Avaliando o forecast content e o forecast horizon de modelos ARMA para os principais agregados de arrecadação de ICMS do Estado do Ceará
The ARMA model is a simple tool which is widely used in research for predicting future values of time series. For this reason, this dissertation's main goal is to evaluate the efficiency of Forecast Content and Forecast Horizon of ARMA models of total tax revenue from ICMS in Ceara. In other wo...
| Author: | |
|---|---|
| Format: | master thesis |
| Status: | Published version |
| Publication Date: | 2019 |
| Country: | Brasil |
| Institution: | Universidade Federal do Ceará (UFC) |
| Repository: | Repositório Institucional da Universidade Federal do Ceará (UFC) |
| Language: | Portuguese |
| OAI Identifier: | oai:repositorio.ufc.br:riufc/50123 |
| Online Access: | http://www.repositorio.ufc.br/handle/riufc/50123 |
| Access Level: | Open access |
| Keyword: | Arrecadação de ICMS Modelo ARMA Forecast Content Forecast Horizon Séries Temporais Previsão |
| Summary: | The ARMA model is a simple tool which is widely used in research for predicting future values of time series. For this reason, this dissertation's main goal is to evaluate the efficiency of Forecast Content and Forecast Horizon of ARMA models of total tax revenue from ICMS in Ceara. In other words, we investigate how well ARMA model preserves its ability to predict ICMS better than average as we increase the time horizon. Our dataset was extracted from Boletim de Arrecadacao Mensal dos Estados e do Distrito Federal between January 1997 and May 2019, available at CONFAZ's official website. Our results show that ARMA keeps good performance in a horizon of 6-12 months. This should be taken as a reference (or standard model) when predicting the state's tax revenue. |
|---|