Highlights
- Evidence shows women often demonstrate higher loan repayment rates than men, suggesting they are lower-risk borrowers yet face higher barriers to loan approval. This suggests an opportunity to refine lending models for more accurate risk assessment and increased financing to women.
- A gender-disaggregated approach can help lenders more accurately measure portfolio risk; such approaches not only can reduce the gender gap in access to credit, but they can make good business sense, by allowing providers to increase their portfolios or reduce their losses.
- The guide introduces a gender-lens analytical framework for lenders to determine whether lending decisions and outcomes in their portfolios differ by gender and, if so, how. For lenders using credit scoring models, the guide presents different gender-intentional techniques for adjusting their credit scoring models to more accurately reflect risk. It also offers implementation strategies—such as setting other decision threshold policies for women and men.
- The guide uses actual loan application and repayment data from AB Bank Zambia and TymeBank to demonstrate how gender-intentional scorecard development and implementation strategies can work. In addition, sample data and codes (in Excel, R, and Python) are available in the appendix for readers who wish to apply the analysis.
Executive Summary
Evidence shows that women tend to have, on average, better loan repayment rates than men. The fact that the women receiving loans are, on average, lower risk than the men implies that women are being subjected to a comparatively higher bar for loan approval than men. These conditions present an opportunity to adjust lending models to more accurately assess risk and, as a result, increase financing to women.
This guide shows that a disaggregated gender analysis of a loan portfolio can unveil potential gender-intentional strategies to grow both the total loan book and the share of women borrowers without increasing the portfolio’s credit risk. Because a gender-intentional approach can help lenders more accurately measure portfolio risk, such approaches not only can reduce the gender gap in access to credit, but they can make good business sense, by allowing providers to increase their portfolios or reduce their losses.
This guide presents a gender-lens analytical framework that lenders can use to determine whether lending decisions and outcomes in their portfolios differ by gender and, if so, how. For lenders using credit scoring models, the guide presents different gender-intentional techniques for adjusting their credit scoring models. It also presents implementation strategies—such as setting different decision threshold policies for women and men.
Finally, the guide uses actual loan application and repayment data from two banks to demonstrate how gender-intentional scorecard development and implementation strategies can work in practice:
- AB Bank Zambia (ABZ) incorporated gender into its microloan scorecard, resulting in more women receiving credit within the bank’s existing risk appetite and business model.
- A Buy Now Pay Later (BNPL) product offered by digital TymeBank in South Africa is used to illustrate how gender-intentional model development and decisioning strategies could be used to increase both the total number of loans and the share of loans to women for a given portfolio risk level. It also shows how the model could be used for risk-based pricing, to lower interest rates for women borrowers.
These examples show that a gender-intentional approach can result in a larger total portfolio and a larger number of loans given to women for a given portfolio risk target. Even the simplest strategies result in significant improvement over a gender-blind approach. More sophisticated approaches, like developing separate credit scoring models, are likely to have the greatest impact in terms of increasing both total lending and lending to women.