Forecasting brazilian inflation

Inflation forecasting is an extremely important task, as it can help economic agents make decisions. The aim of this study was to use ARIMA and ETS models to forecast monthly variation in the IPCA, the main inflation index used in Brazil. The ARIMA model presented residuals with satisfactory behavior, whereas the ETS presented non-independence. However, both showed very low predictive performance.

Important files contained in this repository:

• data: raw data file.
• Results: final paper as a PDF file.
• script: R script that produced the results.
• utils: R script with functions used in the main R script.