Volume 8, No. 1, April 200 9
Asymmetries in Macroeconomic Time Series
in Eleven Asian Economies


Khurshid M. Kiani*
Department of Economics, University of Nottingham, Ningbo, China


Abstract


We investigate business cycle asymmetries in the real GDP of eleven selected Asian economies using nonlinear switching time series models and artificial neural networks. Results based on neural network linearity tests show evidence of business cycle asymmetries in all series. Results based on switching and augmented time series models also reveal business cycle asymmetries in most series studied.


Key words : real GDP growth rates; fat tails; stable distributions; neural networks; out-of-sample forecasts; long memory; nonlinearities; business cycles
JEL classification : C22; C32; C45; C53

Back