Credit Reporting: What Role for the State? - World …siteresources.worldbank.org/.../Credit_Reporting_Presentation.pdf · Credit Reporting: What Role for the ... of scale to avoid

Credit Reporting: What Role for the State?

Miriam Bruhn (DECFP), Subika Farazi (FPDCE) Martin Kanz (FPDCE)

1

Global Financial Development Report (GFDR) Seminar

• Why credit information matters

– Why share credit information? (i) Risk-management and prudential regulation asymmetric information (ii) Promotion of credit discipline moral hazard

• Credit information and the crisis

– New focus on credit information sharing for risk-management and prudential regulation – Need for comprehensive credit reporting. Identify barriers to the entry

of smaller players, non-regulated financial institutions (important rapid growth of consumer finance in emerging markets)

Introduction Introduction Roadmap Background Evidence: Bank Concentration and Credit Reporting Policy Implications

2

• Evidence on the real effects of credit reporting • Reduced default risk, lower interest rates (Padilla and Pagano, 2000) • Improved financial access (Galindo and Micco, 2010, Brown et al., 2009,

Love and Mylenko, 2003, and Galindo and Schiantarelli, 2003) • Improved stability of banking sector (Houston et al., 2010, Simovic, 2009) especially important in rapidly growing consumer finance markets

• Credit reporting and financial access • Reputational collateral for marginal borrowers (Miller, 2003) • Inclusion of non-regulated lenders (NBFIs, MFIs)


3

• Today How is the performance of credit reporting institutions affected by the degree of competition and concentration in the banking sector?

• Policy implications: what role for the state?

– Transparency in credit reporting is closely linked to bank structure

– The state can play a useful role overcoming market failures and coordination problems arising from bank concentration

– Smart regulation needs to provide incentives for transparency where the market does not provide them (increasing social returns to scale)

-- Where registries and bureaus coexist their roles should be distinct and non- overlapping to realize economies of scale to avoid crowding out -- Support extension of credit reporting to non-regulated institutions, lower bar -- Set and monitor standards for the type and quality of information provided


4

1. Introduction -Why credit information matters

2. The challenge of building transparent credit reporting systems -When and why the market may not do it -Bank concentration and credit reporting

-What role for governments in credit reporting?

3. Empirical overview - The nature and reach of credit reporting systems

- Bank concentration and credit reporting

4. New Evidence

5. Policy implications

Roadmap Introduction Roadmap Background Evidence: Bank Concentration and Credit Reporting Policy Implications

5

• Monopoly rents in the market for credit information – Banks (especially large banks) can capture monopoly rents and

prevent entry by not sharing information – Increasing social returns of credit information sharing (Japelli and

Pagano, 1993, 2002) – …but decreasing private returns of credit information sharing – Coordination problems in information sharing among private lenders

• Implications – Mere existence of credit reporting institutions may not be enough; underlying bank concentration matters (closed user groups etc) – Especially important barrier for the extension of credit reporting to

non-regulated/nontraditional lenders such as MFIs and NBFCs

Problems in credit information sharing Introduction Roadmap Background Evidence: Bank Concentration and Credit Reporting Policy Implications

6

The role of the state

• What role for the state?

– Public good character of credit information sharing may justify involvement of the state

• Role of the state (i) Regulator (ii) Promoter (iii) Operator of credit information systems

• What should be the objective?

– Fix market failures (coordination problems among private lenders) – Realize economies of scale where the market does not provide the right incentives

(lower barriers of entry for smaller players NBFCs, MFIs) – Set data quality and reporting standards

The role of the state Introduction Roadmap Background Evidence: Bank Concentration and Credit Reporting Policy Implications

7

• How does bank concentration affect credit reporting?

(i) Emergence of credit reporting institutions (ii) Institutional participation in credit reporting (iii) Extent of information collected and distributed • Can state involvement help? • Are there complementarities between public and private credit reporting?

