Can Economically Intuitive Factors Improve Ability of Proprietary Algorithms to Predict Defaults of Peer-to-Peer Loans?

Economically intuitive macroeconomic factors and borrower characteristics predict peer-to-peer loan defaults beyond what proprietary algorithms predict. Using county-level unemployment data, we find that loans originated in high unemployment areas are more likely to default. In addition, we develop novel measures of a borrower's job instability, inattention to detail and evasiveness and show that borrowers with these characteristics are also more likely to default on their loans. Our results are robust to controlling for other loan and borrower characteristics previously used to model default risk