The purpose of this dissertation is to demonstrate the role that standard risk-assessment techniques can play in increasing access to formal sector funding for micro-level entrepreneurs in developing countries. More specifically, it focuses on developing a tool that provides precise measurement of factors that differentiate between potentially 'good' and potentially 'bad' micro-level borrowers To begin the development of this tool, the credit applications from three microenterprise lending programs in Ecuador were examined. While Ecuador was chosen as a case study, the intent was to develop a technique for selecting potentially superior micro-level customers from non-traditional populations in other developing countries and regions as well Credit applications from each organization were standard and used uniformly in the application process, but the information requested from individual applicants varied dramatically among programs. In addition, the completeness of the applications varied within each program. Another data problem related to the fact that credit applications for 'rejected' applicants were not kept on file by two of the three lending programs For each of the two organizations unable to provide data for 'rejected' applicants, a univariate probit model was developed. These probability models can be used by credit officers to screen credit applicants. In addition, these models, if incorporated into the loan decision process, can improve repayment rates by more than 10% in each case. For the program that was able to produce 'rejected' applicant data, a bivariate probit with selection was estimated to take into account the potential correlation between disturbances of the credit and repayment equations. The results of the bivariate model suggest that there were no statistically identifiable determinants of repayment or loan issuance. This finding suggests that some variable outside the credit application may influence the decision processes more than those factors included This research extends previous research on the transformation of 'grant-driven' non-governmental organizations (NGOs) involved in microenterprise lending into the self-sufficient financial organizations required by the decreasing availability of donor funds. This research suggests a tool, adapted from the banking sector, that is a viable complement to guarantee funds in terms of securing more domestic funding for micro-level borrowers. Credit scoring models provide a mechanism for lowering the perceived risk through more reliable credit assessments of microenterprise borrowers. In fact, banks may be willing to work directly with microentrepreneurs rather than through intermediaries once a reliable risk-assessment tool exists. A second conclusion of this research is to suggest a cost-reducing tool for microenterprise lending programs. A credit scoring model can offer opportunities to streamline the lending process. It can decrease the need for extensive loan applications by providing a reliable means of assessing an applicant's repayment probability with a minimal amount of information. The ability to assess the creditworthiness of applicants through a more cost-efficient process not only lowers loan processing costs but also reinforces the status of these organizations as financial institutions. In addition, the current donor funds could be re-lent more often if the lending process were less expensive and more reliable. These conclusions suggest this study could become the basis for the development of a more sophisticated credit scoring tool unique to microenterprise borrowing groups and could result in the expansion of the domestic capital available to these micro-level lending programs. (Abstract shortened by UMI.) / acase@tulane.edu
| Identifer | oai:union.ndltd.org:TULANE/oai:http://digitallibrary.tulane.edu/:tulane_23879 |
| Date | January 1996 |
| Contributors | Cunliffe, Charlotte Conrad (Author), Morrison, Andrew (Thesis advisor) |
| Publisher | Tulane University |
| Source Sets | Tulane University |
| Language | English |
| Detected Language | English |
| Rights | Access requires a license to the Dissertations and Theses (ProQuest) database., Copyright is in accordance with U.S. Copyright law |
Page generated in 0.1231 seconds