Countries with low bank concentration

Countries with high bank concentration

Credit Registry? .56 .37

Credit Bureau? .80 .39

Credit bureau or registry? .92 .53

Bank concentration and credit reporting Introduction Roadmap Background Evidence: Bank Concentration and Credit Reporting Policy Implications

8

Institutional Participation in Credit Reporting

Introduction Roadmap Background Evidence: Bank Concentration and Credit Reporting Policy Implications

• Large variation in type of credit reporting institutions across regions • Legal origin and credit institutions: Anglo-Saxon countries typically stronger creditor rights, less state involvement

020

4060

80

EAPECA

LAC

MENAOECD SA

SSA

by Type and RegionCredit Reporting Institutions

Credit bureau Credit registry

9



• Different ways of measuring the coverage of the credit reporting system • Financial stability: volume of credit listed • Financial access: number of individuals covered

020

4060

8010

0

EAPECA

LAC

MENAOECD SA

SSA

as % of GDPTotal Credit Listed

Credit listed in largest bureau Credit listed in credit registry

10



Functional Differentiation

• Private bureaus collect more information from non-bank/non-regulated lenders • Potential limitations of registry for risk-management in emerging markets with large NBFI share • Strong case for private bureau if target is incorporation of MFIs, NBFIs

SSA

SA

OECD

MENA

LAC

ECA

EAP

Who Provides Information to Credit Bureaus?

Banks NBFCs MFIs

SSA

SA

OECD

MENA

LAC

ECA

EAP

Who Provides Information to Credit Registries?

Banks NBFCs MFIs

11

Depth of Credit Information


It’s not just ownership that matters…

• Variation in the type and quality of information collected • Bureaus do more active monitoring… • …and collect more repayment information

