Striking a Balance: Credit Penetration and NPA Management - Role of Information Sharing by Anand Sinha - ആർബിഐ - Reserve Bank of India
Striking a Balance: Credit Penetration and NPA Management - Role of Information Sharing by Anand Sinha
Striking a Balance: Credit Penetration and NPA Management – Role of Information Sharing* Anand Sinha Mr. Nair, Chairman, Credit Information Bureau (India) Limited (CIBIL) and CMD, Union Bank of India; Mr. Mallya, CMD, Bank of Baroda; Mr. Siddharth Mehta, President and CEO, TransUnion; Mr. Thukral, MD, CIBIL and other distinguished guests. It is indeed an honour for me to address you all here during the fourth annual Credit Information Conference. 2. Banking is a risky business. Banks, or more generally the credit institutions, in their role as intermediaries take upon themselves significant credit risks, while interposing between lenders (depositors) and borrowers. Credit risk is, by far, the largest risk faced by banks. The risk transformation, where they become borrowers to lenders and lenders to borrowers, leads to warehousing of risks in credit institutions. In fact, this risk transformation offered by credit intermediaries is the cornerstone of credit intermediation which facilitates pooling of savings for being lent for various economic activities. 3. In credit markets, the lenders’ knowledge of a borrower’s likelihood to repay (risk profile) is imprecise and has to be estimated based on available information. Borrowers, obviously, would have superior information than lenders about their own past actions and future intentions although it is the lender who bears the negative consequences of the risk. Moreover, borrowers have incentive to misrepresent their risk profile. In any case, lenders cannot rely solely on the information provided by the applicant but must evaluate borrowers’ claims even when borrowers are truthful. This entails significant cost and time. Such information asymmetry has consequences for lending as it results in misallocation of credit. Let me explain. 4. Before the advent of credit information models, traditional neo-classical economics assumed that markets were ‘informationally efficient’. This implied that one cannot consistently achieve returns in excess of the average market returns on a risk-adjusted basis, given the information available. More recent work by Stiglitz and others has shown that only under exceptional conditions are markets perfect and even under competitive equilibrium, credit markets may ultimately witness credit rationing owing to insufficient information, i.e., information asymmetry. When lenders cannot assess the risk profile of borrowers adequately, they mistake good risks as bad ones and vice versa. Lenders typically respond to insufficient information about borrowers by increasing interest rates. Since credit is now expensive for low-risk borrowers, they are driven out from the market whereas high risk borrowers finding credit cheap are brought in. This makes the portfolio riskier and interest rates rise to reflect this. Higher interest rates create incentive even for good risks to take up riskier projects so as to have higher returns to compensate for higher cost of loan. Furthermore, some borrowers may have incentive not to repay, because without information sharing they can still obtain loans from other lenders. Thus, with the problems of moral hazard (lack of penalty for non-payment) and adverse selection (i.e., higher interest rates attract riskier borrowers or make borrowers take more risks), lenders ration credit, i.e., given two entities with identical risk profiles, one will get a loan and the other will not. 5. Credit bureaux/Credit Information Companies (CICs) which enable sharing of credit information among lenders are an institutional arrangement to deal with moral hazard and adverse selection problems on account of information asymmetry. Shared information allows better assessment of risk profiles of potential borrowers (i.e., deals with adverse selection) and creates incentive for borrowers to pay on time by limiting borrowers’ ability to access credit from other lenders (i.e., deals with moral hazard). 6. Credit information bureaux have thus come into existence to deal with the information asymmetry issues and have come a long way since the first credit registry started functioning in the USA in the mid-19th century. 7. CICs deal with the problems associated with asymmetric information in the following ways, by: i. Mitigating adverse selection by allowing loans to be extended to safe borrowers who had previously been priced out of the market, resulting in higher aggregate credit. Studies analysing the effect of CICs on credit markets found a positive effect on the volume of bank credit as a percentage of GDP and a decrease in credit risk. It was also found that firms in countries with better credit information systems were less credit constrained. ii. Mitigating moral hazard by either incentivising borrowers to repay or increasing the borrowers’ cost of defaulting as non-payment with one institution would result in sanctions from all others. This would encourage debt repayments and a more efficient recycling of credit. A study showed that without a CIC, borrowers had a tendency to repay loans only when they intended to maintain their