Introduction of Credit Scoring in AccessHolding network banks

2012-10-31 - 2013-01-31
Germany, Azerbaijan, Madagascar, Tanzania, Nigeria, Liberia, Tajikistan, Zambia

In its role as manager of the AccessBank network, LFS launched in November 2011 a credit scoring project with its largest and oldest affiliate, AccessBank Azerbaijan (ABA). Credit scoring is increasingly being recognized as a suitable tool in microfinance to lower transaction costs, enhance credit risk management and to ultimately improve both outreach and sustainability. For ABA, LFS initiated the project as it is expected that credit scoring will help the bank to increase efficiency in its credit processes and workflows and enhance the competitiveness of its credit products. With the technical expertise of an external scoring specialist, a “segmented” scorecard for ABA’s Express and Loyal Loans was developed. This scorecard follows a “hybrid” approach and combines factors selected and weighted based on statistical analysis of historical data as well as factors that were added based on the working group’s expertise and judgement. Following the completion and design of the scorecards, a pilot project will be launched before the scoring model will be fully integrated into ABA’s credit process. With the experience and knowledge gained in the ABA project LFS is prepared to roll-out credit scoring systems as well to other AccessHolding network banks. LFS is currently identifying the next network bank in which credit scoring will be implemented.
Credit Scoring with AccessBank Azerbaijan: Design and overall coordination of the project Reviewing consistency of data and the statistical model in the analytical software program “R” Reviewing proposed scorecards and discussing the selected risk factors as part of the work group Defining credit workflows that include credit scoring Coordinating the pilot phase in one of ABA’s branches Periodical performance review of the model Credit Scoring with other AccessHolding network banks (planned): Identifying next partner bank to roll out credit scoring Definition of scope and project design Extracting data from the LFS MBS core banking system Developing the statistical model based on historical data Selection and weighting of risk relevant factors and development of scorecards Coordinating the implementation of credit scoring in the partner bank Training of employees in the partner bank Periodical performance review of the model