SSA

SA

OECD

MENA

LAC

ECA

EAP

Credit BureausInformation Collected

Personal Loan Repayment

SSA

SA

OECD

MENA

LAC

ECA

EAP

Credit RegistriesInformation Collected

Personal Loan Repayment

12

b= -.48, R2=.12

ALBALG

AGO

ANU

ARG

ARM

AUS

AUT

AZE

BHS

BHR

BGD

BRB

BLR

BLZ

BEN

BOL

BIH

BWA

BRA

BGR

BFABDI

CMR

CAN

CPV

TCD

CHL

CHN

COL

COR

CRI

CIVHRV

CYP

CZE

DEN

DOMECU

EGY

SLV

EST

ETH FIN

GAB

GEO

GER

GHA

GTM

GUY

HND

HKG

HUN ISLIND

IDN

IRN

IRL

ISR

ITA

JPN

JPN

KAZ

KEN

KOR

KWT

KGZ

LVA

LEB

LSO

LBY

LTU

MAC

MDG

MYS

MLI MRT

MRU

MEX

MDV MNG

MAR

NPL

NLD

NIC

NERNGA

NOR

PAK

PAN

PNG

PAR

PER

PHL

POL

PRTQAT

ROM

RUSRWA

SAU

SEN

SLE

SIN

SVK

SVN

ZAF

ESPLKA

SDN

SWZ

SWE

CHETHA

TTO

TUN

TUR

UGA

UKR

ARE

GBRURY

YEM

ZMB

-.4-.2

0.2

.4.6

Res

idua

l, pa

rtici

patio

n in

dex

.25 .5 .75 1Bank concentration

Participation in Credit Reporting


13

b= -.47, R2=.10


ALBALGAGO

ANU

ARG

ARM

AUS

AUT

AZE

BHS

BHR

BGD

BRB

BLRBLZ

BEN

BOLBIH

BWA

BRA

BGR

BFABDI

CMR

CAN

CPV

TCD

CHL

CHN

COL

COR

CRI

CIV

HRV

CYP

CZE

DEN

DOM

ECU

EGY

SLV

EST

ETH

FIN

GAB

GEO

GER

GHA

GTM

GUY

HND

HKG

HUN

ISL

IND

IDN

IRN

IRL

ISR

ITA

JPN

JPN

KAZ

KEN

KOR

KWTKGZ

LVALEB

LSO

LBY

LTUMAC

MDG

MYS

MLIMRT

MRU

MEX

MDV MNG

MAR

NPL

NLD

NIC

NERNGA

NOR

PAK

PAN

PNG

PARPER

PHLPOL

PRT

QAT

ROM

RUS

RWA

SAU

SENSLE

SIN

SVK

SVN

ZAF

ESP

LKA

SDN

SWZ

SWE

CHETHA

TTO

TUN

TUR

UGA

UKR

AREGBR

URY

YEM

ZMB-.4

-.20

.2.4

.6R

esid

ual,

info

rmat

ion

colle

cted


Information Collected

14

b= -.48, R2=.12

ALBALGAGO

ANU

ARG

ARM

AUS

AUT

AZE

BHS

BHR

BGD

BRB

BLRBLZ

BEN

BOL

BIH

BWA

BRA BGR

BFABDI

CMR

CAN

CPV

TCD

CHL

CHN

COL

COR

CRI

CIV

HRV

CYP

CZE

DEN

DOM

ECU

EGY

SLV

EST

ETHFIN

GAB

GEO

GER

GHA

GTM

GUY

HND

HKG

HUN

ISL

IND

IDN

IRN

IRL

ISR

ITA

JPN

JPN

KAZ

KEN

KOR

KWT

KGZ

LVALEB

LSO

LBY

LTUMAC

MDG

MYS

MLIMRT

MRU

MEX

MDV MNG

MAR

NPL

NLD

NIC

NERNGA

NOR

PAK

PAN

PNG

PAR PERPHL

POL

PRT

QAT

ROM

RUS

RWA

SAU

SENSLE

SINSVK

SVN

ZAF

ESP

LKA

SDN

SWZ

SWE

CHETHA

TTO

TUN

TUR

UGA

UKR

ARE

GBR

URY

YEM

ZMB

-.4-.2

0.2

.4.6

Res

idua

l, in

form

atio

n di

strib

uted


Information Distributed


15

Pr[Credit Registry]=1 Pr[Credit Bureau]=1

Probit Probit Probit Probit Probit Probit Probit Probit

Bank concentration -0.312 -0.204 -0.191 0.076 -0.547** -0.502** -0.457** -0.529*** (0.220) (0.230) (0.275) (0.299) (0.226) (0.233) (0.203) (0.200)

Per capita GDP -0.041 -0.030 -0.018 -0.053 0.037 0.044 0.070** 0.088** (0.037) (0.037) (0.040) (0.053) (0.033) (0.032) (0.034) (0.038)

Private Credit/GDP -0.003* -0.002* -0.002 -0.002 0.002* 0.002 0.002 0.001 (0.001) (0.001) (0.001) (0.002) (0.001) (0.001) (0.002) (0.002)

Contract enforcement 0.131 0.152 0.027 -0.076 -0.104 -0.175 (0.106) (0.118) (0.133) (0.093) (0.096) (0.111)

Creditor rights -0.042 0.044 0.025 0.012 (0.042) (0.053) (0.036) (0.041)

French legal origin 0.601*** -0.211** (0.098) (0.106)

German legal origin 0.348*** -0.331 (0.079) (0.211)

Transition legal origin 0.036 -0.396* (0.193) (0.234) Number of observations 135 132 115 111 135 132 115 111 R-squared 0.070 0.072 0.072 0.281 0.139 0.141 0.249 0.286 Note: *** p<0.01, ** p<0.05, * p<0.1

Bank concentration and the emergence of credit reporting


16

• Going from the 25% to the 75% percentile in bank concentration (from 56% to 89%) is associated with a 17 percentage point lower likelihood of having a credit bureau • Countries with French and German legal origin are more likely to have a credit registry, but less likely to have a credit bureau, compared to countries with British legal origin

• Results remain unchanged when using an instrumental variables strategy with the following instruments for bank concentration and institutions (i.e. creditor rights, contract enforcement) – Banking applications denied – Legal origins – Region

