1

The value of positive credit information sharing

Importance of expanding access to finance, ensuring financial stability and guaranteeing data protection: the international experience

Oscar Madeddu, Bucarest – May 24, 2007

2

. . . .1. . . .2. . . .3

Global credit bureaus status

The benefits of positive information sharing

How can regulators help? International best practice

Agenda

3

0

10

20

30

40

50

60

70

Pre-1970

1974 1977 1981 1985 1989 1992 1994 1996 1998 2001 2003 2005

OECD Latin AmericaAsia Africa Middle East and North AfricaEastern Europe and Central Asia

Source: World Bank - Doing Business 2006

Private Credit Bureaus continue to grow

•Most of the countries have established some kind of information sharing mechanism•Most advanced areas remain OECD and Latin America (old info-sharing tradition)•Asia and Central Europe catching-up fast (2000-2005)•Sub-Saharian Africa and MENA still a long way to go

4

Who contributes data to Private Credit Bureaus?

65

43

2833

57 5747

4231

18

6

010203040506070

Priv

ate

Com

mer

cial

Ban

ksPu

blic

Com

mer

cial

Ban

ksD

evel

opm

ent

Ban

ks

Cre

dit U

nion

/ Coo

ps

Fina

nce

Cor

pora

tions

Cre

dit C

ard

Issu

ers

Firm

sPr

ovid

ing

Loan

sR

etai

l /M

erch

ants

Util

ities

Prov

ider

s

Oth

er C

redi

tB

urea

us

Mic

rofin

ance

Inst

itutio

ns

Sources of Information

# of

PC

Bs

Source: World Bank - Doing Business 2006

•All lending sectors do contribute positive information to CBs•The importance of utilities, telecoms, and microcredit is growing steadily•This info is tremendously useful for the poor, underserved, amd informal economy

5

Who are the users of Private Credit Bureaus?

Banking supervisors

Other private firms / agencies/

consumers

15%

24%

38%

36%

Supervised FIsOther non-supervised FIs

Source: World Bank - Doing Business 2006

•Non-regulated entities account for 74% of inquiries while regulated only for 24%•Banks generally concentrate on large business / corporate (volumes)•Retailers and NBFIs on consumer credit (number of customers)•Reluctance of banks towards consumer credit, retailers “pioneers” in many countries

•New trend: consumers are becoming “buyers” of information

6

When are records erased from the bureaus?

Defaults/cancelled debts are erased from the records

0%

10%

20%

30%

40%

50%`

60%

Immediately Within 1 year Within 5years

Within 10years

Over 10 years Never

Repaid Not repaid

Source: World Bank - Doing Business 2006

•Vast majority of countries / bureasu holds data for 5 years or more•Best protection for good payers

7

Ensuring data access / right to consumers: a fundamental need

•Consumers’s right to be clearly and umnistakably enforced by law / regulation •Consumers’s awareness is keyl: unknown rights = no rights•Role of regulators, lenders, and CB is key for the cultural change

35

42

3438

7

0

5

10

15

20

25

30

35

40

45

Routine checks with other data provided by FIs

Borrower complaints

Statistical checks

Software program to identify

abnormalities in data

Other

Source: World Bank - Doing Business 2006

8

. . . .1. . . .2. . . .3

Global credit bureaus status

The benefits of positive information sharing

How can regulators help? International best practice

Agenda

9World Bank Credit Reporting Systems Project - February 2002

•Credit is linked to some kind collateral, but customers cannot always offer collateral

Credit Bureaus and the “democratization” of creditThe shift from collateral to ¨Reputational collateral¨

•In a low protected environment creditors raise prices, collateral, or reject credit•Also the value of collateral is proportioned to the insolvency regime efficiency

Credit bureaus help to build a “reputational collateral“ based on payment performance which is more effective than legal rights, it is free for all, protects consumers, allows

broader credit access, hence democratizes credit lending.

10

•Credit in Bolivia (1999-2004) - Banks huge decrease (credit mostly with collateral) - Microcredit rocketed ( without collateral)

Collateral a real problem or mainly an alibi?

