A B S T R A C T

The purpose of the research is to compare methods and approaches used in forecasting macroeconomic indicators and to explore the potential of forecasting macroeconomic indicators using nowcasting.

The methodology of the research is based on combines general and specific research methods, the application of machine learning technologies in management decision-making, and mathematical and statistical methods of analysis and forecasting. The study's hypothesis is that nowcasting will become an alternative method of macroeconomic analysis in the near future.

The practical importance of the research lies in the fact that the application of machine learning technologies in economic research in the context of digital transformation represents a new paradigm. The nowcasting approach, which allows for the identification of important macroeconomic trends in real time, may have significant practical implications for the development of highly accurate forecasts.

The results of the research in statistical practice, most important macroeconomic indicators are systematized quarterly. Due to the prevalence of quarterly data, quarterly econometric models based on aggregated data are more common. The practical application of such models faces a number of challenges, including delays in entering new values for macroeconomic indicators, loss of information on variable dynamics due to frequent failures, and revisions of source data. Therefore, economic research places particular emphasis on issues related to the revision of source data (vintage data) and assessing the impact of real-time data on forecasts, which also play a significant role in improving forecast accuracy.

The originality and scientific novelty of the research lie in the description and systematization of successful applications of the short-term forecasting method as a new methodological approach aimed at forecasting macroeconomic indicators with near-real-time accuracy, as well as in the substantiation of its core principle-the possibility of implementing more accurate forecasts of macroeconomic indicators as new data becomes available.

Keywords: macroeconomic indicators, forecasting, machine learning technologies, realtime mode, short-term forecasting, nowcasting.

https://doi.org/10.30546/2707-2037.052.2.2026.1038