Mutual Influence of Macroeconomic Indicators Of Kazakhstan and its Trading Partners Bulat Mukhamediyev, al-Farabi Kazakh National University, Almaty, Kazakhstan, bulat.mukhamediyev@kaznu.kz Azimzhan Khitakhunov, al-Farabi Kazakh National University, Almaty, Kazakhstan, xit8_89@mail.ru Abstract The paper investigates interdependence of real and financial indicators of members of the Eurasian Economic Union and other countries. First of all, attention is paid to the study of their impact on the dynamics of macroeconomic indicators of Kazakhstan taking into account its linkages to the global economy. The study is based on the method of global vector autoregression. Individual country vector autoregression and error correction models for 33 countries are constructed. On this basis, the global vector autoregression model was estimated and calculations for the forecast of macroeconomic indicators responses in each country on the different economic shocks (including world oil price shock) were performed. The results may be of interest to the different governments in terms of rapidly changing world of interconnected economies. Keywords: Interdependence of indicators, global vector autoregressive model, Eurasian Economic Union, economic shocks, functions of pulse return. Introduction The dynamics of macroeconomic variables in each country is affected by the behavior of macroeconomic variables of other countries, especially its main trading partners. To investigate the interdependence of macroeconomic variables, it would be natural to apply the method of vector auto regression, which does not impose any a priori assumptions about the relationships of the investigated processes. However, there appears a problem of so-called "curse of dimensionality", due to which its decision is almost impossible, even for a relatively small number of countries and indicators taken into consideration. It became possible to overcome this problem in the framework of the global vector auto regression. This approach is carried out by decomposition of large models to individual country models, which are then summarized in a correct manner in the global model. Global vector autoregression method was proposed in Pesaran et al. (2004). Model of great dimension in this approach is decomposed into individual country vector auto regression model and error correction model, based on which the estimates are obtained for the initial global model. For the correctness of the results it is necessary to conduct a series of statistical tests, such as checking the variables on stationarity, determining the order of integration, weak exogenous external variables, determine the number of cointegrating relations and others. Further development of the method of global vector auto regression was obtained in the works of DØes, di Mauro, Pesaran, and Smith Sustainable Economic Growth, Education Excellence, and Innovation Management through Vision 2020 2105