BANCOS (1)

2.000.000

2.500.000

3.000.000

3.500.000

4.000.000

31/12/99 31/12/00 31/12/01 31/12/02 31/12/03 31/12/04

DEPOSITOSCARTERA

BANKS ASOFIN

116.289142.572

167.426

213.778

270.399

184.295208.858

233.689257.902

318.361

414.815

84.836

0

50.000

100.000

150.000

200.000

250.000

300.000

350.000

400.000

450.000

31/12/99 31/12/00 31/12/01 31/12/02 31/12/03 31/12/04

DEPOSITOS

CARTERA

MICROFINANCE

11,58

8,30

5,626,69

9,41

11,58

6,91

4,55

11,55

3,6

2,97

16,25

14,50

17,9417,14

6,56

0,00

3,00

6,00

9,00

12,00

15,00

18,00

Dic-99 Dic-00 Dic-01 Dic-02 Dic-03 Dic-04

ASOFIN(2) ASOFIN(3) BANCA(1)

1999-2004 Non Performing Loans in Bolivia (by lending sector)

BANKS

MICROCREDIT

11

•Success factors? •Loan applications accepted by Internet, phone, mail (or branch)•No tax return or financial / income statements required; •Collateral not required about 94% of loans is unsecured•70% of decision totally automated (credit bureau +bureau scores + scoring)•Lowest processing costs (US$ 30) and time (15minutes) to underwrite a loan•By end 2004 over 1,5 million applications processed automatically•Automated portfolio monitoring (every month 100 pieces of info from CB, also bureau score)•Risk based loan price and automatic credit line increases (with behavioral scoring and CB)

•Government merits: eradicate legal constraints, support infrastructure building •strong support to good Credit Bureaus since decades•sound payment and banking system (e.g. interest rate flexibility for risk based price)•good infrastructure ( e.g. postal system, communication)•balanced legal and political environment (e.g. contract enforcement laws)

•In 1994 Wells Fargo started Business Direct (new SME division)•Average customer profile: sales US$ 325.000 year; employ > 5 people•Maximum loan amount US$ 100.000 average loan amount US$15.000•Portfolio: by 1999 US$ 9.9 billion; by 2003 US$ 15bn•In 2002 WF rated number 1 SME lender in the USA

Lending to SME:with reputational collateral and credit card information

Small Business Lending by Selected U.S. Banks in 1999

3.7

5.4

7.4

9.9 9.9

0

2

4

6

8

10

12

National City U.S.Bancorp

Bank One Bank ofAmerica

Wells Fargo

US$ billion

Wells Fargo Online Small Business Customers 1998-2003

25 3262

200

297

415

0

50

100

150

200

250

300

350

400

450

98 99 00 01 02 03

(thousands)

12

% Bad rate (decreases with higher % of positive information)

0.00%

5.00%

10.00%

15.00%

20.00%

% increase of positive info

% ba

d ra

te

accept rate 20% 8.46% 5.91% 4.66% 3.52%accept rate 40% 13.85% 7.52% 6.67% 4.89%accept rate 60% 15.30% 9.25% 8.49% 7.20%

0 25 50 100

Benefits of full positive information sharing: bad debt reduction

Source: Infopolicy 2006 – information sharing in Colombia

13

% Accepted clients (increases with a higher % of positive information)

0.00%

20.00%

40.00%

60.00%

80.00%

100.00%

% positive info

% a

ccep

ted

clie

nts

3% bad rate 2.56% 4.80% 6.64% 10.00%5% bad rate 13.60% 25.71% 19.28% 41.35%12% bad rate 72.26% 75.06% 76.49% 77.80%

0% 25% 50% 100%

Source: Infopolicy 2006 – information sharing in Colombia

Benefits of full positive information sharing: increased access to credit

14

Risk detection capacity(increases with a higher % of positive information)

0.00%

1.00%

2.00%

3.00%

4.00%

5.00%

6.00%

7.00%

8.00%

% positive info

% d

ecre

ase

False positives 3.31% 2.22% 1%

False negatives 7.53% 5.32% 3.81%

25% 50% 75%

Source: Infopolicy 2006 – information sharing in Colombia

• There are two types of wrong credit decision

Benefits of full positive information sharing: more responsible decision making

2. Refuse good customers1. Accept bad customers

15Source: Infopolicy 2006 – information sharing in Colombia

• Responsible lending: the lower the amount of positive information the higher the discrimination

