Fundamentální analýza pro automatické obchodní systémy

This thesis deals with the creation of automatic trading systems which are able to predict market trends for stocks selected in advance. Proper trading strategy of this system is mainly created from the elements of fundamental analysis, such as annual returns of company, it's gains, level of shareholder's equity or total debt. All the stocks are classified by these fundaments, where result of this classification determines whether to buy or sell the stock. For the purpose of this thesis, 5 autamatic trading systems were created in order to compare different approaches to the stock evaluation, managment or diversification of business portfolio. Created systems were properly tested on historical data and, in order to determine their level of complexity, tests were executed in both periods of economic recession and expansion too. All the created systems reported great returns and most of them have potential to generate long-term gains. On the basis of received results, it is possible to make conclusion that fundamental analysis has a high value in the field of automatic trading systems, and it increases the chances of generating a profit.

Automatic trading systems, stock exchange, stocks, fundamental analysis, long-term investment, decision trees, XGBoost algorithm