Paper: arXiv 2409.00296

Abstract

Credit scores are critical for allocating consumer debt in the United States, yet little evidence is available on their performance. We benchmark a widely used credit score against a machine learning model of consumer default and find significant misclassification of borrowers, especially those with low scores. Our model improves predictive accuracy for young, low-income, and minority groups due to its superior performance with low quality data, resulting in a gain in standing for these populations. Our findings suggest that improving credit scoring performance could lead to more equitable access to credit.

Complexity vs Empirical Score

  • Math Complexity: 6.5/10
  • Empirical Rigor: 7.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced machine learning (e.g., AUC metrics, interpretability techniques) and statistical modeling, qualifying as high math complexity, while its empirical rigor is high due to the use of a large, proprietary dataset (1M households), backtesting on historical data (2004-2015), and clear performance benchmarking against a real-world credit score.

Research Flowchart

  flowchart TD
  A["Research Goal: Benchmark Credit Score<br/>Performance & Equity"] --> B["Data: Consumer Credit Records<br/>(e.g., TransUnion)"]
  B --> C["Method: Compare vs.<br/>Machine Learning Model"]
  C --> D["Process: Evaluate Misclassification<br/>& Predictive Accuracy"]
  D --> E{"Key Findings"}
  E --> F["Performance Gap:<br/>Significant misclassification<br/>in low-score borrowers"]
  E --> G["Equity Gain:<br/>ML model improves standing<br/>for young/low-income/minority groups"]
  E --> H["Outcome:<br/>Better scoring leads to<br/>more equitable credit access"]

Cited References (263)

