Sparse and Robust Factor Modelling
Factor construction methods are widely used to summarize a large panel of variables by means of a relatively small number of representative factors. We propose a novel factor construction procedure that enjoys the properties of robustness to outliers and of sparsity; that is, having relatively few nonzero factor loadings. Compared to more traditional factor construction methods, we find that this procedure leads to better interpretable factors and to a favorable forecasting performance, both in a Monte Carlo experiment and in two empirical applications to large data sets, one from macroeconomics and one from microeconomics.
|Keywords||dimension reduction, forecasting, outliers, regularization|
Croux, C., & Exterkate, P., P.. (2011). Sparse and Robust Factor Modelling (No. TI 2011-122/4). Tinbergen Institute. Retrieved from http://hdl.handle.net/1765/25712