18.10 Bureau submission file format

Lenders must submit monthly data on all borrowers and loans to all 4 CICs per CICRA. This page is the format reference.

Authoritative source: CIBIL TransUnion publishes the Consumer Credit Information Bureau (CCIB) format specifications; CRIF / Experian / Equifax follow similar / aligned formats. Always verify the current version with the CIC.

Submission cycle

Frequency: Monthly. NPCI/RBI direction permits weekly; many lenders submit monthly. Best practice is weekly to keep bureau data fresh.
Deadline: Typically by 15th of the next month for the previous month’s data.
Submission: Via CIC’s SFTP / portal.
Acknowledgement: CIC returns ack file with success / errors per record.

File types

File	Direction	Purpose
Header file	Lender → CIC	Metadata about the submission
Member submission	Lender → CIC	Loan + borrower data
Acknowledgement	CIC → Lender	Success / error per record
Rejection	CIC → Lender	Records that failed validation

Consumer (individual) submission — CIBIL CCIB format

File-level structure

Header record (1 per file).
Consumer Segment (CN) records — borrower demographic + loan summary.
Address Segment (CA) records — borrower addresses.
Telephone Segment (CT) records — borrower phones.
Email Segment (CE) records — borrower emails.
Account Segment (AC) records — per loan account.
Trailer record (1 per file).

Header record

Position	Field	Description
1 – 6	Record type	`HEADER`
7 – 36	Member name	Lender’s registered name
37 – 76	Member ID	CIC-assigned member ID
77 – 84	Submission date	`YYYYMMDD`
…	Other metadata

Consumer segment (CN)

Per borrower:

Field	Width	Description
Consumer name	80	As per KYC
DOB	8	`YYYYMMDD`
Gender	1	M / F
PAN	10
Aadhaar	12	(Optional; per consent + per CIC permitted)
Voter ID	20	(Optional)
Passport	12	(Optional)
Driving licence	20	(Optional)
…

Address segment (CA)

Per address:

Field	Width	Description
Address line 1	80
Address line 2	80
Address line 3	80
Address line 4	80
Address line 5	80
State	2	State code
PIN	6
Address category	1	Permanent / Current / Office / Other
…

Account segment (AC)

Per loan account (the heart of the file):

Field	Width	Description
Account number	25	Lender’s internal loan ID
Account type	2	Type code (e.g., `06 = Auto`, `13 = Credit Card`, `20 = Business Loan – Priority Sector`, `21 = Business Loan – General`, `22 = Business Loan – Term Loan`, `51 = Property Loan`, etc.)
Ownership indicator	1	`1` = Single / `2` = Joint / `3` = Guarantor
Date opened	8	`YYYYMMDD`
Date of last payment	8	`YYYYMMDD`
Date closed	8	`YYYYMMDD` (if closed)
Date reported	8	`YYYYMMDD` (data as-of date)
High credit / sanctioned amount	9	`₹`
Current balance	9	`₹` outstanding
Amount overdue	9
Tenure	3	months
Days past due	3
Payment history	90	36 months × 3 chars; e.g., `000` (no DPD), `030` (30 DPD), `060`, `090`, `XXX` (NPA), etc.
Asset classification	3	`STD` (Standard), `SUB` (Sub-standard), `DBT` (Doubtful), `LSS` (Loss), `WRO` (Written off), `SET` (Settled)
Credit limit	9	(For revolving)
Cash limit	9	(For revolving)
Rate of interest	5	(e.g., `19.50`)
Repayment tenure	3	(months)
EMI amount	9
Written-off amount	9
Settlement amount	9
Value of collateral	9	(For secured)
Type of collateral	2
…

(Format above is illustrative; exact CIBIL CCIB spec is authoritative.)

Payment history — encoding

The 36-month payment history is the most-watched field:

000 = paid on time / 0 DPD.
001 to 029 = 1 – 29 DPD.
030 = 30+ DPD.
060 = 60+ DPD.
090 = 90+ DPD.
STD / SUB / DBT / LSS for the relevant classification.
SET for settled.
WO for written off.