  1. Racial Disparity in Debt Collection Lawsuits: A Study of Three Metro Areas (2015)
  2. Modeling the Revolving Revolution: The Debt Collection Channel (2017)
  3. Prejudice and wages: an empirical assessment of Becker’s The Economics of Discrimination (2008)
  4. The cost of racial animus on a black candidate: Evidence using Google search data (2014)
  5. Wealth of two nations: The US racial wealth gap, 1860-2020 (2022)
  6. The financial restitution gap in consumer finance: insights from complaints filed with the CFPB (2021)
  7. A profile of 2013 mortgage borrowers: Statistics from the National Survey of Mortgage Originations (2017)
  8. Double/debiased machine learning for treatment and structural parameters (2018)
and 255 more
  1. Consumer-lending discrimination in the FinTech era (2021)
  2. Racial Disparities in the Auto Loan Market (2022)
  3. The wild bootstrap for few (treated) clusters (2018)
  4. Racial Segregation in Housing Markets and the Erosion of Black Wealth (2019)
  5. Consumer-lending discrimination in the FinTech era (2019)
  6. Are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination (2004)
  7. Discrimination in metropolitan housing markets: National results from Phase I HDS 2000 (2002)
  8. Do racial disparities in private transfers help explain the racial wealth gap? New evidence from longitudinal data (2014)
  9. Dealing with limited overlap in estimation of average treatment effects (2009)
  10. Racial disparity in unemployment (2011)
  11. Collecting Consumer Debt in America (2007)
  12. Broke but Not Bankrupt: Consumer Debt Collection in State Courts (2008)
  13. Debt Collection Agencies and the Supply of Consumer Credit (2015)
  14. How do consumers fare when dealing with debt collectors? Evidence from out-of-court settlements (2019)
  15. Delinquent debt in America (2014)
  16. Using publicly available information to proxy for unidentified race and ethnicity (2014)
  17. Payday loans and consumer financial health (2014)
  18. Do state regulations affect payday lender concentration? (2016)
  19. Predicting Consumer Default: A Deep Learning Approach (2019)
  20. Unobservable Selection and Coefficient Stability: Theory and Evidence (2019) — link
  21. XGBoost : A Scalable Tree Boosting System (2016) — link
  22. The Impact of Jury Race in Criminal Trials (2012)
  23. Testing for Racial Prejudice in the Parole Board Release Process: Theory and Evidence (2015)
  24. Racial Bias in Bail Decisions (2018)
  25. Assessing Racial Disparities in Parole Release (2015)
  26. Racial Disparity in Federal Criminal Sentences (2014)
  27. Racial Bias in Motor Vehicle Seraches: Theory and Evidence (2001)
  28. A new look at racial profiling: evidence from the Boston Police Department (2009)
  29. A Test of Racial Bias in Capital Sentencing (2014)
  30. Do Judges Vary in Their Treatment of Race? (2012)
  31. An Empirical Analysis of Racial Differences in Police Use of Force (2016)
  32. A framework for understanding sources of harm throughout the machine learning life cycle (2021)
  33. Recovering from biased data: Can fairness constraints improve accuracy? (2019) — link
  34. To split or not to split: The impact of disparate treatment in classification (2021)
  35. Risk-based pricing of interest rates for consumer loans (2006)
  36. The Arity of Disparity: Updating Disparate Impact for Modern Fair Lending (2024)
  37. Unpacking the black box: Regulating algorithmic decisions (2021) — link
  38. How costly is noise? Data and disparities in consumer credit (2021) — link
  39. An introduction to the frbny consumer credit panel (2010)
  40. Analysis of Differences between Consumer- and Creditor-Purchased Credit Scores (2012)
  41. Does Credit Scoring Produce a Disparate Impact? (2012)
  42. An overview of consumer data and credit reporting (2003)
  43. A century of consumer credit reporting in America (2005)
  44. Insolvency After the 2005 Bankruptcy Reform (2018) — link
  45. Predictably unequal? The effects of machine learning on credit markets (2022)
  46. Predicting Consumer Default: A Deep Learning Approach (2019) — link
  47. Credit growth and the financial crisis: A new narrative (2022)
  48. Creditworthy: A history of consumer surveillance and financial identity in America. (2017)
  49. A Human Algorithm: How Artificial Intelligence is Redefining who We are (2019)
  50. Predictably unequal? the effects of machine learning on credit markets (2018)
  51. How to Incorporate Monotonicity in Deep Networks While Preserving Flexibility? (2020)
  52. Racial disparities in debt collection (2019) — link
  53. Deep lattice networks and partial monotonic functions (2017)
  54. Credit card redlining (2011)
  55. Racial disparities in access to mortgage credit: Does governance matter? (2011)
  56. The foreclosure crisis and racial disparities in access to mortgage credit 2004-2009 (2011)
  57. Ethnic diversity grows, neighborhood integration lags (2000)
  58. Report to the congress on credit scoring and its effects on the availability and affordability of credit (2007)
