Murga N. Information technologies of intraday-trading processes

The thesis is dedicated to actual theme of modern information technologies use for automatic traiding on financial markets. Necessity of automatic system of traiding (AST) is caused by solid influence of human factor on this process. Moreover, at the trading pro-cess trader makes a lot of routine actions, which could be automated and be made without of traded using specialized program applications. In the thesis possibility of intraday-trading automation is studied. In the thesis AST uses fuzzy portfolio optimization theory and multiagent AST are designed and compared with well-known systems. Creation of AST is connected with solutions of three main tasks: the task of financial instruments prices forecasting, the task of funds distribution between financial instruments and finding of optimal time-periods for opening/closing positions on financial instruments. For solving the task of funds distribution, theory of fuzzy portfolio optimization and mul-tiagent approach have been used. For forecast five forecasting methods have been used. Experimental investigations show the necessity of using of all these methods because there is no the best method in all traiding situations. For this task solving two new methods of known neural networks train and new forecast approach were proposed. New indicator for determining time-periods of opening/closing position on financial instruments was created.