Claims bordereaux monitoring is the regular review of the claims listings a reinsurer, captive or programme carrier receives from cedents, coverholders and TPAs. The reviewer compares each listing with the last one and picks out the claims that need attention: new large losses, reserve movements, reopened claims and claims that match a watch-list of hazard terms.
Some teams call the listing a claims bordereau. Some call it a loss run. In parts of the reinsurance market the bulk version is called a claims laundry list. The work is the same.
Anyone who carries risk on business they do not handle themselves:
Reviewing what cedents report under treaties.
Overseeing claims handled by a TPA.
Overseeing claims handled by an MGA or TPA under delegated authority.
Overseeing coverholder claims.
They do it for four reasons. Contracts usually require notice of any loss above a set amount. Reserves have to be adequate, and a pattern of late increases is an early warning. Some claims threaten a layer or an aggregate long before any single row looks alarming. And paying for claims that are duplicated, miscoded or outside the contract is money lost.
Eight things. The first four need this period compared with the last.
Claims appearing for the first time above your notification threshold.
Increases, and sudden releases, beyond a tolerance you set. Total incurred is the usual measure.
Closed last period, open again now.
On the last listing, missing from this one, with no closure recorded.
Narrative language from your own list of hazards: fatality, amputation, traumatic brain injury, abuse, a named product. Often misspelled.
Jurisdictions and courts your claims team treats as high severity.
A long gap between the date of loss and the date first reported.
Totals and headers counted as claims; one claim listed twice.
A watch-list here means your own vocabulary of concern. It is not sanctions or denied-party screening, which is a separate compliance check.
In a spreadsheet, by one experienced person. Each sender uses its own layout, so the first job is lining the columns up. Then a lookup on claim number against last month’s file, a formula for the change in incurred, a filter on the large ones, and a read through the narrative column for anything worrying.
It works at small volume. It fails in three ways. Claim numbers change format between periods, so the lookup misses. Narrative review is slow, so it gets skimmed. And the method lives in one person’s head, which is a risk of its own.
Match each claim to itself in the prior listing, compute the change in paid, outstanding and incurred, and flag anything beyond your tolerance. The hard part is the matching, because senders re-key claim numbers, split claims and merge them. Set the tolerance as an amount and a percentage together, so a large claim moving slightly and a small claim tripling are both seen.
For illustration, here is how that looks on a typical listing. A listing has 1,840 rows. Against last month, 61 are new, 9 have gone, and 212 changed in incurred. Applying a tolerance of 50,000 or 25 percent leaves 14 movements. Three of the new claims exceed a 250,000 notification threshold. Watch-list terms hit 11 narratives. After duplicates, about 25 claims need a person. That is the purpose of monitoring: 1,840 rows reduced to 25 worth reading.
No product category is named for it yet. Four kinds of software each cover part of the job.
| Kind of software | What it covers | Examples |
|---|---|---|
| Bordereaux management platforms | Ingest, standardise and validate claims bordereaux alongside premium and risk | VIPR Intrali, Charles Taylor Tide, distriBind, Vellum |
| TPA and risk-management claims systems | Hold the claims themselves, with reserve history and bordereau reports out | VCA Software; risk-management systems such as Origami Risk |
| Document AI for reinsurance claims audit | Reads cedent claim files and bordereaux together to find leakage and duplicates | Nomad Data |
| Insurance AI Analysts | Read the listing in any layout, compare it with the prior version, and screen it against your watch-list | Brisc |
Bordereaux platforms give you clean, standard data; the monitoring rules are then yours to build. A TPA’s claims system shows reserve history only for claims that TPA handles, which does not help the reinsurer receiving listings from twenty of them. Nomad Data cross-references cedent claim files, payment records, bordereaux and treaty documents to find duplicate payments and misallocations, which is audit of a cedent more than routine monthly monitoring. We have not found another vendor that describes version comparison and watch-list screening of claims listings as one product.
Brisc covers this with two jobs.
The Bordereaux Analyst reads the file in the layout the sender uses and compares every row with the last file you accepted. It flags reserve movements, new large losses, reopened claims and claims that vanished without being closed, with the before and after shown as evidence.
The Claims Analyst screens every row against your own hazard vocabulary and problem venues. Matching runs at three levels, exact, synonym and fuzzy, so a misspelled term is caught and a lookalike word is not. Each claim gets a risk band and each flag shows why it fired. The screening engine is deterministic. AI adjudicates borderline matches with written reasoning and stops for a person when unsure. The Analyst also tracks each insured’s total incurred loss against their layer attachment point, banded below, approaching or reached.
It does not set or propose reserve figures. It does not handle or pay claims. It does not run sanctions screening. And the thresholds and vocabulary are yours: Brisc does not decide what counts as a large loss for your book.
By comparing each period’s bordereau with the last and flagging new claims above the notification threshold in the treaty, plus existing claims whose incurred has moved past it. Most do this in spreadsheets. Software reads every cedent’s layout, matches claims across periods and produces the exception list.
A TPA’s own claims system tracks reserve history for its claims. To track movements across listings from many TPAs you need something that reads each listing and compares versions: a bordereaux platform with rules you configure, or an Analyst such as Brisc’s.
Yes. Brisc’s Claims Analyst screens claim narratives against your own list of hazard terms and venues and explains every flag. This is a watch-list of severity signals, not sanctions screening.
The market’s name for a bulk claims register sent by a cedent or broker with a renewal, a treaty discussion or periodic reporting. See how to review a claims laundry list.
Start from the notification clause in the contract. Consider an internal trigger set below that figure, so a claim is seen before it becomes reportable.
30 minutes, no slide deck. Tell us where the volume hurts and we show you the Analyst on it. Bring a bordereau, statement or submission if you have one; you don’t need to.
Book the walkthroughNot ready for a call? Get the MGA operations report →