Abstract
The article analyzes Kazakhstan’s main macroeconomic indicators, namely gross domestic product (GDP), inflation, and unemployment, using ARIMA and VAR econometric models. The country’s exposure to external factors makes accurate forecasting and interpretation of these indicators especially important. The purpose of the study is to assess how well classical time series methods model key macroeconomic variables in Kazakhstan and reveal links between them. ARIMA and VAR models are compared using annual data for 1994-2023, and the calculations are carried out in R studio. The results show that classical time series models are suitable for short-term forecasting and for examining the main dynamic relationships between the variables. At the same time, their limitations become more visible dur-ing structural breaks, such as the COVID-19 pandemic. For this reason, time series models should be used with cau-tion in such periods and, where necessary, combined with alternative approaches. The findings may be useful for economic policy planning, anti-crisis measures, and future work on improving forecasting accuracy.