Nowcasting GDP using mixed-frequency based composite confidence indicators

Composite confidence indicators are widely used to nowcast GDP. In this paper, we aim to construct a new composite confidence indicator which weighting scheme reflects the impact of consumer and business confidence on economic conditions. While GDP is quarterly measured, confidence indicators are monthly recorded. Our approach allows us to deal with data sampled at different frequencies. In particular, we propose a weighting scheme estimation based on U-MIDAS regression techniques.