Analysis of the Factors that Determine the Regularities in the Errors in the Formation of Macroeconomic Forecasts

The construction of macroeconomic forecasts is an important issue that has various theoretical and practical aspects. Its primary role is based on the needs of economic agents to understand the future dynamics of the market in the process of making strategic decisions. The main objective of this study is to identify the main factors explaining the quality of macroeconomic forecasts. The research method is represented by regression analysis, where the dependent variable is the deviation of the forecast of the growth rates of GDP and the CPI from the actual value. The regressors are the qualitative characteristics of the forecasting organization: nationality, dependence on the state and specialization (financial institution, analytical agency, ministry), and also the date of publication of the forecast. It is assumed that the medium-term forecasts will be less accurate than the short-term forecasts, and the best forecast results will be generated by representatives of financial institutions. This work consists of four chapters and has the following structure. The first section briefly describes the main goal and objectives of the study, as well as its novelty and relevance. The second section consists of a detailed analysis of domestic and foreign literature on the topic of the study of the quality of macroeconomic forecasting. Description of the research methodology of this investigation, the structure of the model and the anticipated results are presented in the third section. In the fourth section, the forecasting of the growth rates of GDP and the CPI in the framework of short-term and medium-term planning with the use of the VAR, BVAR and ARIMA models, including comparison with the forecasting results of economic agents, is implemented. Finally, the reasonable conclusions with meaningful interpretation, as well as possible ways of practical applicability, are formulated.