In a groundbreaking study that promises to reshape the financial landscape for billions of unbanked individuals worldwide, researchers from the University of Notre Dame and their collaborators have unveiled how alternative retail transaction data can be harnessed to construct reliable credit scores for those without formal credit histories. This innovative approach offers a powerful tool to surmount the pervasive challenge that has long plagued traditional financial institutions: how to assess creditworthiness in the absence of a conventional credit record.
The World Bank estimates that approximately 1.4 billion people globally remain unbanked, unable or unwilling to engage with financial services largely because they lack the credit history required by conventional lenders. This exclusionary cycle historically entrenches economic inequality, as individuals without documented credit histories find themselves locked out of loans, credit cards, and other financial instruments essential for fostering economic growth and personal financial security.
The new research, detailed in a forthcoming article in the Journal of Marketing Research, titled “Who Benefits from Alternative Data for Credit Scoring? Evidence from Peru,” elucidates how rich data derived from everyday retail shopping behaviors can be transformed into predictive credit metrics. Spearheaded by Joonhyuk Yang, assistant professor of marketing at Notre Dame’s Mendoza College of Business, alongside Jung Youn Lee from Rice University and Eric T. Anderson from Northwestern University, the study demonstrates that retail transaction data substantially elevates the ability of lenders to extend credit to “no history” applicants, increasing approval rates from a modest 16 percent to a striking 48 percent under certain conditions.
The crux of the challenge lies in the classic lending paradox: a credit history is needed to receive credit, yet credit is essential to build a credit history. The researchers propose that everyday consumer purchasing patterns — the cadence of grocery store visits, responsiveness to promotions, product choices, and returns behavior — encapsulate nuanced behavioral signals indicative of financial responsibility. These markers effectively serve as alternative data points that traditional scoring models overlook.
This innovative methodology builds on the team’s prior work, which focused on individuals with existing credit files. Their earlier research revealed that grocery shopping choices, such as selecting healthier foods and demonstrating disciplined shopping routines like adhering to budgets or consistently visiting stores on specific days, strongly predict timely credit card payments. Conversely, purchases of items such as cigarettes and ready-to-eat processed foods correlated with higher default risk. This behavioral insight provides a more granular understanding of consumer financial health beyond simple repayment records.
However, the new study expands the scope significantly by analyzing a larger cohort of over 45,000 consumers in Peru, including those with no formal credit history. The team utilized customer loyalty data across multiple retail sectors, tracking consumer responses to promotions, product return frequencies, and purchase types over a two-year timeframe. When combined with conventional credit-related data — including utility bill payment histories and national credit registry records — the integrated model enabled the creation of alternative credit scores applicable even to those traditionally excluded from credit markets.
Technically, merging heterogeneous data streams presented challenges related to data privacy, reliability, and the predictive validity of novel inputs. Yet, through sophisticated econometric modeling and machine learning techniques, the research team skillfully demonstrated that the incorporation of retail transaction data enhances credit scoring models’ predictive power without compromising risk thresholds. Importantly, the inclusion of such data allowed financial institutions to maintain or even lower default risks while expanding approval rates for previously underserved populations.
The implications are profound: whereas conventional credit scores remain the backbone for lending decisions among individuals with established credit histories, they offer little room for expansion in reaching the unbanked. Retail data, by contrast, acts as a catalyst to break the credit invisibility barrier, providing lenders with a richer, more contextual portrait of potential borrowers’ financial behaviors. As Yang notes, “Retail data barely moves the needle for people who already have credit scores, but it’s a game changer for those who don’t. That’s where inclusion really happens.”
Moreover, the study’s simulations illustrate that decision-makers can calibrate lending policies to prioritize risk mitigation or accessibility goals, with alternative data offering superior flexibility. Governments, non-governmental organizations, and private fintech companies aiming to promulgate financial inclusion stand to benefit from adopting such innovative scoring mechanisms, which promise to democratize credit access without increasing systemic risk.
Policymakers increasingly recognize the urgent need to close the global credit gap, echoing calls from financial inclusion advocates to deploy data-driven solutions that are both scalable and equitable. Hypothetically, in emerging markets where traditional financial infrastructure is underdeveloped, leveraging routine retail transactions — an ubiquitously available data source — can leapfrog conventional barriers, propelling millions into the formal credit economy.
This research marks a pivotal juncture in the evolving discourse around alternative credit scoring, signaling a departure from overly rigid models toward more holistic and behavioral-centric frameworks. As fintech firms race to innovate, the Notre Dame team’s work provides a robust empirical foundation and a scalable blueprint for practical implementation.
For the unbanked population, this development is more than technical progress; it is a potential lifeline. By recognizing and validating everyday financial discipline through purchase data, lenders can extend credit confidently and fairly to those the system has routinely overlooked. This opens new avenues for entrepreneurship, homeownership, education finance, and overall economic empowerment.
In closing, the integration of retail transaction data into credit evaluation systems exemplifies the power of interdisciplinary collaboration — merging marketing insights, data science, and economics — to solve entrenched societal problems. As this methodology gains traction, the era where billions remain invisible to credit systems may finally give way to a more inclusive and just financial ecosystem.
Subject of Research: Alternative Data in Credit Scoring for the Unbanked Population
Article Title: Who Benefits from Alternative Data for Credit Scoring? Evidence from Peru
News Publication Date: Not specified in the content (article publication date is 17-Jul-2025)
Web References: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4852032
References: Notre Dame study on grocery shopping and credit default, prior research published by the same team
Keywords: Alternative credit scoring, unbanked individuals, retail transaction data, financial inclusion, credit risk modeling, consumer behavior, credit approval rates, fintech, developing countries, data-driven lending
