India’s fintech boom is loaded with moral hazard and could trigger a crisis at the slightest shock

Unsecured retail loans account for 53.1% of total retail slippages at scheduled commercial banks, even as their retail gross non-performing assets (GNPA) ratio sits at a comforting 1.8%. This is finance’s oldest optical illusion: fixate on the average while the distribution quietly sharpens its knives. It always works—right up until it doesn’t.

Crises do not arrive waving banners marked “excess credit” but disguised as innovation. As Adair Turner warns in Between Debt and the Devil, credit can grow to socially useless or dangerous levels long before it gets inflationary. Stable prices can end up as camouflage, as they often do. India’s household credit boom fits this description.

Unsecured retail credit is growing at high double-digit rates, comfortably ahead of nominal GDP. Personal loans, credit cards and gold loans are doing the heavy lifting, while credit that builds productive capacity jogs behind. Banks have retreated to the retail foxhole not by choice, but because corporate credit demand is anaemic.

In a world of thin margins, the ‘high-yield’ siren song of a 24% annual rate personal loan is irresistible to a balance sheet starved of industrial capital expenditure. This does not lead to capital formation. It is economic output in rented clothing. The bill is payable by households.

None of this is new. In 1920s America, instalment credit allowed households to buy cars long before incomes could support them. Consumption surged. Growth looked unstoppable. When wages failed to cooperate, debt did what it does: it quit pretending. The collapse was not psychological but mechanical.

Fintech has turbocharged India’s version of that story. Its share of non-bank consumer loans has jumped to 8.9% from 7.3% barely a year earlier. More than 70% of fintech loans are unsecured. Over half go to borrowers under 35.

Such credit is instant, modular and embedded. Borrowing is more of an app feature than a considered decision. And for platforms paid on loan origination, repayment is someone else’s problem. The industry hails the ‘Account Aggregator’ framework as the answer to information asymmetry.

Yet, even the best data is a rear-view mirror. An algorithm can predict a borrower’s intent based on past UPI swipes, but it cannot forecast the borrower market’s capacity to pay if underlying drivers like the gig economy begin to cool off.

Household debt has grown at roughly twice the pace of nominal GDP, reaching about $1.5 trillion. This is leverage, sliced thin and spread wide, reassuring everyone that no single default matters—until millions of them do. Risk has been atomized, which history suggests does not disappear, but synchronises.

For young Indians, credit bridges the gap between aspiration and income. Many iPhones, vacations, consumer durables and lifestyle spending are financed upfront. Repayment is deferred.

As financial historian David Graeber noted, once debt is moralized, default becomes a personal failure rather than a design flaw. That is when trouble begins. In Korea before 1997, high leverage was tolerated because growth was strong and defaults were low. When capital flows reversed, balance sheets that looked resilient proved brittle.

Digital lending removes friction but preserves consequences. Algorithms trained only on good times allocate capital with exquisite confidence and no memory. Unsecured credit does not fail gracefully. It suddenly drops off a cliff—like America’s subprime mezzanine tranches did. We in India are witnessing a ‘democratization of distress.’ By the time the GNPA ratio moves from 1.8% to 4%, the liquidity bridge will have already collapsed for non-bank financial companies that feed fintech firms.

Unsecured credit has thin recovery values and a nasty habit of herd defaults. A mild shock like a hiring slowdown, higher interest rates and marginally looser underwriting need not be dramatic. It merely needs to arrive.

Our regulatory framework treats unsecured retail credit as individually risky but systemically trivial. Risk weights were tightened, partly loosened and broadly remain indulgent. Provisioning is backward-looking. Capital buffers are calibrated for bank-level solvency, not for correlated household failure. Fintech lenders are lightly regulated and banks chase yield.

This is macroprudential policy stuck in a micro-prudential mindset. History suggests the cycle will not wait. Debt is a claim on the future. Credit is useful only up to a point after which it stops feeding growth and starts feeding on it.

Regulatory comfort in “small ticket sizes” ignores a basic law: a million pebbles falling together is still an avalanche. Should the brakes fail, the impact won’t be cushioned by ‘averages.’ It will be felt all at once by a generation that spent its future before it earned it.

The author is a former executive director, Nomura and currently a guest faculty at various India B-schools.