Global Credit Data Coverage
Which credit bureaus actually exist in each market, who owns them, and what scale their scores use. Built from an IFC private-bureau directory covering 103 countries and 196 bureaus, cross-referenced against published score constructions.
Coverage by market
Red edge marks a single-bureau market. Tags show the multinational owner where there is one.
Score scales and what they measure
Highlighted rows all publish on 1–100 — and measure three different things.
| Score | Scale | What it measures |
|---|
What this data says
Half the world’s credit markets have exactly one bureau. Fifty of 103 countries list a single provider. In those markets there is no second opinion, no way to arbitrage a disputed file, and no competitive pressure on data quality. If that bureau has thin coverage of your applicant, you have no alternative — you have a gap.
The second finding is the one that breaks global credit policies. Experian, Equifax and TransUnion appear in only 43% of these markets. A multinational lender whose underwriting standard reads “pull a bureau score” and whose vendor contracts cover the big three has, in practice, no standard at all across most of the world. Over half the bureaus on this list are local-only names that never appear in a group credit policy.
A third, more surprising one: on this directory Creditinfo has the widest footprint of any group — present in more countries than Experian or TransUnion. It is an Iceland-headquartered specialist in frontier and emerging markets, and it is largely invisible in Western credit-risk conversations. Where a global lender assumes no data exists, there is often a bureau; it simply is not one anybody in London or New York has an account with.
The 1–100 trap
The scale table above is the part worth internalising. D&B PAYDEX, Experian Intelliscore Plus and Creditsafe Risk Rating all publish on 1–100. They are not interchangeable and they are not measuring the same thing.
PAYDEX is a description of past behaviour — dollar-weighted days beyond terms, where 80 means paying exactly on terms. Creditsafe is a forecast: probability of 90+ days beyond terms or insolvency within twelve months. Intelliscore is a forecast of serious delinquency. A 75 in each of those three is three unrelated statements about a company.
Different scales are an obvious problem. Identical scales measuring different constructs is the dangerous one — because the numbers line up in a spreadsheet and nothing looks wrong. Normalising them to a common percentile without first checking they answer the same question produces a model that is worse than no model, because it is confidently wrong.
How this connects to model performance
The global equipment finance scorecard found that the same fitted model achieved a Gini of 0.70 in the United States and 0.49 in Nigeria. This directory supplies the structural reason. Discrimination is a function of data depth, and data depth is a function of how many bureaus exist, how long they have operated, and what share of the adult and business population they cover.
The practical implication for anyone running a cross-border book: a single global score cut-off silently exports credit policy to whichever market has the thinnest files. Either calibrate the cut-off per market, or accept that you are lending on a different standard in each one without having decided to.
Provenance and staleness — read this before using it
This directory is an IFC / World Bank Group list of private credit bureaus, and it is dated — roughly 2010–2014. Several entries pre-date well-known changes: Veda Advantage became Equifax Australia in 2016, CallCredit was acquired by TransUnion in 2018, CIBIL is now TransUnion CIBIL, Highmark is now CRIF High Mark. Where I could identify a change with confidence I have annotated the entry in the explorer above; I have not attempted to silently update the whole list, because a partially-refreshed directory presented as current is worse than a clearly-dated one.
Treat the structural findings — monopoly share, big-three coverage, the long tail of local providers — as robust, since consolidation since 2014 has if anything increased concentration rather than reduced it. Treat any individual bureau name as a starting point for verification, not a fact. Score scales in the second table are separately sourced from current published documentation and are not subject to the same vintage problem.
Reference tool assembled from a public IFC bureau directory plus published score documentation. Bureau names reflect the source list’s vintage and may be superseded. Not an endorsement of any provider, not a substitute for a vendor due-diligence process, and not credit advice. Runs entirely in your browser — nothing is stored or transmitted.