Assessing Banking Sector Stability with Special Reference to Montenegro and Central and Eastern Europe

The aim of this research is to investigate the main weaknesses of the banking sector in Montenegro and, more generally, in Central and Eastern Europe. By contributing to the understanding of banking sector fragility, we seek to help regulatory authorities craft more effective regulations and policy interventions in order to minimize the costs that could arise from banking instability. The review of previous research suggests that an increase in the incidence of non-performing loans indicates increasing fragility of the banking sector and/or the lack of efficient banking supervision. Thus, the ratio of non-performing loans is taken as the key indicator of banking vulnerability. To examine the determinants of non-performing loans we use data at the individual bank level. We investigate whether in these countries the ratio of non-performing loans is driven predominantly by macroeconomic developments (i.e. similar factors to those recognized in the theory of financial crises) or by transition-specific factors, such as inadequate risk-assessments, the high risk appetite of banks’ management and the high concentration in the banking sector. This empirical analysis initially focuses on the sensitivity of the ratio of non-performing loans to macroeconomic and bank-specific factors in Montenegro. Subsequently, the analysis is extended to Central and East European countries. Bearing in mind the large presence of foreign banks in these countries, a particular feature of this analysis is that non-performing loans are modelled to capture differences between banks based on ownership structure. In order to account for time persistence in the structure of non-performing loans, a dynamic panel approach is used. However, in the Montenegrin model, given that the lack of cross sectional units precludes GMM Estimators, we investigate the use of FE estimation adjusted to take account of dynamic misspecification. The empirical findings suggest that strong performance in the real economy results in a lower ratio of non-performing loans but there is also a significant positive effect of past rapid loans growth in the second year after the end of the credit boom. The latter finding suggests that aggressive lending coincides with more reckless risk taking. Moreover, the empirical evidence suggests that some bank-specific features, which reflect banks’ management policies, affect the evolution of non-performing loans. The multi-national study suggests that foreign banks are likely to have lower ratios of non-performing loans. Based on our findings, regulatory authorities should expand their monitoring framework to include both macroeconomic and bank-specific indicators when assessing the stability of the banking system. In addition, regulators should be more concerned about any loosening of bank lending criteria in an upturn, since our results suggest a delayed effect of loans growth on the incidence of non-performing loans.