Termiz State Pedagogical Institute

Abstract

This article forecasts sectoral time series of Uzbekistan’s entrepreneurship sector using ARIMA and SARIMA models. Monthly observations for 2022–2026 cover agriculture, industry, construction, wholesale and retail trade, transportation and storage, accommodation and food services, information and communication, and health and social services. ADF and KPSS tests support first differencing (d=1) for all series. According to AIC, SARIMA outperforms ARIMA in all eight sectors, whereas RMSE indicates a closer in-sample fit for ARIMA. Ljung–Box p-values exceed 0.05, suggesting no material residual autocorrelation. By December 2029, growth is projected in agriculture, accommodation and food services, information and communication, and health and social services, while construction shows a declining path. SARIMA is therefore treated as the main forecasting model and ARIMA as a benchmark and control specification.