current relationship with the lender. However, in economies with CIC, borrowers had a higher chance of repaying their loans regardless of whether or not they intended to continue the relationship. Thus, it can be surmised that credit sharing institutions, by documenting borrower behaviour, can positively impact borrower repayment. iii. Reducing information monopoly which would result in lowering of interest costs for the borrower. Lenders with long-standing relationship with borrowers know their credit histories while other lenders do not. This allows the lenders to charge higher interest rates and extract other rents from high quality borrowers. Sharing of credit-related information reduces the monopoly that a lender has on its borrowers. This reduces the extra rents that lenders can charge their borrowers. Moreover with CICs, credit institutions can screen away more risky borrowers and charge interest more aligned to their risk profile. They thus extend lower rates to less risky borrowers who would not have borrowed otherwise. Thus, the average interest cost to the borrower gets lowered. Additionally, credit reporting can play a key role in improving the efficiency of lenders by reducing loan processing costs as well as time taken to process loan applications all of which, like interest rates, have a positive impact on pricing. iv. Reducing over-indebtedness as overly extended borrowers would now receive less credit. Information sharing among lenders reveals borrowers’ debt level to all lenders. This results in highly indebted borrowers receiving less credit and in turn, results in lowering the aggregate indebtedness. A recent study demonstrated that, after establishment of a CIC, lenders were more likely to issue smaller and shorter term loans and to require more guarantees. This could indirectly provide evidence that sharing information allowed lenders to see and factor in the indebtedness of their borrowers. v. Lowering poverty and improving distribution of income. Constraints in allocation of credit in situations of poor credit information are likely to hurt the poor relatively more and increase inequality. By reducing credit constraints, credit reporting may be expected to facilitate credit to the poor and thereby help reduce poverty. Evidence demonstrates how credit sharing increased the quantity of small business loans in the US and, more importantly, served to increase credit to the marginal borrowers, i.e., firms which in the absence of credit information sharing mechanism would probably not have received any credit. It would stand to reason that such marginal borrowers would have been unable to fulfill the normal collateral requirements imposed on them by lenders. This evidence is very relevant for our financial inclusion drive which is a major policy plank. 8. All in all, international evidence suggests that information sharing increases access to credit. Studies have also shown that sharing of only negative information results in higher default rates while systems that include both positive and negative information result in lower default rates. A CIC serves the purpose of storing and sharing more comprehensive borrower information both negative and positive. This would enable lenders to achieve lower default rates. Regimes which have tough data protection norms, implying restricted access to credit information, have a detrimental effect on credit allocation. It would be a useful exercise to employ some of the models to carry out simulations based on country data on the impact of information sharing, say, on such variables as access to credit, especially to the MSME sector and default rates in lending. 9. Lack of credit information among lenders and financial market participants can have negative and sometimes serious consequences for the smooth functioning of the financial and credit markets and the general economy. Several instances bear this out. 10. In our country, we have recently seen indebtedness levels which made families in rural India very vulnerable like in Andhra Pradesh, due to multiple borrowing. This would not have reached the proportions that it did if there was information sharing. According to the Report of the sub-committee of the Board of the Reserve Bank to study the issues and concerns in the Micro Finance sector (Chairman: Shri Y.H.Malegam), an essential element in the prevention of multiple lending and over-borrowing is the availability of information to the Micro Finance Institutions (MFIs) of the existing outstanding loan of a potential borrower. According to the sub-committee, a Credit Information Bureau needs to be established expeditiously and where more than one bureau discharges the role, adequate co-ordination among bureaux will need to be established. With the establishment of 3 more CICs in addition to Credit Information Bureau of India Ltd. (CIBIL), it is hoped that the gap in credit information in Microfinance would soon be filled up and Microfinance can operate on a sounder and more sustainable basis than hitherto. Already some CICs are working on these lines. It may also be mentioned that state governments have an important role to play in financial education, particularly on the issue of debt sustainability of households. 