Impact of updating weights on
tracking performance and volatility:
Industry survey
G. Bruno, L. Crosilla, P. Margani, A. Righi
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
Outline
 Aim of the study
 The ISTAT Industry survey: methodological aspects
 The updating weights
 The impact of updating weights :
 Analysis of volatility
 Analysis of performance of the series
 Concluding remarks
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
Aim of the study
 This study is aimed to provide evaluations on the impact of
updating weights on ISTAT Industry time series in terms of volatility
and tracking performance
 As the weighting is used to improve the quality of the estimates,
ISTAT recently decided to update the weights at stratum level without changing the overall sampling scheme - for two reasons:
 In order to make the reference year of the weights as close as
possible to the year of the survey
 To improve the quality of value added data using a new
available source: the ISTAT Statistical Information System
“SBS Frame”
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
The ISTAT Industry survey: methodological aspects
 The sample is a fixed panel of about 4000 enterprises with at least 5
employees selected from the ISTAT Statistical register of active
enterprises (ASIA)
 The sample strata are identified by:
 firm size (5-9; 10-49; 50-249; 250-999; 1000 and more)
 geographical areas (N-W; N-E; C; So-Is)
 economic sectors (two-digit sectors of NACE rev.2 from 10th to 33
and three-digit sectors of divisions 10,13,20,25,26,27,30,32)
 The sampling method provides:
 a random sampling scheme for firms with less than 1000
employees
 a census sample for the ones with 1000 or more employees
 for firms with less than 1000 employees the sample allocation is
based on the Robust Optimal Allocation with Uniform Stratum
Threshold (ROAUST)
 for larger firms a census survey is designed
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
The ISTAT Industry survey: methodological aspects
 A weighting system is used to process the results
 The weighting scheme considers:
 at firm level, size weights (the number of employees)
 at stratum level, a weighting coefficient reflecting the relative
significance of each stratum in the population (the value added
is used) in order to reflect the structure of the economy
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
The updating weights
 In the data processing, the current weights based on National
Accounts information referring to year 2005 are replaced by a set of
weights updated to year 2012
 The new source (based on micro level information) :
 For small and medium enterprises - new ISTAT SBS Frame an
integrated statistical information system for estimating structural
economic variables on business accounts, based on micro data from
survey, different administrative sources integrated with Statistical
register of active enterprises (ASIA)
 For large enterprises - Survey on enterprise accounting system
integrated with the administrative sources
 Benefits:
 higher reliability of information for main economic aggregates (v.a.)
 higher levels of consistency between business statistics and NA
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
The updating weights
 Main results are synthetized in the following comparison of the percentage
composition of the weights in terms of value added according to the NACE
rev.2 two-digit sectors
 Small changes for all the divisions; Div.28 (Manufacture of machinery and
equipment n.e.c.), presents the bigger change being around +3%
 This is probably due to the better comparative economic performance
shown by Italian mechanical firms during the post-2008 period economic
crisis
The updating weights
NACE(REV.2)
VALUE ADDED 2005
(%)
VALUE ADDED 2012
(%)
DIFFERENCES
DIV. 10
DIV. 11
DIV. 12
DIV. 13
DIV.14
DIV. 15
DIV. 16
DIV.17
DIV.18
DIV. 19
DIV.20
DIV. 21
DIV. 22
DIV. 23
DIV.24
DIV. 25
DIV. 26
DIV. 27
DIV.28
DIV. 29
DIV. 30
DIV. 31
DIV. 32
DIV.33
TOT
8.1
1.2
0.1
3.8
4.0
2.6
2.3
2.1
2.3
1.9
4.6
3.1
4.3
6.1
4.9
12.6
3.6
4.5
12.4
3.5
2.5
3.4
2.4
3.5
100.0
9.2
1.8
0.0
2.7
3.3
3.1
1.6
2.3
1.7
1.1
4.8
4.1
5.2
4.4
3.9
11.7
3.7
5.1
15.6
4.4
3.0
2.4
1.9
2.9
100.0
1.1
0.6
-0.1
-1.1
-0.7
0.5
-0.7
0.2
-0.6
0.0
0.2
1.0
0.9
-1.7
-1.0
-0.9
0.1
0.6
3.2
0.9
0.5
-1.0
-0.5
-0.6
0.0
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
The impact of updating weights
 The Confidence climate (COF) and component variables series (seasonally
adjusted data) are considered in the analysis:
 Time span: Jan.2000-Aug. 2014
 Disaggregation: total Industry and Main Industrial Groupings
 Since January 2009, the series are rebuilt with the new weights
 The Month of Cyclical Dominance (MCD) (This indicator shows the number of
months to be taken into account in order to ensure that the cyclical component is
dominant on the irregular one) is calculated to compare the volatility of the new
and outdated series
 The correlation between Confidence climate and the reference series (that is the
production index - year-on year percentage change) is calculated to compare
the performance of the updated and outdated series
The impact of updating weights
Analysis of volatility : MCD
MCD - weights
2005
MCD - weights
2012
Current level order book
2
2
Inventories
4
4
Production expectations
3
3
Confidence climate
2
2
4
4
2
2
2
2
Series
Total sector
Consumer goods - Confidence
MIG (Main
climate
Industrial
Groupings) Capital goods - Confidence climate
Intermediate goods- Confidence
climate
MCD remains unchanged for all the considered series
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
The impact of updating weights
Analysis of performance: Visual inspection
0.18
120
0.09
110
0.00
100
-0.09
90
-0.18
80
-0.27
70
-0.36
60
Jul-14
Jan-14
Jul-13
Jan-13
Jul-12
Jan-12
Jul-11
Jan-11
Jul-10
Jan-10
Jul-09
Jan-09
Jul-08
Jan-08
Jul-07
Jan-07
Jul-06
Jan-06
Jul-05
Jan-05
Jul-04
Jan-04
Jul-03
Jan-03
Jul-02
Jan-02
Production index- Delta 12 log
Confidence climate s.a.data - weights 2005
Confidence climate s.a. data - weights 2012
The new COF series is more in line with the evolution of the
industrial production index
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
Reference
series:
production index
(year-on-year
percentage
change)
The impact of updating weights
Analysis of performance: Correlation COF - reference series
production
index t+3
production
index t+2
production
index t+1
production
index t
Confidence climate
(weights 2005)
0.73
0.78
0.80
0.80
Confidence climate
(weights 2012)
0.76
0.81
0.83
0.83
 Reference series: production index (year-on-year percentage change)
 The cross-correlation analysis highlights a light improvement at lead -1
(months) for the confidence climate weighted with value added 2012
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels
Concluding remarks
In the case of the Industry survey a substantial qualitative improvement and
updating of the weights shows
 a small positive effect on tracking performance of COF series witnessed
by an improvement of the cross-correlation analysis at lead-1 (months)
 No effect on the volatility of the series: MCD remain unchanged
EU Workshop on Recent Developments in BCS, 13-14 November 2014, Brussels