Abstract (EN)

The aim of this paper is the identification of structural shocks which affect the dynamics of sectoral inventories, through an analysis of the forecasting error variance decomposition. This identification is achieved using the American and French time series of aggregate production, sectoral production and sectoral inventories. The results are similar for both countries : aggregate production is mainly lead by aggregate shocks. The sources of fluctuations in sectoral time series change with time. First, the idiosyncratic supply shock is dominant in the short run of sectoral production, then aggregate shocks become significant. Conversely, sectoral inventories are influenced in the short run by the idiosyncratic demand shock and in the long run by the supply shock.