What the Warning Signal Measures
The lived-use evidence shows why account ownership is not enough to judge credit risk. This chapter changes scale and reconstructs how the IMF study measures the warning signal: it separates rapid growth in the borrower base from rapid growth in total private credit, then asks what happens when the two occur together. The value of this reconstruction is methodological as much as factual. It lets us read the paper’s results as a carefully defined predictive pattern rather than as a loose claim that financial inclusion causes crises.
Two kinds of credit growth
Credit can expand in at least two different ways. Existing borrowers can receive larger loans or draw more heavily on existing facilities. That is growth on the intensive margin: more credit for roughly the same set of borrowers. Or the pool of borrowers can widen, bringing new households or firms into formal credit. That is growth on the extensive margin, which the IMF paper calls credit inclusion.
The difference matters because the two forms of growth expose a financial system to different questions. Intensive-margin growth may increase leverage among people already known to lenders. Extensive-margin growth changes who is being assessed, how much information lenders have about them, and how a portfolio is distributed across borrower types. A system that suddenly adds many first-time borrowers may be broadening opportunity. It may also be relaxing standards, using new distribution channels, or reaching people whose repayment behavior has not yet been tested through a full cycle.
The paper measures credit inclusion with two supply-side indicators from the IMF Financial Access Survey: the number of borrowers from commercial banks per 1,000 adults and the number of loan accounts with commercial banks per 1,000 adults. These are not the same as the World Bank’s survey measures of whether an adult borrowed in the previous year. The Findex asks people about their experience. The IMF measure uses provider-side counts that can be tracked annually and compared with macroeconomic credit data.
The denominator matters. Reporting borrowers per 1,000 adults makes countries with different population sizes more comparable, but it does not make their banking systems identical. A borrower counted by a commercial bank may have a different product, income, or legal status from a borrower elsewhere. Loan accounts and borrowers also answer different questions: the first describes the number of credit accounts, while the second describes the people receiving credit. Their movement together can strengthen the picture of expanding access; divergence can be informative too, because it may indicate changes in repeat borrowing or account structure rather than a simple increase in people reached.
The annual frequency is another reason to use the FAS data for this exercise. The Global Findex provides rich demand-side detail but is collected at intervals, while the IMF study needs a sequence of observations that can be aligned with credit cycles and later outcomes. The trade-off is breadth for timing. The Findex can show what people report doing across 141 economies; the matched IMF panel can follow narrower indicators across years for the countries and periods where all required measures are available.
The study combines the borrower and loan-account indicators into a composite financial-inclusion measure. In concept, the composite protects the analysis from making the result depend on only one of two closely related ways of counting access.