Evaluating Financial Impact of Misclassifications in Credit Scoring Through IRR Analysis

Peer-to-peer (P2P) lending platforms create direct connections between borrowers and lenders, providing an alternative to conventional banking systems. Accurately assessing credit risk and loan profitability is essential for the ongoing success of these platforms. This research explores models for credit scoring and profit scoring within P2P lending domain. Classifiers like Logistic Regression, neural networks, and ensemble methods, are employed to estimate the likelihood of borrower default. To quantify the financial impact of classification errors, the Internal Rate of Return (IRR) was computed for misclassified loans in each model. The effectiveness of the proposed approach is assessed using Bondora P2P lending dataset. The results demonstrated that the proposed approach gives better insight about model selection for credit risk assessment. The analysis highlights the importance of a data-driven approach to refining lending practices.