Nonlinear Forecasting Using Factor-Augmented Models


Bruno Cara Giovannetti, Av. Prof. Luciano Gualberto, 908, Prédio FEA 2, Cidade Universitária, São Paulo - SP, CEP: 05508-010. E-mail:


Using factors in forecasting exercises reduces the dimensionality of the covariates set and, therefore, allows the forecaster to explore possible nonlinearities in the model. For an American macroeconomic dataset, I present evidence that the employment of nonlinear estimation methods can improve the out-of-sample forecasting accuracy for some macroeconomic variables, such as industrial production, employment, and Fed fund rate. Copyright © 2011 John Wiley & Sons, Ltd.