Introduction

  • Ulrich Woitek
Part of the Contributions to Economics book series (CE)

Abstract

The study has two related purposes: First, I want to extend and deepen the empirical knowledge regarding the stylized facts of business cycles. Particular emphasis is placed upon the stability of empirical regularities by means of international and historical comparison. Moreover, a spectral analysis method relatively novel in the application to economic time series is presented and implemented. This method is Maximum Entropy (ME-) spectral analysis, which was developed during the 1960s and 1970s in the natural sciences. ME-spectral estimation is especially suitable for econometric purposes, since it efficiently extracts the information on the cyclical structure contained in very short time series. The application of the ME-spectrum to the analysis of business cycles was pioneered at SEMECON, University of Munich, but thus far only for univariate series. The present study extends this approach to the multivariate case.

Keywords

Business Cycle Stylize Fact Cyclical Structure Empirical Knowledge Time Domain Method 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    The term Kitchin- cycle refers to Joseph Kitchin (1923), who discovered a short cycle with a length of 40 month (= 3.33 years) in time series of interest rates, commodity prices and clearings for the United States and the United Kingdom in the period 1890–1922.Google Scholar
  2. 2.
    Kuznets (1958) found a cycle with a length of 22 years in economic variables. Other authors report similar results (15–25 years), see e.g. the overview in Abramovitz (1964).Google Scholar
  3. 3.
    See e.g. the citation of Alvin Hansen (Samuelson (1961), p. 289–290):Google Scholar
  4. The American experience indicates that the major business cycle has had an average duration of a little over eight years. Thus, from 1795 to 1937 there were seventeen cycles of an average duration of 8.35 years. [...]Google Scholar
  5. Since one to two minor peaks regularly occur between the major peaks, it is clear that the minor cycle is something less than half the duration of the major cycle. In the one hundred and thirty-year period 1807 to 1937 there were thirty-seven minor cycles with an average duration of 3.51 years. [...] [...] it appears that the building cycle averages somewhere between seventeen and eighteen years in length, or almost twice the length of the major business cycle. [...]Google Scholar
  6. 4.
    The problematic nature of statistical testing in econometrics and the natural science alternative are discussed in Hillinger (1992a) and Reiter (1995), pp. 9–12.Google Scholar
  7. 5.
    Two recent German textbooks on business cycles (Maußner (1994) and Tichy (1994)), which consider the traditional as well as the modern approaches, are noteworthy for their emphasis on stylized facts in relation to explanatory theories.Google Scholar

Copyright information

© Physica-Verlag Heidelberg 1997

Authors and Affiliations

  • Ulrich Woitek
    • 1
  1. 1.Seminar für Wirtschaftsgeschichte, SEMECONUniversität MünchenMünchenGermany

Personalised recommendations