11. Similarly, lack of effective credit information sharing has increased the risk in multiple banking, to which we had moved sometime back from consortium lending with a view to introducing flexibility in the credit delivery system by providing more choice to the borrowers. Taking advantage of information asymmetry, several unscrupulous borrowers had manipulated the multiple banking arrangement and defrauded banks of substantial funds. More recently, inadequate sharing of information among lending banks, had led to some borrowers acquiring multiple exposures in exotic derivatives which amplified the downside risk when they were seriously impacted due to adverse rate movements. This led to several defaults as well as disputes. An elaborate system for exchange of credit information among lending banks was devised to deal with these deficiencies while three new Credit Information Companies (CICs) were licenced and were in the process of being set up. 12. The subprime crisis is an example of the extremely serious consequences of lending based on inadequate credit information (risk profiling). These high risk subprime assets were securitised through a complex and long chain of intermediaries. The investors relied largely on ratings, though it should have been obvious to them that ratings are no substitute for careful due diligence before investing. However, it is doubtful whether given a long chain of intermediaries with likely progressive loss of information along the chain, willing investors would have been able to carry out meaningful due diligence. The subsequent freezing of interbank markets and hoarding of liquidity by banks was also caused by lack of information, leading to mistrust about banks’ holding of toxic subprime assets and its probable impact on the financials of counterparty banks. II. Indian Context: Steps taken by RBI to build a ‘credit information’ environment in India Arrangements for sharing credit information prior to the Credit Information Companies (Regulation) Act 13. Pursuant to the then Finance Minister’s Budget Speech made in Parliament on February 28, 1994, the Reserve Bank had put in place a scheme to collect details about borrowers of banks and Financial Institutions (FIs) with outstanding aggregating 10 million and above which were classified as ‘Doubtful’ or ‘Loss’ or where suits had been filed, as on March 31 and September 30 each year. This was done with the objective of alerting the banks and FIs and putting them on guard against the defaulters to other lending institutions. In February 1999, the Reserve Bank had also introduced a scheme for collection and dissemination of information on cases of willful default, of borrowers with outstanding balance of `2.5 million and above.1 It is to be noted that these lists are negative in nature. While this information would, no doubt, be of use for aiding banks and FIs in taking decisions in regard to sanctioning of loans, the requirement of an adequate, comprehensive and reliable information system on borrowers through an efficient database was felt by the Reserve Bank, Central Government, credit institutions and other players in the banking and financial sector. The Reserve Bank constituted a Working Group2 under the chairmanship of N H Siddiqui to explore the possibilities of setting up a credit information bureau in India. The Working Group observed that it would not be possible to set up a world class credit information company within the existing legal framework. The legal prohibition on disclosure of information contained in various laws relating to banking did not permit banks and financial institutions to share credit information with the credit information companies. Besides, the power vested with the Reserve Bank of India under Chapter III-A of the Reserve Bank of India Act, 1934, to collect and furnish credit information from banks or financial institutions cannot be delegated by it to another institution. 14. The Working Group, therefore, observed that amendments would be required in various enactments. The Working Group in its report, submitted in October 1999, recommended (a) enactment of a legislation for facilitating collection and sharing of information by the proposed bureau and (b) to amend certain Acts relating to banking to permit sharing of information with the said bureau. On the recommendation of the Working Group in January 2001, a company by the name of CIBIL was promoted by SBI in collaboration with HDFC and two foreign technology partners. The company commenced operations by pooling information on suit-filed accounts as also transactions on which the borrower had given consent for sharing amongst the user group. The Credit Information Companies (Regulation) Act, 2005 was thus enacted3 with a view to providing necessary legislative support to the business of credit information and for the regulation of credit information companies and to facilitate efficient distribution of credit. Further, the Credit Information Companies Rules, 2006 and the Credit Information Companies Regulations, 2006 framed under the Act were notified on December 14, 2006 and the Act was brought into force. Credit Information Companies (Regulation) Act, 2005 15. In this regard, the Act provides that credit information companies are companies formed and registered under the Companies Act, 1956 and which have been granted