Combining forecasts under structural breaks using Graphical LASSO

In this paper we develop a novel method of combining many forecasts based on Graphical LASSO. We represent forecast errors from different forecasters as a network of interacting entities and generalize network inference in the presence of common factor structure and structural breaks. First, we note that forecasters often use common information and hence make common errors, which makes the forecast errors exhibit common factor structures. We separate common forecast errors from the idiosyncratic errors and exploit sparsity of the precision matrix of the latter. Second, since the network of experts changes over time as a response to unstable environments, we propose Regime-Dependent Factor Graphical LASSO (RD-FGL) that allows factor loadings and idiosyncratic precision matrix to be regime-dependent. The empirical applications to forecasting macroeconomic series using the data of the European Central Bank’s Survey of Professional Forecasters and Federal Reserve Economic Data monthly database demonstrate superior performance of a combined forecast using RD-FGL.