Second stage First stage

iiii nsinstitutioionconcentraty εξββα ++++= i21 X'

ii regionlegoriappdeniedionconcentrat εξγγγφ +++++= i3211 X'_

iiiii regionappdeniedlegornsinstitutio εξγγγφ +++++= i3212 X'

Robustness check: Instrumental variables strategy


17

Credit listed/GDP, credit registry Credit listed/GDP, credit bureau

OLS OLS OLS OLS OLS OLS OLS OLS

Bank concentration -0.024 0.060 0.174 0.392 -0.517* -0.444* -0.368 -0.383 (0.152) (0.163) (0.217) (0.245) (0.289) (0.260) (0.293) (0.346)

Per capita GDP 0.049 0.057* 0.070* 0.068* 0.032 0.042 0.060 0.093 (0.031) (0.032) (0.038) (0.039) (0.056) (0.057) (0.071) (0.090)

Private credit/GDP 0.002 0.002 0.002 0.002 -0.001 -0.001 -0.002 -0.002 (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.003)

Contract enforcement 0.086 0.136 0.097 0.173 0.230* 0.252* (0.078) (0.090) (0.094) (0.107) (0.124) (0.149)

Creditor rights 0.028 0.049 0.086* 0.056 (0.036) (0.037) (0.047) (0.040)

French legal origin 0.319*** -0.262 (0.113) (0.223)

German legal origin 0.515** -0.384 (0.240) (0.236)

Scandinavian legal origin -0.343** -0.300

(0.159) (0.307) Transition legal origin 0.150 -0.118

(0.116) (0.197) Constant -0.189 -0.853 -1.390* -1.545* 0.295 -0.946 -1.599* -1.704* (0.242) (0.584) (0.763) (0.797) (0.429) (0.646) (0.970) (1.025)

Number of observations 130 127 110 110 101 99 85 85

Adjusted R2 0.045 0.047 0.074 0.173 0.023 0.037 0.063 0.079 Note: *** p<0.01, ** p<0.05, * p<0.1

Bank concentration and the reach of credit reporting institutions


18

• Higher bank concentration is associated with a lower volume of credit listed in the credit bureau

Credit registry Credit bureau Overall Personal Loan Repayment Overall Personal Loan Repayment


Bank concentration -0.012 -0.042 0.028 -0.098 -0.378* -0.420* -0.355 -0.534 (0.134) (0.138) (0.171) (0.178) (0.182) (0.176) (0.209) (0.228)

Contract enforcement -0.060 -0.037* -0.075* -0.081* -0.007 -0.050 0.008 -0.035 (0.057) (0.059) (0.068) (0.073) (0.096) (0.094) (0.108) (0.134)

Creditor rights 0.020 0.009 0.034 0.033 0.047 0.048 0.030 0.073 (0.020) (0.021) (0.026) (0.026) (0.030) (0.029) (0.031) (0.040)

Per capita GDP -0.032 -0.045 -0.021 -0.047 0.062 0.037 0.051* 0.106* (0.022) (0.023) (0.026) (0.028) (0.027) (0.026) (0.035) (0.034)

Private credit/GDP -0.000 -0.000 -0.001 0.000 -0.002 -0.002 -0.002* -0.003 (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)

Crisis 1970-2000 -0.086 -0.118 -0.072 -0.099*** -0.037 -0.070 -0.001 -0.014 (0.055) (0.056) (0.065) (0.077) (0.064) (0.063) (0.075) (0.077)

French legal origin 0.268 0.264 0.311 0.355** -0.046 -0.049 -0.017 -0.066 (0.057) (0.064) (0.070) (0.069) (0.074) (0.071) (0.087) (0.097)

German legal origin 0.199 0.206 0.253 0.264** 0.032 -0.023 0.148 0.056 (0.074) (0.070) (0.089) (0.110) (0.095) (0.088) (0.116) (0.128)

Scandinavian legal origin -0.024 0.012 -0.080 0.035 -0.266 -0.106 -0.318 -0.445 (0.087) (0.088) (0.108) (0.115) (0.124) (0.151) (0.118) (0.157)

