1. Credit Bureau Perspectives
for Developing Markets
Presented at:
Peking University International Academic Conference
On Personal Credit System 2008
Presented by:
Frank Lenisa, MBA (Director)
Email: frank@compuscan.co.za
2. Agenda:
Background To Compuscan
Experiences And Considerations For A Developing Market
CB Systems/Technologies
Telecommunications
Data
Reporting
Analytics
Value Added Products
CB Staffing Needs
Awareness of CB
Legal and Regulatory
3. Introductory Statement:
The establishment of a Credit Bureau (CB) entails much more than
technical considerations
Business acumen, strategic vision and strong management are required
4. Background to Compuscan:
Due to South Africa’s history of racial oppression, it has what may be
called a dual economy
One extremity is comparable to industrialized nations and the other to
developing markets
The developed economy operates in information rich environments,
with formal procedures, higher-value added products and linkages
The developed economy was served by a international CB
5. Background to CompuScan (Cont’d):
Prior to 1992, South African consumers in the developing market wanting
to borrow small amounts of money, had to rely on the unregulated
informal credit sector
informal township based moneylenders (also knows as Mashonisas)
loan sharks
Pawnbrokers
During 1992, new rules governing the provision of credit evolved, which
saw the rapid evolution of a new credit industry, known as the micro
credit industry
Legislative changes saw the growth of this sector during 1992 to 1999
From 1999 to date, an era of commercialization continues in a regulated
environment
6. Background to CompuScan (Cont’d):
By 1994, there was still a distinct lack of a credit reporting system for the
micro credit industry in South Africa
Absence of information sharing, caused weakening performance of the
institutions and a high default rate prevailed
Lenders perceived the micro credit sector risky, because lenders could not
properly price the risk involved due the state of asymmetric information
The founders of Compuscan saw the opportunity in the market to create a
credit reporting system
Full-file (positive and negative information) was collected from the start
7. Background to CompuScan (Cont’d):
Today, CompuScan is a full service CB, servicing formal banks, credit
institutions, and retailers whilst remaining the leader in the micro credit
industry
CompuScan currently employs more than 100 staff members
Delivering more than 1 million credit reports per month
Managing more than 58.2 million accounts
With more than 20 million borrowers registered on the databases
CompuScan has 3500 subscribing lenders; including banks, public and
private companies, NGO’s and Cooperatives/Credit unions
CompuScan currently operates in South Africa, Namibia, Botswana and
Uganda
8. Background to CompuScan (Cont’d):
The CB data available consists of:
Demographic information on borrowers to ensure proper identification
and analytics
Information that is pertinent to the borrower’s creditworthiness,
(accounts history, repayment profiles, judgment information, fraud
information, collections information)
Contains indications of the overall risk relating to an applicant with regard
the repayment of newly established credit, inquiries by other lenders with
a permissible purpose and credit scores
An avenue for the verification or validation of any information that may
be questioned or disputed by the borrower is available
9. Experiences And Considerations for a Developing
Market:
Lenders in developing markets often lack formal procedures, appropriate
technology, internal structures, and methodical approaches when
compared to a developed institutions in industrialized countries
The modus operandi of a CB therefore needs to adapt to this
environment, as attempts to import CB methodologies from other
contexts may have mixed results
10. CB Systems/Technologies:
The CB system need to be robust, world-class and highly customizable /
adaptive
The CB system needs constant refinement to ensure ongoing
improvement in the collection, verification, hosting and distribution of
credit data/information
Consider managing the IT requirements of the CB system in-house; this
provides control over the entire process and enables the rapid evolution
of functionalities and features
11. CB Systems/Technologies (Cont’d):
Functionalities and features include:
Configurable business rules
Data loading engine that validates data against business rules, loads
data into staging servers, then production servers, and provides
administrators/lenders with full data loading reports
Member/Subscriber/Contract management module
User administration module to allow permissible access to features
based on their roles and authorization levels
Audit trails facilities
Custom MIS reports
Data packets (XML, CSV, others) can be delivered and received by
online (CPU-to-CPU) and dialup (internet) services.
