Quantifying Drivers of Forecasted Returns Using Approximate Dynamic Factor Models for Mixed-Frequency Panel Data

This article considers the estimation of Approximate Dynamic Factor Models with homoscedastic, cross-sectionally correlated errors for incomplete panel data. In contrast to existing approaches, presented method comprises two expectation-maximization algorithms and uses conditional factor moments in closed form. To determine unknown dimension autoregressive order, we propose a two-step information-based model selection criterion. The performance our procedure criterion is investigated within Monte Carlo study. Finally, apply Model real-economy vintage data support investment decisions risk management. For this purpose, an estimated span mixed-frequency as exogenous variables maps behavior weekly S&P500 log-returns. We detect main drivers index development define dynamic trading strategies resulting from prediction intervals subsequent returns.