Building a model to predict return on investment (ROI) of consumer loans issued through the company Lending Club. Predictions are used to simulate a historical investment portfolio and analyze performance of the model.
Consumer loans are a relatively new asset class available for private citizens to invest in. They potentially offer returns similar to traditional investments while offering lower portfolio volatility. However, a significant number of consumer loans are defaulted on by the borrowers. A significant amount of capital can be lost to defaults if a portfolio invests in loans based off of poor criteria. The goal of this project is to construct a model to predict the ROI of a loan in order to determine if it would be a good investment. Once a model is built it can be run through a portfolio simulation of historical loan payment data to analyze how the model's strategy would have performed.
Please see Lending Club's statistics page for data on loans that have been issued.Files containing all payments that borrowers have made on their loans can be found at Lending Club's additional statistics page.
Supplemental data comes from the Federal Reserve Economic Database, FRED. The supplemental data currently used are the monthly values for the bank prime loan rate (MPRIME), the 30-year fixed rate mortgage average (MORTGAGE30US), as well as the University of Michigan inflation expectation rate (MICH).