Eurostat tools for benchmarking and seasonal adjustment j_demetra+ and jecotrim_buono_final_20130820

Practical issues in benchmarking and seasonal Adjustment

1. Eurostat tools for Benchmarking and Seasonal
adjustment JDemetra+ and JEcotrim
by
Dario BUONO, Ph.D.
Eurostat, European Commission, Luxembourg
Macroeconomic Imbalances Procedure team
IASC satellite conference on
"Big data and computational statistics"
August 22
nd
2013, 13:45, Seoul, Korea
SESSION SS2R5: Practical Issues in Chain Linking and Benchmarking

2. Objective
Inform the about the Eurostat IT tools available for
Seasonal Adjustment and Benchmarking and
Temporal Disaggregation
No methodological issues will be addressed
Summary information on JDEMETRA+ with briefing on
the ESS guidelines on Seasonal Adjustment
Summary information on JECOTRIM
User support provided by Eurostat
Q & A

3. Content
Recall of seasonal adjustment issue;
An overview of the seasonal adjustment software;
IT solutions for SA software;
The role of Eurostat;
JDemetra+;
–Aims of the project, Functionalities, Advantages;
JEcotrim: a plugin of JDEMETRA+

4. What is Seasonal Adjustment?
SEASONAL
ADJUSTMENT
Fluctuations observed during the year (each
month, each quarter) and which appear to
repeat themselves on a more or less regular
basis from one year to the other
Remove Seasonality
Seasonality:

5. What happens after SA?
Original and Seasonally Adjusted series
Original Series

6. What happens after SA?
Original and Seasonally Adjusted series
The series has been cleaned!!
Seasonally Adjusted series

7. What happens after SA?
Growth Rates
Original Series
1
1
-=
-t
t
t
X
X
G

8. What happens after SA?
Growth Rates
Seasonally Adjusted series
1
1
-=
-t
t
t
X
X
G

9. Leading seasonal adjustment software – a
quick review
The main SA programs are:
–TSW – the Windows application, developed by Bank of
Spain, that integrates the TRAMO and the SEATS
programs;
–X-12-ARIMA and X-13ARIMA-SEATS – the programs
produced by the U.S. Census Bureau, that include X-12-
ARIMA method
•(X-13ARIMA-SEATS is also capable to generate
ARIMA model-based SA).
Both written in a FORTRAN language.

10. Seasonal adjustment software from an IT
perspective
The algorithms written originally in FORTRAN might be
applied to solve time series related issues, but are not
designed for reusability;
In case of introduction of the new functionality, the actual
programs are modified;
Uncertain future of the FORTRAN language;
–Lack of developers;
–Not a strictly object-oriented language.

11. Eurostat’s scopes in area of SA
Eurostat aims to:
–Promote the idea of seasonal adjustment;
–Ease an access of non-specialists to TRAMO/SEATS and
X-12-ARIMA (X-13ARIMA-SEATS);
–Converge towards a harmonised process for seasonally
and calendar adjustment practices.

12. SA software promoted by Eurostat
Demetra (2002);
–Initially successful because of its user-friendliness;
Demetra+ (2010);
–Implementation of the ESS Guidelines on SA;
–Provides graphical interface and common input/output
diagnostics for TRAMO/SEATS and X-12-ARIMA;
–Includes complex technical solutions. Cannot be used
under IT environments other than Windows (.NET)
technology;
JDemetra+ (2012);
–Fortran codes re-written in JAVA using NetBeans.
JECOTRIM (2013) as plug in of JDEMETRA+

13. What is JDEMETRA+?
JDEMETRA+ is a new tool for
Seasonal and Calendar Adjustment
developed by NBB and EUROSTAT
Identify more components
Trend-Cycle Component
Outliers
Irregular Component

14. New tool new issue!!
Maintenance of the tool in the long-term;
Integration of the libraries in the IT environments of
many institutions (portability issue of Demetra+.NET);
Re-use of the modules/algorithms for other
purposes.

15. Aims of the current project
Provide a tool for SA which:
–Is flexible, i.e.:
•encompasses the leading SA algorithms;
•could evolve independently when improvements or
alternative methods appear.
–Is versatile, i.e. can be:
•used in a rich graphical interface (JDemetra+ itself);
•integrated in other (in-house) developments.
–Consists of modules that can be reuse in the other
circumstances;
–Is an open source, and therefore may increase the
transparency of statistical computation and contribute to a
better sharing of the statistical knowledge.