  59. Use of indirect measures of race/ethnicity and socioeconomic status in managed care settings to identify disparities in cardiovascular and diabetes care quality (2005)
  60. Using the Census Bureau’s surname list to improve estimates of race/ethnicity and associated disparities (2009)
  61. Access to credit by small businesses: How relevant are race, ethnicity, and gender? (2012)
  62. A quantitative theory of information and unsecured credit (2012)
  63. Analysis of Differences between Consumer-and Creditor-Purchased Credit Scores (2012)
  64. Consumer bankruptcy: A fresh start (2007)
  65. Machine learning methods that economists should know about (2019)
  66. Machine learning: an applied econometric approach (2017)
  67. The roles of alternative data and machine learning in fintech lending: evidence from the LendingClub consumer platform (2018)
  68. Alternative Data and the Unbanked (2017)
  69. Regulating consumer financial products: Evidence from credit cards (2014)
  70. Consumer behavior and the stickiness of credit-card interest rates (1995)
  71. The failure of competition in the credit card market (1991)
  72. Do Banks Pass Through Credit Expansions to Consumers Who Want to Borrow? (2015) — link
  73. Predicting mortgage default using convolutional neural networks (2018)
  74. A novel image classification method with CNN-XGBoost model (2017)
  75. Hybrid Gradient Boosting Trees and Neural Networks for Forecasting Operating Room Data (2018) — link
  76. Employer Credit Checks: Poverty Traps versus Matching Efficiency (2018) — link
  77. A theory of credit scoring and competitive pricing of default risk (2011)
  78. A quantitative theory of unsecured consumer credit with risk of default (2007)
  79. Credit Growth and the Financial Crisis: A New Narrative (2017) — link
  80. The Mortgage Rate Conundrum (2017) — link
  81. The Role of Non-Owner-Occupied Homes in the Current Housing and Foreclosure Cycle (2010)
  82. What Have They Been Thinking? Home Buyer Behavior in Hot and Cold Markets (2012) — link
  83. A New Look at the U.S. Foreclosure Crisis: Panel Data Evidence of Prime and Subprime Borrowers from 1997 to 2012 (2015) — link
  84. The Housing Boom and Bust: Model Meets Evidence (2017) — link
  85. Regional Redistribution through the US Mortgage Market (2016) — link
  86. Why is the US unemployment rate so much lower? (1998)
  87. Cheap credit, collateral and the boom-bust cycle (2012)
  88. The Financial Crisis at the Kitchen Table: Trends in Household Debt and Credit (2014)
  89. The ins and outs of mortgage debt during the housing boom and bust (2015) — link
  90. Interest rates and equity extraction during the housing boom (2016)
  91. Renters: The Innocent Victims of the Foreclosure Mortgage Crisis (2009)
  92. Recourse and residential mortgage default: evidence from US states (2011)
  93. Macroeconomic effects of bankruptcy and foreclosure policies (2016)
  94. House Prices and Consumer Spending (2015) — link
  95. Quantitative Modeling of the Financial Crisis A Simple Model of Subprime Borrowers and Credit Growth (2016)
  96. Liquidity constraints in the US housing market (2017)
  97. The Subprime Mortgage Crisis (2016)
  98. Loan Product Steering in Mortgage Markets (2016) — link
  99. Real estate investors, the leverage cycle, and the housing market crisis (2011)
  100. Credit Cycles (1997)
  101. The Great Recession: Lessons from Microeconomic Data (2010)
  102. Consumption, house prices, and collateral constraints: a structural econometric analysis (2004)
  103. Household Debt and Defaults from 2000 to 2010: Facts from Credit Bureau Data (2015)
  104. The Consequences of Mortgage Credit Expansion: Evidence from the U.S. Mortgage Default Crisis (2009)
  105. HOUSEHOLD DEBT AND DEFAULTS FROM 2000 TO 2010: FACTS FROM CREDIT BUREAU DATA (2015)
  106. Optimal personal bankruptcy design under moral hazard (2010)
  107. A boosted decision tree approach using Bayesian hyper-parameter optimization for credit scoring (2017)
  108. Interpretable Machine Learning (2019)
  109. The Regulation of Non-Judicial Debt Collection and the Consumer’s Choice among Repayment, Bankruptcy and Informal Bankruptcy (2009)
  110. Subprime foreclosures and the 2005 bankruptcy reform (2012)
  111. Staff Reports (2007)
  112. Saving up for Bankruptcy (2009)
  113. An anatomy of us Personal bankruptcy under chapter 13 (2014)
  114. The homeownership experience of households in bankruptcy (2011)
  115. Saving Your Home in Bankruptcy (2008)
  116. What Do We Know About Chapter 13 Personal Bankruptcy Filings? (2007)
  117. Liquidity constraints and consumer bankruptcy: Evidence from tax rebates (2012)
  118. Consumer bankruptcy: A fresh start (2007)
  119. Accounting for the Rise in Consumer Bankruptcies (2010)
  120. The Consumer Bankruptcy Fee Study: Final Report (2012)
  121. Welfare implications of the bankruptcy reform act of 1999 (2002)
  122. The Impact of the Bankruptcy Reform Act of 1978 on Consumer Bankruptcy (1993)
  123. The household bankruptcy decision (2002)
  124. An empirical analysis of personal bankruptcy and delinquency (2002)
