Bidyuk P.I.

Formimg hedge-fund portfolios using the quadratic approximation of loss function

This paper describes the optimization model for the hedge funds portfolio search. We provide the analysis of quadratic and linear loss function implementation. Furthermore, we study how the loss function type influences the behavior of different risk measures. The obtained results help us to evaluate the correctness of quadratic approximation for the given problem.

Forecasting financial processes at the foreign exchange market using technology indicators

In this paper we propose a method for constructing the regression models in order to forecast currency changes at the Foreign exchange market (FOREX). This method is based on utilizing the technology indicators as the independent variables. According to statistical data, we determine that our forecasts enjoy high quality and do not delay as compared to actual values.

The system approach to the credit risk assessment using Bayesian networks

this paper, we consider the basic facets of the credit risks assessment and the methods of their decreasing. The reasonability of Bayesian Networks (BN) application for the credit risk analysis and estimation of the client’s default probability is grounded. The methodology of BN building and using is proposed and tested for solving the problem of the credit capacity analysis.

On the specificity of ssdm designing and realization for financial and economic processes forecasting

This paper reports on the decision-making support system for financial and economic processes forecasting. The system is implemented in the environment of Visual Studio 5.0 in programming language C#. On the experimental side, the given software product is applied to the price forecasting for oil products in Ukraine. Moreover, we’re going to continue this thread of research to further modify the given software product by new methods for non-stationary and nonlinear processes forecasting.

The algorithms for selecting the portfolio stocks and determining their shares in the investment capital

The article considers the development of new algorithms for selecting the portfolio stocks. A new algorithm is proposed for solving the problem of stock selection for the investment portfolio and the problem of the investment capital distribution between the stocks taking into consideration returns, diversification risk and liquidity.