Transcript Slide 1

Introduction to Ecotrim and
DEMETRA+
Lunch time presentation 14th of April 2011
Dario BUONO and Enrico INFANTE
Unit B2 Methodology and Research
14/04/2011
Objectives
 Inform the production units about the Eurostat IT tools
available for Temporal Disaggregation and Seasonal
Adjustment
 Identify audience for a possible internal courses
(including demo session) at ESTAT on the subject
 No methodological issues will be addressed
 No software demo, only screenshots
14/04/2011
Introduction
Contents
 Summary information on ECOTRIM (30 mins)
 Summary information on DEMETRA+ with briefing on
the ESS guidelines on Seasonal Adjustment (30 mins)
 User support provided by Unit B2 Methodology and
Research (10 mins)
 Q & A (20 mins)
14/04/2011
Introduction
Why Ecotrim and DEMETRA+ together?
 Target users are the same
ECOTRIM
TOOL FOR TEMPORAL DISAGGREGATION OF TIME SERIES
Dario BUONO
14/04/2011
Topics
 Ecotrim in the framework of data production
 Techniques available in Ecotrim
14/04/2011
ECOTRIM
Forewords
 Ecotrim is a program written for Windows and has been
developed by Eurostat
 There is a project within Eurostat (UNIT C2) for the
development/maintenance of the program
 The main aim of this presentation is to demonstrate the
potential of the application
 User manual is available in EN and FR
http://circa.europa.eu/Public/irc/dsis/ecotrim/library
14/04/2011
ECOTRIM
Why Eurostat developed Ecotrim?
ESA 95 paragraph 12.04
 “The statistical methods used for compiling quarterly accounts
……..can be classified in two major categories: direct
procedure and indirect procedure”
 ... indirect procedures are based on Temporal
Disaggregation of the annual accounts data in accordance
with mathematical and statistical methods using reference
indicators that permit the extrapolation for the current year
14/04/2011
ECOTRIM
Definitions
 Temporal Disaggregation
– Process of deriving high frequency data from low
frequency data and, if available, related high frequency
information
 ECOTRIM
– Supplies a set of mathematical and statistical techniques
to carry out Temporal Disaggregation
14/04/2011
ECOTRIM
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
14/04/2011
ECOTRIM
Other short-term statistics
 Short-term industrial statistics
 Employment
 Money and banking statistics
14/04/2011
ECOTRIM
Techniques available in Ecotrim
14/04/2011
ECOTRIM
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
14/04/2011
ECOTRIM
Techniques available in Ecotrim
 Temporal Disaggregation of univariate time series by
using or not related series and fulfilling temporal
aggregation constraints
– Mathematical approach and the optimal, in the least
squares sense, approach
 Temporal Disaggregation of multivariate time series
with respect of both temporal and contemporaneous
aggregation constraints
 Flash estimates of current year observations by using
or not available information on related series
14/04/2011
ECOTRIM
Temporal and Accounting constraint
 Statistical offices are often faced with the estimation of a
set of quarterly series linked by some accounting
relationship
 Temporal Disaggregation methods can also be used in
such cases, to give a solution consistent with both
temporal and accounting aggregation constraints
14/04/2011
ECOTRIM
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)
14/04/2011
ECOTRIM
Temporal Disaggregation methods
 Univariate Approach
–
–
–
–
Smoothing methods
Two steps adjustment methods
Time series methods
Regression based methods
• Static models
• Dynamic models
 Multivariate Approach
– Two steps adjustment methods
– Regression based methods
14/04/2011
ECOTRIM
The classical univariate Temporal Disaggregation
 Given the annual observations for an account aggregate
and the quarterly values for a related series, the
unknown quarterly/monthly values of the aggregate are
estimated
 Chow and Lin (1971) worked out a least-squares optimal
solution on the basis that a linear regression model
involving the quarterly aggregate series and the related
quarterly series will hold
14/04/2011
ECOTRIM
Statistics






