| Summary/Abstract |
Lending by big technology companies (BigTechs) is an important new financial innovation in the digital era. This paper attempts to evaluate robustness and special features of BigTech's credit risk assessment. Using 1.8 million loan transactions for online merchants of a leading Chinese virtue bank, we carry out a horse race analysis between the BigTech approach (i.e., big data and machine learning models) and the bank approach (i.e., traditional financial data and scorecard models) in predicting loan defaults. We show that the BigTech approach better predicts loan defaults, reflecting information and modeling advantages. Though bank approach do well for the firms which have records in credit registry, BigTech's proprietary information can complement or, where necessary, substitute for credit history in predicting defaults, especially for the unbanked borrowers. We further discuss inclusiveness feature of the BigTech approach and the implications for financial inclusion, financial intermediaries' businesses and regulators' policy.
|