By Sohaib Jamali

Pakistan is soon going to have digital only banks. These would be important additions to the country’s growing fintech industry. Globally, digital banks and other fintech actors have contributed to growth in financial inclusion. However, as far as credit to the private sector is concerned, the impact of digital banks would be enhanced if the country also had well-functioning credit bureaus.

Despite multiple financial and banking sector reforms since the 1990s, formal credit to the private sector in Pakistan has declined both in absolute and relative terms. It is currently among the lowest for emerging market and developing economies (EMDEs). It also has a skewed distribution, as information asymmetries have made commercial banks averse to lending to segments such as housing, agriculture, and small and medium enterprises (SMEs).

Much international experience shows credit bureaus reducing information asymmetries in both advanced economies and EMDEs via the assessment and disclosure of credit scores.  This is not only because credit scores reduce the administration expense of banks and other lenders, but also because of its disciplinary effect: borrowers with good credit scores reap benefits whilst those with bad scores are penalized.

Countries where credit bureaus offer a broader range of data and value-added services have significantly more private credit (especially for small and riskier firms) than countries where such services are not available. The presence of credit bureaus is also found to be associated with lower cost of credit to firms. The information provided through credit bureaus can help reduce the funding gap between small and large enterprises. Moreover, credit to GDP ratio is not only closely associated with the presence of credit bureaus but also whether or not credit bureaus have access to positive credit information, alternate information (such as utility bills and court cases), and information on very small loans.  

Steps have recently been taken to improve the credit reporting system, including the introduction of the Credit Bureaus Act 2015 and the granting of licenses to two private credit bureaus. However, a number of legal, operational and policy level challenges continue to impede private credit bureaus from realizing their potential, as a result of which credit bureau coverage in Pakistan is still only marginally better than the minimum threshold of 5 percent as defined by the World Bank.[1]    

Regulatory framework and an enabling environment

The Credit Bureaus Act 2015 makes it mandatory for every credit institution to become a member of “at least” one private sector credit bureau in the country.  In spirit, this was meant to be a floor and not a ceiling of membership. While many Pakistani credit institutions are members of both credit bureaus in the country, many have not yet enrolled in either.  As a result, the problem of information asymmetry continues to persist in the country’s credit market.

Considering that neither the membership of two or more credit bureaus, nor the sharing of credit information is an expensive proposition, all commercial banks and other credit institutions need to be nudged to become members of all credit bureaus. Credit institutions could then choose the bureau(s) of their liking to purchase various value-added services.  This would mean that there would be no competition between the bureaus on data accumulation, leading instead to greater competition on value-added services that actually hold the potential of minimizing information asymmetries in the credit market. It is pertinent to note that India had also faced the challenge of limited coverage in its early years. But it later changed the regulation to require all banks and other credit institutions to be a member of all credit bureaus and to submit data to them, including historical information set.

Similarly, the scope of information coverage by both credit bureaus has been confined to individual loans, be it consumer loans, or loans taken by sole proprietorships. Loans taken by corporations and SMEs are not yet adequately covered by the bureaus. In part, this is because of the continuing role of public credit registry i.e., the eCIB.

At the one end, the eCIB has the legal mandate to collect credit information from banks, DFIs and microfinance banks for regulatory and supervisory purposes. And at the other end, the eCIB also processes raw credit information to produce credit information reports (CIR), which lenders are legally obligated to use before giving out corporate and commercial loans, whereas microfinance banks have the option to get CIR from the e-CIB. This implicitly substitutes and even eclipses the role of credit bureaus as credit information providers. The government, therefore, needs to notify a sunset clause for eCIB, leaving the market space open for private sector credit bureaus to compete on a full spectrum of loan segments and credit information services.

In addition to the challenge of coverage, Pakistan’s credit bureau industry faces several operational issues including delays in information sharing by credit institutions and limited historical data that form the basis of bureaus’ credit scores.  Concerted efforts are needed to streamline reporting formats, reporting time, data retention, and operational bottlenecks.  India provides an example.  India’s original regulations did not offer standardized formats for reporting corporate, consumer and microfinance lending data.  But it later streamlined various aspects of the credit reporting framework, including standardized data formats for different credit segments.  These factors contributed towards an increase in the credit information coverage on commercial lending in India from 1 million firms in 2010 to more than 23 million firms in 2019.

