Innovations in the Use of Information for SME Lending

Download Report

Transcript Innovations in the Use of Information for SME Lending

Credit Reporting Systems
Around the Globe
Policy Seminar
Inter-American Development Bank
Washington, D.C.
May 9, 2001
Margaret Miller, Senior Economist
Corporate Strategy Group
World Bank
Organization of today’s
presentation
I. The state of the art in credit reporting results of 1999-2000 World Bank survey
II. The importance of credit registries in
financial markets - empirical evidence
III. Public policies to support credit reporting
- emerging elements of “good practice”
IV. Conclusions
Elements of a Credit Reporting System
• the public credit registry, if one exists
• private credit registries, including chambers of commerce,
and banking associations
• the legal framework for credit reporting
• the legal framework for privacy, as it relates to this activity
• the regulatory framework for credit reporting
• the characteristics of other pertinent borrower data
available in the economy
• the use of credit data in the economy, by financial
intermediaries and others
• the cultural context for credit reporting
I. Results of 1999-2000 World
Bank survey of public and
private credit registries
Survey sample by country
Region
Public CIR Private CIR
Number of obs by:
Country
Country
Firms
Latin America
Africa
26
13
17
1
29
1
12
4
6
5
66
7
6
4
1
36
7
8
5
2
52
(includes 8 nations in BCEAO)
W. Europe
E. Europe
Asia Pacific
Other
TOTAL
The credit reporting industry is in
transition, with many new entrants
• The median age of private registries in the
survey sample is 10 years. Thirty percent of
the private registries were established since
1995.
• Of the 29 private registries from Latin
America, 13 were established since 1993.
• Latin America led all other regions in the
1990s in the establishment of public credit
registries.
Public vs. Private
Credit Registries
Feature
Public
Private
Source of
information
Participation
mandatory?
Positive Info?
Supervised
institutions
Yes
Varied sources
Yes
In some cases
No
Borrowers assigned Yes
No
a rating?
Minimum loan size In some countries No
Fee for service
No charge or
minimal charge
Yes
Institutional Arrangements for
Private Credit Registries
Institutional Type
Pros
Cons
Private firm w/no
bank ownership
Private firm w/ bank
ownership
All types of data,
independence
All types of data,
Special access to
bank data
Access to bank data
Integrity
Retail & non-bank
data, broad cover,
historical record
In-depth data on
commercial sector
No automatic
access to data
Independence may
be questioned
Bank association
Chamber of
Commerce
Commercial &
Credit insurance
firms
Only bank data,
only bank access
No bank data,
Limited funds for
modernization
Limited coverage,
High cost per entry
Who submits information to
public and private registries?
Public
(of 29,
w orldw ide)
35
30
No. of registries
25
Private
(of 28 in
LAC)
20
15
10
5
0
i
pr
om
c
v
nk
a
b
p
ub
m
co
nk
a
b
p
ub
vt
e
d
nk
a
b
o
oo
c
/
n
p
sin
a
/le
g
ue
s
is
rs
l
g
'
d
ns
a
o
rd
rp
a
ov
o
c
r
c
d
p
d
e
e
il &
s
e
r
c
a
r
c
t
m
c
an
re
fir
if n
i
un
s
nt
a
h
rc
e
m
Firm data collected by
public and private registries
Public
Credit
Registries
(30
worldwide)
Private
Credit
Bureaus
(26 in Latin
America)
30
25
20
15
10
5
0
l
)
n
n
e
n
fo ee t ent
ID oan ity r al era rity ess ate r(s .
io lo a a m loa
n
r
t
v
i
e
l
i
t
r
u
t r n e gl.. ax
n of
sh em
ct l a t lla atu d d
ye of
s
ts it t of
a
a
t
o
a
m
re
p ng
ow con
ce tat
e
s
co f c
in u n
e
x
n
f
p
s
s
t
i
f
e
g
ta ra t
la
ty
o eo
o
in e o o r
e
n
n
a
i
i
m
s
b
rt
m up
pe alu
am
o
u
o
y
a
c
b
t
v
n
ro
in
ep
g
r
f
s
o
s
e
e
n
si
am
u
N
B
Lo
ca
lg
er
ov
e
O
th
rn
m
en
es
ie
s
nc
ci
t
ct
or
m
en
ge
n
ge
la
fo
rc
e
lle
s
es
rie
ne
ss
co
si
ta
de
ra
en
x
bu
ist
a
n.
