Bordereaux management for MGAs: own the book

If you run an MGA, MGU, program administrator, or Lloyd’s coverholder, you already know what a bordereau is: a periodic schedule of written premium, claims, or exposure sent to a carrier or reinsurer under delegated authority. The practical problem is different. A bordereau is a report you file. A ledger is a record you can defend. Many MGAs have the first and not the second.

That gap matters because your largest variable income line, especially profit commission, is often calculated from a version of the book the carrier controls. If premium, claims, cash, producer commissions, and treaty-year accruals do not tie back to your own records, you are left arguing from spreadsheets instead of from a reconciled book. This article is about the reverse problem: how an MGA owns its own ledger even when every carrier wants its own bordereau format.

What a carrier does with your bordereau

On the carrier side, bordereaux management usually means ingesting the file, validating the fields, matching it to policies and claims, and using it to calculate loss ratio, commission, and profit commission. That is why many products in this market focus on extraction, workflow, and carrier reporting. They help the recipient consume your file.

For the MGA, that is not enough. You still need to prove what was written, what was bound, what was remitted, what commission was paid, what was returned after cancellation, and what claims have been incurred or developed. If the carrier’s file is the only place those numbers live in a joined form, then the carrier owns the version that counts.

Where reconciliation breaks

The breakage is usually quiet, not dramatic. The same policy can appear under slightly different identifiers in premium and claims files. A mid-term endorsement can change the premium base, but the claims bordereau still points at the old policy number. A return premium after cancellation may be posted in cash but never reflected in the producer commission ledger. A claim may arrive in a different currency or with a class code that does not map cleanly to the premium file.

These are not edge cases. They are the normal result of delegated authority reporting across multiple carrier templates, different policy administration systems, and different operational teams. If you rely on spreadsheet joins, the book can look close enough to move money while still being wrong enough to distort profit commission.

The most common failure points are:

• Policy numbers formatted differently in premium and claims files
• Endorsements and mid-term cancellations that change written premium after binding
• Return premium and chargebacks that are settled in cash but not in the producer ledger
• Multi-currency reporting with inconsistent exchange dates
• Coverage or class-code mappings that vary by carrier template
• Period cut-offs, especially around treaty year boundaries

The three ledgers an MGA should own

If you want an independent record, you need three linked ledgers, not one master spreadsheet. Each answers a different question, and each should survive carrier-specific reporting formats.

1. Written and bound premium

This is the book as sold and bound, not just the amounts remitted. It should include policy number, insured name, producer, effective and expiration dates, coverage class, carrier or program, treaty year, written premium, bound premium, endorsements, cancellations, and currency. If you do not preserve the history of endorsements and cancellations, you will not be able to explain why the carrier’s premium bordereau differs from your own.

2. Cash, commission, and producer splits

This ledger should show what the carrier remitted, what commission was earned, what was paid to each producer, and what was clawed back through chargebacks or return premium. In an MGA with a retail network, this is where the operational truth lives. A policy can be bound, then cancelled, then reinstated, and the producer still needs the correct net commission after all three events.

3. Claims and loss development

This ledger should track claim number, policy number, date of loss, report date, incurred amount, paid amount, case reserve, recovery, claims handler, and treaty year. If claims are delegated to a TPA, or third-party administrator, you still need the ability to connect claims movement to the premium book and to the period in which profit commission is measured.

Each ledger can exist on its own. The point is that they should also be joinable. If one file uses a policy number with a leading zero and another drops it, or one file keys off the endorsement number and another does not, the join has to be normalized in your system, not improvised at month-end.

Profit commission is an accrual problem before it is a payment problem

Profit commission is the payment, often from a carrier or syndicate, that depends on the profitability of the program or treaty year. The exact formula varies, but the ingredients are usually some combination of earned premium, claims incurred, expenses, ceding commission, and agreed profit share. The key point is that you should model the accrual yourself, even if the carrier ultimately makes the payment.

That means you need to know the carrier’s formula, the thresholds, the exclusions, and the accounting basis. Is loss ratio measured on incurred claims or paid claims? Are IBNR reserves, incurred but not reported, included? Are there sliding scales? Is profit commission measured by treaty year, underwriting year, or calendar year? Does the calculation reset after a portfolio transfer or termination?

A simple internal model does not need to be perfect to be useful. It needs to be reproducible. If the carrier says the accrual is 8.2 percent and your book says 6.9 percent, you should be able to identify whether the gap comes from a missing endorsement, an unposted return premium, a claims reserve movement, or a period mismatch.

The practical test is this: can you explain your own accrual from source records without asking the carrier to rebuild it for you? If not, you do not own the number, even if it appears on your income statement.

How to treat carrier templates as configuration

Many MGAs start with seven separate workbook habits because each carrier asks for a different bordereau layout. That works until the eighth carrier arrives, or until one carrier changes its template and every downstream formula breaks.

