Paper ID
3418
Paper Type
short
Description
Artificial intelligence (AI) has become ubiquitous in the consumer finance industry. One of the major AI applications in this industry is AI-based credit scoring models. We investigate whether AI applications improve financial inclusion, as measured by three seemingly contradictory metrics, i.e. approval rate, default rate, and false rejection rate. We cooperate with an AI solution provider whose AI-based credit scoring models are widely used by online lenders in China. Using data obtained from these online lenders, we find that AI-based credit scoring models increase approval rate and reduce default rate simultaneously, which enhances both the magnitude and the quality of financial inclusion. AI-based credit scoring models also tend to reduce false rejection rate, suggesting that they can help provide access to capital to a previously underserved population. We plan to collect more data and conduct additional analyses in the future to enrich our current findings and explore for underlying mechanisms.
Recommended Citation
Wang, Hongchang; Li, Chunxiao; Gu, Bin; and Min, Wei, "Does AI-based Credit Scoring Improve Financial Inclusion? Evidence from Online Payday Lending" (2019). ICIS 2019 Proceedings. 20.
https://aisel.aisnet.org/icis2019/blockchain_fintech/blockchain_fintech/20
Does AI-based Credit Scoring Improve Financial Inclusion? Evidence from Online Payday Lending
Artificial intelligence (AI) has become ubiquitous in the consumer finance industry. One of the major AI applications in this industry is AI-based credit scoring models. We investigate whether AI applications improve financial inclusion, as measured by three seemingly contradictory metrics, i.e. approval rate, default rate, and false rejection rate. We cooperate with an AI solution provider whose AI-based credit scoring models are widely used by online lenders in China. Using data obtained from these online lenders, we find that AI-based credit scoring models increase approval rate and reduce default rate simultaneously, which enhances both the magnitude and the quality of financial inclusion. AI-based credit scoring models also tend to reduce false rejection rate, suggesting that they can help provide access to capital to a previously underserved population. We plan to collect more data and conduct additional analyses in the future to enrich our current findings and explore for underlying mechanisms.