Abstract: The rapid development of China's Fintech online lending platforms has provided borrowers with more accesses to funds, espe-cially by “robbing Peter to pay Paul”. Using micro data from one of the largest cash loan platforms in China, this paper empirically stud-ies the paying behavior of borrowers “robbing Peter to pay Paul” from two perspectives:payday loans' order level and borrower level. This study shows borrowers who are “robbing Peter to pay Paul” have higher probability of default or late payment. The more frequently a borrower uses payday loan “huabei”, the more likely the borrower is to default or delay payment. Default or late payment is more likely to occur in male than in female, and in those who live in small cities than those who live in big cities. This paper tries to explain the bor-rowers' default from the aspect of “robbing Peter to pay Paul”, and the conclusions are helpful for supervising China' online lending market, improving personal credit system and regulating the cash loan market.

Key words: online lending, cash loan, robbing Peter to pay Paul, paying behavior