Transcript KOF Business Situation Indicator

New Composite Indicators Based on
KOF Business Surveys
Richard Etter, Sibylle Gübeli, Michael Graff
November 13, 2007
Overview
We present three different composite indicators
 Financial Sector Indicator
 Without reference series
 Coincident indicator
 Business Situation Indicator
 GDP as reference series
 Coincident indicator
 Employment Indicator
 Employment as reference series
 Leading indicator
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
2
KOF Financial Sector Indicator
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
3
KOF Financial Sector Indicator
Aim




Situation in the financial sector: There are plenty of monetary indicators
but few are value added oriented for the financial sector (banks). The
Swiss Federal Statistical Office publishes only yearly data on the
„production account“.
A quarterly or even a monthly reliable reference series for the value
added does not exist
 Our financial indicator will therefore not be measured or scaled on a
reference series
We want to find a simple and easy to understand indicator for financial
activities which is early disposable
The financial sector indicator will be constructed along the following lines
 Referring only to questions that directly relate to the value added
 Transparent computation and aggregation
 Timely provision of information on economic coincident activity of
banks
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
4
KOF Financial Sector Indicator
Selected Items





Questions from the business survey of banks
No screening of the data
Three items as candidates that refer to value added on a coincident basis
 Judgement of the business situation
 Change in demand
 Expected change of employment
For quantification: traditional balance indicator
There is only a minor seasonality but some erratic movements, therefore
a low pass filter is applied to separate cycle components from noise
 Plausible decomposition of the questions on the business situation
and on expected employment but not on demand. Therefore,
demand will not be considered further
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
5
KOF Financial Sector Indicator
Aggregation


Aggregation procedure:
 Responses are weighted by employees
 Calculation of the balance indicator
 Arithmetic mean of the low pass filtered balance indicators of the two
selected questions
Quarterly data of the indicator from 2000Q1 until 2007Q3
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
6
KOF Financial Sector Indicator
First Indication of Performance
60
42'000
40
40'000
20
38'000
0
36'000
-20
34'000
-40
32'000
-60
30'000
00
01
02
03
KOF Financial Sector Indicator
04
05
06
07
Value added banks
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
7
KOF Financial Sector Indicator
First Indication of Performance
60
42'000
40
40'000
20
38'000
0
36'000
-20
34'000
-40
32'000
-60
30'000
00
01
02
03
KOF Financial Sector Indicator
04
05
06
07
Value added banks
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
8
KOF Financial Sector Indicator
Results and Summary


The KOF financial indicator captures the main features of the dynamics
that characterise the value added series
This is very encouraging. It leads us to conclude that in due time we may
be able to refer to a more adequate – quarterly – data corpus with
success too
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
9
KOF Business Situation Indicator
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
10
KOF Business Situation Indicator
Aim






Monthly coincident indicator to GDP quarterly yoy-growth rate
Data base is the KOF BTS
Monthly data: manufacturing and retail trade
Balance indicators
Low erratic movements
 Judgement
Item which is easy to understand for the public
 Business situation
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
11
KOF Business Situation Indicator
Judgement of the business situation




Business situation is a common term
It goes with the growth-cycle concept
 Level
 No trend
Natural reference series is the (detrended) GDP-Level, but
Best known reference series of the economic development is the
yoy-change of GDP
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
12
KOF Business Situation Indicator
Calculations


Answers weighted by employees of the firms for manufacturing
and retail separately
Stratification by firm-size and weighting according to the universe
for manufacturing and retail trade
Balance indicator of the two sectors
Aggregation of both sectors by employees according to the
universe
– This indicator is compatible with growth-cycles not with
acceleration and deceleration
Yoy-differences of the aggregated balance indicator

Original values as well as smoothed values (trend cycle)



OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
13
KOF Business Situation Indicator
KOF Business Indicator and GDP (original data)
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
14
KOF Business Situation Indicator
KOF Business Indicator and GDP (trend cycle data)
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
15
KOF Business Situation Indicator
Summary




Easy to communicate
High correlation (0.80) with yoy-growth rate of GDP
Coincident indicator
Monthly data with low erratic movements
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
16
KOF Employment Indicator
(Work in Progress)
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
17
Introduction

