The Quarters for Cyclical Dominance (QCD) diagnostic and non-monotonic behaviour of the I/C ratio.
by
Valeria Kazakova
Coauthors: Alistair Gray

The Quarters for Cyclical Dominance (QCD) and non-monotonic behaviour of the I/C ratio.

Alistair Gray, Valeria Kazakova To judge the quality of the trend estimation for the quarterly time series the X-12 seasonal adjustment method uses the QCD statistic. The QCD statistic is equal to the quarter's span for which the I/C ratio falls below one for the first time. The problem is that for many very irregular series, the I/C ratio has a non-monotonic behaviour. In this case, the X-12 program often states the QCD value to be equal to the quarter's span for which the I/C ratio goes above one for the second time.

Possible explanation of the non-monotonic behaviour of the I/C ratio are either that a short-term business cycle is left in the trend or that outliers are influencing the I/C ratio. Motivated by these explanations we examined the effect of selecting a 5-Henderson filter (usually, for such series, a 7-Henderson filter is automatically selected for the trend) and using prior adjustment for outliers. This paper presents the results, which suggest that we can produce improved trend estimation by adopting this strategy.

Date received: August 16, 2000

Copyright © 2000 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cadt-22.