A Scoring Model of the Risk of Costly Arrears at a Microfinance Lender in Bolivia

Publication Date: 1999
Published by: Microfin Technical Support Page
Document Type: Paper

View this Document   Email this Document

Can scoring models help microlenders in poor countries as much as they have helped credit-card lenders in rich countries?

Credit scoring is used to estimate the likelihood that loans from a microlender in Bolivia had arrears of 15 days or more. Although arrears in microfinance depend on many factors difficult to include in statistical models, it is found that inexpensive data does have some predictive power. In microfinance, computer models will not replace loan officers, but they can flag the highest risks and act as a crosscheck on human judgement

Concludes:

• both credit-card lenders in rich countries and microfinance lenders in poor countries make large numbers of small, short, unsecured loans
• can scoring help microfinance? A scoring model of costly arrears at a lender in Bolivia suggests that it can
• how should scoring be used? It is difficult to collect data and then to use the estimates of risk
• the model is probably best used as a super-fail filter that flags cases which have passed the normal evaluation but still have very high estimated risk and thus deserve a more careful review
• even lenders who do not score each borrower can still use the results from scoring models to make informed policy changes