The inclusion paradox
Across Latin America's six largest economies, financial institutions face a paradox that has defined the industry for decades. The market opportunity is enormous. Chile, Peru, Argentina, Brazil, Mexico, and Colombia together have over 400 million adults. A significant portion of them — estimates vary by market but the regional average sits between 40% and 60% — are underbanked or entirely outside the formal financial system. They earn income, spend money, save informally, and manage financial risk in ways that demonstrate real economic capability. They are not poor credit risks. They are invisible ones.
The paradox is this: the customers financial institutions most need to reach to grow are precisely the customers that traditional credit infrastructure was not built to see. Bureau data captures formal credit history. For customers who have never had a formal loan, it records nothing. A blank bureau file is not evidence of bad credit risk. It is evidence of exclusion from the formal system. But traditional models treat it as the former, perpetuating the exclusion it reflects.
What behavioral data actually is
The term "alternative data" is broad enough to obscure more than it reveals. Behavioral data refers specifically to signals generated by how people interact with digital systems — signals that are increasingly available across LATAM as smartphone penetration rises and digital financial activity expands.
Digital payment behavior. How frequently does a person make digital payments? Through which channels? How consistent is the pattern over time? Irregular payment behavior — not necessarily missed payments, but erratic timing and amounts — is predictive of financial stress in ways that bureau data captures only after the fact.
App usage patterns. The apps a person uses, how frequently they use them, and how those patterns change over time contain information about financial behavior. A person who consistently uses budgeting apps, savings apps, and digital banking tools exhibits a financial management orientation that is meaningfully correlated with creditworthiness.
Transactional patterns. For customers who have any history of digital transactions — even informal ones through mobile money or payment apps — the pattern of those transactions is highly informative. Income regularity, spending stability, the ratio of incoming to outgoing transactions — these patterns capture financial behavior in real time, not retrospectively.
Telco data. In markets where mobile penetration is high and formal banking penetration is low — which describes Peru, Colombia, and significant portions of Brazil and Mexico — telecommunications data provides signals that are uniquely valuable. Airtime top-up frequency, data usage patterns, and payment regularity for mobile services are among the most predictive alternative signals available for thin-file populations.
Why behavioral data works where bureau data doesn't
Bureau data is backward-looking by design. It records what happened. For thin-file customers, the retrospective record is empty — not because their behavior is unpredictable, but because they have not participated in the formal system that generates the record.
Behavioral data captures present and recent behavior. It answers questions the bureau cannot: Is this person's income stable right now? Are their payment patterns consistent? Are there signals of financial stress emerging in their digital behavior? These questions are more relevant to predicting future credit performance than historical bureau records — especially in markets like Argentina where macroeconomic volatility means that behavior from three years ago may be a poor predictor of behavior today.
The result is that models trained on behavioral data can differentiate creditworthiness within thin-file populations in ways that bureau-based models simply cannot. They can identify the creditworthy customers that bureau-based models reject by default, without increasing default risk.
The inclusion impact
The business case for behavioral data is straightforward: it expands the addressable market without expanding risk. But its social impact deserves attention in its own right.
In Peru, behavioral data enables lenders to extend credit to populations that have been systematically excluded from formal financial services. In Mexico, behavioral data is the primary mechanism by which fintechs have expanded credit access to self-employed workers, gig economy participants, and informal traders. In Colombia and Brazil, behavioral data is enabling micro-lending and BNPL products that reach customers in smaller cities and rural areas.
This is not philanthropy. It is a commercial opportunity that happens to align with social good. The 400 million adults across these six markets who are underbanked represent an enormous and largely untapped credit market. The institutions that build the capability to reach them responsibly will capture a disproportionate share of the next decade of growth in Latin American financial services.
The responsible use imperative
Behavioral data is powerful. That power requires responsibility. The most important principle is validation: a behavioral signal that correlates with creditworthiness in aggregate may be a proxy for a protected characteristic at the individual level. Institutions using behavioral data need to validate their models not just for predictive accuracy but for fairness across segments.
The second principle is transparency. Customers whose credit applications are evaluated using behavioral data have a right to understand, in general terms, what signals were used and how they influenced the decision. Regulators across LATAM — the CMF, SBS, and Superintendencia Financiera — have all signaled growing interest in the explainability of alternative data models.
The signal was always there
The customers that behavioral data helps financial institutions reach were never actually invisible. They were always there — earning, spending, managing money, demonstrating financial behavior in their daily lives. What was missing was not the signal. It was the infrastructure to capture it, process it, and translate it into credit decisions.
That infrastructure now exists. Across Chile, Peru, Argentina, Brazil, Mexico, and Colombia, the combination of rising smartphone penetration, expanding digital payment ecosystems, and maturing alternative data networks has created conditions where behavioral data can be collected, validated, and used at scale. The institutions that use it well will expand their markets, improve their models, and advance financial inclusion in ways that create durable value.