Abstract:
Using a recently introduced nonparametric test, I investigate two important and distinct asymmetries in cross-country quarterly macroeconomic time series. Asymmetries are suggested by many theories (old and new), and those discovered aid in the selection of the appropriate nonlinear time series representation (useful, for example, in both forecasting and policy guidance). Further, asymmetries can help determine underlying economic mechanisms. The key findings: positive growth rate asymmetry is nearly ubiquitous in price level data (but is not caused by money growth asymmetry); and the pattern of asymmetries varies dramatically across countries (making widespread reliance on US data to study fluctuations worrisome).