Welcome to the fascinating world of seasonal adjustment with JDemetra+ (JD+). Please note that this site is still under construction.
This repository is basically a collection of course material used in various Bundesbank trainings on seasonal adjustment. These trainings are hosted by the Centre for International Central Bank Dialogue (CIC) and cover the annual introductory seminar Elementary seasonal adjustment of economic data with JDemetra+ (Level I) and the biennial advanced seminar Advanced topics in seasonal adjustment (Level II). The folder structure is as follows:
Year Level I/II: PDF presentation slides and additional files, such as data files provided by the participantsLiterature: selected literature on seasonal adjustment topics covered in at least one of the two seminars
Working exclusively with monthly and quarterly data, the introductory seminar provides basic information about the following topics:
- Concept of seasonality and aims of seasonal adjustment
- Introduction to JD+
- Data pretreatment (log transformation, prior adjustment for outliers and calendar variation)
- Seasonal adjustment with the X-11 approach
- Seasonal adjustment with the ARIMA model-based (AMB) approach
- Quality diagnostics
- Direct versus indirect seasonal adjustment of composite time series
- Overview of revision policies
Focusing more on specific data issues and proper streamlining of the entire seasonal adjustment process, the advanced seminar covers the following topics:
- Recap of introductory seminar
- Additional tools for JD+ and access through R
- Outlier modelling in times of strong economic changes
- User-defined exogenous variables for data pretreatment
- Details on the AMB approach
- Seasonal adjustment of daily time series
- Direct versus indirect seasonal adjustment of composite time series
- Revision policies: the controlled current adjustment strategy
Do you have a question or want to leave a comment? No problem, just send an e-mail to the Bundesbank's JD+ development team or create an issue directly on GitHub, which is our preferred communication channel.