Economic Computation and Economic Cybernetics Studies and Research, vol.49, no.2, pp.1-22, 2015 (SCI-Expanded, SSCI, Scopus)
The Greek banks have to limit their operations in the Southern and Eastern Europe, because the European Commission requested to stop transfer capital to their subsidiaries in this region. This means that the Greek banks have to sell some of their subsidiaries in the region, especially the non-profitable ones. The aim of this paper is to forecast the profitability ratios of the Greek capital owned banks activating in Romania for 2013-2014 to see what would be their best option regarding these subsidiaries in Romania against the other subsidiaries of the Greek banking groups in the region. In the strategy of a multi-national bank, two categories of macroeconomic indicators forecasts are important: the ones in the countries where it has subsidiaries, and it’s own forecasts, in order to monitor the future evolution of their business and financial stability. In this paper, the authors focus on forecasting the performance indicators of four Greek banks and their Romanian branches, based on artificial neural networks (ANN). The study uses various implementations of the neural network algorithm offered by GMDH Shell software and the balance sheet of the four Greek banks and their Romanian branches, during 2006-2012, in order to obtain a short term forecasting of performance indicators.