Personal Data in Bordereaux: A Field-by-Field Decision Guide
On this page
- Key Takeaways
- Why Does Personal Data in Bordereaux Need a Decision Rule?
- What Are the Four Options for Each Field?
- Which Bordereaux Fields Should You Send, Minimise, Pseudonymise or Exclude?
- How Do You Decide When a Reinsurer Asks for More?
- What Are the Pitfalls When Handling Personal Data in Bordereaux?
- Where Does Insurnest Fit, and Where Does It Not?
- What Should the DPO and COO Agree This Quarter?
- Key Terms
- Sources
- Frequently Asked Questions
Last reviewed: November 2026
Bordereaux built from policy and claims extracts often carry names, birth dates and medical details that no reinsurer asked for.
Once that file leaves your systems, you cannot take the data back.
Personal data in bordereaux should be limited, field by field, to what each reinsurer needs for pricing, reserving, claims control and settlement.
For each field, decide whether to send, minimise, pseudonymise or exclude it, and record why.
This guide describes practical options and cites the GDPR text. It is not legal advice, so confirm your obligations with your DPO or legal counsel.
Key Takeaways
- Start from purpose, not from the extract. Each field should map to a reason the reinsurer needs it.
- Four options cover almost every field. Send, minimise, pseudonymise or exclude.
- Pseudonymised is still personal. Data you can re-link remains personal data under the GDPR.
- Free text is the biggest leak. Claim notes often hold health and family details.
Why Does Personal Data in Bordereaux Need a Decision Rule?
Personal data in bordereaux needs a decision rule because bordereaux are usually built by exporting whatever the policy or claims system holds.
Without a rule, the default is "everything". Yet reinsurers mainly need risk and money figures, and identity rarely matters for pricing or accounting.
Which bordereaux carry the most personal data?
The bordereaux that carry the most personal data are life, health, personal accident and group life claims bordereaux, plus motor and liability claims listings.
These often include dates of birth, causes of death or diagnoses and claimant names.
Risk bordereaux for individual life or group business can also hold full member census data.
What does the GDPR actually say?
The GDPR sets principles rather than a bordereaux rulebook.
Article 5(1)(c) says personal data must be "adequate, relevant and limited to what is necessary" for the purposes of processing.
Article 9(1) treats "data concerning health" as a special category, which can only be processed under stricter conditions.
Recital 26 adds that pseudonymised data, which could be attributed to a person using additional information, is still information on an identifiable person.
Other jurisdictions have their own laws, so check every rule in this guide locally.
What Are the Four Options for Each Field?
The four options for each field are send, minimise, pseudonymise and exclude. Each one trades the reinsurer's analytical needs against privacy risk.
| Option | What it means | Use it when | Trade-off | Pitfall |
|---|---|---|---|---|
| Send | Field shared as held | The value itself drives price, reserve or payment | Most useful to the reinsurer | Sending "just in case" |
| Minimise | Field generalised, such as birth date to age band | The reinsurer needs the pattern, not the exact value | Less detail for analysis | Bands too narrow to protect anyone |
| Pseudonymise | Identifier replaced by a code only the cedent can reverse | Records must link across periods or files | Key management work | Treating it as anonymous |
| Exclude | Field removed from the bordereau | No reinsurer purpose exists | None for the reinsurer | Removing a field settlement needs |
When is pseudonymisation better than minimisation?
Pseudonymisation is better than minimisation when the reinsurer must follow the same policy or claim over time, such as claim development across monthly bordereaux.
A consistent code lets records link up without a name.
Minimisation suits fields used only in aggregate, such as age or region.
Why is pseudonymised data still personal data?
Pseudonymised data is still personal data because someone holds the key that can re-link it.
The EDPB's Guidelines 01/2025 on Pseudonymisation were adopted for consultation in January 2025.
They say this holds even when the data and the key sit with different parties.
So a pseudonymised bordereau still needs a lawful basis, proper security and a clear data sharing agreement.
Which Bordereaux Fields Should You Send, Minimise, Pseudonymise or Exclude?
The field-by-field table below is a starting default for claims and risk bordereaux. Adjust it to the treaty, the class of business and your own legal advice.
| Field | Reinsurer purpose | Default decision | How |
|---|---|---|---|
| Policy number | Linking premium and claims records | Pseudonymise | Consistent code held by the cedent |
| Claim reference | Tracking claim development | Pseudonymise | Same code across every bordereau |
| Insured or claimant name | Rarely needed | Exclude | Share only for a named large claim |
| Date of birth | Age-based pricing and reserving | Minimise | Age at inception or five-year band |
| Gender | Life and health pricing | Send only if rated | Exclude where not a rating factor |
| Full address | Accumulation and exposure | Minimise | Postcode district or zone |
| National ID or tax number | None in normal reporting | Exclude | Never in a bordereau |
| Sum insured, premium, claim amounts | Pricing, accounting, settlement | Send | Core financial fields |
| Dates of loss, notification, payment | Reserving and settlement | Send | Keep as dates |
| Cause of loss or diagnosis | Pricing and claims review | Minimise | Coded category, not free text |
| Free-text claim notes | Occasional claims review | Exclude | Replace with codes |
| Payee bank details | None for the reinsurer | Exclude | Stay with the cedent |
| Medical reports | Large or disputed claims only | Exclude from bordereau | Share case by case, securely |
How should free-text claim notes be handled?
Free-text claim notes should be replaced with coded fields before the bordereau is built.
Free text often holds diagnoses, family details and names that no banding rule will catch.
If a reinsurer needs narrative on a large claim, it can be shared for that claim alone.
What changes for large or disputed claims?
Large or disputed claims change the decision because a treaty may give the reinsurer claims control or review rights. Identity and documents may then be needed.