a Certificate of Registration by the Reserve Bank. The Act also provides that the Reserve Bank may, having regard to the potential and scope for expansion of the existing credit information companies and other relevant factors, determine the total number of credit information companies which may be granted the certificates of registration for carrying on such business. Section 14 of the Act enumerates the forms of business, which a credit information company may undertake and stipulates that a credit information company shall not engage in any other form of business. Section 17 of the Act, by and large, stipulates that credit information collected by a credit information company under that section shall be provided to its specified users and that the credit information company and its specified users shall not disclose the credit information received by them to any other person or for any other purpose other than as permitted or required by any other law for the time being in force. 16. These statutory restrictions on disclosure answer the question which is often asked as to whether the Act violates the privacy principles laid down in Tournier v. National Provincial and Union Bank of England [1923] All ER 550: [1924] 1 KB 461 and followed in India. The exceptions to the rule of confidentiality laid down in the Tournier’s case include disclosure, where there is a duty to the public to disclose and where the interests of the bank require disclosure. Since disclosure of information relating to borrowers will enable other credit institutions to take an informed decision regarding lending to such borrowers and the disclosing bank itself benefits by such disclosure as the borrowing of the borrowers concerned may not be expected to increase beyond limits once the other credit institutions also come to know of the existing debt of the borrower, disclosure to credit institutions may be regarded as a duty to the public or a disclosure in the interests of the bank under the Tournier principles. 17. Thus the provisions of the Act may be regarded as striking a balance between the dual objective of putting in place an efficient system of sharing of credit information in the larger interest of financial system on one hand, and maintaining the privacy of such information, on the other. In other words, the Act serves as a means to provide an effective and reliable information system on borrowers (through the database maintained by credit information companies) and, at the same, time strives to maintain customer confidentiality, to the extent permissible, and privacy as envisaged in Tournier’s case mentioned earlier. III. Challenges and the Road Ahead 18. Based on stakeholders’ feedback it is observed that CICs in India face issues relating to data accuracy, timeliness and completeness. Credit institutions in India need to substantially improve the quality and completeness of data as well as timeliness while furnishing credit information to the CICs. They also need to use the credit information data in a more wholesome manner. Our impression is that the data on corporates is not being used as much as for retail exposures. This needs improvement. 19. Let me now provide some direction that this sector would need to take in the future. First, credit institutions will need to be more proactive in cases of credit denials and unfailingly ensure that their proposed borrowers are supplied with a copy of their credit reports as provided under the Act. In fact, one of the most effective mechanisms for maintaining the quality and accuracy of information is to notify borrowers when their credit applications are refused. This notice of refusal of credit serves as an important tool to cross-check the accuracy of the credit information and also informs proposed borrowers of the need to build and maintain a good credit history. 20. The quality and scope of credit data collected and disseminated is the basic building block of an effective credit reporting environment. Accuracy of data implies that such data is free of error, truthful, complete and up-to-date. In this regard, the Act provides that credit institutions shall ensure updation of credit information on monthly basis or more frequent intervals, as mutually agreed upon with the CIC. Inaccurate data may lead to numerous problems, including unjustified loan denials or higher borrowing costs. Quality also means that data are sufficient and adequate, implying that: i) relevant detailed information is captured, including negative as well as positive data; ii) information from as many relevant sources is gathered, within the limits established by law and iii) information is sufficient in terms of the period over which observations are available. The World Bank in its General Principles for Credit Reporting, has interalia, set out that credit reporting systems should have relevant, accurate, updated and sufficient data – including positive data – collected on a systematic basis from all reliable, appropriate and available sources, and should retain this information for a sufficient period of time. 21. Incorrect data may result from human error or other causes. For example, incorrect data provided by the data subject (i.e., individuals and businesses whose credit histories and identification data are stored in these