Transcript 3 MB

EU Workshop on Recent Developments
in Business and Consumer Surveys
Hotel Bedford, 135-137,
Rue du Midi, 1000 Brussels
15-16 November 2012
Eurostat – Unit C4 Key indicators for the European policies
On-going international work and prospects in
the field of BCS
A statistical overview of the
economic situation in the
euro area
By
Gian Luigi Mazzi
Filippo Moauro
Rosa Ruggeri-Cannata
Eurostat – Unit C4 Key indicators for European policies
Outline
 Introduction
 PEEIs' nowcasts: a bridge modelling approach
real time simulation exercise
 PEEIs' nowcasts: an unobserved component based
approach
performance of EuroMIND based on survey data
 The system of turning point detection
 Real time monitoring and the early warning system
 Conclusions
Eurostat – Unit C4 Key indicators for European policies
Background
 PEEIs: a tool for short term economic monitoring based
on official statistics
 Picture for the euro area rather incomplete for 2012
- few monthly data
- quarterly GDP only for the second quarter
- employment provided with delay
 Updated picture for the euro area
- combining official statistics and econometric
techniques
Eurostat – Unit C4 Key indicators for European policies
Most recent evolution of main PEEIs
2011q2
2011q3
2011q4
2012q1
2012q2
2012q3
% (Q/Q-1)
0.2
0.1
-0.3
0.0
-0.2
-0.1
% (Q/Q-4)
1.7
1.3
0.3
0.3
-0.9
-0.6
% (Q/Q-1)
0.2
-0.1
-0.3
-0.3
0.0
-0.2
% (Q/Q-4)
0.5
0.3
-0.1
-0.5
-0.6
-0.7
2012m5
2012 m6
2012m7
2012m8
2012m9
2012m10
% (M/M-1)
-0.1
-0.1
-0.5
0.4
0.7
0.2
% (M/M-12)
2.4
2.4
2.4
2.6
2.6
2.5
% (M/M-1)
-0.5
-0.5
0.3
0.9
0.2
0.0
% (M/M-12)
2.3
1.8
1.6
2.7
2.7
2.6
% (M/M-1)
1.0
-0.5
0.6
0.6
-0.7
-
% (M/M-12)
-2.6
-2.0
-2.8
-2.9
-1.0
-
Unemployment rate %
11.3
11.4
11.5
11.5
11.6
-
ESI
90.5
89.9
87.9
86.1
85.2
84.5
GDP
Employment
HICP
PPI
IPI
Index
Note: forecasts (in red) obtained by a bridge modelling approach
Eurostat – Unit C4 Key indicators for European policies
GDP growth by country
2011q2
2011q3
2011q4
2012q1
2012q2 2012q3
Belgium
0.2
0.0
-0.1
0.2
-0.5
0.0
Netherlands
0.0
-0.3
-0.6
0.2
0.2
0.1
Spain
0.2
0.0
-0.5
-0.3
-0.4
-0.5
Italy
0.3
-0.2
-0.7
-0.8
-0.8
-0.8
France
0.0
0.3
0.0
0.0
0.0
0.1
Germany
0.5
0.4
-0.1
0.5
0.3
0.2
Euro area
0.2
0.1
-0.3
0.0
-0.2
-0.1
Note: forecasts (in red) obtained by a bridge modelling approach
Eurostat – Unit C4 Key indicators for European policies
-6.0
-8
Eurostat – Unit C4 Key indicators for European policies
2007Q1
2006Q1
2005Q1
2004Q1
2003Q1
2002Q1
2001Q1
2000Q1
2011Q1
Belgium
2011Q1
Netherlands
2010Q1
Spain
-6
2010Q1
Italy
-4.0
2009Q1
4
2009Q1
6
2008Q1
variations over same quarter of prevoius year
2008Q1
France
2007Q1
Germany
-4
2006Q1
-2
2005Q1
-2.0
2004Q1
0
2003Q1
0.0
2002Q1
2
2001Q1