Alternate data & key demo-geographic indicators

Alternate credit data is critical for individuals as well as micro and small businesses, that have no credit history i.e. information relating to credit from formal financial sector and have no or insufficient collateral. Alternate data, such as telephone bills, utility payments, rental information, court judgments, and other types of transactional and reputational data, becomes a proxy of income and expenditure, as well as risk profile. It also serves as a reputational collateral, which is critical for low income borrowers.   

The link between alternate data and financial inclusion is well documented.  For example, Guyana had experienced significant increase in financial inclusion (from 2.4 percent in 2015 to 16.4 percent in 2016) after adopting a regulatory provision that obligated utility providers to share data with credit bureaus.  Similarly, in Kenya, the number of borrowers increased significantly after mobile lenders were made to provide data to credit bureaus.

Artificial intelligence and machine learning are also helping bureaus analyze trends using Big Data.  For instance, Equifax has worked with an American tech-based startup Cignify to construct a Prediction Inclusion Score in Chile for individuals who either have limited or no formal credit history.  Individuals who opt in for the service have their telecommunications usage patterns analyzed and a credit worthiness score assigned to them, which is shared with financial institutions to offer them loans.  Recent research has also shown that even simple digital footprints such as information that users leave online by accessing different websites, complement the information content and analytical framework of credit bureaus, and leads to substantial increase in access to finance and reduction in default rates.

In Pakistan, the coverage of alternate data is currently non-existent because the data is not available, even though three notifications to this effect have been passed since 2015. As per a 2016 notification, court judgements relating to debtors were to be included into the public records as credit information.  However, data pertaining to court judgements are not yet available in a format that is easily accessible to credit bureaus for the purpose of credit information reports and credit scores.

In 2020, the government directed all electricity and gas distribution companies to become members of credit bureau and advised them to furnish requisite information to the bureaus. However, those datasets are not yet reported to credit bureaus. Even if utility data is reported, it is currently prone to factual inaccuracies, and therefore unreliable. This is because many consumers of utilities are not registered as official users of those utilities. For instance, billing meters are mostly in the name of the owner of the property rather than the tenant. Or in many cases it is often in the name of the parent or grandparent of the current owner of the property, and in some cases even the previous owner of the property.

Earlier in 2019, the government had instructed mobile operators to become members of credit bureaus and furnish telecom users’ billing information. However, telecom datasets are not being reported to the credit bureaus. This is because telecom companies maintain that they are restricted to share consumers’ billing data under telecom sector regulations.

Legal interpretation of what constitutes telecom data is contested. Telecom firms maintain that even if they could share consumers’ billing data, they can only share data of post-paid customers since no credit is being furnished to pre-paid telecom users. This presents a challenge to lenders who might have appetite for giving nano or micro loans based on transactional information, where telecom spending is an important proxy to assess creditworthiness of lower-income households.

Some branchless banks and fin-tech players backed by telecoms may have access to telecom data their own clients. Similarly, start-ups such as Creditbook and DigiKhata that aim to digitize personal and small business payments and keep a log of credit transactions (receivables and payables) may also have useful credit history in their respective databases.

However, the utility of such datasets is limited and does not address the problem of industry-wide information asymmetry. Instead, such information needs to be made available to all credit bureaus, so they take a holistic view by looking at various types and sources of alternate data to determine the creditworthiness of individuals and enterprises, especially those who lack adequate collateral or previous banking history.

Sohaib Jamali works as Senior Economist at the State Bank of Pakistan (SBP). 

---

[1] This article is a summary of key findings of the SBP’s Special Section on credit bureaus titled: “Private Credit Bureaus in Pakistan – Enhancing Credit Penetration by Addressing Information Asymmetries”. The chapter is a part of the SBP’s State of Pakistan’s Economy report for Q3-FY21. The complete report is available here.