Determinants of the credit cycle : a flow analysis of the extensive margin

We use monthly data on individual loans from the Italian Credit Register over the period from 1997 to 2019 and show that bank credit expansions in the non-financial private sector are mostly explained by variations in the extensive margin calculated either in credit ows or headcount of new borrowers. We then build on a ow approach to decompose changes in the net creation of borrowers into gross ows across three states: (i) borrowers, (ii) applicants and (iii) others (neither debtors nor applicants). The paper investigates the macroeconomic dimension of these gross ows and documents three key cyclical facts. First, entries in the credit market by new obligors ("inflows") account for the bulk of volatility in the net creation of borrowers. Second, the volatility of borrower in ows is two times as large as the volatility of obligors exiting from the credit market ("outflows"). Third, borrower in ows are highly pro-cyclical, lead the economic cycle, and their uctuations are mainly driven by the probability of getting a loan from new banks. We read these results in light of the macrofinance literature on search frictions and on competition with lender-lender informational asymmetries. Overall, our findings support theoretical predictions of these models, but search frictions seem to play a major role in shaping movements along the extensive margin.