systems) or human error from creditors or other sources when inputting data, will result in incorrect data being transmitted to the credit reporting system, affecting the quality of reports. In addition, data pertaining to a certain data subject may erroneously be associated to another data subject due to inadequate identification mechanisms (e.g., improper matching of names, lack of identification keys for individuals and/or businesses, the inability of such keys to provide a unique identifier or the absence of robust matching mechanisms). If no proper definitions, tools and controls are in place, execution of such processes may result in distortions which would then lead to incorrect inferences about the data subject due to, for example, underestimation or overestimation of the data subject’s outstanding liabilities. 22. Unique Customer identification (UCI) across the banking system becomes of paramount importance. While some of the Indian banks have established a unique customer identification process, there is no unique identification across the banking system based on a shared database. This could enable borrowers to circumvent the risk profiling guidelines and obtain multiple facilities across banks. UCI should eventually graduate to a Legal Entity identifier (LEI). LEI are a unique ID associated with a single corporate entity which assists in easy identification of the entity across multiple financial market utilities and aids in tracking settlement activity and exposures. Financial Stability Board has been supporting the work by financial regulators and industry to establish a global system for identifying parties to financial transactions. G-20 has also expressed support for this initiative. The LEI solution would be used by both firms and regulators to facilitate better counterparty and systemic risk management as well as regulatory reporting. As regards individuals, the Aadhaar system devised by the Unique identification Authority of India (UIDAI) will seek to minimise or eliminate duplicate identity using the biometrics feature to do so. This will ensure that no person can get more than one identity and at the same time no one can assume the identity of another person as is possible with paper-based documents. The effect of the new systems through unique identification will enhance the quality of data with CICs. This will also ease the number of complaints received from concerned consumers. All this would ultimately have a beneficial effect on the efficient allocation of credit, especially among the poor. 23. Second, it is incumbent on CICs to address customer grievances in a transparent and timely manner and also put in place a responsive public grievance redressal mechanism. In case, a CIC or the lender discover of their own or are informed about any inaccuracy, error or discrepancy in respect of the data, the CIC or the lender has to take immediate steps to correct the error as provided under the Act. Further, in case of any dispute raised by the borrower on the content of information in his credit report or on any matter pending before a court of law, which is likely to take further time, the CIC, before furnishing data to anyone, would need to include an appropriate remark to reflect the nature of the inaccuracy. 24. Third, as with all technology-intensive organisations dealing with multiple parties, the potential for operational risk and associated reputational risk on account of errors and unauthorised access to the information, either from inside the CIC or from outside, is significant. CICs have to manage these risks. 25. Fourth, CICs must carry out campaigns to educate the public on the advantages accruing to them through enhanced access to finance through information sharing through CICs. Consumers of CICs would need to be educated about the benefits of credit sharing, what role credit reports have in facilitating their access to credit or how they could contribute to a more competitive credit market. 26. Campaigns targeted to the general public should, therefore: i. Explain the role and nature of credit information sharing to mitigate general concerns about sharing of personal information. ii. Inform about the function of a CIC and its obligations for respecting the privacy of consumer data. iii. Discuss the measures that a CIC would take to ensure the security of consumer data and the way mistakes can be corrected. 27. Fifth, as the principle of a successful credit sharing regime is based on reciprocity, it is important for credit institutions who are the primary providers of data to the CICs to use the data, in turn, for taking decisions on loan applications. Besides banks, the Act, which includes financial institutions, NBFCs, housing finance companies and companies engaged in the business of credit cards or dealing with the distribution of credit in any manner, provides that such lenders or credit institutions are required to be members of at least one CIC. It is possible that the credit information in respect of the borrowers of a particular lender, who contributes data to only one CIC, is available only with that CIC leading to an ‘information silo’. In India, in order to reduce the gap in the generation of credit information on a specific borrower and to enhance the quality of the credit information