  125. Consumer debtors ten years later: A financial comparison of consumer bankrupts 1981-1991 (1994)
  126. Less stigma or more financial distress: An empirical analysis of the extraordinary increase in bankruptcy filings (2006)
  127. The Costs of BAPCPA: Report of the Pilot Study of Consumer Bankruptcy Cases (2010)
  128. The Affordability Paradox: How Consumer Bankruptcy’s Greatest Weakness May Account for Its Surprising Success (2010)
  129. Bankruptcy reform and credit cards (2007)
  130. Abuse or protection: Economics of bankruptcy reform under BAPCPA (2006)
  131. The Effect of BAPCPA on Credit Card Industry Profits and Prices (2009)
  132. Informal bankruptcy (2004)
  133. A comparison of neural networks and linear scoring models in the credit union environment (1996)
  134. Statistical classification methods in consumer credit scoring: a review (1997)
  135. Benchmarking state-of-the-art classification algorithms for credit scoring (2003)
  136. Random forests (2001)
  137. Risk and risk management in the credit card industry (2016)
  138. A decision-theoretic generalization of on-line learning and an application to boosting (1997)
  139. Two-level classifier ensembles for credit risk assessment (2012)
  140. Deep learning in neural networks: An overview (2015)
  141. Algorithms for hyper-parameter optimization (2011)
  142. Neural network credit scoring models (2000)
  143. Deep learning (2016)
  144. Empirical evaluation of gated recurrent neural networks on sequence modeling (2014) — link
  145. The Handbook of Brain Theory and Neural Networks (1998) — link
  146. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling (2018) — link
  147. Backpropagation applied to handwritten zip code recognition (1989)
  148. Phoneme recognition using time-delay neural networks (1990)
  149. Fully convolutional networks for semantic segmentation (2015)
  150. Deep Learning for Mortgage Risk (2018) — link
  151. Multilayer feedforward networks are universal approximators (1989) — link
  152. Approximation capabilities of multilayer feedforward networks (1991) — link
  153. Scaling learning algorithms towards AI (2007)
  154. On the Number of Linear Regions of Deep Neural Networks (2014) — link
  155. Dropout: a simple way to prevent neural networks from overfitting (2014)
  156. A Unified Approach to Interpreting Model Predictions (2017) — link
  157. Adam: A method for stochastic optimization (2014) — link
  158. The NumPy array: a structure for efficient numerical computation (2011)
  159. Data structures for statistical computing in python (2010)
  160. Matplotlib: A 2D graphics environment (2007)
  161. Keras: Deep learning library for theano and tensorflow (2015)
  162. Tensorflow: a system for large-scale machine learning. (2016)
  163. Scikit-learn: Machine learning in Python (2011)
  164. Batch normalization: Accelerating deep network training by reducing internal covariate shift (2015) — link
  165. Measures of diversity in classifier ensembles and their relationship with the ensemble accuracy (2003)
  166. Proxy Fairness (2018)
  167. Man is to computer programmer as woman is to homemaker? debiasing word embeddings (2016)
  168. Two-player games for efficient non-convex constrained optimization (2018) — link
  169. Discrimination aware decision tree learning (2010)
  170. Fairness-aware classifier with prejudice remover regularizer (2012)
  171. CFPB and DOJ Order Ally to Pay $80 Million to Consumers Harmed by Discriminatory Auto Loan Pricing. (2013)
  172. Unsafe at Any Bureaucracy: CFPB Junk Science and Indirect Auto Lending. (2015)
  173. Unsafe at Any Bureaucracy, Part II: How the Bureau of Consumer Financial Protection Removed Anti-fraud Safeguards to Achieve Political Goals. (2016)
  174. Unsafe at Any Bureaucracy, Part III: The CFPB’s Vitiated Legal Case Against Auto-Lenders. (2017)
  175. Statement of the Bureau of Consumer Financial Protection on enactment of S.J. Res. 57. (2018)
  176. Fair Lending: Implications for the Indirect Auto Finance Market. (2014)
  177. Fairness under unawareness: Assessing disparity when protected class is unobserved (2019)
  178. A methodology for direct and indirect discrimination prevention in data mining (2012)
  179. Strategic mortgage default: The effect of neighborhood factors (2015)
  180. Credit Invisibles and the Unscored (2016) — link
  181. Mortgage-Backed Securities and the Financial Crisis of 2008: A Post Mortem (2016)
  182. Anatomy of the Beginning of the Housing Boom: U.S. Neighborhoods and Metropolitan Areas, 1993-2009 (2011) — link
  183. Why Did So Many People Make So Many Ex Post Bad Decisions? The Causes of the Foreclosure Crisis (2012) — link
  184. Negative equity and foreclosure: Theory and evidence (2008)
  185. Misinformed Speculators and Mispricing in the Housing Market (2014) — link
  186. Housing Demand, Regional House Prices and Consumption (2017)
  187. Housing Booms, Manufacturing Decline, and Labor Market Outcomes (2013) — link
  188. Endogenous gentrification and housing price dynamics (2013)
  189. Education and unemployment (1991)
