Essays on the Determinants and the Macrofinancial Feedback Effects of Non-Performing Loans

Abstract

The first chapter of the paper analyzes the feedback effects between aggregate non-performing loans (NPLs) and macroeconomic factors by estimating a panel vector autoregression model in economies of 183 countries and areas over the period 1997 to 2019. For the testing purpose, new country-level dataset on NPL ratio constructed from various sources contributes to the increase of sample size. Granger causality tests indicate that the feedback effect between NPLs and real GDP growth as well as inflation is not found in this global model, while the feedback from NPLs to unemployment is verified. However, the estimation of orthogonalized impulse response functions demonstrates the impact of change in NPLs on macroeconomic factors. Although a positive shock to NPL ratio does not have an immediate effect on macroeconomic factors, they are gradually affected over a short period of time before a recovery back to the initial level. The analysis also confirms the effects from macroeconomic factors to NPLs, which is consistent with other related literatures. In addition, orthogonalized impulse response functions present time-consuming recovery from shocks to NPLs and macroeconomic factors in economies of EU countries. This chapter analyzes how a sharp rise in non-performing loans works under the economic environments using a panel data in economies of 183 countries and areas over the period 1997 to 2019. The first part of the chapter identifies the determinants of a sharp rise in NPL ratio under the financial and banking environments using a two-part model, a mixture model of the binomial and the continuous distribution, and the second part employs an observational study to examine how a sharp rise in NPL ratio affects macrofinancial activities using average treatment effect on the treated model based on propensity score matching technique to compare treatment group and control group designed to consist of episodes in which NPL ratios hold below 7 percent without any sharp rise. based on two-part model analysis, both financial regulatory framework and banking conditions tend to cause significant, albeit small, effects on a sharp rise in NPLs. The ATT estimation reveals that the sharp rise of more than 5 percentage points in NPL ratio significantly causes extra fall in GDP growth rate and extra increase in unemployment rate over time, compared with the cases of low NPL ratios.