Trend und Zyklus im Bruttoinlandsprodukt der Bundesrepublik Deutschland / Trends and Cycles in the Gross Domestic Product of the Federal Republic of Germany: Zur ökonometrischen Ermittlung der empirischen Basis der Konjunktur- und Wachstumstheorie / On the Econometric Identification of the Empirical Components of Business Cycles and Growth Theory

For a long time economists were in broad agreement that gross domestic product develops with stationary cyclical fluctuations around a deterministic trend. The long-run smooth movement is viewed as being in the domain of growth theory while the cyclical component is a prominent issue of business cycles theory. The firm statistical foundation of the deterministic trend hypothesis was questioned by the work of Nelson and Plosser (1982) whose findings suggest that macroeconomic time series are of stochastic rather than deterministic nature. Then the decomposition of macroeconomic time series in deterministic trend and stationary cyclical fluctuations must be misleading. To decide whether the time series of gross domestic product in Germany shows a deterministic or stochastic trend a regression which embeded both hypotheses is estimated from deflated annual post war data and the unit root test is carried out. In addition the test is repeated allowing under both hypotheses for the presence of two exogenous structural breaks. To quantify the ability of the unit root test to reject a false hypothesis specifically calibrated difference stationary and trend stationary models are estimated to simulate small-sample distributions for the various unit root test statistics under both hypotheses. The results of that analysis and the evidence reported in other recent works suggest that the time series of the gross domestic product of Germany and of other economies is not characterized by the presence of a unit root. It may be more advantageous to adopt the trend stationary hypothesis with exogenous breaks and detrend a deterministic trend to provide the cyclical component for empirical purposes.