Structural Change in Swedish and Finnish Monthly Industrial Output Series

  • Timo Teräsvirta
Conference paper

Summary

This chapter considers the hypothesis of no structural change in Swedish and Finnish industrial output series. The alternative hypothesis to a linear regression model is a rather flexible parametric form of structural change called smooth transition regression. Its parameters may thus be estimated if the null hypothesis is rejected. The null hypothesis may be tested without actually estimating the alternative using a simple F-test. The test procedure is derived in this chapter. The null hypothesis is rejected for both Swedish and Finnish autoregressions. There may be other reasons for rejection than structural change; however, a smooth transition model corresponding to the alternative is successfully estimated by nonlinear least squares for both countries. The results indicate that the structural change has been similar in both countries but has occurred in Sweden about a decade earlier than in Finland. The two time series are also found not to be cointegrated.

Keywords

Explosive Autocorrelation OECD 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Box, G.E.P. and Jenkins, G.M. (1970), Time Series Analysis, Forecasting and Control. San Francisco: Holden-Day.Google Scholar
  2. Brock, W.A. and Sayers, C.L. (1988), Is the business cycle characterized by deterministic chaos? Journal of Monetary Economics, 22, 71–90.CrossRefGoogle Scholar
  3. Chan, K.S. and Tong, H. (1986), On estimating thresholds in autoregressive models. Journal of Time Series Analysis, 7, 179–190.CrossRefGoogle Scholar
  4. Dickey, D.A. and Fuller, W.A. (1979), Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association, 74, 427–431.Google Scholar
  5. Farley, J.V., Hinich, M., and McGuire, T.W. (1975), Some comparisons of tests with a shift in the slopes of a multivariate linear time series model. Journal of Econometrics, 3, 297–318.CrossRefGoogle Scholar
  6. Goldfeld, S.M. and Quandt, R.E. (1973), The estimation of structural shifts by switching regressions. Annals of Economic and Social Measurement, 2, 475–485.Google Scholar
  7. Granger, C.W.J, and Lee, H.S. (1990), An introduction to time varying parameter cointe-gration, in P. Hackl and A.H. Westlund (eds.), Economic Structural Change. Analysis and Forecasting. Berlin: Springer-Verlag.Google Scholar
  8. Granger, C.W.J, and Lee, T.-H. (1989), The effect of aggregation on nonlinearity. Economics Discussion Paper No. 89–43, University of California, San Diego.Google Scholar
  9. Krämer, W. and Sonnberger, H. (1986), The Linear Regression Model Under Test. Heidelberg: Physica-Verlag.CrossRefGoogle Scholar
  10. Kunst, R. (1989), Cointegration in macroeconomic systems: Seasonality and explosive roots. Research Memorandum No. 255, Institute for Advanced Studies, Vienna.Google Scholar
  11. Luukkonen, R. and Teräsvirta, T. (1990), Testing linearity of economic time series against cyclical asymmetry. Discussion Paper No. 262 (revised version), Research Institute of the Finnish Economy.Google Scholar
  12. Luukkonen, R., Saikkonen, P., and Teräsvirta, T. (1988), Testing linearity against smooth transition autoregressive models. Biometrika, 75, 491–499.CrossRefGoogle Scholar
  13. MacKinnon, J.G. (1990), Critical values of cointegration tests. Economics Discussion Paper No. 90–4, University of California, San Diego.Google Scholar
  14. Tsay, R.S. (1986), Nonlinearity tests for time series. Biometrika, 73, 461–466.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1991

Authors and Affiliations

  • Timo Teräsvirta

There are no affiliations available

Personalised recommendations