Imagine a file on your desk with three printouts for the same applicant: a SLIK report, an IdScore report and a CLIK report. They look alike, with names, facilities and numbers in tables, so it is tempting to read them as one combined verdict. That is how a bureau score ends up treated as if it were a collectibility grade.
How do you review them so each report keeps its own meaning? The four steps below follow the order in which mistakes usually enter the file.
Step 1: identify the source
Interpretation depends on who issued the report.
- SLIK is OJK’s debtor-information system. OJK explains its role in supporting supervision and financial information services.
- CLIK is the brand of PT CRIF Lembaga Informasi Keuangan, a private credit bureau with its own report format and definitions.
- PEFINDO Biro Kredit is a private credit bureau. Its IdScore website, published under PT PEFINDO Biro Kredit’s name, presents IdScore as a credit-scoring information service.
- CBI (Credit Bureau Indonesia, PT Kredit Biro Indonesia Jaya) is also a private credit bureau. OJK lists it with PEFINDO Biro Kredit and CLIK among the licensed credit information bureaus (LPIP).
A facility’s reported collectibility grade and a bureau score are different measures. Read each report with its own definitions; the difference between a record and a score is explained further in what a credit report tells a lender.
Step 2: confirm the borrower and reporting position
A correct reading of the wrong borrower or the wrong month is still wrong. Before reading any figure, establish:
- Whether the report identifies the intended person or entity.
- Which institution or bureau issued it.
- When the report was obtained and which data positions it covers.
- Which material facilities or periods need clarification.
A mismatch between reports can come from identity, timing or interpretation. Keep both records until the cause is understood. For reading the SLIK report itself, see how to read a SLIK OJK report.
Step 3: check extracted tables against the original
When report data is extracted into a table, the table becomes a second document that can contain its own errors. Extraction makes a long report quick to scan, and checking the material fields against the source makes the table safe to rely on. Verify material balances, currencies, facility status and ambiguous fields against the source.
Example: the fictitious trader UD Berkah Tani has a working-capital facility shown in SLIK with an outstanding balance of IDR 450.000.000. Because the balance is material, the reviewer opens the linked source page and confirms it in seconds. The collectibility cell in the same row is unreadable in the scan; a well-built extraction marks that grade as unavailable instead of filling it in, so the reviewer knows to confirm it from the source.
Keep missing values distinct from zero. For derived amounts, distinguish an estimate from a figure the creditor explicitly reported; material repayment obligations may need confirmation from agreements or lender statements.
Protecting access and keeping the report in context
Credit reports contain confidential information, so how they are obtained and used matters as much as how they are read. Use authorised channels and apply the institution’s confidentiality and data-handling controls. Sharing a report for review does not justify unrestricted reuse.
The report is also only part of the picture. OJK’s iDebku service states that debtor information is one input to an assessment, not its sole basis. Read credit records together with the borrower’s financial position and other evidence relevant to the application.
Checklist for a multi-bureau file
- Label every figure with its source and data position.
- Confirm identity before comparing reports.
- Check material balances and grades in any extracted table against the original.
- Mark unreadable or missing values as unavailable, never zero or Lancar.
- Keep conflicting records until the cause is known.
Sxored reads iDeb reports into facility tables per borrower and related party, and marks an unreadable collectibility grade as unreadable instead of counting it as Lancar. Next, see how to combine credit information in a 5C analysis.
