Direct approximation of the trend-cyclical-component versus seasonal adjustment at the current end of a time series

This paper compares the efficiency of two methods explaining the cyclical movement of a time series at its current end. It refers to a newly developed low-pass filter and the seasonal adjustment method, represented by ASA-II. The empirical analysis starts with a visual comparison of the results of the alternative methods applied to five select rime series. These examples point to a superior informative value of the directly estimated cyclical development. Subsequent attempts to underpin these results using stochastic simulations support the conclusion that the new filter is an interesting alternative to conventional seasonal adjustment methods, especially for short time series.