Chandra Shekhar Ghosh’s argument that microcredit is “business credit, not personal loan” goes to the heart of a difficult transition in Indian finance: how to extend formal credit to people whose livelihoods are too informal, too small or too new to be assessed through conventional banking models. His account of Bandhan’s evolution shows that financial inclusion is not simply a question of putting lending on an app. It is also about understanding how households earn, save, borrow and build businesses.
Ghosh, who founded Bandhan and served as its managing director and chief executive before becoming group chair, said the organisation’s move from a microfinance institution to a universal bank was necessary to expand its ability to serve low-income customers. In his account, the banking structure provides a risk-mitigation framework and allows customers to access products beyond small loans, including affordable housing finance.
That progression matters because the economic lives of borrowers do not remain fixed. Ghosh described a customer who had been selling tea on the street a decade ago and now owns a restaurant employing five people. Her husband, previously a taxi driver, works under her management. The example illustrates the central institutional challenge in inclusive finance: a customer may begin with a very small credit requirement but later need more sophisticated products as her enterprise grows.
A lender that treats such a borrower as permanently dependent on microcredit can limit her economic mobility. A lender that advances larger sums without understanding the business can increase the risk of over-indebtedness. The balance between those two failures is where the quality of microfinance is ultimately tested.
## The limits of a credit-score-only model
Ghosh’s criticism of purely digital underwriting is built around the difference between a personal loan and a business loan. In his view, a lender should assess how the money will be used, whether the borrower has the capacity to repay and how the business is performing. That requires physical visits before and after lending, as well as continuing group meetings.
This approach is labour-intensive, but it addresses a problem that conventional credit data cannot fully solve. According to Ghosh, about 60 per cent of Bandhan’s customers already have credit histories that can be checked through credit bureaus. The remaining 40 per cent are new to credit. For these borrowers, the absence of a substantial formal record does not necessarily mean the absence of repayment capacity. It means the lender must gather information differently.
The distinction is particularly important for women entrepreneurs operating in the informal economy. Their businesses may have regular cash flows without formal accounts, registered premises or extensive documentation. A credit bureau can show previous borrowing and repayment, but it may not reveal whether a borrower’s stall, shop, service business or household enterprise is expanding, declining or being used for a different purpose.
Ghosh therefore argues that digital tools should support, rather than replace, the relationship-based model. Loan paperwork and collections can move online. Voice-to-text systems could allow staff to record answers directly into loan applications. But the group meeting, he said, remains important because customers may stop attending once they can make payments digitally.
The issue is not resistance to technology. It is the risk of confusing transaction digitisation with full underwriting intelligence. A digital payment confirms that money moved. It does not, by itself, explain whether the loan financed productive activity, household consumption or the repayment of another loan.
## Why scale creates a governance problem
Bandhan’s institutional history also reveals the tension between scale and identity. Ghosh said the bank is no longer operationally managed by him and that it independently determines its policies, while the holding company sets strategic values and objectives. This separation is significant because a bank serving millions of customers cannot rely indefinitely on the personal systems of its founder.
At the same time, the processes that made microfinance effective can become difficult to maintain as the institution expands. Ghosh said no other lender’s staff meet customers 50 times a year. Such contact may improve information about borrowers, but it also carries a cost. A universal bank is expected to diversify, control risk, mobilise deposits and operate within regulatory requirements. It cannot remain a niche lender focused on one product or one customer segment.
The result is a structural trade-off. More secured lending and broader banking products can strengthen the balance sheet, but a shrinking share of microfinance may weaken the institution’s connection with its original inclusion mission. Ghosh said secured lending accounts for about 57 per cent of the bank’s portfolio, while also stressing that loans classified as unsecured under regulation are business loans rather than personal or consumption loans.
That distinction matters for public policy. Financial inclusion is often measured by access: whether a customer has a bank account, receives a loan or can make a digital payment. But access alone does not show whether credit is affordable, productive or appropriately sized. A system can expand formal borrowing while leaving borrowers vulnerable if lending decisions are based on incomplete information or if institutions compete aggressively for the same customers.
