Прогнозирование показателей стабилизационных процессов экономики России на основе моделей векторной авторегрессии

The article presents the results of construction, estimation and use of vector autoregression models (VAR) for short-term forecasting of indices, reflecting stabilization processes, taking place in Russia's economy. To construct VAR models economic indices corresponding to picked out indications of Russia's stability have been selected. A preliminary statistical processing of time series for indications reviewed have been made (seasonal component has been eliminated, time series have been transformed into stationary form, the length of variables lag included into the model has been defined and others). A model of vector autoregression describing the dynamics of changes of stabilization processes indices has been constructed and the short-term forecasting of these indices values for the following time periods has been made on its basis. A test forecast has been obtained and a comparison of forecast characteristics of the VAR model constructed has been made as well as a system of simultaneous equations (SSE), which was constructed in the previously published authors' works. The forecasts obtained with the help of SSE don't exceed forecast accuracy, obtained on the basis of VAR model.