Business Cycle Extraction of Euro-Zone GDP: Direct versus Indirect Approach

Abstract

SummaryMost of the Euro-zone economic short-term indicators are computed through aggregation from Member States data. The seasonally adjusted figures can be calculated by seasonally adjusting the aggregate (direct approach) or aggregating the seasonally adjusted national data (indirect approach). Statistical and practical considerations to choose the right strategy are given in the paper. An application to the Euro-zone GDP is presented. The same aggregation problem encountered in the case of seasonal adjustment will persist when extracting the business cycle. Moreover, since raw figures imply problems in terms of excessive noise of the series, analysts generally prefer the use of seasonally adjusted time series. As a consequence, the problem of choosing between direct and indirect both in seasonal adjustment and in business cycle extraction appears to be closely linked. In fact, the approach chosen to seasonally adjust the data can in theory lead to different results when the cyclical component has to be extracted from seasonally adjusted data. After a review of different filters widely used in the literature, we extracted the cycle indicator for the Euro-zone employing the Baxter-King filter to data coming from both direct and indirect seasonally adjustment approach and then compared the relative results.

Detecting Cyclical Turning Points

“…If dating the classical business cycle is not so easy, then dating the growth cycle is even more difficult since the series must first be de-trended. Moreover, the way the series is seasonally adjusted (directly or indirectly for geographic aggregates like Eurozone indicators) and previously adjusted for calendar effects may impact on the datation (see for example Astolfi et al, 2001 andLommatzsh andStephan, 2001). It may therefore happen that different estimates are available on the market: we refer, for instance, to Anas (2000) or Krolzig (2001) for Eurozone business cycle datations.…”

A Comparison Between Direct and Indirect Seasonal Adjustment of the Chilean GDP 1986–2009 with X-12-ARIMA

“…The question about how we know if it is better to use direct or indirect adjustment is similar to that treated in Astol…, Ladiray, and Mazzi (2001), Hood and Findley (2001), Otranto and Triacca (2002), among other papers. Nevertheless, to the author's knowledge, no similar study has been carried out with Chilean GDP data.…”

“…These capabilities include spectral plots, sliding-spans-based diagnostics and revision history diagnostics, all of them simple checks that cast for both quality and stability of results. 3 By performing this exercise, I will be able to provide an informed opinion about which scheme provides the most stable seasonal adjustment for the mentioned Chilean GDP vintage. Obviously, the same strategy can be applied to another dataset to investigate the same problem.…”