a b s t r a c tIn this paper we resort to singular spectrum analysis to disentangle the US GDP into several underlying components of interest. The business cycle indicator yielded through this method is shown to bear a resemblance with band-pass filtered output, and our results suggest it possesses better revision performance than other commonly applied filters.

We explore a new approach for nowcasting the output gap based on singular spectrum analysis. Resorting to real-time vintages, a recursive exercise is conducted so to assess the real-time reliability of our approach for nowcasting the US output gap, in comparison with some well-known benchmark models. For our applied setting of interest, the preferred version of our approach consists of a multivariate singular spectrum analysis, where we use a Fisher g test to infer which components, within the standard business cycle range, should be included in the grouping step.We find that singular spectrum analysis provides a reliable assessment of the cyclical position of the economy in real-time, with the multivariate approach outperforming substantially the univariate counterpart.

The modeling of multivariate extremes has received increasing recent attention because of its importance in risk assessment. In classical statistics of extremes, the joint distribution of two or more extremes has a nonparametric form, subject to moment constraints. This paper develops a semiparametric model for the situation where several multivariate extremal distributions are linked through the action of a covariate on an unspecified baseline distribution, through a socalled density ratio model. Theoretical and numerical aspects of empirical likelihood inference for this model are discussed, and an application is given to pairs of extreme forest temperatures.

No abstract