Alina Stundžienė


There are several institutions that constantly announce their predictions of general domestic product (GDP) and lots of institutions that use this information. Frequently the forecasts of different institutions vary because they use different methods, but all the institutions for which this indicator is relevant cannot make the predictions themselves because the models are too sophisticated. The main purpose of this research is to create simple enough but also accurate model for prediction of Lithuanian GDP that can be used by all the institutions that need this indicator. The research is based on the economic data that are measured and published quarterly or monthly by Statistics Lithuania. 154 economic indicators were analysed as possible independent variables for regression model creation. The analysis showed that the regression model with twelve lag independent variables can be quite accurate for a short-term prediction of Lithuanian GDP. It can be forecasted by such indexes as the number of immigrants, the turnover of wholesale and retail trade and repair of motor vehicles and motorcycles, the number of overnight stays in the accommodation establishments, an average hourly earnings, the rate of change in the producer prices of all the industry (except construction) of Lithuanian market, the imports and seasonally adjusted imports, the seasonally adjusted exports, the projected number of employees in the trade enterprises for the next 2–3 months, the industrial production (of all the industry except construction).



Lithuanian GDP; forecasting; model creation; regression models; lag models

Full Text: PDF

Print ISSN: 1392-2785
Online ISSN: 2029-5839