Transition legal origin 0.048 0.013 0.054* 0.130* 0.019 -0.100 0.106* -0.022* (0.114) (0.116) (0.148) (0.147) (0.149) (0.140) (0.172) (0.180)

Constant 0.762 0.809 0.775 0.999 0.217 0.714 0.163 0.287 (0.456) (0.458) (0.550) (0.616) (0.803) (0.761) (0.926) (1.085) Number of observations 115 115 115 115 107 107 103 107 Adjusted R2 0.151 0.180 0.143 0.119 0.115 0.077 0.065 0.196 Note: *** p<0.01, ** p<0.05, * p<0.1

Bank concentration and the quality of collected credit information


19

• Higher bank concentration is associated with a lower quality of the information collected by the credit bureau

Credit registry Credit bureau

Overall Personal Loan Repayment Overall Personal Loan Repayment


Bank concentration -0.104 -0.131 -0.071 -0.160 -0.368* -0.383* -0.391 -0.509 (0.123) (0.135) (0.146) (0.167) (0.169) (0.185) (0.195) (0.225)

Contract enforcement -0.053 -0.041* -0.062* -0.056* 0.010 -0.068 0.037 -0.013 (0.057) (0.057) (0.062) (0.073) (0.090) (0.094) (0.104) (0.132)

Creditor rights 0.029 0.021 0.040 0.026 0.050 0.046 0.036 0.082 (0.020) (0.021) (0.024) (0.025) (0.028) (0.029) (0.030) (0.040)

Per capita GDP -0.031 -0.057 -0.016 -0.041 0.059 0.038 0.049* 0.104* (0.021) (0.022) (0.023) (0.026) (0.025) (0.027) (0.033) (0.034)

Private credit/GDP -0.000 -0.000 -0.001 0.000 -0.002 -0.002 -0.002* -0.003 (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001)

Crisis 1970-2000 -0.087 -0.125 -0.066 -0.107*** -0.023 -0.050 0.016 -0.012 (0.052) (0.056) (0.059) (0.075) (0.060) (0.069) (0.069) (0.075)

French legal origin 0.244 0.264 0.237 0.313** -0.061 -0.066 -0.033 -0.074 (0.058) (0.066) (0.067) (0.069) (0.069) (0.073) (0.080) (0.095)

German legal origin 0.184 0.202 0.191 0.252** 0.009 -0.029 0.143 0.030 (0.071) (0.070) (0.079) (0.105) (0.090) (0.092) (0.115) (0.129)

Scandinavian legal origin -0.003 0.058 -0.064 0.039 -0.275 -0.111 -0.274 -0.469 (0.088) (0.096) (0.099) (0.111) (0.111) (0.133) (0.114) (0.148)

Transition legal origin 0.041 0.023 0.020* 0.127* 0.051 -0.129 0.157* 0.034* (0.113) (0.120) (0.129) (0.140) (0.145) (0.139) (0.166) (0.185)

Constant 0.761 0.943 0.703 0.861 0.079 0.788 -0.029 0.111 (0.450) (0.447) (0.501) (0.606) (0.746) (0.780) (0.862) (1.061) Number of observations 115 115 115 115 107 104 103 107 Adjusted R2 0.129 0.171 0.098 0.093 0.138 0.068 0.097 0.203 Note: *** p<0.01, ** p<0.05, * p<0.1

Bank concentration and the quality of distributed credit information


20

• Higher bank concentration is associated with a lower quality of the information distributed by the credit bureau

Registries and bureaus:

Credit registries and bureaus: complements or substitutes?


