Economic Shocks and Non-Performing Loans in Nigerian Banking System: A Panel Data Analysis

This paper estimates the economic shocks from the key macroeconomic variable to the growth of nonperforming loans in the Nigerian banking system. The significant determinants amongst the macroeconomic variables were confirmed to impact both positively (real gross domestic product growth rate) and negatively (consumer price index) on the growth of non-performing loans. A key innovation was the modeling of the impact with individual banks heterogeneity. Most studies assume homogeneity in the data generating process, thus losing information due to aggregating the different bank's data into a single data source. Furthermore, this paper determined the key determinants of the nonperforming loans via a system of panel data estimation; one-step and two-step GMM methods. This has the attendant advantage of overcoming bias and inconsistency in the parameter estimates, which OLS estimation suffers from. A panel vector autoregression (pvar) model has employed a second layered technique to trace the shock pathway via the impulse response functions. The cause-effect relationship between the economic variables and NPLS was determined through a panel Granger causality test. The results showed; the lags of most of the macroeconomic variables impact the quantum and growth of nonperforming loans mostly in the long run. There is a presence of a long-run relationship among the variables. The macroeconomic shocks to the nonperforming pathways do not appear to fizzle out in the short or medium term as depicted by the IRF.