Macroeconomic Forecasting Using Data of Different Frequency

Macroeconomic GDP forecasting is a study of great significance due to the importance of this key indicator and its efficient estimations for decision-making process while working out strategies by governments, central banks, companies and other economic agents. However, there is no an econometric model that would appear to be universally preferable due to the social, cultural, political peqularities of each country. Moreover, there is an issue caused by the different periods for which the GDP and its macroeconomic indicators are published. Generally stated, the purpose of the proposed research is to identify which of the two methods, MIDAS-AR which employs data of different frequency, or VAR, which uses figures of the same frequency, can better foresee the dynamic of Russian GDP growth rate. The core method is a comparative econometric analysis, which will enable to give findings on the relative forecast power of the models. This study will throw light on the factors which influence the choice towards each of the models. Then, based on this information the strategy of identifying the most suitable model for forecasting Russian GDP growth will be outlined.