Each month’s status is one 3-char field; 36 months × 3 = 90 chars.

Trailer record

Position	Field	Description
Total records	10	Count of all records (excl header / trailer)
Total accounts	10	Count of AC records
Total amount	15	Sum of current_balance

Commercial (entity) submission — CIBIL CCMR / similar

Different format for entity borrowers. Key fields:

Entity name, entity type, PAN, GSTIN (where allowed), CIN / LLPIN.
Director list (each director’s PAN + name + DIN).
Industry code (NIC).
Address(es).
Loan details per facility:
- Facility type (CC / OD / Term / Bills / etc.).
- Sanctioned / outstanding.
- Asset classification.
- Suit-filed status.
- Date of NPA (if applicable).

Commercial bureau reports are also pulled differently — typically by entity PAN.

Submission process

1. Lender generates the file end-of-cycle (typically end of month for previous month's data).
2. Lender validates the file format (Schema validation).
3. Lender PGP-encrypts the file with CIC's public key.
4. Lender uploads via SFTP / portal.
5. CIC processes file:
   a. File-level validation.
   b. Per-record validation.
6. CIC returns acknowledgement (success / partial-success / failure with errors).
7. Lender processes acks:
   a. Records accepted → confirmed.
   b. Records rejected → fix and re-submit.

Common rejection reasons

Invalid PAN format.
Date in future for date-opened / date-reported.
Asset classification inconsistent with DPD pattern.
Sanctioned < current balance (impossible).
Missing mandatory fields.
Duplicate record (same account submitted twice).

What the LMS must generate

From LMS internal state, for each borrower / loan:

Consumer or commercial segment.
Address segment (current + permanent).
Telephone / email segments.
Account segment with current outstanding, DPD, payment history (last 36 months derived from classification_snapshot history).
Classification as of period-end.

Per file:

Header with member ID + submission date.
All segments in correct order.
Trailer with counts + totals.

File generation algorithm

For each lender:
    For each CIC:
        For each borrower with at least one active or recently-closed loan:
            Generate consumer / commercial segment(s).
            For each loan:
                Generate account segment with:
                    - Account ID (lender's internal)
                    - Type code
                    - Dates
                    - Balances
                    - Payment history (collect last 36 months of classification_snapshot)
                    - Asset classification
        Generate header.
        Generate trailer.
        Validate file structure.
        Encrypt + upload.

Performance

For ~50,000 active borrowers:

Generation time: < 30 minutes.
File size: ~50 – 100 MB.
Submission time (SFTP): ~5 minutes.

Frequency considerations

While monthly is the minimum, weekly submission is best practice:

Catches data corrections faster.
Bureau data is more current for borrowers’ future credit applications.
Reduces month-end rush.

What happens to bureau ack

Accepted records → CIC adds / updates borrower’s bureau record.
Rejected records → lender fixes and re-submits.
Persistent rejection → CIC may downgrade lender’s data quality rating.

CIC data-quality monitoring

CICs publish quarterly data-quality reports ranking member lenders on:

Submission timeliness.
Data completeness.
Accuracy.
Error rate.

Poor performers may attract RBI attention.

Submission frequency variation

Lender	Typical frequency
Large bank	Daily
Mid-size bank	Weekly
NBFC (mid-large)	Weekly
NBFC (small)	Monthly
NBFC (new)	Monthly

What the LMS must enforce

Generation per CIC in their specific format variant.
Validation before submission.
Submission tracking with acks.
Re-submission workflow for rejected records.
Audit trail per submission.
CIC data quality metric tracked internally.

Compliance touchpoints

CICRA, 2005 — mandate.
RBI directions to CIs on data quality.
Per-CIC member agreements + format specs.
DPDP — borrower data shared with CIC (covered by initial consent).