Transcript (I) Euro area

Revision Analysis for
General Economic Statistics
Henning Ahnert
European Central Bank
OECD Short-term Economic Statistics Working Party
26-28 June 2006, Paris
2
Motivation
ECB discussion February 2006:
• Revision of initial (low) GDP growth estimate for
2005Q4 cannot be excluded; volatility of short-term data
Recent press quotes :
• Initial estimate may underestimate 2005Q4 GDP growth
(JP Morgan)
• Europe’s statisticians too gloomy (The Economist)
 ECB Revision analysis
2
3
Basic features of the analysis (I)
Selected key indicators:
• GDP & expenditure components & compensation
• Employment & Unemployment
• HICP and Labour Cost Index
• Industrial production & Retail trade turnover
• Euro area, six largest euro area countries, UK,
US, Japan
• Data sources: Eurostat, ECB, OECD
3
4
Basic features of the analysis (II)
• Revisions are necessary !
• Low revisions do not prove accurate measurement
• Revisions may incorporate new information,
conceptual changes, changes in seasonal and
working day adjustment factors etc.
• Euro area statistics: harmonisation, improved
timeliness, changes in country coverage have led to
welcome changes and revisions
4
5
Timeliness and coverage
Euro area: timeliness and country coverage rates
Indicator
Timeliness
Country coverage rate
Early-1999
End-2005
Early-1999
End-2005
GDP
75
45
77
94
Industrial production
104
47
85
97
Retail trade turnover
90
33-36
70
97
5
6
Main results for euro area statistics
• more stable than euro area country data
 main reason: revisions ‘cancel out’
• but: small GDP revisions coincide with low
growth
• monthly indicators (e.g. retail trade) less
reliable than quarterly indicators
6
7
Main results by country
Six largest euro area countries:
•
revisions in 1999/2000 higher (new statistics)
•
revisions vary from relatively low (for GDP) to
rather high (for retail sales)
UK: as largest euro area countries
US: higher revisions for GDP and employment,
comparable revisions for other variables (but
typically higher growth rates!)
JP: higher revisions for GDP and retail sales,
employment and unemployment
hardly revised
7
8
GDP
Euro area GDP volume growth
(seasonally adjusted; quarter-on-quarter growth rate)
 (Recently) stable results and no bias (t+45)
Current data minus first release
First release
Current data
1.3
1.3
1.1
1.1
0.9
0.9
0.7
0.7
0.5
0.5
0.3
0.3
0.1
0.1
-0.1
-0.1
-0.3
-0.3
-0.5
-0.5
05Q3
05Q1
04Q3
04Q1
03Q3
03Q1
02Q3
02Q1
01Q3
01Q1
00Q3
00Q1
99Q3
99Q1
8
9
Euro area and country GDP revisions
Euro area revisions smoother
than country revisions
percentage points
Euro area
France
Italy
Germany
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0.0
0.0
-0.2
-0.2
-0.4
-0.4
-0.6
-0.6
-0.8
-0.8
01Q1
01Q3
02Q1
02Q3
03Q1
03Q3
9
04Q1
04Q3
05Q1
05Q3
10
HICP
Euro area annual HICP
(non-adjusted; annual growth rate)
 Very reliable flash estimate (t+0)
Current data minus first release
First release
Current data
3.0
3.0
2.5
2.5
2.0
2.0
1.5
1.5
1.0
1.0
0.5
0.5
0.0
0.0
-0.5
-0.5
Jul-05
Jan-05
Jul-04
Jan-04
Jul-03
Jan-03
Jul-02
Jan-02
10
11
Unemployment
Euro area unemployment rate
(seasonally adjusted)
 Revisions due to conceptual changes (t+30)
Current data minus first release (left axis)
First release (right axis)
Current data (right axis)
1
9
percentage points
0.5
8.5
0
8
-0.5
7.5
-1
7
Jul-05
Jan-05
Jul-04
Jan-04
Jul-03
Jan-03
