Estimating the Influence of Different Shocks on Macroeconomic Indicators and Developing Conditional Forecasts on the Basis of BVAR Model for the Russian Economy

In this paper, we investigate the influence of internal and external shocks on macroeconomic indicators of Russian economy using Bayesian vector autoregression (BVAR) model. We develop conditional medium-term forecasts (scenarios, up to 2017) and then compare the forecasting outcomes achieved in BVAR under these scenarios with respective official forecasts of the Ministry of Economic Development (MED) of the Russian Federation. Our results indicate that within the similar scenario conditions our proposed BVAR predicts (1) a deeper and (2) more prolonged recession on the medium-term forecasting horizon as compared to the MED’s forecasts. Our comparative analysis allowed us to reveal the bottlenecks in the forecasting methodologies applied both in the MED’s model and in our BVAR model, which seriously worsen the quality of forecasts.