..
da
t
id
i
ov
in
g
tr
eg
pr
Ta
er
ed
i
w
fe
La
O
th
cr
O
T
id
Private
at
e
N
pr
ov
Public
Pr
iv
s
s
nk
nk
ba
ba
at
e
at
e
Pr
iv
Pr
iv
Percent
Distribution of data by
public and private registries
100
80
60
40
20
0
Consumer Attention:
Survey Responses
Comparing Private and Public Registries
35
30
25
20
15
10
5
0
Private
Public
Access to
own data
Consumer
Relations
Department
Complaints
taken by
phone
Protocol for
taking
complaints
Comment on
record
II. The importance of credit
registries in financial markets empirical evidence
Percentage of Banks which consult a
credit registry for consumer loans
no
16%
yes
84%
Percentage of Banks which consult a
credit registry for small business loans
no
7%
yes
93%
Importance of Registry information relative to
other sources of creditworthiness
35
number of firms
30
25
20
15
10
5
0
Collateral
Financial Standing of Borrower's History
the Borrower
with the bank
Information from a credit registry is more important
Information from a credit registry is less important
II. Importance of credit registries in
financial markets
Can Credit Registries Reduce Credit Constraints?
Empirical Evidence on the Role of
Credit Registries in Firm Investment Decisions
Arturo Galindo
Margaret Miller
February 2001
Empirical Evidence of the Importance of
Credit Registries for Credit Markets
• Jappelli & Pagano (1999)
– relationship between registries characteristics
(age, type of data) and credit / GNP
• Barron & Staten (2000)
– greater availability of information reduces
default rates, improves access to credit
• Kallberg & Udell (2000)
– data from D&B has greater predictive power
than firm financial statements
• Galindo & Miller (2001)
0
ECDR
MEX
COL
PERU
ARG
CHL
BRZ
LAC
OTHER DEV
U.S.
CREDIT REGISTRY INDEX
100
80
60
40
20
Median Firm (Debt/Capital) v.
Bureau Index (Galindo & Miller)
coef = .28506701, se = .10285224, t = 2.77
UNITED S
.202021
IRELAND
AUSTRALI
Median (Debt/Capital)
BELGIUM
JAPAN
MEXICO
CHILE
SWEDEN
ARGENTIN
BRAZIL
AUSTRIA
ITALY
GERMANY
PORTUGAL
COLOMBIA
VENEZUEL
PERU
THAILAND
SPAIN
MALAYSIA
RUSSIA
TURKEY
-.244506
-.41669
.373786
e( igen4 | X )
Bureau Index
Source: World Bank survey, World Scope and authors’ calculations
Empirical Model
I it
I it 1
 c  1
  2 MPK it   3 FIN it 
K it
K it 1
 Credit 
 4 FIN it * INDEX c   5 FIN it * 
   i ,t
 GDP  tc
Regression Results for
Credit Registry Variables
CF/K*Index:
-0.046 (95%)
CF/K*(Pos/Neg): -0.022 (90%)
CF/K*Quantity: -0.102 (95%)
CF/K*Access:
-0.044 (90%)
Type of loans, type of report not significant
Estimated sensitivity of firm investment
to cash flows, with and without registries
(Galindo & Miller)
Sensitivity of
Sensitivity of
Investment to
Investment to
Cash Flows
Cash Flows given
Assuming
current bureau
Bureau Index=0
index
ARGENTINA
BOLIVIA
BRAZIL
CHILE
COLOMBIA
COSTA RICA
DOMINICAN REPUBLIC
ECUADOR
GUATEMALA
MEXICO
PANAMA
PERU
URUGUAY
0.078
0.072
0.078
0.071
0.079
0.079
0.078
0.076
0.080
0.080
0.062
0.078
0.074
0.047
0.047
0.042
0.039
0.052
0.045
0.051
0.053
0.053
0.054
0.037
0.047
0.048
Percentage
Reduction in
Coefficient
39.7%
35.3%
46.2%
44.2%
33.7%
43.4%
34.5%
30.9%
32.9%
32.4%
40.8%
39.3%
35.4%
Regression Results
Dependent Variable: Private Credit/GDP
Constant
Income per capita (logs)
Average economic growth
Effective creditors rights
Year credit registry
R2
Number of observations
*** Significant at 1%.
** Significant at 5%.
* Significant at 10%.
