This paper proposes a high-frequency coincident economic activity indicator, constructed according to the dynamic factor model methodology introduced by Stock and Watson (1998a,b). The principal component analysis is used to summarize the information contained in a large dataset in a limited number of common factors, capable of capturing the main features of local business cycles. The EM (Expectation Maximization) algorithm then allows to obtain an estimate of the year-to-year variation of regional GDP or provincial value added. The exercise is applied to Lombardy and to the provinces of Milan and Pavia.
Keywords: Coincident economic activity indicators, italian regions, diffusion indexes.
Jel Code: E32, C32, C82.