Reporting utility · Bordereaux exposure check

Your bordereau was accepted. Your exposure can still be wrong.

Format validation tells you the file was well formed. It does not tell you whether the exposure inside it describes what was genuinely on risk. We review a sample of your risk and premium bordereaux and show you which figures can be reconstructed from your own movement history — and which cannot.

Book a demo See how bordereaux are produced

Run as a working session on a sample you provide · anonymised extracts accepted · not an audit or a compliance opinion

Why an accepted file can still misstate exposure

The Lloyd's Coverholder Reporting Standards model movements properly. Transaction type covers new business, renewal, endorsement, cancellation and reinstatement, with reasons and instalment handling. That is not where this breaks.

It breaks because sum insured, limit of indemnity, deductible and excess are risk-level fields, while each row you submit is a transaction. When a policy administration system overwrites the risk record as an endorsement is processed instead of versioning it, the row carries the value as it stands today rather than the value in force on the transaction date. Every validation rule passes. Every column total ties. And recovering the real exposure means recomputing each row against a movement history the system never kept.

One commercial property risk over one reporting year: movements that occurred against the exposure a snapshot bordereau reports.
Movement	Date	Actually on risk	Reported in a snapshot bordereau
New business	15 Jan	£2,000,000	£5,000,000The current risk value, not the value at inception
Endorsement (MTA)	3 Mar	£5,000,000	£5,000,000Correct on this row only, and correct by coincidence
Cancellation	20 Jun	nil from this date	£5,000,000No indication in the row that cover ceased mid-year
Reported exposure at year end	31 Dec	—	£15,000,000Three rows at the present value, summed. Overstated, and unusable for aggregate.
Time-weighted position	31 Dec	£1,647,000	£2m for 47 days, £5m for 109 days, nil for the remaining 209. Recoverable only from the movement history.

Worked illustration on a single risk, shown so the arithmetic is checkable. The same distortion compounds across a book, and it runs in both directions — exposure is understated wherever cover increased late in the period.

What the review checks

Fifteen checks in three groups. Some are mapping problems you can fix this month. Some are structural, and no reporting layer can fix them because the movement was never recorded. The review tells you which is which.

Exposure reconstruction

Whether the exposure you reported can be rebuilt from the movements behind it.

Risk-level values — sum insured, limit of indemnity, deductible, excess — reported at their present value rather than the value in force at each transaction date.
Endorsement rows carrying no recoverable pre-endorsement position, so the change in exposure cannot be derived from the bordereau alone.
Cancellations and lapses reported without an effective date, or reported in a way that leaves the risk counted for the full period.
Reinstatements that re-add exposure already counted, producing double-counting across periods.
Period-end aggregate that cannot be time-weighted, because the row set describes states rather than a movement sequence.

Premium and cash integrity

The breaks that strand settlement and age your premium receivable.

Instalment transactions reported as full premium, or full premium reported where only an instalment was collected.
Rate of exchange applied at the reporting date rather than the transaction or settlement date, so the net premium to London will not tie.
Coverholder commission, sub-producer commission and Lloyd's brokerage netted inconsistently between the coverholder block and the block the London broker completes.
Terrorism premium and other separately reportable components folded into gross written premium.
Premium and claims submissions carrying the same field with different values, which the collect-once principle is meant to prevent but which routinely happens where the two travel to different bureaux.

Structure and mapping

Where the standard's shape and a spreadsheet's shape disagree.

Repeating tax and levy blocks — the standard allows several per transaction — flattened into a single set of columns, losing jurisdiction, basis or rate.
Repeating intermediary blocks collapsed, so the distribution chain cannot be reconstructed.
Claims with several movements in one reporting period aggregated on a basis that was never agreed in writing with the syndicate.
Free-text where the standard expects a coded value, and coded values carried at the wrong granularity.
Territory overlays not applied — the additional and redefined fields that apply to risks in the US, Australia, Hong Kong, Singapore, South Africa, Spain, the EU and through Lloyd's Brussels.

Why this lands on the finance director

Settlement is gated on reporting

Premium and claims settlement moves when the reporting is clean. A bordereau that cannot be reconciled converts directly into cash sitting unallocated and premium receivable ageing past where anyone is comfortable.

Restatement risk at year end

Exposure figures that cannot be reconstructed are exposure figures that cannot be defended when someone asks how they were derived — and correcting them late means restating numbers that have already been reported.

Decisions made on the number

Aggregate returns, reinsurance structures and capacity conversations all rest on reported exposure. An overstated aggregate is not a reporting inconvenience; it is a number that shaped a commercial decision.

How it runs

You send a sample

A few reporting periods of risk and premium bordereaux across one or two binders, with the matching policy or transaction extract where you have it. Anonymised extracts are fine.

We run the checks with you

A working session rather than a report thrown over the wall. We walk the findings row by row so you can see the arithmetic and challenge it, and so the ones that are actually your syndicate's agreed treatment get discarded.

You keep the findings

A written list of the specific problems found, where each one originates, which are fixable in mapping or process and which are caused by the underlying system not recording the movement at all.

On your data

The checks operate on dates, amounts, transaction types and file structure, so anonymised or pseudonymised extracts work for the review. Samples are handled under a written agreement covering scope, handling and deletion, agreed with you before anything is shared.

Frequently Asked Questions

What is a bordereaux exposure check?

A guided review of a sample of your existing risk and premium bordereaux that tests whether the exposure reported in them can be reconstructed from the underlying policy movement history. It looks past format validation — which your syndicate already checks — at whether the numbers describe what was actually on risk.

My bordereaux are accepted by the syndicate. Why would anything be wrong?

Acceptance tests format and completeness, not whether risk-level values were correct at the date of each transaction. Sum insured, limit of indemnity, deductible and excess describe the risk, while each row describes a transaction. If a policy administration system overwrites the risk record when an endorsement is processed instead of versioning it, rows carry present-day values rather than the values in force at the time. The file validates and the exposure is still overstated.

What do I need to provide?

A sample of risk and premium bordereaux — typically a few reporting periods across one or two binders — and, where available, the corresponding policy or transaction extract. Anonymised or pseudonymised extracts are fine; the checks operate on dates, amounts, transaction types and structure rather than on policyholder identity.

Is this a self-service tool?

Not today. The review is run as a working session with Facio against your sample, and the checks are executed by our team rather than by you in a browser. If enough coverholders want it as something they can run themselves, self-onboarding is the next step — telling us you want that is genuinely useful.

Does this replace a Lloyd's audit or a compliance review?

No. It is an operational data review, not an audit, an assurance engagement or a compliance opinion. It has no standing with Lloyd's, a syndicate or a regulator, and it does not determine whether you meet your reporting obligations. Findings should be interpreted alongside your own binder terms and, where the consequences are material, reviewed by an appropriately qualified professional.

Do I have to move to Facio to act on the findings?

No. The output is a list of specific, reproducible problems in your own data, and most of them can be addressed inside whatever system you run today — some with a mapping change, some with a process change. Where the cause is that the underlying system never recorded the movement, no reporting layer can recover it, and that is a platform question rather than a reporting one.

Bring one binder and a few periods.

The fastest way to find out whether this applies to you is to run it on something real. Book a demo and we will walk the checks against your own sample rather than a slide.

Book a demo Coverholder software guide

This review is an operational data check, not an audit, assurance engagement or compliance opinion, and it carries no standing with Lloyd's, a syndicate or a regulator. Where findings have material consequences, have them reviewed by an appropriately qualified professional. See the other Facio utilities.