Grocery purchasing patterns revealing clues about credit reliability
Financial institutions are exploring innovative ways to help those who are 'unbanked' or have limited access to resources in an emergency. A study led by Eric Anderson, a professor of marketing at the Kellogg School, suggests that grocery shopping behaviors can serve as an alternative data source for assessing the creditworthiness of individuals without traditional credit scores.
The study, which overlays data from a grocery store and a credit-card company, found that regular, consistent purchasing of essential goods like groceries indicates stable income and budgeting habits. Conversely, erratic or impulsive buying might signal financial stress or poor money management, potentially predicting a higher risk of defaulting on credit payments.
In the current tightened credit market, traditional credit models often exclude many consumers, particularly the underbanked or those without credit history. Lenders are increasingly turning to alternative data—including payment behaviors on Buy Now, Pay Later services and daily spending habits—to fill this gap.
By analyzing grocery shopping frequency, spending consistency, and payment methods, lenders can better gauge repayment capacity beyond traditional credit reports. This approach supports reaching underbanked populations who may lack formal credit history but demonstrate financial responsibility through everyday transactions.
The study used a well-known machine-learning model, XGBoost, to explore the relationship between grocery behaviors and other financial behaviors. Consistent grocery shopping behaviors, such as shopping at regular times, making repeat purchases, and steadily relying on promotions, tend to correlate with timely credit-card payments.
However, the usefulness of the grocery data diminished over time as the consumer began building a credit score and repayment history. Spending a significant amount on vinegar salad dressing, for instance, is the single greatest indicator of a non-defaulter. On the other hand, buying pre-processed foods like mortadella beef, energy drinks, and canned fish also indicates a higher risk of defaulting.
The study by Anderson and his colleagues found that the grocery data effectively filtered out defaulters among credit-card applicants without credit scores, leading to a 1.46 percent increase in per-person profits. This suggests that habits in one domain (grocery shopping) can predict habits in another domain (credit repayment).
Worldwide, over a billion people are "unbanked." This method supports tapping underserved markets and reducing information gaps, allowing for more efficient and inclusive lending decisions. As the study demonstrates, grocery shopping behaviors provide valuable alternative data that reflects an individual's financial habits and can predict their likelihood of repaying loans, making it a promising tool for financial institutions.
Financial institutions can use grocery shopping behaviors as alternative data to assess the creditworthiness of the 'unbanked' or those with limited access to resources, especially in an emergency. By analyzing factors like shopping frequency, spending consistency, and payment methods, lenders can better gauge the repayment capacity of these individuals, thereby supporting more inclusive lending decisions and tapping into underserved markets.
In the realm of business and technology, this innovative approach merges traditional credit models with machine learning, offering a novel solution for evaluating the financial habits of individuals who may lack formal credit history, ultimately predicting their likelihood of repaying loans.