variations over same quarter of prevoius year
1999Q1
Billions of euro, chain-linked volumes, reference year 2000
2000Q1
2.0
1998Q1
-2.5
-3.0
1999Q1
1500
1997Q1
-1.5
-2.0
1998Q1
1600
1996Q1
1700
1997Q1
1800
1995Q1
1900
1996Q1
4.0
2011Q3
2010Q4
2010Q1
2009Q2
2008Q3
2007Q4
2007Q1
2006Q2
2005Q3
2004Q4
2004Q1
2003Q2
2002Q3
2001Q4
2001Q1
2000Q2
1999Q3
1998Q4
1998Q1
1997Q2
1996Q3
1995Q4
2000
1995Q1
2011Q1
2010Q1
2009Q1
2008Q1
2007Q1
2006Q1
2005Q1
2004Q1
2003Q1
2002Q1
2001Q1
2000Q1
1999Q1
1998Q1
1997Q1
1996Q1
1995Q1
1400
1995Q1
Quarterly GDP of the euro area
variations over prevoius quarter
1.5
1.0
0.5
0.0
-0.5
-1.0
Some comments: the euro area in the first
3 quarters of 2012
 3 negative signs of GDP growth in last 4 quarters
 Persisting slowdown of employment accompanied by a
rising unemployment rate
 Pretty stable inflation
 Negative perception of the economic situation by
producers and consumers as shown by the ESI
Eurostat – Unit C4 Key indicators for European policies
Some comments: economic growth at
country level
 Germany shows a recovery after the negative result in
2011Q4
 The growth in France is rather weak
 Belgium shows a stop and go development in last
quarters.
 Netherlands shows a positive growth in all quarters of
2012 after two consecutive negative quarters.
 Italy and Spain continue in their negative trend
 Trend in France is flat and never negative in last
quarters
Eurostat – Unit C4 Key indicators for European policies
PEEIs' nowcasts: the Eurostat BM approach (1)
p
k
q
yt  c   i yt i    jl x j ,t l  ut ,
i 1
t  1,..., T
j 1 l 0
 yt is the dependent variable observed over the sample
period t = 1, …, T taken in its first (log-) differences, with
Δ the difference operator such that Δyt = yt - yt-1;
 c is an intercept;
 αi and βjl are regression coefficients respectively related
to Δyt-l and to a set of k predictors xjt, j=1, …, k, ;
 ut is a mean zero disturbance with variance σ2.
 general specification adaptable to the case of
cointegration among implied variables imposing suitable
coefficient restrictions error correction (EC) terms.
Eurostat – Unit C4 Key indicators for European policies
PEEIs' nowcasts: the Eurostat BM approach (2)
BM approach in 3 steps:
a) xit are projected over a forecast horizon by means of a
univariate time series technique;
b) the indicator is temporal aggregated at the same time
span of the target variable;
c) parameter estimation of model (1) is carried out to
determine the coincident indicator over the full sample
period t=1,2, …, T.
Variable selection is carried through the least angle
regression (LARS) algorithm by Bai and Ng (2008) able to
selecting a set of targeted predictors to obtain the most
accurate nowcasts.