• Some anecdotal evidence of crowding out

Bosnia and Herzegovina: public registry has no cutoff

0

10

20

30

40

50

60

70

80

2003 2004 2005 2006 2007 2008 2009 2010

Public registry coverage (% of adults) Private bureau coverage (% of adults)

Source: Doing Business and Mylenko 2011

21



Credit bureau

Exists=1 Credit listed/GDP 2005-2010

Institutional participation

Information collected

Probit OLS OLS OLS

Registry=1 -1.048** -0.249* -0.085 -0.146** (0.411) (0.127) (0.059) (0.067)

Bank concentration -2.191** -0.353 -0.406** -0.334** (0.975) (0.336) (0.168) (0.168)

Contract enforcement -0.635 0.264* 0.051 0.020 (0.509) (0.150) (0.067) (0.092)

Creditor rights 0.102 0.071* 0.025 0.054** (0.188) (0.042) (0.028) (0.028)

Per cpita GDP 0.351** 0.081 0.065*** 0.058** (0.147) (0.087) (0.022) (0.027)

Credit/GDP 0.004 -0.002 -0.000 -0.002** (0.008) (0.003) (0.001) (0.001)

French legal origin -0.298 -0.127 0.027 0.026 (0.475) (0.180) (0.071) (0.081)

German legal origin -0.528 -0.285 -0.098 0.085 (0.635) (0.201) (0.087) (0.096)

Transition legal origin -1.066* -0.095 0.035 0.033 (0.646) (0.198) (0.100) (0.141)

Scandinavian legal origin -0.345 -0.188 -0.299** (0.305) (0.128) (0.123)

Constant 4.560 -1.660 -0.214 0.058 (4.001) (1.067) (0.566) (0.749) Number of observations 111 85 107 107 R2 0.328 0.216 0.289 0.232 note: *** p<0.01, ** p<0.05, * p<0.1

22 • Evidence suggests credit registries may crowd out credit bureaus



Credit bureau

Exists=1 Credit listed/GDP 2005-2010

Institutional participation

Information collected

Probit OLS OLS OLS

RegistryXConcentration 0.570 1.623** -0.717** -0.061

(0.428) (0.751) (0.282) (0.292)

Registry=1 -0.517* -1.302** 0.391* -0.106

(0.291) (0.579) (0.202) (0.203)

Bank concentration -0.862** -1.297* 0.014 -0.299

(0.370) (0.682) (0.247) (0.215)

Contract enforcement -0.108 0.274* 0.035 0.019

(0.092) (0.148) (0.064) (0.093)

Creditor rights 0.021 0.083* 0.018 0.053*

(0.033) (0.043) (0.027) (0.028)

Per capita GDP 0.065* 0.094 0.058** 0.057**

(0.035) (0.085) (0.023) (0.028)

Credit/GDP 0.001 -0.003 0.000 -0.002**

(0.001) (0.003) (0.001) (0.001)

French legal origin -0.077 -0.187 0.043 0.028

(0.093) (0.183) (0.072) (0.083)

German legal origin -0.166 -0.332 -0.079 0.086

(0.200) (0.214) (0.081) (0.098)

Transition legal origin -0.381 -0.192 0.066 0.036

(0.247) (0.218) (0.105) (0.144)

Scandinavian legal origin -0.167 -0.270** -0.306**

(0.293) (0.135) (0.121)

Number of observations 111 85 107 107

R2 0.253 0.291 0.339 0.232

note: *** p<0.01, ** p<0.05, * p<0.1

23 • But: existence of a registry may also help break banks’ information monopoly

Policy implications


• Transparency in credit reporting matters – Financial access and inclusion – Financial stability and prudential regulation

• Degree of state involvement varies – …but it’s not only the ownership of credit reporting systems that matters

• Promoting credit information cannot be isolated from credit market competition more broadly

• Bank concentration also determines quality of credit reporting – Smart regulation needs to understand concentration of the banking sector – Private credit reporting may not emerge without state intervention if bank

concentration is high

• Country examples: market failure, capture of credit reporting systems by small number of banks if bank competition not taken into account

24

Additional research


• Would like to study how different types of government interventions can promote the emergence of a private credit bureau when bank concentration is high (i) Regulation (ii) Mandatory information sharing

• Currently do not have good cross-country data on credit information

regulation…

• …suggestions for data sources and additional research questions?

25

Credit Reporting: What Role for the State? - World …siteresources.worldbank.org/.../Credit_Reporting_Presentation.pdf · Credit Reporting: What Role for the ... of scale to avoid

Documents