Can be integrated into institutions lending system
12. Telecommunications:
Cost of telecommunications is high in developing countries is high and
Of limited bandwidth
ISP’s infrastructure limited
Power supply issues
CB system and technologies need to cater for these limitations:
Employ data compression technologies
Limited unnecessary real-time reporting of data
Use batch processing
Back-up power systems
13. Data:
Obtaining data presents major challenges to the CB
Lenders often do not have the required data easily available
Data in different formats, database structures, and not necessarily
available in electronic format
Identifying information, such as ID numbers, names, addresses and dates
of birth is often not available or reliable (e.g. same names, mis-spelt)
CB’s make use of sophisticated matching algorithms to correctly match
and merge separate records belonging to the same individual
14. Data (Cont’d):
These algorithms traditionally use a combination of name, address and
date of birth and other such indentifying information
The ability to use such algorithms is however significantly restricted in
emerging markets where other crucial information, such as names,
addresses and date of birth are unreliable
CB needs to make use of a unique identifier – consider Biometrics
In Uganda, Compuscan implemented a biometric Identification system as
well a financial card for all borrowers, as Uganda does not have a
national ID system
15. Reporting:
Need agreement with all institutions on data layout. Develop data
manuals and data submission guidelines
Receiving the required data timely, regularly and complete are not
without its challenges, despite putting the appropriate systems in place
Loans terms shorter in developing countries, therefore data reporting
needs a higher frequency – consider daily/weekly
Institutions and their branches in almost every region/state, therefore
data collection may be need to be online (CPU-to-CPU) or dial-up
(internet); If done manually it would be become exceedingly cumbersome
16. Reporting (Cont’d):
Reporting methods include:
Dial-up (Internet)
System needs to be effective and fast. Careful consideration in what
data is reported – consider reporting only needed data. Consider
reporting only new data e.g. reporting only payment data and update
accounts details on database with payment data received
Online (CPU-to-CPU)
Preferred method; seamless/automated, and less prone to data
capture errors, but implimentation needs a lot of support from CB
Many lending institutions lack the required development capacity to
integrate the CB reporting system. CB need to support institution
technically. CB also needs knowledge institutions lending system
17. Reporting (Cont’d):
Providers of commercially available lending systems can be target to
integrate the CB reporting system. However, its not without problems
Providers do not always perform the required integration properly
Release software upgrades sometimes results in the system not reporting
the data correctly
Providers often do not inform the institution of system changes, nor train
institutions staff on CB functionalities
Compuscan has to provide the technical support capacity in many
cases
Audits of integration prior to the release of the software, or an
upgrade
CB needs knowledge of lending system to support institution
18. Reporting (Cont’d):
A way Compuscan managed to reduce these isssues was to purchase a
lending system. System is commercially available to institutions and fully
supported by CompuScan. Available as a value add product
Manual data extracts – data extracted by institutions and submitted to
CB. Best for large volume
CB supplies a Data Quality Summary Report for lenders to sign-off before
data is loaded into production database – encourages lenders to get
involved in data quality
19. Analytics:
CB data elements of developing market versus developed markets are
very similar. However, in developing markets, the emphasis is far more
towards data being recent
A judgment 2 years ago holds little weight, were in a more developed
market it might hold weight
Since the loan terms are generally shorter, the trends on variables will
differ – a worst arrears of 3 on a short term loan has a different weight
of evidence to a worst arrears of 3 on a longer term loan
Data quality and analytic value from developing markets is not great
Status codes are often not updated
Payment information is often not supplied past the loan being granted
20. Analytics (Cont’d):
I do not think lenders have a full appreciation of the value of the data,
which is why they tend not to bother reporting correctly. The lenders
need to be educated to appreciate the potential value of the data for
them to report correctly
An ongoing challenge we face is in educating lenders to report correctly
and for them to understand the analytical value of the data they report
Consider making credit score easy to interpret
22. Technical Support and System Training:
Institutions lack hardware, software, and networking capabilities
Lack of strong, competent human capacity
Therefore, support and training of institutions is key to the business
model of the CB
Common problem is that institutions often neglected to report timely
and/or regularly through
Neglect
Not knowing how
Concern about predatory lending practices of other institutions
23. Technical Support and System Training (Cont’d):
Compuscan provides ongoing on-site support and training. Each branch
visited by CB staff every 6 weeks
Support and training on CB system
Support and training of insitiutions computer and network system
Ongoing training seminars are regularly scheduled in the different regions
24. Technical Support and System Training (Cont’d):
Call centre is operational 6 days week.