The answer is a separate, secure route for those claims, not a richer bordereau for every claim.
How Do You Decide When a Reinsurer Asks for More?
A reinsurer asking for more personal data in bordereaux should trigger a short, recorded decision rather than an automatic yes.
These five rules keep the answer consistent.
- Purpose: Which pricing, reserving, claims or accounting purpose needs the field?
- Contract: Does the treaty or a data sharing agreement cover it?
- Less identifying version: Would an age band, region or code serve the same purpose?
- Special category: Is it health data or another special category needing extra conditions?
- Record: Log the decision, the approver and a review date.
Who should approve a field decision?
A field decision should be approved by the DPO, with the reinsurance operations lead and the treaty owner confirming the purpose.
Claims leads should join for claims bordereaux. The COO owns the process across all treaties.
What Are the Pitfalls When Handling Personal Data in Bordereaux?
The pitfalls when handling personal data in bordereaux are mostly about false comfort. Each option can look safe while still exposing people.
Why can minimised data still identify someone?
Minimised data can still identify someone when few people share the same combination of values.
A group life claim with an age band, a small town and a date of death may point to one family.
Check small groups before sharing, and widen bands or regions where counts are low.
Three other pitfalls appear often enough to check for:
- Sending everything to avoid questions. It feels efficient but spreads data no one needs.
- One pseudonym code for every reinsurer. Shared codes let recipients link records together.
- Excluding a field settlement needs. The result is ad hoc emails carrying more data than the bordereau did.
Where Does Insurnest Fit, and Where Does It Not?
Insurnest's Bordereaux Calculation & Reconciliation does not mask, band or pseudonymise data.
The field decisions in this guide need to happen in your own data preparation before files are shared.
What the product does is calculate premium and claim allocation per reinsurer from participation shares.
It then reconciles those figures against bordereaux, master data, premium and claim registers, and settlement advices.
Its calculations centre on premiums, claims and participation shares, which this guide marks "send". Confirm that its inputs follow the same field standard.
For cross-border sharing, see our guide to data residency in reinsurance.
For wider controls, read about protecting PII and PHI in insurance.
What Should the DPO and COO Agree This Quarter?
The DPO and COO should agree one field standard for personal data in bordereaux and one exception route this quarter. Both can start with a spreadsheet.
- List every bordereau template sent to reinsurers and brokers.
- Apply the field table and record a decision for each field.
- Replace free-text fields with codes.
- Use separate pseudonym codes per recipient unless linkage is needed.
- Set up a secure, case-by-case route for large claim documents.
- Review the standard at each renewal, alongside bordereaux quality.
Key Terms
- Data minimisation: Limiting personal data to what is necessary for the purpose.
- Pseudonymisation: Replacing identifiers so data cannot be linked to a person without separate information.
- Special category data: Sensitive personal data, including health data, that needs extra conditions.
- Data sharing agreement: A contract setting out what data is shared, why and how it is protected.
- Claims control clause: A treaty term giving the reinsurer rights over handling of certain claims.
Sources
- EUR-Lex: Regulation (EU) 2016/679 (General Data Protection Regulation) (27 April 2016)
- European Data Protection Board: Guidelines 01/2025 on Pseudonymisation (adopted for consultation 16 January 2025)
- Insurnest: Bordereaux Calculation & Reconciliation
Frequently Asked Questions
What personal data appears in bordereaux?
Personal data in bordereaux often includes policyholder and claimant names, dates of birth, addresses, policy and claim references, causes of loss or death and free-text claim notes. Life, health, personal accident and group life bordereaux usually carry the most sensitive fields, including health details that need extra care.
Why should cedents minimise personal data sent to reinsurers?
Cedents should minimise personal data sent to reinsurers because the GDPR requires personal data to be adequate, relevant and limited to what is necessary for the purpose. Reinsurers mainly need risk and financial figures, so names and identifiers rarely justify the extra privacy risk.
How do you decide which bordereaux fields to share?
Deciding which bordereaux fields to share starts with the purpose each field serves, such as pricing, reserving, claims control or settlement. For each field, choose to send, minimise, pseudonymise or exclude it, check special category data, and record the decision with an approver and a review date.
Who should approve personal data decisions for bordereaux?
Personal data decisions for bordereaux should be approved by the data protection officer, with the reinsurance operations lead and the treaty owner confirming each purpose. Claims leads should join for claims bordereaux, and the COO should own the process so the same standard applies across treaties.
Is pseudonymised bordereaux data still personal data?
Pseudonymised bordereaux data is still personal data under the GDPR when it can be re-linked to a person using additional information, such as a key the cedent holds. The EDPB's draft pseudonymisation guidelines say this remains true even when the data and the key are held by different parties.
Can software mask personal data in bordereaux automatically?
Software can apply masking, banding and pseudonymisation rules to bordereaux, but the decisions about which fields to treat and how must come from the data protection officer and the business. Insurnest's bordereaux reconciliation product does not mask data, so these rules sit in your own data preparation.
What is the difference between minimisation and pseudonymisation?
Minimisation and pseudonymisation solve different problems. Minimisation reduces detail, such as turning a birth date into an age band. Pseudonymisation replaces an identifier with a consistent code that only the key holder can reverse, so records can still be linked across periods without names.
Should a reinsurer receive medical reports in a claims bordereau?
A reinsurer should normally not receive medical reports in a claims bordereau. Where a treaty gives claims control or review rights on large or disputed claims, documents can be shared for that claim alone through a secure route, with the purpose and recipient recorded.

Hitul Mistry
CEO, Insurnest
An InsurTech leader with more than a decade of experience across insurance and technology, focused on solving business problems with the help of technology. Has worked with brokers, insurance carriers, and reinsurance firms across the India, UAE, and US markets.
View LinkedIn profile →