WIFO Leading Indicator

The WIFO Leading Indicator is a monthly composite indicator designed to provide timely indication of business cycle turning points in the Austrian economy. It consists of a combination of 10 different indicators, each having a constant lead to the overall Austrian economic activity of up to 3 quarters to macroeconomic activity in Austria. These individual indicators relate to about half each to the Austrian and European economic area and, in addition to real economic indicators and stock indices, mainly comprise survey data.

 

Methodological description

Publication

This paper describes the methodologies used for constructing a composite leading indicator for the Austrian economy (CLI-AT). First, a selection of those monthly indicators which overall fare best in showing a "steady" leading behaviour with respect to the Austrian business cycle was performed. The analysis was carried out by means of statistical methods out of the timeseries domain as well as from the frequency domain. Thirteen series have been finally classified as leading indicators. Among them, business and consumer survey data form the most prevalent group. Second, I construct the CLI-AT based on the de-trended, normalised and weighted leading series. For the de-trending procedure I use the HP filter and the weights have been obtained by means of principal components analysis. Further, idiosyncratic elements in the CLI-AT have been removed along with checking the endpoint-bias due to the HP filter smoothing procedure. I find that the "real-time" smoothed CLI-AT does not exhibit severe phase-shifts compared to a full-sample estimate. Next, I show that the CLI-AT provides a useful instrument for assessing the current and likely future direction in the Austrian business cycle. Over the period 1988-2008, the CLI-AT indicates cyclical turns with a "steady" lead in the majority of cases. Finally, in using an out-of-sample forecasting exercise it is shown that the CLI-AT carries important business cycle information and that its inclusion in a forecasting model can increase the projection quality of the underlying reference series.