Macroeconomic forecasting using penalized regression methods

We study the suitability of lasso-type penalized regression techniques when applied to macroeconomic forecasting with high-dimensional datasets. We consider performance of the lasso-type methods when the true DGP is a factor model, contradicting the sparsity assumption underlying penalized regression methods. We also investigate how the methods handle unit roots and cointegration in the data. In an extensive simulation study we find that penalized regression methods are morerobust to mis-specification than factor models estimated by principal components, even if the underlying DGP is a factor model. Furthermore, the penalized regression methods are demonstrated to deliver forecast improvements over traditional approaches when applied to non-stationary data containing cointegrated variables, despite a deterioration of the selective capabilities. Finally, we also consider an empirical application to a large macroeconomic U.S. dataset and demonstrate that, in line with our simulations, penalized regression methods attain the best forecast accuracy most frequently.