Click here to load reader
Upload
george-kershoff
View
65
Download
0
Embed Size (px)
DESCRIPTION
The need for a financial sector survey stems from the fact that the financial sector’s contribution to GDP has increased significantly over the last few decades and information on the sector’s performance is published with long time lags and is incomplete. The business tendency surveys of only a few countries currently cover the financial sector.
Citation preview
Conducting Financial Sector Surveys in South Africa
George KershoffDeputy Director
Bureau for Economic Research (BER), Stellenbosch University, South Africa
Overview
1. The need for a financial sector survey2. The survey method3. The validity of the results
The need for a financial sector survey
To improve the coverage and cyclical characteristics of business tendency surveys (BTS)
To produce timely and consistent information on the performance of firms in the financial sector
The survey method
1. The international experience2. Sample design
a) Target populationb) Sampling method and panel creationc) Sample size and representativeness
3. Questionnaire4. Processing of the results
a) Weightingb) Treatment of non-responses
The international experience
BTS coverage expanded to the other service sectors, but not the financial sector.
Why not? Few firms No reference series of an invisible activity
The target population
Universe
Panel
Retail banksTransac-tional banking
Private banking, micro-lending, retail banking, business banking, corporate banking, asset & fleet financing
33 15
Investment banks
Financial market activity
Corporate finance, private equity, project finance, treasury, stock broking
49 11
Investment managers
Institutional (pension), retail (unit trust) and private client business
64 31
Life insurers Individual life and group life & employee benefit business
21 12
Short term insurers
General insurers 16 8
Total 183 77
The target population
Universe
Panel
Retail banksTransac-tional banking
Private banking, micro-lending, retail banking, business banking, corporate banking, asset & fleet financing
33 15
Investment banks
Financial market activity
Corporate finance, private equity, project finance, treasury, stock broking
49 11
Investment managers
Institutional (pension), retail (unit trust) and private client business
64 31
Life insurers Individual life and group life & employee benefit business
21 12
Short term insurers
General insurers 16 8
Total 183 77
The target population
Universe
Panel
Retail banksTransac-tional banking
Private banking, micro-lending, retail banking, business banking, corporate banking, asset & fleet financing
33 15
Investment banks
Financial market activity
Corporate finance, private equity, project finance, treasury, stock broking
49 11
Investment managers
Institutional (pension), retail (unit trust) and private client business
64 31
Life insurers Individual life and group life & employee benefit business
21 12
Short term insurers
General insurers 16 8
Total 183 77
The target population
Universe
Panel
Retail banksTransac-tional banking
Private banking, micro-lending, retail banking, business banking, corporate banking, asset & fleet financing
33 15
Investment banks
Financial market activity
Corporate finance, private equity, project finance, treasury, stock broking
49 11
Investment managers
Institutional (pension), retail (unit trust) and private client business
64 31
Life insurers Individual life and group life & employee benefit business
21 12
Short term insurers
General insurers 16 8
Total 183 77
The target population
Universe
Panel
Retail banksTransac-tional banking
Private banking, micro-lending, retail banking, business banking, corporate banking, asset & fleet financing
33 15
Investment banks
Financial market activity
Corporate finance, private equity, project finance, treasury, stock broking
49 11
Investment managers
Institutional (pension), retail (unit trust) and private client business
64 31
Life insurers Individual life and group life & employee benefit business
21 12
Short term insurers
General insurers 16 8
Total 183 77
The target population
Universe
Panel
Retail banksTransac-tional banking
Private banking, micro-lending, retail banking, business banking, corporate banking, asset & fleet financing
33 15
Investment banks
Financial market activity
Corporate finance, private equity, project finance, treasury, stock broking
49 11
Investment managers
Institutional (pension), retail (unit trust) and private client business
64 31
Life insurers Individual life and group life & employee benefit business
21 12
Short term insurers
General insurers 16 8
Total 183 77
Sampling method and panel creation
Compiled a sampling frame / business register The response units = head of business units
(kind-of-activity units) Purposive (non-random) sampling Establish a fixed panel
Sample size and representativeness
Panel vs. universe Number of firms Size of firms
Response rate Importance of the representativeness of the
sampling units of a qualitative survey vs. that of a quantitative survey Ordinal scaled data Missing at random assumption
Questionnaire
Compare SA, UK and Swiss questionnaires Customised per sector Length Variables covered in all surveys and countries
Rating of the overall business situation Change in income Change in employment Change in profitability
Processing of the results
Firm, sector and sample weights Treatment of non-responses Revisions
The validity of the banking survey results Compare the survey data (in net balance
format) and the corresponding quantitative data (calculated as year-on-year growth rates)
Draw charts to see how closely the lines and turning points fit
Calculate the correlation coefficient Log transform the data to provide for non-linearity, a
not normal distribution, different formats and outliers
0
10
20
30
40
50
60
70
80
90
02Q
1
02Q
3
03Q
1
03Q
3
04Q
1
04Q
3
05Q
1
05Q
3
06Q
1
06Q
3
07Q
1
07Q
3
Net
%
0
5
10
15
20
25
% c
h Y
oY
Survey (lhs) SARB (rhs)
Operating expenses
-60
-40
-20
0
20
40
60
80
100
02Q
1
02Q
3
03Q
1
03Q
3
04Q
1
04Q
3
05Q
1
05Q
3
06Q
1
06Q
3
07Q
1
07Q
3
Net
%
-100
-50
0
50
100
150
200
250
% c
h Y
oY
Survey (lhs) SARB (rhs)
Investment income
0
10
20
30
40
50
60
70
80
90
02Q
1
02Q
3
03Q
1
03Q
3
04Q
1
04Q
3
05Q
1
05Q
3
06Q
1
06Q
3
07Q
1
07Q
3
Net
%
0
5
10
15
20
25
% c
h Y
oY
Survey (lhs) SARB (rhs)
Operating expenses
-60
-40
-20
0
20
40
60
80
100
02Q
1
02Q
3
03Q
1
03Q
3
04Q
1
04Q
3
05Q
1
05Q
3
06Q
1
06Q
3
07Q
1
07Q
3
Net
%
-100
-50
0
50
100
150
200
250
% c
h Y
oY
Survey (lhs) SARB (rhs)
Investment income
Correlation High (80%): investment income, non-interest income Medium (40%): total income, operating expenses,
employment Compared to other countries and surveys
The way forward
Produce a reference series Fine tune weights and method of substitution