Modelos de previsão para o ICMS da mesorregião noroeste do estado do Ceará

Abstract

This study consists of a time series analysis for quarterly forecasting of the tax on circulation of goods and on rendering of services on an interstate and intermunicipal level (ICMS) for the Northwestern region of the State of Ceará. The objective is to address different time series analysis methodologies, in order to provide different models for public managers for prediction and consistent with satisfactory predictive power, which can contribute to help the development of public policies and increased the municipal and regional economic welfare for the state of Ceará. To achieve this goal, the following techniques were used: ARIMA models, Exponential Smoothing, Transfer Function and the combination of forecasts for prediction. This work begins with an analysis of the ICMS tax, followed by a review of the literature which addresses the theoretical models used, as well as the empirical results obtained in the study. Lastly, it is observed that the results achieved corroborate the postulates of similar studies conducted on the subject, which confirm the relevance of the models used in time series analysis and its importance as instruments of prediction.