A Signal Is Not Yet an Explanation
The IMF result tells us when to look harder, not what the final diagnosis must be. This chapter examines the unresolved mechanisms behind a concurrent credit and credit-inclusion boom, then adds the user-level failure surfaces that headline stability measures miss. The practical goal is to decide what evidence should be gathered before a warning signal becomes a restrictive response, while keeping digital safety, distributional gaps, and financial health inside the stability discussion.
The same pattern can have several causes
A rapid expansion of the borrower base during a credit boom is an observed relationship. It is not a complete explanation. The IMF paper identifies several plausible channels: lending standards may weaken, oversight may be insufficient, the new borrowers may be progressively riskier, or an asset-price upswing may make credit look safer before a reversal exposes losses. These explanations can coexist, but the paper does not establish which one is responsible.
The first candidate is a change in standards. A lender that wants to grow quickly may reduce documentation requirements, rely more heavily on automated assessments, accept weaker collateral, or price risk too optimistically. New borrowers then enter the system not simply because access improved, but because the threshold for receiving credit changed. If that is the mechanism, the important evidence would include underwriting rules, approval rates, borrower risk distributions, repayment schedules, and how those measures changed during the boom.
The second candidate is portfolio composition. Even if lenders apply the same formal rule, a larger borrower pool may contain more people with irregular income, less collateral, or no prior credit history. A portfolio can become riskier because its mix changes, not because every loan is badly designed. This is the extensive-margin issue in practical form: the number of borrowers rises, but the average information available about each borrower may fall. The relevant question is not whether new borrowers are inherently risky. It is whether the system has enough information, pricing discipline, and repayment flexibility to serve them sustainably.
The third candidate is weak oversight or a competitive race for growth. Digital distribution can make it cheaper to acquire customers and easier to issue small loans at scale. That can expand useful access, but it can also create incentives to prioritize origination volume over repayment quality. The same infrastructure that solves a reach problem may intensify a supervision problem if regulators cannot observe product terms, repeat borrowing, or the concentration of exposure across providers quickly enough.
The fourth candidate is asset-price inflation. A credit boom can raise the value of collateral or create expectations that prices will continue to rise. Lenders and borrowers may then behave as though repayment risk is lower than it really is. If prices reverse, the apparent safety of the loans can disappear at the same time that borrowers face weaker income or demand. In this story, rapid credit inclusion is part of a broader boom-bust process rather than an isolated source of losses.
These are hypotheses to discriminate among, not conclusions to choose by intuition. A useful diagnosis would compare countries or periods with similar borrower growth but different lending standards, collateral conditions, borrower composition, or supervisory practices. It would examine whether distress first appears in new borrowers, particular sectors, certain products, or the balance sheets of specific providers.