Statistical Analysis of the Dynamics of the Main Macroeconomic Indicators of Russia in Real Terms Since 1993

Statistical analysis of the dynamics of the main macroeconomic indicators (8 indicators) of Russia in real terms since 1993 has been carried out. In the course of the work, the following problem was solved: time series turned out to be short, data were presented in different periodicity (quarterly / annual), and for some variables data was missing for certain years. In order to solve this problem, annual data was reduced to quarterly, using methods of temporary disaggregation. Moreover, data on some variables for certain years have been recovered. As a result, Bayesian vector autoregression model with apriori distribution of Minnesota was constructed. Model with 328 coefficients was obtained. Also functions of impulse response for each of 8 variables were constructed, which allowed to draw a conclusion about the adequacy of parameterization of the model. In the last part of the work, forecasts were made for each variable until 2022, and a comparison between BVAR and VAR was made in terms of forecast accuracy.