Academic Research for Credit Bureau

• Credit bureau policy and sustainable reputation effects in credit markets, Vercammen, J. A. (1995). Economica, 461-478. This paper discusses how credit markets generate welfare-increasing reputation effects when adverse selection results in borrower reputation formation incentives with address moral hazard challenges. The author notes that welfare that stems from reputation effects declines over time as private borrower information is revealed to lenders. Borrower welfare may, therefore, decrease with time unless reputation effects are sustained. One way to sustain reputation effects is by restricting access of lenders to a borrowers credit history through credit bureau policy.
• Credit Scoring: Statistical Issues and Evidence from CreditBureau Files, Avery, R. B., Bostic, R. W., Calem, P. S., & Canner, G. B. (2000). Real Estate Economics, 28(3), 523-547. This article discusses the relevance and effect of credit scoring on the reputation of borrowers. The author asserts that although credit scores is beneficial to both lenders and borrowers, it raises significant statistical issues which can affect the ability of scoring systems to accurately quantify the credit risk of a consumer. A sample of credit-bureau records have been used to show the concerns about omitted-variable bias. Local factors have also show significant correlations with the credit scores.
• Predictive value of credit bureau reports, Chandler, G. G., & Parker, L. E. (1989). Journal of Retail Banking, 11(4). According to this article, almost all creditors use reports from credit bureaus to evaluate credit risk of new applicants. Most of the time, they do not know the real predictive value of the credit reports information. Few of them are always aware of fact that variation in the predictive value of the reports varies with the level of the detailed examination. The author further denotes that understanding this information is critical in designing a more accurate credit-risk evaluation process and in determining the value of the reports compiled by the bureaus. Consequently, the article measures the predictive power of information from credit bureaus with or without application information. In addition, the article measures changes in the predictive power as reports become more detailed. Findings have been presented for a retail revolving card, a bank card, and a nonrevolving card.
• Assessing credit card applications using machine learning, Carter, C., & Catlett, J. (1987). IEEE expert, 2(3), 71-79. This article discusses the process of credit card assessment and how it influences credit card application. Credit assessment refers to the process of predicting applicant profitability and reliability. The author has adopted a machine learning approach to understand and how credit card assessment is conducted. Using Quinlans ID3 algorithm was found plausible due to the fact that the approach does not require any domain-specific knowledge and is a cheaper and quicker approach than building an expert system using conventional approach. Finally, the author speculates the future techniques and financial applications of credit card assessment.
• Monitoring the Household Sector with Aggregate Credit Bureau Data: A New Source of Data Provides a Firmer Foundation for Credit Analysis and Decisions, Barron, J. M., Elliehausen, G., & Staten, M. E. (2000). Business Economics, 63-76. This paper illustrates how aggregated credit bureau data can be used to benchmark portfolio performance and modeling trends in the case of household borrowing and payment behavior. The authors utilize a database build from a sample of US consumer histories dated back to 1992. The comprehensive data used provide an accurate picture of borrowing behaviors at different levels including local, state, and regional levels compared to aggregate statistics that are available from the industry association and the federal government. The data used also has a predictive power which is apparent in models used to explain county-level patterns in personal bankruptcies. Further, the author associates the relevance of the data with 3 types of consumer loan delinquencies witnessed between 1993 and 1998.
• The use of credit bureau information in the estimation of appropriate capital and provisioning requirements, Falkenheim, M., & Powell, A. (1999). Central Bank of Argentina, mimeo. This article by Michael Falkenheim and Andrew Powell discusses the use of credit bureau information to estimate the appropriate capital and provision requirements. In-depth information has been provided on the primary roles of credit bureaus in collecting information on the larger debtors of financial systems and understanding those debtors including their risks. The authors also elaborate on how the bureaus have become powerful tool for sharing information between financial institutions at affordable costs. Furthermore, the bureau databases have been used to study the relationships between lenders and borrowers. Thus, the authors have used a simple technique to estimate the losses and potential losses variance of portfolio in the context of Argentine loans. They advocate for provisions and capital requirements to cover potential losses subject to some statistical level of tolerance.
• Household debt and defaults from 2000 to 2010: Facts from Credit Bureau Data, Mian, A., & Sufi, A. (2015). (No. w21203). National Bureau of Economic Research. This study reviews existing literature and offer evidence that support the view of credit supply which holds that increasing credit supply that is unrelated to fundamental improvements in productivity or income was the shock that caused the household debt boom and bust. This credit supply boom, according to the authors, has been supported by 4 theories: expansion of mortgage credit supply between 2002 and 2005 which was independent of economic circumstances; mortgage credit supply expansion which increased house prices; response of homeowners to the rising housing prices by borrowing out of home equity; and the fact that default crisis resulted from lower credit score individuals.
• Information sharing in credit markets, Pagano, M., & Jappelli, T. (1993). The Journal of Finance, 48(5), 1693-1718. This article presents a model with adverse selection as it shares information among lenders endogenously. According to the authors, the model reveals that there is a positive relationship between lenders incentive to share information about borrowers and the mobility and heterogeneity of the borrowers. The model also found that level of sharing information has a positive correlation with the size of the credit market and advances in information technology. Further, the model revealed that information sharing among the lenders increases with the volume of lending when there is severe adverse selection which makes safe borrowers to drop out of the market. These findings have been supported by historical and international evidence in regard to the consumer credit market.
• The impact of the functional characteristics of a credit bureau on the level of indebtedness per capita: Evidence from East European countries, Simovic, V., Vaskovic, V., Rankovic, M., & Malinic, S. (2011). Baltic journal of economics, 11(2), 101-130. This paper discusses the role of credit bureau in credit assessment. It outlines the role of credit bureaus in controlling indebtedness levels of populations. According to the authors, credit bureaus have key functional characteristics that influence the development of indebtedness. To prove this assertion, the authors attempt to identify the key characteristics of credit bureaus as well as quantify the characteristics and identify the casual relationships between credit bureau characteristics and indebtedness trends. The Credit Bureau Functional Index has been introduced as a quantified value that reflects the characteristics of a credit bureau. Consequently, a correlation between index and indebtedness per capita is established.
• The reaction of consumer spending and debt to tax rebatesevidence from consumer credit data, Agarwal, S., Liu, C., & Souleles, N. S. (2007). Journal of political Economy, 115(6), 986-1019. Using consumer credit data, this study presents the reaction of consumer spending and debt to tax rebates. The authors apply new credit accounts panel data set to analyze how consumers respond to tax rebates with respect to the 2001 federal income tax rebates. A monthly response of credit card payments, debt, spending, and unique exploitations have been estimated including randomized rebate disbursement timing. It is found that consumers initially saved some of their rebates due to increased credit card payments, thus, paying down their debts. However, their spending increased afterward, a phenomenon that was contrary to the model of canonical permanent-income. It is further revealed that spending among liquidity constrained consumers increased, whereas there was a debt decline among majority of unconstrained consumers. These findings suggest that there are important dynamics that influence consumers response to lumpy increases income such as tax rebates and liquidity mechanisms.
• Better Lending and Better Clients: Credit Bureau Impact on Microfinance, McIntosh, C., Sadoulet, E., & de Janvry, A. (2006). BASIS, Madison, WI Brief, 45. This paper provides a comprehensive review of the role of credit bureau on microfinance. The authors reveal that, based on administrative records, credit bureaus has large impact on microfinance and loan performance. Before the introduction of bureau, individual and group loans in arrears was not stable and group loans were the ones dominating the lending market. However, after credit bureau information became common among credit agents, the number of individual loans with at least one lay payment reduced from 67.2% to 52.8%.