Framework:
employment indicator

Objectives:
Early quarterly signaling of the newest development of the Swiss
employment by the use of KOF business surveys

Employment as a reference series:
 Relevant for the business cycle
 Interesting for economic policy
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
18
Proceedings







Introduction
Data base
Analysis, selection criteria and combinations
Results: single indicators
From the single to the composite indicators
Results: composite indicators
Summary
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
19
Data Base: KOF Business Survey Results



Business surveys in different Swiss sectors
Questions about employment and general situation
 Earlier development
 Assessment
 Expected prospective development
Most important analysed transformations
 Trend cycles
 Balance statistics
 Monthly data: Average of three months, values of the 1st, 2nd or 3rd
months of a quarter
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
20
Analysis, Selection Criteria and Combinations

Reference series: employment growth rate relative to the quarter of the
previous year (PY)

Correlation Analysis: KOF business survey results (single indicators) and
the reference series

Selection criteria
 Economically plausible correlation to the employment growth rate (PY)
 Strong correlation (maximal Correlation > 0.75) at a
 Lead of two quarters

Combining the selected single indicators > composite indicators

Correlation Analysis: composite indicators and the reference series
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
21
Results: Single Indicators
>> Correlation of selected single indicators and Swiss employment growth rate (PY)
Item
Assessment of Employment
Retail trade
Hotels and restaurants
Expected Employment
Construction
Retail trade
Manufacturing
Business Situation
Construction: exp. incoming orders (3 months)
Architects and engin.: exp. business situation
Retail trade: exp. business situation
Hotels and restaurants: exp. overturn (3 m.)
Corr.
Lead
max. (in quarters)
Transformation From
5
4a + 3a
0.895
0.818
2 tc, 3rd month of quarter 94q2
2
tc 88q4
9
6c
8d
0.922
0.892
0.800
2
2
2
tc 94q3
tc 94q2
tc, average 99q2
8a
9
0.965
0.921
2
2
tc 94q3
tc 96q1
10
7+6
0.885
0.776
2
2
tc, average 94q2
tc 88q4
tc … trend cycle
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
22
From the Single to the Composite Indicators

Selected single indicators use information of one sector only

Objective: bundle information

Bundled information using Principal Components Analysis (common
variance)

Two reasonable combinations to compare:
 Combination of the employment-based items (five items)
 Combination of the employment-based and the business-situationbased items (nine items)
>> Two composite indicators
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
23
Results: Composite Indicators
>> Correlation of two composite indicators and Swiss employment growth rate (PY)
Employment: Assessments and
Expections
five Items
Employment and Business Situation
nine Items
Sample (to 2003q3)
Corr. max.
(Lead of two quarters)
Out-of-Sample (to 2007q2)
Corr. max.
(Lead of two quarters)
0.972
0.888
0.981
0.946
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
24
Possible Employment Indicators
Q1 2007
Q1 2006
Q1 2005
Q1 2004
-2.0
Q1 2003
-5
Q1 2002
-1.0
Q1 2001
-3
Q1 2000
0.0
Q1 1999
-1
Q1 1998
1.0
Q1 1997
1
Q1 1996
2.0
Q1 1995
3
Q1 1994
3.0
Q1 1993
5
Swiss Employment, Growth Rate (PY)
Employment Indicator (9 Items)
Employment Indicator (5 Items)
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
25
Summary
 Given the Swiss employment growth rate (PY) the composite
indicators perform very well:
 They show a strong correlation
 They signal turning points in advance
 Given the selected single indicators the composite indicators are
 Can be commented comprehensibly because of employmentrelevant items of the KOF business survey results
 Broadly based on items of the different Swiss sectors
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
26
Thank You for Your Attention
Richard Etter
Sibylle Gübeli
KOF Swiss Economic Institute
KOF Swiss Economic Institute
E-Mail: [email protected]
E-Mail: [email protected]
Tel.: +41 44 632 51 56
Tel.: +41 44 632 53 60
Free registration for our new monthly newsletter (KOF Bulletin):
www.kof.ethz.ch/publications/bulletin/
OECD Workshop on Business and Consumer Tendency Surveys
November 13, 2007
27