spans for the terms you applied, and a human sign-off on the residual. If prior-year development arrives as a restatement with no bridge, that is a gap in the accounts pack. If currency converts with a rate from a website on the day of booking rather than the rate basis in the wording, that is a conflict waiting for audit.

IBNR language belongs with triangles and as-at dates, not with a model sentence that "reserves look adequate." The claims bordereau quality is the input. Garbage in, fluent IBNR out, is still garbage.

Reinstatement premium is accounting, not a pricing footnote. When a layer exhausts and the wording requires a paid reinstatement, the premium due is a calculated field that still needs the clause span and the loss that exhausted the limit. Booking a round number because "we always reinstate at 100 percent" is the same failure as filling a missing hours clause: it is a completion, not a citation. If the clause is free reinstatement, the accounts should show capacity restored and no extra premium, with the same span.

Profit commission, sliding scale, and loss corridors are formula fields. Extract the formula. Apply it to cited premium and cited losses. If the bordereau used a different earned basis than the clause, the profit commission is a conflict, not a true-up you can bury in cash.

Children of this hub

Use the children. Do not fork new definitions on this page.

• What is treaty reinsurance?
• Treaty reinsurance in the glossary
• What is quota share reinsurance?
• Quota share versus excess of loss
• Attachment point
• Reinstatement clause
• Treaty Agent
• Pricing Agent
• Actuarial Agent
• Treaty pricing guide
• Bordereaux automation guide
• Treaty pricing illustration
• Facultative reinsurance when the file is a risk, not a class

Treaty is standing capacity. Facultative is a pack per risk. Bordereaux are the rows. Wordings are the terms. Pricing is cited inputs plus a human. Accounting is the book against those terms. If a vendor offers a treaty chatbot that will not show the attachment span, you do not have treaty operations software. You have a summary of a PDF.