Why are Beveridge-Nelson and Unobserved-Component Decompositions of GDP so Different?
Two widely used methods of decomposing GDP into trend and cycle yield starkly different results. The unobserved component approach implies smooth trend with large, persistent cycle. In contrast, the Beveridge and Nelson (1981) approach implies most of the variation is attributable to trend. This conflict has been widely noted. It should surprise us that the two approaches produce very different trend-cycle decompositions since both are model-based. This paper attempts to find out why we do not, after decades of research, have a consistent picture of how variation in a series like real GDP should be allocated between trend and cycle.