February 2009

The study South Africa’s political history has lead to a large section of the population falling into the low income group where there is insufficient access to credit. This has marginalised a large part of the South African population where they have little or no credit history at all, which in turn denies them access to credit and a better living standard.

The low income and marginalised population are not perceived key to the mainstream credit providers as these consumers are regarded as non-profitable and too risky to be seen as valued consumers.

The information within credit reports is a major factor in determining the amount and terms of credit to be extended to a consumer, if credit is extended at all. This information therefore plays an important role in shaping the lives of individual South Africans. The result of poor credit reports, and the subsequent denial of credit, often leads to individuals who require credit to seek out and use unregulated and informal lenders who provide the credit at a high cost and often employ unlawful and unfair practices.

The question is, can the problem of financial exclusion of marginalised borrowers be changed through using non-traditional credit assessment data in SA?

FinMark Trust, an independent African Trust which aims to promote and support the objective of increasing access to financial services for the un- and under-banked in Africa, decided to fund a research study in SA to answer the above question and Compuscan was extremely proud to be chosen to conduct the research study in collaboration with the Political and Economic Research Council (PERC), an American non-profit research institution that specialises in credit access, alternative data and information-led economic development.

Currently, mainstream credit providers use automated credit assessment processes and utilise additional information through using central sources such as credit bureaux. Non-traditional data on payment obligations such as rent, prepaid data, petrol and insurance is currently not submitted and stored at a credit bureau.

It is generally considered that, for alternative information to be used effectively in credit assessment processes, it firstly has to be gathered centrally where the controls are in place in terms of the quality and frequency of how the information is updated. This will also require all involved stakeholders seeing the quantified benefit of using the information in their businesses, acknowledging the investment costs to submit and store the information centrally and the long value add for all involved.

The objectives of the study will be as follows:

• Identifying the availability of central sources of alternative data in South Africa.
• Establishing the current status of alternative data at credit bureaux in South Africa.
• Investigating the operational challenges of regular updates of alternative data.
• Investigating industry specific regulation and/or legislation implications of sharing alternative data.
• Assessing possible infrastructure challenges for credit providers to use alternative data in credit assessment processes.
• Determining the information flow cycle considered by credit providers in their application for credit assessment purposes.
• Propose possible short, medium and long term plans of how to use alternative data in credit decision assessment processes.
• Establish and quantify the impact of alternative data if used in credit assessment processes for low income/marginal societies in terms of:

Traditionally credit bureaux collect positive and negative data on previous payment histories where credit is involved. Traditional credit reports reveal how well consumers pay individual credit accounts and include reports of any loans, judgments, bankruptcies, requests for credit and other data. Credit bureaux also provide each credit consumer with a credit score - a numerical analysis of a consumers creditworthiness. The score considers payment histories, level of indebtedness, credit levels, credit use, available credit, type of credit, and other related data.

The traditional data sets used by credit bureaux are as follows:

• Detailed Account Information
• Identifying Information
• Public Domain Information
• Default/Adverse Orders
• Fraudulent Data
• Collection Data
• Disputes Data
• Debt Restructuring Data

The majority of marginalised South Africans, who have limited credit data and are therefore not scored by typical traditional credit scoring models, do in fact engage in activities where regular periodic payments are required; which could be considered to be similar to typical credit agreements. The payment behaviour displayed by consumers could be used to demonstrate a sense of responsibility.

Examples of areas where consumers typically make regular payments are:

• Energy Payments (Water, Electricity, Gas, Petrol) Data
• Club/Loyalty Cards
• Municipalities
• Telecommunications (prepaid cellular and landline)
• Rental/Leasing Payments (House, Apartment, motor vehicle, truck hire)
• Child Care Payments
• School Fees
• Medical Aid Payments
• Saving Accounts
• Lay buys
• Television accounts (DSTV, M-Net, TV License)
• ‘Stokvel’ type savings clubs
• Insurance payments (Funeral and Life policies)
• Payment switches
• Remittance Data
• SME Trade Data

Data on these types of transactions, where some of these data sets are not currently included in credit bureau data, is referred to as non-traditional or alternative data.

If this study finds that non-traditional data is a valid resource for determining credit risk for mainstream and non-mainstream credit consumers, one can safely assume that non-traditional data can be used to assess credit risk associated with consumers. This will in turn help marginalised consumers access credit easier as their profiles will be similar in terms of mainstream credit consumers. This study, if the results are positive, could help change how the credit industry collect and submit positive and negative data for the use in credit risk assessment.

Other benefits which can be derived from this study are:

• That non-traditional data could decrease credit risk and increase access
• Create a more equitable distribution of credit
• Improve acceptance rates for mainstream credit providers
• Scoring models could be improved with the addition of alternative data
• NPL’s and defaults will decrease with even more comprehensive data sets.

We will keep you updated with regards to the progress of the research project, which will commence during March 2009.

Should you need any further information about this project, please do not hesitate to contact us at Tel: 021 888 6000 or e-mail us at This e-mail address is being protected from spambots. You need JavaScript enabled to view it