When borrowers are connected, their risks can be too.
A bank may lend to several companies without seeing that they share an owner—or that one has promised to repay another’s loan.
CreditGraph brings those relationships into the same view. You can see which borrowers are linked, how much they owe, and what deserves a closer review.
Start with shared ownershipThe first example follows three companies and one owner.
Why look at the loans together?
Imagine three companies borrowing from the same bank. Each has its own loan, but all three depend on one owner. Counting them as separate borrowers can hide that shared dependency.
An analyst can review the combined loan balance and ask how dependent the companies are on that person. The connection gives them a reason to investigate.
See the companies and their loans →A few terms before you start
- Borrower
- A person or company that owes money on a loan.
- Guarantee
- A promise to cover another borrower’s loan, subject to the contract.
- Connected exposure
- The outstanding loan balances in the group you are reviewing.
Does knowing the relationships improve predictions?
A separate experiment compares models with and without relationship data. The results show how their performance changes under different network assumptions.