Eurostat – Unit C4 Key indicators for European policies
BM specification: GDP, Employment and IPI
Δln(GDPt)
Δln (Emplt)
Δln(IPIt)
GDP
Employment
ln(yt-1)
-
ln(yt-1)
IPI
Δln(xt)
ln(xt-1) Δln(xt-1), Δln(xt-2)
ln(xt-1), Δln(xt)
ln(xt-1), Δln(xt)
Δln(xt)
Δln(xt)
Δln(xt-1), Δln(xt-2)
Business and consumer surveys
Present business situation
Δ(xt)
Major purchases over next 12 months
xt-1
Retail Sales
Exports
Construction output
Unemployment rate
Exchange rate euro/$
-
Δln(xt)
-
-
-
x t,
Δxt-h·|Δxt-h|
with h=1,2,3
-
Industrial confidence
-
Δ(xt)
Construction confidence
Construction opinion on activity
Construction employment expectation
-
Δ(xt-1)
Δ(xt)
xt-2
Note: Δ denotes the difference operator, ln the logarithm and Δxt-h·|Δxt-h| the squared difference with sign
transformation
Eurostat – Unit C4 Key indicators for European policies
Real time experiment and error statistics
2011
q1
2011
q2
2011
q3
2011
q4
2012
q1
2012
q2 MAE RMSE
Employment
0.63
0.44
-0.02 -0.40
0.00
0.00
0.26
0.30
(-0.20) (0.27) (-0.17) (-0.10) (0.00) (0.20)
0.22
0.13
-0.01 -0.07 -0.08 -0.16
0.18
0.19
(0.17) (-0.18) (0.11) (0.15) (0.11) (-0.19)
2012
2012
2012 2012 2012 2012
Febr. March
April May June July MAE RMSE
IPI
0.50
(-0.01)
GDP
0.05
(0.34)
0.07
0.15 -0.43 -0.49
(0.89) (-0.41) (0.22) (-1.09)
0.57
0.73
Note: Nowcasts are in growth rates with, in parenthesis, errors computed as difference from respective first official
Eurostat release. MAE is the mean absolute error and RMSE is the root mean squared error over the quarters from
2010q1 to 2012q2 for GDP and Employment and over the months from January 2010 to July 2012 for IPI
Eurostat – Unit C4 Key indicators for European policies
EuroMind: an overview
 Monthly indicator of economic activity
 Proxy of GDP
 Combining Stock and Watson approach with temporal
disaggregation in the state-space framework
 Monthly EuroMind estimates calculated by aggregation
of all output and expenditure sides component:
recnciliation obtained by weigthed averages of the two
monthly estimates
Eurostat – Unit C4 Key indicators for European policies
EuroMind: main characteristics
1) use of a disaggregate approach  output and expenditure
breakdowns of quarterly GDP;
2) for each component, selection of indicators;
3) monthly and quarterly indicators modelled into the Stock and
Watson single index model;
4) state space form allowing for temporal disaggregation;
5) use of a computational efficient procedure;
6) chain-linking is taken into account;
7) final estimate  combining the estimates from the output and
expenditure sides  optimal weights;
8) benchmarking to official quarterly accounts
9) explicit measure of uncertainty around the indicator available.
Eurostat – Unit C4 Key indicators for European policies
EuroMind extensions
 EuroMind-S: more coincident version of EuroMind based
on a two factor Stock and Watson model where one
factor models the effects of qualitative information
 EuroMind-C: medium size factor model simultaneously
modelling euro area and largest countries
 EuroMind-B: back calculated version of EuroMind to
1970
– No back calculation of components available
 EuroMind-Gap: euro area output estimates derived from
EuroMind by means of a state-space modelling of trend
and cycle
Eurostat – Unit C4 Key indicators for European policies
Eurostat – Unit C4 Key indicators for European policies
Oct-11
Nov-11
Dec-11
Jan-12
Feb-12
Mar-12
Apr-12
May-12
Jun-12
Jul-12
Aug-12
Sep-12
Oct-12
Sep-12
Aug-12
Jul-12
Jun-12
May-12
Apr-12
Mar-12
Feb-12
Jan-12
Dec-11
Nov-11
Oct-11
Sep-11
Aug-11
Jul-11
Jun-11
May-11
630
Apr-11
Mar-11
Feb-11
Jan-11
Dec-10
Nov-10
Oct-10
Sep-10
Aug-10
Jul-10
Jun-10
May-10
Apr-10
Mar-10
Feb-10
Jan-10
EuroMind-S: last 13 releases
650
645
640
635
625
EuroMIND-S: ME, MAE and RMSE statistics
(revisions errors over the quarters 2010q2-2012q2)
ME
MAE
RMSE
1 step
0.04
0.12
0.14
2 steps
0.01
0.11
0.13
Eurostat – Unit C4 Key indicators for European policies
3 steps
0.11
0.15
0.17
Growth rates: comparison among indicators
