An Approach to Identify First Signs of a Development toward a Cyclical Turning Point.

The usual estimation methods provide only a blurred picture of the cyclical development for the current values of a time series, the section that is of most interest for short-term economic analysis. As an alternative, an approach to identify first signs of a development toward a cyclical turning point is suggested. The method is based on the moving adaptation of a parabola to predetermined sections of time series and determines a change in the curve of the consecutive parabolas as a first signal of a turning point. The different features of economic time series call for a time series specific design of the method. Some measures, like the determination of the length of the parabola or the moderation of the irregular movement in advance, allow a certain turning. On principle, the lead and the reliability of the signals decrease with the share of the irregular movement on the variance of the time series and increase with the smoothness of its cyclical development.