Determinants of non-performing loans: Empirical evidence from Republic of Croatia

Development of the Croatian banking sector has its origins in 1990s when capitalism was introduced in the economy. As the transition neared to an end, Croatia managed to stabilise the banking sector which then became the growth engine of the newly reformed economy. From 2000 to 2007, remarkable growth rates were recorded in Croatia with living standard converging to European levels rapidly. However, following a slowdown in global demand, Croatian economy experienced a deep contraction. Rising borrowing costs and decline in exports resulted in a six-year depression during which the non-preforming loans (“NPLs”) reached an all-time high. However, increasing NPLs did not lead to a greater banking instability due to high capitalisation of Croatian banks. Nevertheless, high levels of NPLs have postponed the recovery from the crisis. Banks became reluctant to lend and significantly tightened their lending criteria. The main source of funding in the economy got disrupted leading to a sharp increase in unemployment. Undoubtedly, NPLs in Croatia are a consequence of adverse macroeconomic conditions. However, having in mind that Croatian banks, while operating in the same macroeconomic environment exhibited significantly different rates of problem loans, provided a sign that apart from macroeconomic conditions some others determinants of NPLs exist. While the determinants of NPLs have been widely studied across different countries around the world, studies on NPLs of Croatian banks are indeed limited. Therefore, this study aimed to analyse and quantify the determinants of NPLs in a sample of 21 Croatian commercial banks. The determinants were examined from 2010 to 2018 meaning the sample period coved both the recession and expansion phase of the Croatian economy. Four bank-specific and four macroeconomic variables were linked with the evolution of problem loans in the Croatian banking sector. Employing the System GMM, results of this study suggested that though macroeconomic factors essentially trigger the evolution of NPLs, bank-specific factor are crucial in explaining the differences in levels of problem loans across banks. Furthermore, results suggested that more capitalised and more profitable banks exhibit lower levels of NPLs. Moreover, cutting resources devoted to screening and borrowing consumers is showed to be one of the major reasons of high levels of NPLs. CNB should consider introducing credit risk assessment methodology review as a part of its macroprudential policies in order to asses signs of trouble and enforce particular banks to lend more conservatively. Finally, results suggested that out of all macroeconomic variables, only exports growth had a significant positive impact on NPLs. The government should introduce structural reforms which will make the Croatian economy more competitive leading to an increase in exports