------------------- 2011 --------------------- ---------------------------------------- 2012 -------------------------------------Aug
Sept
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sept
------------------------------------------ variations over previous period -----------------------------------------------Quarterly
GDP
-
0.1
-
-
-0.3
-
-
0.0
-
-
-0.2
-
-
-
EuroMind
0.1
-0.4
-0.1
0.0
-0.2
0.1
-0.2
0.2
-0.4
0.2
0.0
0.2
0.2
-
EuroMind-S
0.0
-0.3
0.0
-0.1
-0.2
0.2
-0.1
0.2
-0.3
0.0
0.1
0.1
0.1
0.1
---------------------------------- variations over same period of previous year ---------------------------------------Quarterly
GDP
-
1.3
-
-
0.3
-
-
0.3
-
-
-0.9
-
-
-
EuroMind
1.5
1.0
0.7
0.6
0.5
0.3
-0.3
0.0
-0.6
-0.5
-0.3
-0.4
-0.2
-
EuroMind-S
1.4
1.0
0.7
0.6
0.5
0.3
-0.3
-0.1
-0.5
-0.5
-0.3
-0.4
-0.3
0.1
Eurostat – Unit C4 Key indicators for European policies
The system of euro area turning point detection (1)
 Euro area dating: business, growth, and
acceleration cycle
– non-parametric dating rule
 Euro area system for turning point detection:
three coincident indicators
– BCCI: Business Cycle Coincident Indicator
– GCCI: Growth Cycle Coincident Indicator
– ACCI: Acceleration Cycle Coincident Indicator
Eurostat – Unit C4 Key indicators for European policies
The system of euro area turning point detection (2)
 BCCI: averaging probability recessions returned by univariate
Markov switching models fitted to each of the three
component series: unemployement, industrial production
index, new car registrations
 GCCI: averaging probability recessions returned by univariate
Markov switching models fitted to each of the five component
series: industrial production index, conctruction confidence
indicator, financial situation of households during the last 12
months, imports of intermediate goods, employement
expectations
 ACCI: recession probability returned by the Markov switching
model fitted to the Economic Sentiment Indicator
Eurostat – Unit C4 Key indicators for European policies
MS-VAR GCCI and MS-VAR BCCI
 MSIH(4) – VAR(0) model fitted on 4 variables
1) Industrial Production Index (differenced over 6 months)
2) Unemployment Rate (inverted diff. over 1 month)
3) New Passenger Car Registrations (diff. over 3 months)
4) Employment Expectations (diff. over 1 month)
 Both indicators jointly obtained as a by-product of model
parameters estimation:
MS-VAR BCCI  filtered probabilities of the first regime of the
state-variable (regimes of the latent Markov-chain are sorted
in ascending order of the state-dependent intercept);
MS-VAR GCCI  sum of filtered probabilities of the first and
second regimes.
Eurostat – Unit C4 Key indicators for European policies
The αABβCD sequence
Classical cycle
A
B
β
α
C
D
trend
series in level
B
Growth cycle
A
A
C
D
Acceleration cycle
α
A
C
β
Eurostat – Unit C4 Key indicators for European policies
Summary of most recent turning points
Economic Cycle
Acceleration
Cycle
Growth Cycle
Classical
Business Cycle
Coincident Indicator
Peak
Trough
Peak
Provisional Dating
2006 Q1
2008 Q1
2010 Q2
ACCI
June 2006
March 2009
December 2010
December 2010
December 2011
March 2012
Provisional Dating
2008 Q1
2009 Q3
2011 Q1
GCCI
March 2007
July 2009
August 2011
Multivariate
December 2007
September 2009
May 2011
Provisional Dating
2008 Q1
2009 Q2
2011 Q3
BCCI
August 2008
October 2009
January 2012
Multivariate
April 2008
September 2009
December 2011
Eurostat – Unit C4 Key indicators for European policies
Some comments
 Euro area growth cycle persists in slowdown phase
started in middle 2011
 Euro area business cycle entered in recession phase at
the end of 2011
 After a long deceleration phase the acceleration cycle
entered in a new acceleration phase since January 2012
Synthesis
 the already achieved trough of the acceleration cycle
could be a positive signal for a quite fast recovery from
the new recession
Eurostat – Unit C4 Key indicators for European policies
Thank you for your attention!
Eurostat – Unit C4 Key indicators for European policies