The work deals with an automatic trading system and adaptive training. Is used both technical and automatic fundamental analyses, therefore as inputs to the neural network is used historical data exchanges and text data from reports. It explains the basics of trading, technical analysis and technical terms. The work deals with technical and fundamental analysis. It contains a description of algorithmic nature, program implementation and experiment with developed trading system. The selected strategy is compared to other approaches.

The dissertation consists of four empirical papers in the field of monetary economics. The first paper examines the extent of real exchange rate misalignment in the selected euro area countries, the next two papers shed light on macroeconomic spillovers in the remaining EU countries which are not part of the single currency area, while the last paper focuses on the exchange rate pass-through in the Czech Republic.