实例模型在投资组合最优化中的应用——基于人人贷平台的实证研究

Application of Instance-based Model in Optimization of Investment Portfolio——An Empirical Study based on Renrendai

如何在P2P网络借贷中实现收益和风险的最优化组合是投资者关注的核心问题。本文将投资者的投资组合最优化决策转化为在一定预期收益率下的风险最小化问题,利用核回归刻画历史贷款与新贷款的相似度,以相似度为基础构建实例模型,支持投资者进行投资组合的最优化决策;进而以在线网络借贷平台人人贷2015年散标交易数据为样本,对使用实例模型的可行性进行实证研究,结果表明,实例模型表现出良好的预测能力,有助于P2P借贷平台的投资者进行投资组合的最优化决策。

In P2P lending, how to realize the optimal combination of income and risk is the core issue of investor concern. In this paper, the optimization of investment portfolio optimization is transformed into a risk minimization problem under a certain expected rate of return. The kernel regression is used to describe the similarity between historical loan and new loan, to con- struct an instance model based on similarity, and support the optimization decision of investor to make the investment portfolio. The feasibility of the example model is empirically studied by taking 2015 data of Chinese online network lending platform Ren- rendai as a sample. Result shows example model shows good forecasting ability, and it is helpful for the investors to make the optimal decision of the investment portfolio of P2P lending platform.