  190. Foreclosures, house prices, and the real economy (2015)
  191. Credit-induced boom and bust (2017)
  192. What explains the 2007–2009 drop in employment? (2014)
  193. The geographic determinants of housing supply (2010)
  194. Deciphering the liquidity and credit crunch 2007–2008 (2009)
  195. Houses as ATMs? Mortgage Refinancing and Macroeconomic Uncertainty (2013) — link
  196. Dynamics of Housing Debt in the Recent Boom and Great Recession (2017) — link
  197. Mortgage financing in the housing boom and bust (2012)
  198. Mortgage rates, household balance sheets, and the real economy (2014)
  199. Leverage and the Foreclosure Crisis (2015) — link
  200. Debt constraints and employment (2016)
  201. Mortgage Debt, Consumption, and Illiquid Housing Markets in the Great Recession (2016)
  202. Household Leverage and the Recession (2011)
  203. Cross-sectional patterns of mortgage debt during the housing boom: evidence and implications (2016)
  204. Owner occupancy fraud and mortgage performance (2015)
  205. Household debt and defaults from 2000 to 2010: The credit supply view (2016)
  206. Non-durable Consumption and Housing Net Worth in the Great Recession: Evidence from Easily Accessible Data (2016) — link
  207. Loan Originations and Defaults in the Mortgage Crisis: The Role of the Middle Class (2015)
  208. Home is where the equity is: Mortgage refinancing and household consumption (2004)
  209. Household Balance Sheets, Consumption, and the Economic Slump (2013)
  210. Credit crises, precautionary savings, and the liquidity trap (2011)
  211. House Prices, Home Equity–Based Borrowing, and the US Household Leverage Crisis (2011)
  212. Factors affecting an individual’s bankruptcy filing decision (1999)
  213. Optimal contracts and competitive markets with costly state verification (1979)
  214. Debtor Protections and the Great Recession (2014)
  215. Informal bankruptcy (2004)
  216. Credit access after consumer bankruptcy filing: new evidence (2014)
  217. Household consumption and personal bankruptcy (2011)
  218. Residential housing and personal bankruptcy (2009)
  219. Household borrowing after personal bankruptcy (2011)
  220. Do Physicians’ Financial Incentives Affect Medical Treatment and Patient Health? (2014)
  221. Determinants of the consumer bankruptcy decision (1999)
  222. Consumer Spending and the Economic Stimulus Payments of 2008 (2013)
  223. Household Expenditure and the Income Tax Rebates of 2001 (2006)
  224. Do Liquidity Constraints and Interest Rates Matter for Consumer Behavior? Evidence from Credit Card Data (2002)
  225. Consumer Bankruptcy and Financial Health (2015)
  226. Debt Relief and Debtor Outcomes: Measuring the Effects of Consumer Bankruptcy Protection (2015)
  227. Did bankruptcy reform fail? An empirical study of consumer debtors (2008)
  228. Racial disparities in debt collection (2024)
  229. Bankruptcy by the Numbers: A Tale of Two Chapters: Financial Data (2002)
  230. Bankruptcy Reform: Dollar Costs Associated with the Bankruptcy Abuse Prevention and Consumer Protection Act Of 2005 (2008)
  231. Consumer credit-risk models via machine-learning algorithms (2010)
  232. Benchmarking state-of-the-art classification algorithms for credit scoring: An update of research (2015)
  233. Long short-term memory (1997)
  234. Parallel networks that learn to pronounce English text (1987)
  235. Deep learning in finance (2016) — link
  236. Deep Learning with Gated Recurrent Unit Networks for Financial Sequence Predictions (2018)
  237. Conditional time series forecasting with convolutional neural networks (2017) — link
  238. Random search for hyper-parameter optimization (2012)
  239. Index for rating diagnostic tests (1950)
  240. The Economics of Discrimination (2010)
  241. Stuck in place: Urban neighborhoods and the end of progress toward racial equality (2013)
  242. The Economics of Debt Collection: Enforcement of Consumer Credit Contract (2015)
  243. A new method for estimating race/ethnicity and associated disparities where administrative records lack self-reported race/ethnicity (2008)
  244. Using the Census Bureau’s surname list to improve estimates of race/ethnicity and associated disparities (2009)
  245. An Alternative Test of Racial Prejudice in Motor Vehicle Searches: Theory and Evidence (2006)
  246. Corporate default forecasting with machine learning (2020)
  247. Forecasting Loan Default in Europe with Machine Learning (2020)
  248. FinTech Credit: A Critical Review of Empirical Research Literature (2020)
  249. Investor Emotions and Earnings Announcements (2020)
  250. Fairness through awareness (2012)
  251. Equality of opportunity in supervised learning (2016)
  252. When worlds collide: integrating different counterfactual assumptions in fairness (2017)
  253. Counterfactual fairness (2017)
  254. Fairness constraints: Mechanisms for fair classification (2017)
  255. Optimization with Non-Differentiable Constraints with Applications to Fairness, Recall, Churn, and Other Goals. (2019)

0 of 263 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.