Previsão de Variáveis Macroeconômicas Brasileiras usando Modelos de Séries Temporais de Alta Dimensão

This paper analyzes the performance of high-dimensional factor models to forecast four Brazilianmacroeconomic variables: two real variables, unemployment rate and industrial productionindex, and two nominal variables, IPCA and IPC. The factors are estimated from a data setcontaining 117 macroeconomic variables. We applied techniques to improve factor models forecasts. Methods of statistical learning are applied aims to increase the performance of factorsmodels. Three types of statistical learning techniques are used: shrinkage methods, forecastcombinations, and selection of preditors. The factors are extracted using supervised and unsupervisedversion. The results indicate that statistical learning improves forecasts performance.The combination of statistical learning and supervised factor models is more accurate than allother models, with exception to the industrial production index which is best forecasted byunsupervised factor model without statistical learning.