-17.64
(-0.220)
0.59
(0.170)
6.18
*
(1.750)
13.45
***
(4.120)
0.41
**
(2.040)
0.57
28
Financial Development vs Years of Credit Registry
80
Coefficient: 0.41
t-statistic: 2.04
R2: 0.57
Private Credit/GDP
60
Panama
40
Bolivia
20
El Salvador
Peru
Colombia
0
Ecuador
Haiti
Uruguay
Guatemala
Brazil
-20
Venezuela
Chile
Costa Rica
Mexico
Dominican Republic
Argentina
-40
-60
-40
-20
0
20
40
Years of Credit Registry
Note: Figures adjusted by effective creditors rights, average GDP growth, and income per capita (log).
60
80
III. Public policies to support
credit reporting emerging elements of “good practice”
III. Emerging elements of “good practice”
Legal and regulatory framework
• Legal framework should encourage information
sharing among lenders
– review bank secrecy laws which can constrict
information flows
• Consideration of privacy issues important
– broad privacy laws may unduly limit credit reporting
• Regulatory framework with enforcement
– consumers have ability to bring complaints outside
judicial system
• Competition policy aspects of credit info.
III. Emerging elements of “good practice”
Data collected and maintained
• Open system, not closed network
– ownership by a limited group of lenders, bank
association, will discourage a broader database
• Collect both positive & negative information
• Maintain data for a reasonable time frame - 5
years minimum
– do not delete data on non-payments when debt repaid
III. Emerging elements of “good practice”
Data collected and maintained
• Data should be inaccessible after a certain amount
of time
– time limits may vary by size of loan, type of inquiry
• Credit reports should not include highly sensitive
information such as sexual orientation, political or
religious affiliation, etc.
• Other identifying information, such as gender,
should be evaluated more carefully
III. Emerging elements of “good practice”
Data distributed
• Integrity and transparency are paramount
– special standing of any group, including owners or
government, will discourage participation
• Open system preferable, reciprocity not necessary
• Access to detailed information preferable
– loans described individually, not aggregates
– institutions providing credit identified
• Restrictions to prevent “cherry-picking”
• Distribution reflects privacy considerations
III. Emerging elements of “good practice”
Credit reporting and bank supervision
• Supervisors include financial institution’s use
of credit information as part of inspections
• Require publicly (government) owned
financial institutions to provide data to
legitimate credit reporting firms, associations
• Encourage all financial institutions to
participate in credit reporting
III. Emerging elements of “good practice”
Public Credit Registry (PCR)
• Clear objectives for PCR
– consult with financial institutions, private credit
reporting firms
• Complement, not compete, with private firms
• Focus on larger loan sizes
• Provide customer service if data is distributed
to financial system
III. Emerging elements of “good practice”
Public Credit Registry (PCR)
• Rating policy carefully considered
– syndicated loans should have uniform rating
– small loans don’t require monthly rating for PCR
– requiring all loans by a borrower to have the same
rating can mask differences in loan types, quality
– distribution of ratings to financial system can
create perverse incentives
III. Emerging elements of “good practice”
Consumer Attention
• Borrowers should have access to their own
data
• Consumer-friendly procedures in place to
challenge erroneous information in reasonable
time frame
• Record who has accessed data as part of report
• Clearly established privacy policy
IV. Conclusions
• Credit registries are an increasingly important part
of modern financial systems
• Empirical evidence supports importance of credit
registries for access to credit, especially by
consumers and small and micro enterprises
• Policy conditions, including the legal and
regulatory framework, are critical in the
development of robust credit reporting systems