USING A VAR MODEL TO INVESTIGATE WHICH FACTORS DETERMINE NON-PERFORMING LOANS: FROM THE PERSPECTIVE OF BANK BEHAVIOR THEORY

Abstract

Purpose: At present, the credit quality of commercial banks has greatly deteriorated. With the simultaneous rising of the non-performing loan balance and non-performing loan ratio (NPLR), a new round of market-oriented reform is needed in commercial banks. This paper builds upon previous studies to investigate the causes of non-performing loans and their impacts on macroeconomics and microeconomics. Methodology: A conceptual model is developed and empirically tested based on 4 waves of a longitudinal survey of 360 banks in China (2007-2018). A Granger causality test is carried out with a vector autoregression (VAR) model empirical analysis for a stationary time series along with an impulse response analysis for the various variables to measure the influence and contribution rate of each of the identified five factors. Main Findings: The empirical results show that the non-performing loan ratio (NPLR) is positively correlated to the commercial bank’s net interest margin (NIM) and its proportions of state-owned holding (GYCG). However, the non-performing loan ratio (NPLR) is negatively correlated to the capital adequacy ratio (CAR), the provision coverage ratio (PC), and the cost-income ratio (CIR). Implications/Applications: The results can provide new insights for bank managers to benefit the long-term effectiveness of decreasing NPLR: bank managers should establish a unified risk supervision mechanism and improve their banks’ risk management system so that the banks can strengthen their credit management systems and their securitization of bad asset certificates. Novelty/Originality of this study: This paper contributes to the theory of non-performing loan ratio (NPLR) by conceptualizing new constructs from bank behavior theory and find that government intervention indirectly affects a bank’s NPLR mainly through the bank state-owned holding (GYCG).