ABSTRACT

A challenge for investors in P2P lending marketplaces is to find an effective allocation of their funds across different loans by assessing the widespread collapse of lending platforms and each loan’s default risk. This paper investigates risk-averse/risk-tolerant/risk-neutral investors’ and borrowers’ transaction behaviours, as well as P2P lending platforms’ operating mechanism. A Heterogeneous Investors Multi-Agent Model (HIMAM) is built to simulate agents’ reactions to variance in platform’s bankruptcy risk. This enables an assessment of changes in investors’ average return rates and investment decisions. This study finds that platform bankruptcy risk influences total investment amounts and average return rates. The correlation between the transaction fee and total investment amount presents as an inverted U-shape. Furthermore, there is a negative correlation between two-part tariffs and both investors’ average return rates and the number of investors. This is because two-part tariffs are dependent on investors’ transaction volumes. High two-part tariffs increase investors’ transaction costs. Additionally, considering bankruptcy and default risk, credit E are the most popular type of borrower because such borrower tradeoff higher risk for higher return; credits B and C are the least popular type of borrower.

摘要

P2P 借贷市场的投资者面临的一个挑战是通过评估借贷平台的普遍崩溃和每笔贷款的违约风险，在不同的贷款中找到有效的资金分配。本文研究了风险厌恶/风险承受/风险中性的投资者和借款人的交易行为，以及P2P借贷平台的运行机制。建立异构投资者多代理模型（HIMAM）来模拟代理对平台破产风险变化的反应。这可以评估投资者平均回报率和投资决策的变化。本研究发现平台破产风险影响总投资额和平均收益率。交易费用与总投资额之间的相关性呈现倒U型。此外，两部分关税与投资者的平均收益率和投资者数量呈负相关。这是因为两部分关税取决于投资者的交易量。高额的两部分关税增加了投资者的交易成本。此外，考虑到破产和违约风险，信用 E 是最受欢迎的借款人类型，因为此类借款人以更高的风险换取更高的回报；信用 B 和 C 是最不受欢迎的借款人类型。信贷 E 是最受欢迎的借款人类型，因为此类借款人以更高的风险换取更高的回报；信用 B 和 C 是最不受欢迎的借款人类型。信贷 E 是最受欢迎的借款人类型，因为此类借款人以更高的风险换取更高的回报；信用 B 和 C 是最不受欢迎的借款人类型。