report, the Act allows one CIC to become a member of another. However, this mechanism has not taken off as much as one would have wanted, probably for reasons of competition among CICs. 28. Last, the coming of age of a credit information system in a country is a bellwether of its financial maturity. The quality of financial infrastructure determines the efficiency of intermediation, the ability of lenders to evaluate risk of consumers to obtain credit, insurance and other financial products at competitive terms. Credit reporting expands access to finance, especially for lower income consumers and MSMEs and plays a key role in improving the competitiveness and efficiency of banks by reducing loan processing costs as well as the time required to process loan applications. Intuitively, the availability of credit information would imply a more efficient allocation of credit at lower interest rates accompanied by higher economic growth and a less lop-sided income distribution. With increasing coverage, which would ultimately subsume the entire population of loans data, the information from CICs may, in future, support both on and offsite supervision. 29. To conclude, information is very critical in financial transactions and is a lubricant that keeps the wheels of the financial machinery rolling. Any gap in information is bound to result in these wheels coming to a grinding halt, as was witnessed during the global crisis where heightened concerns about counterparties’ creditworthiness resulted in freezing of markets. The importance of information in credit intermediation cannot be overemphasised. With our plans of economic recovery post crisis and also in the light of the financial inclusion initiative, envisaging bringing a vast majority of the excluded into the financial fold, the role of information sharing is very critical to keep credit flowing to various segments of the economy in adequate measures and ensuring against over-indebtedness. Over-indebtedness would result in heightened credit risk perception leading to drying up of credit flow and jeopardise the welfare of the poor. A robust credit information sharing framework is thus a very important constituent of financial system and all of us must strive to strengthen it further. 30. I thank you all once again for your attention and I wish the Conference all success. Bibliography Barron John M & Michael Staten (2000), ‘The value of comprehensive credit reports: Lessons from the US experience’ available at http://www1.worldbank.org/finance/assets/images/value_of_comprehensive_credit_reports1.pdf. Bernardo Alberto, Marco Pagano and Salvatore Piccolo (2009), ‘Multiple bank lending, creditor rights and information sharing’ Centre for Economic Policy Research, Discussion Paper No. 7186, February 2009. Brown, Martin & Christian Zehnder (2005), ‘Credit registries, relationship banking and loan repayment’ Institute for Empirical Research in Economics, University of Zurich Working Paper No. 240. Doblas-Madrid Antonio & Raoul Minetti (2009), ‘Sharing information in the credit market: Contractlevel evidence from US firms’, Social Science Research Network, March 17. Jappelli, Tullio, and Pagano, Marco (1999) ‘Information Sharing in Credit Markets: International Evidence’ IDB Working Paper No.106, June. ---, (1999) ‘Information Sharing, lending and defaults: Cross-Country evidence’ CSEF Working Paper No.22. Joseph E Stiglitz (2001), ‘Information and the change in the paradigm in economics’ available at http://www.nobelprize.org/nobel_prizes/economics/laureates/2001/stiglitz-lecture.pdf. Joseph E Stiglitz & Andrew Weiss (1981), ‘Credit rationing in markets with imperfect information’ Credit Rationing in Markets with Imperfect Information, The American Economic Review Vol. 71, No. 3, June (pp. 393-410). Michael A Turner & Robin Varghese (2010), ‘The economic consequences of consumer credit information sharing: efficiency, Inclusion & Privacy’ Background Paper #2 for the roundtable at OECD Conference Centre, December 1. OECD, Facilitating access to finance- Discussion Paper on Credit Information Sharing available at http://www.oecd.org/dataoecd/37/1/45370071.pdf. Padilla A Jorge & Marco Pagano (1999), Centre for Studies in Economics & Finance, ‘Sharing default information as a borrower discipline device’ Working Paper No. 21, July. Robert M Hunt (November 2002), Federal Reserve Bank of Philadelphia, ‘The development and regulation of consumer credit reporting in America’ Working Paper No.02-21, November. World Bank-IMF(2005), ‘Financial Sector Assessment: A Handbook’ , September 29. * Address by Shri Anand Sinha, Deputy Governor, Reserve Bank of India at the fourth Annual Credit Information Conference on March 7, 2012 at Mumbai. Inputs provided by Shri Manoj Sharma, Smt. Rachna Dikshit and Smt. Jayanthi N are gratefully acknowledged. 1 At present, the data relating to non-suit filed accounts for the above lists are being disseminated by RBI to banks and FIs for their confidential use whereas the lists pertaining to suit-filed cases are in the public domain as published by the Credit Information Companies. 2 Report of the Working Group to explore the possibilities of setting up a Credit Information Bureau, Oct 1999. 3 Statement of Objects and Reasons – Credit Information Companies (Regulation) Act 2005. |