Inbound calls support:
Lenders needs on the CB system
Lenders needs computer and network system
Outbound calls are directed at lenders who do not use the CB
system correctly or report correctly
25. Technical Support and System Training (Cont’d):
Insitiutions who do not report correctly are often placed in an incubation
programme, where
Lenders are placed under “high-care” and undergoes intensive training
Lenders are monitored until such time as all requirements are met and
is able to function optimally
Support and training often more than just technical; CB staff are
mentors/coaches on just about anything about credit management,
compliance, computer and networks
26. Credit Skills Training Institute:
Compuscan establiched a credit management skills training instituite
and consultancy to improve the skills, knowlegde and ablities of the
industry
South African Qualifications Authority Accredited
Focus area:
Credit management
Business and personal development
End-users computer training
Consultancy services on credit management issues
e.g. credit policy, best practices, compliance
27. Value Added Credit Information Solutions
Compuscan’s value add products /services where developed to improve
the lenders abilities, but also to strengthen the CB.
E.g. improve the ease of reporting and thus the data quality improved.
Getting lenders to understand the analytical value of data through
credit scores and analytic work
Improving the skills, knowledge and abilities of the institutions staff
through skills training and development
Identity solutions to improve data matching
Value added offerings include:
Credit Decision Engine
Allows credit institutions to easily implement automated decisioning
into their credit granting process
28. Value Added Credit Information Solutions (Cont’d):
Credit Application Processing Software
Automates the origination process into a customer focused, multi-
channel, and cost-efficient operations
Loan Management System
The loan administration software program manages the entire loan
cycle and ranges from capturing new accounts to performing
collections on outstanding loans
Fraud & Identity Management Solutions
Solutions for a wide range of business needs from securing work
stations and sensitive credit information, to the identification of
people
29. Value Added Credit Information Solutions
(Cont’d):
Credit Scoring and Analytics
A range of techniques and processes for the interpretation of credit
data leading to valuable insights to make decisions with confidence.
Creation of customised scorecards
Credit Skills training
Outcomes based skills training that provides learners with the
knowledge, skills and abilities to effectively function within the
lending operations
30. CB Staffing Needs:
The staffing requirements of a CB are vast and required skills typically
include
Information technology; database administrator, systems & network
architect, web & applications developers, network engineers
Data management; statisticians and analysts
Legal and compliance officers
Sales and telesales personnel
Technical support staff; computer and networks
Marketing and research assistants
31. CB Staffing Needs (Cont’d):
Credit Scoring Analysts and statisticians
Finance and administration personnel
Strong management and leadership
Its is important that all staff have knowledge, skills and abilities not only
of the CB environment, but also of credit management, computer and
networks, all application legislation and regulations of the industry
sector
32. Awareness of CB:
Public in emerging markets likely to resist the concept of sharing financial
information
Authorities that are unfamiliar with financial information sharing may also
be resistant for political reasons
Fragmentation, competitive fears and secrecy often characterize such
economies
Lenders fear that other lenders may poach their good clients by learning
of positive financial information
Credit if often a scarcity
33. Awareness of CB (Cont’d):
CB needs to change these perceptions through education of the
benefits of credit and financial information sharing
Lenders need to overcame their distrust of competition
Borrowers must be assured of the security of their financial information
Therefore essential that the CB focus on building awareness among
lenders, their clients, the general public, government officials, policy
makers, regulators and other potential stakeholders in the CB
34. Awareness of CB (Cont’d):
Round tables and workshops were held with all the institutions CEO’s
and the central bank in Uganda to ensure buy-in and to work though
the entire concept of the CB
Suspicions were eased, the benefits were understood and potential
concerns and problematic issues were identified and addressed
An extensive public awareness campaign was launched in conjunction
with the central bank, and is currently under way in order to address
the awareness of the CB
Awareness initiatives are ongoing in South Africa. There are some
would like to close the CB down.
35. Legal and Regulatory:
The primary objective of legislation to enable credit information sharing
and reporting is to balance the ability of lenders to exchange credit
information, whilst at the same time protecting the individual’s rights to
privacy
Positive and Negative Information Sharing - consider mandatory sharing
of data
South Africa – lenders mandated to report, but optional on enquires –
mandated to conduct a affordability assessment; only proper way to
do this is via CB report
Uganda - lenders mandated report, but optional on enquires – lenders
have options, but non adherence means that the debt must be
provisioned as a sub-standard loan immediately
36. Legal and Regulatory (cont’d):
Explicit borrower consent is generally sufficient to allow a lender to share
the information
If borrower does not consent – no loan
Only issue is retro data
Consumer Protection - Grievance and dispute resolution mechanisms
Borrowers are CB clients too
Additional mechanism to improve data quality
Legal and regulatory bodies, two broad approaches - Strong Supervisory
Body - Industry Self Regulation
South Africa has both. Started with self-regulation. More compliance
through supervisory Body (as long as body conducts audits and forces
compliance)