Title: Now-casting macroeconomic trends and cycles
Authors: Filippo Pellegrino - LSE; Now-Casting Economics (United Kingdom) [presenting]
Lucrezia Reichlin - London Business School - Now-Casting Economics - CEPR (United Kingdom)
Giovanni Ricco - London Business School (United Kingdom)
Thomas Hasenzagl - University of Minnesota (United States)
Abstract: Building on recent developments in semi-structural econometric models, a new approach is proposed to nowcast key economic indicators and understand inflation dynamics. A mixed-frequency dataset is used that includes nominal, expectational, and real macroeconomic time series and timely survey data. These variables are processed in real-time, according to the official calendar of publications. As data becomes available, model estimates are updated to perform a timely trend-cycle decomposition and produce a new sequence of now-casts of the observed variables. The parameters are estimated via Bayesian methods with weakly informative priors. The models restrictions are informed by macroeconomic theory encompassing different hypotheses on the Phillips curve, the Okuns law, the permanent income hypothesis and the cyclical behaviour of oil prices.