Динамічний факторний аналіз у прогнозуванні індексів фондового ринку

The application of models of dynamic factor analysis for the study of the dynamics of world stock indices is considered. The developed model combines the methods of factor analysis and auto-regression estimates. In contrast to the classical schemes of vector time series analysis, a system of dynamic factors is more informative in comparison with the initial set of indicators for the description of evolution. Dynamic factor analysis allows to minimize the target's error in the ex post forecast. The influence of the parameters of the developed models, such as the composition of the system, the number of factors and the length of the lag in the autoregressive equations, on the quality of the description of the dynamic change of indicators and the determination of the corresponding predictive values are shown. For the considered systems, interval and recursive predictions of the selected indicators are calculated. A comparison of the results of forecasting with actual data of statistics is made and it is established that the forecast error does not exceed 1-2%, which confirms the high efficiency of the proposed approach.

Citation

Вітлінський B. B. Динамічний факторний аналіз у прогнозуванні індексів фондового ринку / Вітлінський B. B., Катуніна О. С. // Актуальні проблеми прогнозування розвитку соціально-економічних систем : монографія / Київ. нац. ун-т ім. Т. Шевченка [та ін.] ; [за ред. О. І. Черняка, П. В. Захарченка]. – Мелітополь, 2019. – С. 25–38.