Nonlinear permanent - Temporary decompositions in macroeconomics and finance

We suggest a method of decomposing univariate and multivariate nonlinear processes into their permanent and temporary components, extending the analysis of Beveridge and Nelson (1981) and Stock and Watson (1987). We provide applications in the univariate nonlinear case to recent work on nonlinearities in the US business cycle, and in the multivariate nonlinear case to recent work on asymmetric nonlinear adjustment in the term structure of interest rates for the US. The business cycle results suggest that the method may be particularly useful in future research on output fluctuations.