A better approach is to treat carrier templates as configuration, not as separate operating systems. The source data should stay in your own canonical model. Each carrier’s format should be a mapping layer that translates your record into the fields that carrier wants.

The honest test is simple: what does it cost you to add the eighth carrier?

• If the answer is another custom workbook, the process is still manual.
• If the answer is a mapping update with no change to the underlying ledger, you have a system.
• If the answer involves re-keying the same policies in several places, the work is being duplicated instead of normalized.

This is where bordereaux management software, policy administration systems, and carrier portals need to be separated in your thinking. The policy admin system captures the transaction. The carrier portal receives the report. The thin reconciliation layer in the middle is what keeps the MGA’s own book defensible.

Build versus buy: what belongs where

You do not need to build everything. You do need to know which layer must stay under your control.

• Policy administration system: keep this as the system of record for binding, endorsements, cancellations, producers, and policy history.
• Bordereaux vendor or MGA software: useful for template generation, file ingestion, and carrier-specific reporting, especially when you have many programs.
• Carrier portal: accept that this is the carrier’s view, not your ledger.
• Reconciliation layer: this should remain yours. It ties premium, cash, commission, and claims together and preserves the history needed to defend every number.

If you buy only carrier-facing tooling, you may reduce file preparation time while leaving the core problem untouched. You will still have to answer the question that matters most to the owner: what do we believe the book is, independent of the carrier’s statement?

What to keep when a program ends or moves

Program termination and portfolio transfer are where weak records become expensive. Once a program winds down, you may still have open claims, return premium, late endorsements, chargebacks, and a profit commission calculation that closes much later than the last bordereau filing.

At minimum, keep the full transaction history for every policy, all carrier submissions, all accepted remittances, all commission statements, all producer settlements, and all claims movements by treaty year. You also need the mapping logic that explains how old carrier fields were translated into your canonical record. Without that, a future audit or dispute becomes a manual forensic exercise.

The point is not archival for its own sake. It is continuity. If a portfolio transfers, you should still be able to reproduce the book as it stood on the transfer date and reconcile any later movements against it.

A useful way to evaluate your current setup

Ask three questions about each program:

• Can we reproduce written premium, cash, commission, and claims from our own source records without using the carrier’s statement as the starting point?
• Can we explain every difference between our book and the carrier’s by policy, endorsement, claim, or timing?
• Can we model profit commission ourselves, to the same treaty-year basis the carrier uses?

If the answer to any of those is no, the problem is not just bordereaux production. It is bordereaux reconciliation, and it belongs in the core operating model rather than in the monthly reporting scramble.

Why this matters now

Delegated authority business is growing, and so is the pressure for cleaner reporting. TMPAA’s State of Program Business report, as cited by Sikich, put program business at 79.07 billion dollars in 2022, up from 53.8 billion in 2020. That kind of growth does not reduce the reporting burden. It multiplies it. The more programs you add, the more often the gaps between premium, claims, and cash show up.

There is also a practical difference between AI extraction and reconciliation. Extraction helps a carrier ingest your file faster. It does not solve the MGA’s need to own a reconciled book. If your margin depends on data you cannot independently recreate, speed alone will not fix the risk.

For operators who want a broader example of reconciliation-shaped finance work, Better Software has written about finance systems in fintech, including its fintech industry work and a fintech software case study. The specific domain is different, but the core problem is the same: keep the internal record strong enough to challenge a counterparty statement.

What is a bordereau report in insurance?

A bordereau report is a periodic report, usually a spreadsheet or structured file, that lists the premium written, claims reported, exposure bound, or related data for delegated authority business. In practice, carriers and reinsurers use it to monitor the book and calculate what is owed or accrued.

For an MGA, the useful question is not only what a bordereau report is, but what it is not. It is not, by itself, your independent ledger. You still need a record that can be reconciled back to source transactions and defended if the carrier’s version differs.

Bordereau vs bordereaux

Bordereau is singular. Bordereaux is plural. In insurance, people often use “bordereaux management” to describe the process of producing, validating, and reconciling those recurring reports across programs and carriers.

What goes in a premium bordereau versus a claims bordereau?

A premium bordereau usually contains policy and financial data such as policy number, insured, effective dates, premium written, endorsements, cancellations, commission, taxes or fees, producer, carrier, and currency. A claims bordereau usually contains claim-level data such as claim number, policy number, date of loss, date reported, paid amount, reserve, incurred amount, recoveries, and status.

The challenge is that the two files are often built by different teams and keyed differently. That is why the reconciliation layer matters. Premium and claims data have to meet on the same normalized identifiers if you want to understand loss ratio and profit commission on your own terms.

If you are choosing between another bordereau tool and a reconciliation layer you control, the simplest question is this: which one helps you own the book that your income depends on? Start there.