TIME SERIES ANALYSIS FOR THE QUARTERLY GROSS DOMESTIC PRODUCT OF AMAZONAS

• DOI: 10.22533/at.ed.2163212330085

• Palavras-chave: GDP, Econometrics, Time Series, Macroeconomics, Statistics.

• Keywords: GDP, Econometrics, Time Series, Macroeconomics, Statistics.

• Abstract:

  In this work, estimates were made for the GDP of Amazonas, in order to help managers in decision-making, seeing that currently the GDP has a two-year lag. The source of information for the data that make up the GDP is from the Brazilian Institute of Statistics and Geography (IBGE) in partnership with the Secretariat for Development, Science, Technology and Innovation (SEDECTI) using a historical series from the first quarter of 2010 to the fourth quarter de 2020. For data analysis and later making estimates, the sectors were broken down by their activities, Agriculture, Industry and Services. Three models were used for the estimates: ARIMA, SARIMA and ARFIMA, being necessary to verify the most adequate model for each variable. First, the stationarity of the data was verified through graphical analysis and the Said & Dickey (1984) – ADF and Kwiatkowski (1992) – KPSS test. These tests generated four results, providing guidance on which model is appropriate. After the application of the ADF and KPSS tests, the d parameter was estimated, which is the order of the differences necessary to remove the tendency of the series, using the GPH and Reisen methods. The most appropriate model was chosen using the Akaike Information Criterion (AIC). Finally, an estimate was generated for the years 2021 to 2023 and verified whether the selected models adequately describe the dynamics of the data. The result generated three models, namely: ARIMA, SARIMA AND ARFIMA and the estimates had a satisfactory dynamic in relation to the data set, that is, how close the predicted data are in relation to the values later observed. Through GDP estimates, managers can make better decisions, since GDP is the main indicator for the Government to make its budget planning for investments in health, security, education and infrastructure.