Modeling the Probability of Incurring Past-due Payment in Microlending

Microcredit is becoming more common in the Russian market. This area is subject to significant risks of irrevocability of money by borrowers in view of the fact that, most often, individuals who do not meet the requirements of banks resort to microlending. In this paper, modeling and empirical assessment of factors affecting the probability of a borrower failing to return microcredit is carried out. The work evaluated logistic regressions on cross-sectional data on borrowers and microcredits issued to them (August 2012 - September 2018) provided by a microcredit company operating in the Perm Region. It was found that the likelihood of no return of funds after the occurrence of a non-payment situation is lower for pensioners than for the rest of the borrowers. Women were more inclined to defaulting microcredit after the fact of overdue debt. At the same time, it was not possible to empirically confirm the presence of the “infection effect” (the presence of non-payments from related borrowers) and the “debt trap” (repeated taking of money to cover existing debts, which makes the borrower less creditworthy). The findings of the study can be used as a microcredit company that provided data, and other microfinance organizations to evaluate potential borrowers since they reflect the specifics of the Russian microcredit market. Key words: microlending, probability of incurring past-due payment, logistic regression, «infection effect», «debt trap», court proceedings, microcredit non-payment.