In this paper we present a coincident indicator for the Italian economy, Ita-coin. We construct a multivariate filter based on a broad information set, whose dimension is reduced by the Generalized Dynamic Factor Model (GDFM) approach proposed by Forni et al. (2002). A regression based on the least absolute shrinkage and selection operator (LASSO) is used to estimate Ita-coin. Most Italian macroeconomic indicators are characterized by high short-term volatility and the 2008-2009 crisis has affected the volatility of both the high- and low-frequency components and the relationships between the variables have become more unstable. LASSO regression allows us to select recursively the relevant information about the comovement of the variables over time. Our indicator displays a satisfactory performance in the pseudo real-time validation as a timely cyclical indicator.