## The importance of starting small
Ghosh described a gradual lending model in which the initial loan is kept at a minimum relative to the customer’s capacity. Increases are limited to about 10 to 20 per cent a year. The purpose is to build repayment discipline and establish a record over time rather than immediately maximise the loan size.
This model directly addresses the over-indebtedness concern raised in the interview. Ghosh argued that people need formal credit because otherwise they may turn to moneylenders charging 5 per cent a month. But the existence of expensive informal credit does not justify unlimited formal lending. The formal lender must still decide how much risk it can take and how much the borrower can repay.
His position places responsibility on lenders rather than shifting it entirely to regulators. Regulation can establish limits and reporting requirements, but it cannot observe every borrower’s business or determine the correct loan amount for every household. That makes underwriting practice, field supervision and institutional incentives central to the health of the sector.
The same logic applies to savings. Ghosh recalled a recurring deposit model in which customers deposited ₹20 or ₹50 a week until the balance reached ₹5,000, after which it automatically became a fixed deposit. The example shows how small, regular transactions can help bring rural households into formal savings systems. It also suggests that inclusion is not only about extending credit; it is about creating products aligned with irregular and modest incomes.
Ghosh said the rural deposit market remains large and that the bank has not captured much of it. This points to another challenge for inclusive banking. Banks may find it easier to distribute loans than to mobilise stable deposits from customers with limited surplus income. Yet deposits are crucial because they provide a lower-cost and more durable source of funding than constant dependence on wholesale or high-cost capital.
## Technology should follow the borrower’s needs
The interview presents technology as useful when it reduces friction without eliminating judgment. Digital paperwork can lower administrative burdens. Voice-to-text tools can help field staff document customer responses. Credit-bureau data can improve assessment for existing borrowers. But technology cannot substitute for evidence about how a small business operates when that evidence is not already present in formal databases.
This is also why Ghosh was cautious about predicting the long-term role of technology in lending. A customer’s needs change as her business changes. The woman who began as a tea seller and later ran a restaurant would not require the same financial products at both stages. A static algorithm trained on an earlier phase of her economic life could either under-lend or over-lend.
The broader policy question is whether India’s digital financial infrastructure can recognise informal enterprise without forcing it into categories designed for larger, documented businesses. Digital systems may make formal finance faster, but speed is not the same as inclusion if the system excludes borrowers whose economic activity is real but poorly documented.
Ghosh’s comments also extend beyond banking. He identified affordable housing as a product that women entrepreneurs may need as their businesses grow. He said premium income in the group’s life insurance business had risen from about ₹335 crore to ₹1,000 crore in two years, while describing health insurance as a larger opportunity because it provides an immediate benefit to the whole family.
These comments reflect the way household finance and urban development intersect. Small-business income, housing security, insurance protection and access to savings are not separate concerns for low-income households. A shock in one area can destabilise the others. A medical expense can affect loan repayment; an insecure home can limit business expansion; weak savings can force a family back towards informal lenders.
## The larger urban question
Microcredit is often discussed as a rural or financial-sector issue, but its consequences are deeply connected to how towns and cities function. Women-led enterprises support household incomes, create local employment and provide goods and services in neighbourhood markets. As these businesses grow, their owners may need formal premises, housing finance, insurance, working capital and reliable digital and banking services.
The institutional question is therefore not whether relationship-based lending should remain unchanged. It is how banks can preserve the information and accountability created by relationships while using technology to improve efficiency. The answer will determine whether digital finance broadens access or simply accelerates the delivery of products to customers already visible to formal systems.
The evidence in Ghosh’s account confirms that Bandhan’s transformation has created a wider platform for financial products, but it also shows why scale cannot resolve the underlying underwriting problem by itself. The bank’s expanding product mix, the reported 57 per cent share of secured lending, the continuing role of field visits and the large population of customers new to credit all point to a model still negotiating between inclusion and institutional risk.
What remains uncertain is how these principles will be applied as the group expands into insurance, technology and other businesses, and as the bank independently sets its future policies. The developments to watch are not only the growth of digital lending or new product launches, but whether borrowers receive credit that matches the capacity and trajectory of their actual businesses.