Jul-02
Jan-02
11
12
Industrial production (I)
Euro area (s.a.; month-on-month growth)
 Relatively volatile on a monthly basis ... (t+35)
Current data minus first release
First release
Current data
2.0
2.0
1.5
1.5
percentage points
1.0
1.0
0.5
0.5
0.0
0.0
-0.5
-0.5
-1.0
-1.5
-1.0
-2.0
-1.5
Jul-05
Jan-05
Jul-04
Jan-04
Jul-03
Jan-03
Jul-02
Jan-02
12
13
Industrial production (II)
Euro area (s. a. 3 months moving average; month-on-month growth)
 ...less volatile 3 months moving average
percentage points
Current data minus first release
First release
Current
0.7
0.7
0.5
0.5
0.3
0.3
0.1
0.1
-0.1
-0.1
-0.3
-0.3
-0.5
-0.5
-0.7
-0.7
Oct.2005
Jul.2005
Apr.2005
Jan.2005
Oct.2004
Jul.2004
Apr.2004
Jan.2004
Oct.2003
Jul.2003
Apr.2003
Jan.2003
Oct.2002
Jul.2002
Apr.2002
Jan.2002
13
14
Industrial production (III)
Euro area and selected countries
(current data minus first release; seasonally adjusted; month-on-month growth)
 Country data more volatile – cancels out at euro area level
percentage points
Euro area
Germany
Spain
France
Italy
Netherlands
Belgium
6.0
6.0
4.0
4.0
2.0
2.0
0.0
0.0
-2.0
-2.0
-4.0
-4.0
Jan- Feb- Mar- Apr- May- Jun04
04
04
04
04
04
Jul04
Aug- Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun04
04
04
04
04
05
05
05
05
05
05
14
Jul05
Aug- Sep- Oct- Nov- Dec05
05
05
05
05
15
Industrial production (IV)
Revision indicators - euro area and selected countries
Percentage
points
Jan02 –
Dec04
Euro area
Belgium
Germany
Spain
France
Total revisions
(latest minus first release)
Successive
revisions
Memo item:
Average monthly
Average
Average absolute
Range of
Average
cumulative
revision
revision
revisions
absolute revision
growth rate
0.32
-0.7 to
0.9
2.98
0.10
1.28
-2.6 to
5.1
7.46
0.15
0.59
-1.2 to
1.9
2.36
0.16
0.55
-1.6 to
1.3
1.60
0.13
0.77
-1.9 to
1.7
2.43
0.08
0.10
0.40
0.21
-0.04
0.00
 3 groups; broadly corresponding
-1.6 to
to country size ?
Italy
1.3
1.65

Significant0.04
successive0.37
revisions in
euro area
15
-4.6 to
-0.07
16
Retail trade turnover (I)
Euro area (s.a.; month-on-month growth) (t+35)
 No bias, but very unstable (improving in 2005?)
percentage points
Current data minus first release
Current data
First release
3.0
3.0
2.0
2.0
1.0
1.0
0.0
0.0
-1.0
-1.0
-2.0
-2.0
-3.0
-3.0
Jul-05
Jan-05
Jul-04
Jan-04
Jul-03
Jan-03
Jul-02
Jan-02
16
17
Retail trade turnover (III)
Revision indicators - euro area and selected countries
Percentage
points
Jan02 –
Dec04
Euro area
Belgium
Germany
Spain
France
Italy
Netherlands
Total revisions
(latest minus first release)
Successive
revisions
Memo item:
Average
Average absolute
Range of
Average cumulative
Average monthly
revision
revision
revisions
absolute revision
growth rate
0.63
-2.4 to
2.1
9.63
0.05
1.03
-3.4 to
2.6
7.70
0.04
1.56
-6.5 to
3.6
8.75
0.04
0.67
-2.3 to
1.6
2.57
0.30
1.11
-3.5 to
2.4
1.83
0.09
0.43
-1.0 to
1.0
3.50
-0.11
7.87
-0.11
-0.01
-0.10
0.04
-0.08
-0.13
-0.06
0.12
0.64
17 -1.8
to
1.7
18
Final remarks
• Favourable result for most euro area indicators - more
stable than country data (revisions cancel out)
• Experience too short for analysis of “cyclical” revisions,
but can revisions be better anticipated?
• Monthly data very valuable despite higher revisions
(timeliness!)…
• …but can we better understand the reasons for high
revisions ?
• Synchronised national release and revision policies may
further stabilise euro area statistics
18