Benefits of full positive information sharing: more responsible decision making

% Accepted clients(lower discrimination with higher positive information)

0%

20%

40%

60%

80%

100%

% positive info

% ac

cpet

edm

en / w

omen

men 67% 60% 57% 54% 53%

women 33% 40% 43% 46% 47%

0% 25% 50% 75% 100%

`

16Source: Doing Business in 2005, IMF Global Stability Report 2004, Moody’s rating of financial system soundness

26

50

Without C.B. With C.B.

•Ratings of countries’ financial system soundness increases with a more controlled credit risk situation

Benefits of full positive information sharing: improved country’s financial rating

17

. . . .1. . . .2. . . .3

Global credit bureaus status

The benefits of positive information sharing

How can regulators help? International best practice

Agenda

18

•Financials systems collapsed •Deposits were frozen

•NBFIs liquidated, 16 banks (out of 40) closed•GDP fell from US$ 20Bn (1998 ) to US$8 bn (2000)•20% of GDP was lost like many other crisis in LAC

•Debt/GDP ratio went from 81% to 156%

•Consequences

Still leaking wounds from over-indebtedness•In 1992 capital inflows surged from US$200 mn per year to US$780 mn in 1994 (5% of GDP). •Abundance of liquidity relaxed credit constraints for households and enterprises •This led to a lending boom offered without proper risk assessment and information•By 1994 bank credit to the private sector expanded at 60% annual rate•Credit to private sector rose to 30% of GDP, compared to less than 15% (1989-92)•This gave incentive for moral hazard, excessive risk taking, financial corruption •Weak regulation after financial liberalisation encouraged financial fragility •No Credit Bureaus were present at the time

Case study 1 – Ecuador, granting credit with no information: a financial Pearl Harbour

•Solution•Establishment of positve credit bureaus

•Project led by the Superintendence•4 CBs exiting today

•All lenders participating

19

-

2.0

4.0

6.0

Banks gross portfolio growth

US$ bn 3.3 4.3 5.4 6.0

2003 2004 2005 Jul-06

0.00%

10.00%

20.00%

Bad debt decrease year on year

Comm. Banks 7.90% 6.40% 4.90% 4.80%

Public banks 17.90% 15.30% 10.40% 10.10%

2003 2004 2005 Jul-06

0

200

400

600

Banks' microcredit portfolio growth

US$ mn 99 202 354 412

2003 2004 2005 Jul-06

0.0%

10.0%

20.0%

30.0%

Provisions decrease as a % of gross portfolio

Comm. banks 10.0% 7.7% 7.0% 6.8%

Public banks 22.2% 16.6% 12.0% 12.0%

2003 2004 2005 Jul-06

Case study 1 – Ecuadorthe results

20

•In 2001 RBI (Central Bank) requested an assessment of retail banking to McKinsey

•Objective: understand why penetration in all retail credit products was very low?

•Particularly: why low penetration in retail credit?

Case study 2 - India the problem

•Findings: low lending volumes and low access to finance was basically a result of lack of information sharing / risk management tools and automation.

21

Case study 2 – India the solution: regulators must lead the change2002 – Centrak Bank led the reforms strongly promoting the start-up of a new CB• Espousing the concept of a private Credit Bureau and fostering its development• Following project from planning to implementation • Leading legislation drafting • Monitoring consent acquisition: reports to be submitted by banks to RBI• Obstacle: legal restrictions to disclose credit information among lenders

– RBI instructed banks to compulsory obtain borrowers’ consent and share data with CB– Consent to be obtained for all new and old loans to create a critical mass in the DB– Progress report on consent obtained by banks to be submitted to the RBI– On April 5, 2004 the first Consumer Bureau (CIBIL) was launched– Penalties on banks that do not comply with submission of data– Other Credit Bureaus are now ready to be licensed in India

2004 – drafted new legislation enforcing that:•New regulatory framework for Credit Bureau’s operations •Clear consumers’s rights established•No borrower’s consent necessary for sharing (only for inquiry) •Clear financial institutions’ rights and obligations

22

•Credit is limited, collateralized and selective)•Rejection rates are extremely high•Nevertheless bad debt rate is outrageous, particularly for consumer credit •Profitability is weak, risible for public banks, and the systems is vulnerable•The most affluent customers represent the poorest risk•While the best customers (medium income) are penalized

Clearly a system that needs to be reinvented!