16. JDemetra+ functionalities
SA methods:
–TRAMO/SEATS;
–X-13ARIMA-SEATS;
–X-12-ARIMA;
–Structural models;
–Mixed Airline;
–Generalised Airline.
SA tools:
–Seasonality tests;
–Direct/indirect comparison;
–Calendars with weights on holidays.
Other tools:
–Benchmarking (JEcotrim);
–Temporal disaggregation (JECOTRIM);

17. Advantages of JDemetra+
Efficient process of large datasets;
A user-friendly graphical interface.
Possibility for different teams to progressively take over
the software or to contribute to its evolution.
Core engines rewritten in Java, by NetBeans platform:
–supported by almost all IT operating systems;
–that allows for easy extensions (plug-ins) and
improvements;

18. Future plans
Modification and extension of the code, e. g.;
–Modification of existing functionalities;
–New data providers due to genetic serialization
functionality;
–Additional diagnostics and output;
–New seasonal adjustment methods (using batch
processing).
–Plug in for revision analysis
–Plugin for business cycle analysis

19. ESS Guidelines on SA
Introduced in 2009
Chapters subdivided into specific items describing different steps of the SA process
Items presented in a standard structure providing:
1.Description of the issue
2.List of options which could be followed to perform the step
3.Prioritized list of three alternatives from most recommended one to the
one to avoid (A, B and C)
4.Concise list of main references
Added value:
1.Conceptual framework and practical implementation steps
2.Both for experienced users and beginners
http://epp.eurostat.ec.europa.eu/cache/ITY_OFFPUB/KS-RA-09-006/EN/KS-RA-09-006-EN.PDF

20. JECOTRIM: un update of ECOTRIM
ECOTRIM contains procedures based on temporal
disaggregation, benchmarking, reconciliation of low
frequency series and matrix balancing via complex
mathematical and statistical methods.
Ecotrim was developed in C++ (for Windows) by Eurostat
JEcotrim can be defined simply as an upgrade of Ecotrim
–Correction of some bugs
–New methods
–Plugged to Jdemetra

21. Definition
Temporal Disaggregation
–Process of deriving high frequency data from low
frequency data and, if available, related high frequency
information

22. Temporal Disaggregation techniques are useful in
compiling short-term statistics:
Quarterly National Accounts (QNA)
Give a quarterly breakdown of the figures in the annual
accounts
Flash estimates
Use the available information in the best possible way
including, in the framework of a statistical model, the
short-term available information and the low frequency
data in a coherent way
Monthly indicators of GDP
The monthly estimates are derived from the available
information respecting the coherence with quarterly data

23. Basic principles
Distribution
–When annual data are either sums or averages of
quarterly data (e.g., GDP, consumption, indexes and in
general all flow variables and all average stock variables)
Interpolation
–When annual value equals by definition that of the fourth
(or first) quarter (e.g., population at the end of the year,
money stock, and all stock variables)
Extrapolation
–When estimates of quarterly data are made when the
relevant annual data are not yet available

24. Estimates have to be consistent and coherent
Temporal consistency
–Quarterly values have to match annual values (for
example the sum of quarterly values of the GDP must be
equal to the annual value)
Accounting coherence
–Quarterly components of an account should respect the
accounting constraints (for example, the sum of quarterly
values of the GDP expenditure side components should
be equal to the corresponding quarterly value of GDP)

25. Temporal Disaggregation and
Benchmarking available in JEcotrim
Univariate Approach (temporal Benchmarking)
–Modified Denton,
–Chow-Lin, Fernández,Litterman
Multivariate Approach (accounting Benchmarking)
–RAS-PM
–Two-Step reconciliation

26. Webpages on JDEMETRA+ and JEcotrim
You can download the latest version of the tools at
http://www.cros-portal.eu/content/seasonal-adjustment
Together with:
ESS guidelines on Seasonal Adjustment
User Manual and other documents
Help-Desk e-mail: [email protected]
Information about the EUROSTAT ESTP Training Course

27. Questions?