Nonlinear effects of P2P lending on bank loans in a Panel Smooth Transition Regression model

This paper proposes an original framework to determine the impact of Peer to Peer lending on bank loans. Using a Panel Smooth Transition Regression (PSTR) model, the authors, set P2P lending balances as a threshold variable, average P2P lending rate, short-term benchmark lending rate and M2 as control variables, and established province-specific and time-specific coefficients of variables for 8 provinces of China from January 2014 to April 2016. The results show a nonlinear dynamic relationship between P2P lending balances and domestic bank loan balances: there are two threshold values and three regimes. In regime 1 and regime 2, the P2P lending balances are small, P2P lending balances and average P2P lending rate exert a positive impact on domestic bank loan balances, and short-term benchmark lending rate exerts a negative impact on domestic bank loan balances. In regime 3, P2P lending balances are bigger, P2P lending balances and average P2P lending rate exert a negative impact on the domestic bank loan balances, and the short-term benchmark lending rate exerts a positive impact on domestic bank loan balances. M2 has a positive impact on domestic bank loan balances in three regimes. Therefore, the results explained in this paper have important policy implications towards P2P lending in China.