Network-Based Credit Risk in P2P Lending
Swiss National Science Foundation | CHF 350K | 2020-2023 —
Overview
SNSF Mobility Research Grant project developing advanced, interpretable credit risk models for peer-to-peer lending markets using network topology approaches.
Role
Principal Investigator / Co-PI - Leading development of network-based credit risk models.
Research Team
- Lennart John Baals (Bern Business School / University of Twente)
- Branka Hadji Misheva (Bern Business School)
- Yiting Liu (Bern Business School)
Challenge
P2P lending faces unique challenges:
- Higher information asymmetry than traditional banking
- Less regulatory oversight
- Increased risk during economic downturns
- Need for investor trust and platform credibility
Methodology
| Approach | Purpose |
|---|---|
| Network Centrality | Identify connected borrower-lender patterns |
| Machine Learning | Predict default probability |
| Graph Neural Networks | Capture network topology features |
| Explainable AI | Ensure model interpretability |
Key Publications
- Finance Research Letters (2024): Network Centrality and Credit Risk: A Comprehensive Analysis of Peer-to-Peer Lending Dynamics
- Expert Systems with Applications (2024): Leveraging network topology for credit risk assessment in P2P lending
Impact
Models developed help P2P platforms and investors better assess credit risk, enhancing trust in the decentralized lending ecosystem.