Interdependence between Macroeconomic and Financial Stability Indicators: Macro-Feedback Effect

Standard stress tests consider only first round effect from macroeconomic variables to financial stability indicators. However, the occurred shocks in banking sector reflect on macroeconomic indicators throughout different transmission mechanisms, such as expectations of economic agents, expected responses of banking sector to increased credit risk and etc. This creates the necessity of expansion and improvement of existing types of models, which will also include second round (macro-feedback) effects. The study explores the dynamic relationship between macroeconomic variables and indicators of financial stability, proving the relevance of considering second-round effects for better policy analysis. This paper develops a macro stress testing model incorporating feedback effects between financial system and the real economy. The study uses VAR approach to analyze various interactions between indicators through Impulse Response Functions (IRFs) and conducts different stress scenarios on exogenous variables. According to empirical results for the case of Georgia, there is significant relationship between real and financial variables, proving the countercyclical nature of NPLs with respect to different estimates of GDP gap. The signs of the impacts are robust with respect to different estimates of GDP gap. However, the magnitude of the effect of change in NPLs on GDP gap and vice versa varies with different estimate of GDP gap. In addition, using historical decomposition of GDP gap, the study shows that the effects of financial variables on variables of real economy differ from each other depending on the observed time interval (pre-crisis or post-crisis). The transmission of the impact goes though ?credit crunch?. The model proves the fact that change in NPL ratio strongly impacts credit growth represented as change in Credit to GDP ratio. At the same time, change in Credit to GDP ratio explain significant part of output gap forecast error and has significant contribution to business cycle fluctuations, strengthening the impact of NPLs and financial stability as a whole on the real economy. The estimated model can be used for generating different scenarios and shocks for improving systemic risk analysis (effect of banking sector?s solvency on real economy) and for providing better policy recommendations.