For most of banking history, access to credit depended less on a person’s potential than on a bank’s ability to understand them. The challenge was never a shortage of entrepreneurs, farmers, traders, or skilled workers. It was a shortage of information.
Banks could only lend confidently to customers whose financial lives were visible through documents; salary slips, tax returns, audited accounts, property records, collateral documents, and established credit histories. Those who possessed documentation gained access to capital. Those who did not were often excluded, regardless of their ability to repay.
This model was understandable in an era when information was scarce. Banks could only make decisions based on what they could verify. Documentation became a proxy for trust, and collateral became a proxy for certainty. Over time, entire lending systems were built around these assumptions.
Yet the limitations of this approach have become increasingly apparent, particularly in emerging economies where large segments of economic activity occur outside formal documentation frameworks.
Data is becoming the new collateral in the digital age
Pakistan offers a compelling illustration of this challenge. Millions of small businesses, farmers, merchants, freelancers, and micro-entrepreneurs participate actively in the economy while remaining only partially visible to conventional financial institutions.
Small and medium enterprises contribute roughly 40 per cent of GDP and account for a substantial share of employment and exports, yet access to formal credit remains disproportionately low relative to their economic significance. The issue is not a lack of productive activity. The issue is that traditional underwriting models were designed for documented economies, while much of the modern economy remains only partially documented.
What is now changing is not the demand for credit, but the availability of information.
The digital economy generates an extraordinary volume of data. Mobile phone usage, utility payments, transaction histories, merchant relationships, digital payment patterns, savings behaviour, supply-chain activity, and countless other signals collectively create a far richer picture of economic activity than was previously possible.
Many individuals who appear invisible through conventional financial records are, in reality, generating a continuous stream of information about their financial behaviour. The challenge is no longer obtaining data. It is learning how to interpret it.
This shift is giving rise to one of the most important developments in modern banking: the transition from document-based underwriting to data-based underwriting. Increasingly, financial institutions are discovering that the absence of traditional collateral does not necessarily imply the absence of creditworthiness. The critical question is no longer whether a customer possesses a specific set of documents. The question is whether enough information exists to assess risk accurately.
In practical terms, this has led to the emergence of entirely new approaches to lending. Alternative-data models can identify patterns that correlate with repayment behaviour. Income-proxy methodologies can estimate economic capacity when formal income records are unavailable.
Psychometric assessments can measure characteristics associated with financial discipline and business resilience. Machine-learning models can process thousands of variables simultaneously, identifying relationships that would be impossible to detect through conventional methods. Together, these tools are transforming underwriting from a process of document verification into a process of information analysis.
The significance of this transformation extends far beyond banking. Throughout history, economic opportunity has often been constrained by what economists describe as information asymmetry, the inability of one party to accurately assess the quality or reliability of another.
Financial markets function most efficiently when information flows freely and accurately. When information is incomplete, capital becomes cautious. Credit becomes expensive. Investment declines. Productive individuals and businesses remain underserved, not because they lack potential, but because they cannot be evaluated effectively.
Seen through this lens, many of the world’s financial inclusion challenges are fundamentally information challenges. The traditional response has often been to lower barriers or introduce subsidised programmes. While these interventions have value, they do not address the underlying problem.
Sustainable inclusion requires institutions to become better at measuring risk, not less disciplined in managing it. The future of inclusion therefore lies not in relaxing underwriting standards but in improving underwriting intelligence.
This is where technology is beginning to reshape the boundaries of what banks can do. Many financial institutions have spent the past several years developing lending architectures that combine digital identity verification, government record integrations, psychometric scoring, alternative data inputs, AI-enabled analytics, and automated decision engines to assess borrowers who have historically been outside conventional credit frameworks. The objective is not to replace prudent banking. It is to expand the universe of customers who can be assessed prudently.
The implications are profound. Every improvement in risk assessment expands the pool of individuals and businesses that can participate in the formal financial system. A farmer with limited documentation but a strong behavioural profile becomes financeable. A small retailer with no audited accounts but a consistent transaction history becomes visible. A first-time borrower with no prior credit record but demonstrable economic activity becomes assessable. In each case, the constraint is not removed by lowering standards. It is removed by increasing understanding.
Yet this transition also introduces important responsibilities. As data assumes a greater role in financial decision-making, questions of governance become increasingly important. How should alternative data be collected and used? How should automated decisions be explained? How can institutions ensure that algorithms enhance fairness rather than replicate existing biases? Trust cannot simply be transferred from loan officers to machines. It must be embedded within transparent systems, strong governance frameworks, and accountable institutions.
The future of banking will ultimately be shaped by the institutions that solve the information problem most effectively. For centuries, documents served as the primary language through which banks understood risk. Increasingly, that language is being supplemented by data.
The most important asset in banking has always been trust. What is changing is how that trust is measured. In the decades ahead, the institutions that thrive will be the institutions that become best at transforming information into opportunity.
In that sense, the future of finance may be that data becomes the new collateral.
The writer is the Chief Information Officer at the Bank of Punjab
Published in Dawn, The Business and Finance Weekly, July 6th, 2026