Weekly WIFO Economic Index

The WIFO Weekly Economic Index (WWWI) is a measure for calculating the weekly and monthly real economic activity of the Austrian economy. For this purpose, a mixed-frequency approach with weekly, monthly and quarterly indicators is used to estimate weekly and monthly values for real GDP and 18 GDP sub-aggregates (use side 8, output side 10). In the WWWI system, year-on-year growth rates are used to estimate the models and are estimated based on dynamic factor models in a state-space representation using the Kalman filter.

 

Methodological description

The WIFO Weekly Economic Index (WWWI) is a calculation system for calculating the real economic activity of the Austrian economy on a weekly and monthly basis. In addition to the weekly and monthly GDP indicators, ten sub-aggregates of the output side and eight sub-aggregates of the expenditure side are estimated on the basis of the Quarterly National Accounts. The WWWI is obtained by summing the growth contributions of the sub-indicators on the output side, which are fully mapped, whereas the output side has a residual reflecting the joint contribution of changes in inventories and the trade balance with non-tourism services.

For each weekly sub-indicator, two types of econometric models are estimated: first, temporal disaggregation models for historical decomposition of previously published monthly and quarterly data to weekly frequency, and second, a nowcasting model for forecasting weekly values at the current margin based on the estimated historical weekly GDP sub-aggregates and observed weekly indicators. The latter include cashless transactions, freight transport performance, passenger flight volumes, Google mobility data, persons registered as unemployed, electricity consumption and pollutant emissions from industry, and international weekly indicators of economic activity.

The estimated disaggregation models are formulated as dynamic factor models in a state-space representation and transformed as forecasts of unobserved weekly values of a quarterly or monthly time series by treating them as missing data and estimating them with the Kalman filter.

All models are updated after each release of new Quarterly National Accounts data and thus always remain consistent with the current national accounts. Some models are additionally revised after the release of relevant monthly data, which may also lead to a revision of the WWWI.

 

Publications

Sandra Bilek-Steindl, Julia Bock-Schappelwein, Christian Glocker, Serguei Kaniovski (WIFO), Sebastian Koch, Richard Sellner (IHS)
Hochfrequente Konjunkturbeobachtung (High-frequency Business Cycle Monitoring)
Studies, October 2020, 75 pages
Commissioned by: Federal Ministry of Finance
Study by: Austrian Institute of Economic Research – Institute for Advanced Studies
Online since: 23.10.2020 0:00
 
In dieser Studie wird eine große Zahl hochfrequenter Indikatoren gesammelt und hinsichtlich ihres Informationsgehaltes im Hinblick auf die Aktivität der österreichischen Volkswirtschaft analysiert. Dies geschieht anhand von Kreuzkorrelationen dieser und anderer häufig verwendeter Indikatoren für verschiedene Referenzreihen, u. a. für das BIP, das verarbeitende Gewerbe, den Bausektor, Teile des Dienstleistungssektors, Investitionen, Konsum, Importe und Exporte. Auf Basis dieser Erkenntnisse werden dynamische Faktormodelle (DFM) für die Referenzreihen spezifiziert. Dies ermöglicht eine Echtzeiteinschätzung und eine Prognose des BIP sowohl durch einen direkten (wöchentlicher WIFO-Wirtschaftsindex – WWWI) als auch durch einen indirekten Ansatz (Cluster dynamischer Faktormodelle – CDFM). Eine Auswahl an hochfrequenten Indikatoren ist auch über den IHS Economic High-Frequency Monitor abrufbar.