Rhô
R-squared ( next to 1)
Durbin-Watson ( about 2 )
Probability of F
T-stat
Reliability indicators (lower value for the range between
Min and Max)
14/04/2011
ECOTRIM
Reminders
 It must be reminded that the statistics obtained within
Ecotrim outputs are based on the annual data
 It is advisable to use a unique model all along the series
 Structural breaks must be taken in consideration
 It is advisable to have observations covering at least 20
years when disaggregating annual data
 It is advisable to have observations covering at least 15
years when disaggregating quarterly data
14/04/2011
ECOTRIM
ECOTRIM
ECOTRIM is a program that supplies a set of mathematical and
statistical techniques to carry out Temporal Disaggregation
 Structured for Windows 95/98 and Windows NT
 Visual Basic and C++
 User friendly
It can be used according to two different modes:
 Interactive mode
 Batch mode
14/04/2011
ECOTRIM
Interactive mode
Input session
Univariate methods
Multivariate methods
•
Boot, Feibes and Lismann
•
White noise
•
Denton
•
Random walk
•
AR(1)
•
Rossi
•
Fernàndez
•
Denton
•
Litterman
Output session
Graphs and
display
14/04/2011
ECOTRIM
Batch mode
ECOTRIM performs Temporal Disaggregation of several jobs
starting from a batch command file
Batch mode is very useful when handling many series
Batch
command
file
14/04/2011
ECOTRIM
Batch session
• univariate
• multivariate
Output
Work with ECOTRIM
Input Data
Aggregated Series
Choose the Aggregation Order
14/04/2011
ECOTRIM
Work with ECOTRIM
Once you have imported the
Aggregated Series, you will able
to import the Related Series
Use the STATUS bar to check
the currently loaded files
14/04/2011
ECOTRIM
Work with ECOTRIM
Univariate Methods
Univariate Methods
Left click on the series
Choose the Type of Aggregation
14/04/2011
ECOTRIM
Work with ECOTRIM
Choose the Method…
…and the Related Series
14/04/2011
ECOTRIM
Work with ECOTRIM
Compute!!!
…and getting results
14/04/2011
ECOTRIM
Work with ECOTRIM
Display the data
14/04/2011
ECOTRIM
Future of Ecotrim
 Updating the Visual Basic/C++ version and refurbish the
interface of ECOTRIM with the purpose to match
production and users’ requirements
 Design of the application as a Building Block, CVD
compliant
 Make system Windows 7 compatible, compilation as a
64-bit application (former was 16-bit)
 Inclusion of more recent methods like:
– Univariate Regression based Temporal Disaggregation
– Reconciliation by Two-Step procedures (Quenneville &
Rancourt 2005 – Di Fonzo & Marini 2009)
14/04/2011
ECOTRIM
DEMETRA+
TOOL FOR SEASONAL ADJUSTMENT
Enrico INFANTE
14/04/2011
What is DEMETRA+?
DEMETRA+ is a tool for Seasonal and Calendar
Adjustment developed by National Bank of
Belgium for EUROSTAT
Trend-Cycle Component
Identify more components
Outliers
Irregular Component
14/04/2011
DEMETRA+
What are Seasonal and Calendar Adjustment?
SEASONAL
ADJUSTMENT
Seasonality:
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
CALENDAR
ADJUSTMENT
Remove Calendar Effects
14/04/2011
DEMETRA+
Calendar Effects:
Any economic effect which appears to be
related to the calendar (one more Sunday in
the month can affect production)
Why DEMETRA+?
 ESS Guidelines
 User friendly!!!
 It will be Open-Source (java)
 Based on X12 ARIMA / Tramo-Seats
14/04/2011
DEMETRA+
Documents about DEMETRA+
You can download the latest version
of DEMETRA+ at the following link:
http://circa.europa.eu/irc/dsis/eurosam/info/data/demetra.htm
At the same link you can find:




