At present, China’s P2P platform has emerged a group of people known as net loan scalper. They borrowed money at low interest rates and invested the high interest rates loans in order to earn profit margins. This paper intends to discuss the factors influencing the loan interest rate into the scalper arbitrage environment from the microcosmic point of view, and focuses on the construction of multiple regression model of 35,970 real transaction data from Hongling Capital, and confirmed that the amount of loan, the term of loan, the mode of repayment, the type of the subject and the lowest bid limit have a significant impact on the setting of the loan interest rate, as well as in the scalper arbitrage environment and no scalper arbitrage environment, the impact of these variables are different. Therefore, platform and users should treat the scalper arbitrage dialectically and learn to identify the risks behind it so as to avoid them.