Case study 3 – Moroccoa problem for the Central Bank ( BAM)

23

Case study 3 – MoroccoHow was it reinvented? The “Decalogue”

• A first class credit bureau systems established1. Every month, all regulated entities must supply BAM with all information concerning all

loans typologies, regardless of the amount (consumer’s consent not required)

2. On a monthly basis BAM • consolidates, processes, and updates the data received• supplies all the licensed Credit Bureaus with the same information base

3. Historical data must be part of the database ( ideally 5 years)4. Non–regulated lenders allowed to share data with CBS (customer’s consent required)5. Inquiring a CB before granting credit becomes mandatory for lenders6. After first CB starts operations lenders no longer allowed to consult BAM’s registry7. BAM is the only supervising and licensing authority for CBs 8. Code of Conduct and other norms to be enforced through a simple BAM’ regulation9. Through supervision BAM ensures compliance with legislation and Code of Conduct10. BAM can access any CB database for supervision/systemic risk prevention purposes

24

0

500

1,000

1,500

2,000

2,500

000s

WesternUnion

Bankbranches

Postoffices

ATMs ETFPOS

Source: WB 2005; VISA; UPU 2005

25m

Case study 4 - the importance of information on mobile phones /utilities

3 bnMobile telephones: the revolution!A payment system mainly in the hands of the poor!

25

•US consumers carry mobiles twice as much as cash or credit card•This proportion is four times among those consumers aged 18-34•Mobile phone payments will climb from $ 3.2 to $37 billion in 2008(*)

•The two GIANTS:•China: 450 million handsets sold (end 2006)•India: 140 million sold (end 2006)•In January 2007 India overtook China: 8 million mobiles sold•By 2010 more than 1.2 billion handset only in China / India

Source: The Economist – February 17 2007

•The social / business impact will be dramatic•Cash to start, will be replaced by mobile phones then cards•Small shopping, short commuting, entertainment, even remittances, particularly impacted•TelCos will be in the virtual position to decide if becoming financiers themselves•Traditionally distribution and systems have been controlled by banks/payments systems•Speculations are that in a few years we may assist to the biggest payment revolution of all:

-the death of credit (plastic money) as we know it today

Case study 4 –mobile phones the great leap forward

How can this information help to grant more and more responsible credit?

26

•Tuition•Child care•Rent•Payday loans

•Gas - Energy •Water •Telecoms Cable TV •Auto insurance

•Rental payments

FEW MANYSUPPLIERS

MANY

FEW

USER

S

•Who are the unscorables?•Poor consumers, informal economy, micro business•They pay cash and have no credit history •They have difficulty in accessing credit•They have a “thin-credit history” or “no history at all•In the US estimated in 35-50 million•In Mexico up to 50% of the population•Their transaction are defined as “credit like / cash-like”

Case study 4 - score the “unscoreables”

•These include: utilities, rentals, cable TV, pay-day loans, childcare, insurance•Regularity on these payments is a strong risk indicator, and a shortcut to build a credit history•This data, normally referred as “non-traditional” or “alternative”, is very predictive•Utilities (public ones in particular), are the most helpful since universal, concentrated, reliable

•However this data rarely reaches the CBs because of regulatory hurdles to share it•Problem not due to restrictive laws but to lack of legislation regulating its sharing•In Panama the new CB law allows sharing of “ non-traditional” data (mainly utilities)•In Brazil the new CB law project (Medida Provisoria) allows the sharing of non traditional data

Source: Infopolicy 2006 – Give credit where credit is due

27Source: Infopolicy 2006 – Give credit where credit is due

Case study 4 -Positive impact of alternative data on acceptance rates•Increases credit access, decreases NPL, reduces discrimination

28

The value of positive credit information sharing

Thanks! Questions?

Oscar Madeddu – omadeddu@ifc.org