ESS guidelines on Seasonal Adjustment
DEMETRA+ User Manual and other documents
Help-Desk e-mail: [email protected]
Information about the next ESTP course (at EUROSTAT, 3-5 May 2011)
and the EUROSTAT Training Course
14/04/2011
DEMETRA+
How to work with DEMETRA+
First overview
 Browser panel
14/04/2011
DEMETRA+
presents the available
time series
How to work with DEMETRA+
First overview
 Workspace panel
14/04/2011
DEMETRA+
shows information used or
generated by the software
How to work with DEMETRA+
First overview
 Central empty zone will contain the actual
analysis
14/04/2011
DEMETRA+
How to work with DEMETRA+
First overview
 TSProperties panel
14/04/2011
DEMETRA+
contains the current
time series
How to work with DEMETRA+
First overview
 Logs panel
14/04/2011
DEMETRA+
contains logging
information
How to work with DEMETRA+
Importing data from Excel
Browser panel
Excel tab
Add command
•
Empty top-left cell
•
True dates in the first column
•
Title of the series in the first row
•
Empty cells for missing values
14/04/2011
DEMETRA+
Open the excel workbook
How to work with DEMETRA+
Displaying the data
Open one of the containers
DRAG & DROP
14/04/2011
DEMETRA+
How to work with DEMETRA+
Single Seasonal Adjustment
Choose and activate a specification
Getting results
Seasonality tests
14/04/2011
DEMETRA+
What happens after SA?
Original, Calendar Adjusted and Seasonally Adjusted series
Original Series
14/04/2011
DEMETRA+
What happens after SA?
Original, Calendar Adjusted and Seasonally Adjusted series
Calendar Adjusted Series
14/04/2011
DEMETRA+
What happens after SA?
Original, Calendar Adjusted and Seasonally Adjusted series
Seasonally Adjusted series
The series has been cleaned!!
14/04/2011
DEMETRA+
What happens after SA?
Growth Rates
Original Series
14/04/2011
DEMETRA+
Xt
Gt 
1
X t 1
What happens after SA?
Growth Rates
Calendar Adjusted Series
14/04/2011
DEMETRA+
Xt
Gt 
1
X t 1
What happens after SA?
Growth Rates
Seasonally Adjusted series
14/04/2011
DEMETRA+
Xt
Gt 
1
X t 1
ESS Guidelines on SA
http://epp.eurostat.ec.europa.eu/cache/ITY_OFF
PUB/KS-RA-09-006/EN/KS-RA-09-006-EN.PDF
 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
14/04/2011
DEMETRA+
ESS Guidelines on SA (example 1)
Calendar Adjustment
Calendar Effects typically include:




Different number of working days in a specific period
Composition of working days
Leap year effect
Moving holidays (Easter, Ramadan, etc.)
A) RegARIMA approach, fully checks the significance and plausibility of the effects
B) Regression approach based on the irregular component
C) Proportional Adjustment; other adjustments; no adjustment
A) Best alternative B) Acceptable C) To be avoided
14/04/2011
DEMETRA+
ESS Guidelines on SA (example 2)
Validation of Seasonal Adjustment
SA is a complex data treatment which needs accurate
monitoring before the results are acceptable
Descriptive
Graphical
Parametric and non- parametric tests
A) Use a detailed set of criteria
B) Use only default criteria already defined within the different tools
C) Use only the basic set of criteria or do not validate
A) Best alternative B) Acceptable C) To be avoided
14/04/2011
DEMETRA+
USER SUPPORT PROVIDED BY UNIT B2
Dario BUONO
14/04/2011
EUROSTAT UNIT B2
Methodology and Research
Define, position and co-ordinate all statistical
actions in the field of research
Provide the EU with a service on the statistical
methodologies in order to improve the data
quality and reliability for the use of the EU policies
Achieve a higher degree of harmonization of the EUROSTAT and ESS practices on SA:




Definition of detailed guidelines on SA
Maintenance and development of reliable tools for SA at ESS level (DEMETRA+)
Fostering the exchange of practices among NSIs
Organisation of training on SA
14/04/2011
EUROSTAT UNIT B2
Methodology and Research
Define, position and co-ordinate all statistical
actions in the field of research
Provide the EU with a service on the statistical
methodologies in order to improve the data
quality and reliability for the use of the EU policies
 Consultative role in the ECOTRIM project
 Partial user support also on methodological aspects by EUROSTAT methodology
helpdesk
14/04/2011
Questions?
14/04/2011