India’s Credit-Card Economy: Rewards, Aspiration and the Cost of Credit

India’s credit-card business is better understood as a combination of short-term credit, rewards, customer relationships and payment acceptance than as a contest for transaction count. Real-time payments can dominate everyday transfers while cards retain value in discretionary spending, financing and premium services. The sector’s profitability nevertheless depends on acquisition cost, funding, credit losses and the net cost of rewards.

This update preserves the original issuer-economics framework and moves its governing question to Indian consumer-credit policy. Unverified card-stock totals and contradictory FY 30 forecasts have been removed from the public current-data argument. It makes no issuer-specific profitability claim.

Why this belongs in India’s policy discussion

Consumer protection and profitable credit supply are linked. Transparent cost disclosure, accurate affordability assessment and timely correction of unauthorised transactions can sustain trust without treating every increase in digital payment volume as an increase in welfare. The policy trade-off is between easier access, borrower comprehension, lender incentives and the household’s capacity to repay.

Four indicators should stay separate: payment adoption, credit access, debt service and losses. A growing first indicator does not validate the other three. Issuer economics explain how the service is financed; they do not by themselves establish whether borrowing is affordable across income groups. A credible sector assessment needs household and vintage evidence before making that wider claim.

Payment volume is not the same as economic value

The 7 October RBI statement records UPI transaction-volume growth of 22.2% in Q 2 FY 2026–27. It does not imply equivalent growth in consumer credit or issuer profit. The source’s 118.6 million cards and ₹23.62 trillion FY 26 spending could not be independently reconciled to a freshly retrieved RBI card table in this refresh, so they are not presented as updated facts.

NPCI’s RuPay credit-card-on-UPI documentation confirms that a linked card can fund permitted merchant payments through UPI, with excluded transaction categories. The rail and the funding instrument can therefore coexist. A UPI transaction count, a card balance and a rupee of revolving debt answer different questions.

A payment rail moves money. A card can also provide a revolving balance, an interest-free period, rewards, dispute handling and access to associated services. Those functions can be bundled into other products over time, but the relevant competitive question is which economic component is being replaced.

Figure 01 · Conceptual framework

Who finances the rewards?

How the parts connect

1. ACustomer behaviourdrives incomeRevenue
2. BCustomer behaviourdrives redemptionReward expense
3. CCustomer behaviourdrives risk and servicingOther costs
4. DRevenueaddsIssuer contribution
5. EReward expensesubtractsIssuer contribution
6. FOther costssubtractsIssuer contribution

The issuer’s profit equation

Issuer revenue comes from interest, interchange, fees and related customer economics. Costs include funding, rewards, acquisition, operations, fraud and credit losses. The balance between transactors who repay in full and borrowers who revolve affects both revenue and risk.

An issuer concentrated in affluent transactors can generate strong spending and interchange with lower credit losses, but may earn less revolving interest and face high reward expectations. A mass-market issuer may earn more interest while carrying greater underwriting and collection risk. Neither model should be judged using a single industry-average margin.

Historical interchange and reward-cost ranges

These are source estimates with different portfolio and merchant mixes, not a universal fee schedule.

The comparison shows why gross interchange alone is misleading. A premium programme can generate more revenue per rupee spent while also paying more for rewards. Profitability depends on the net contribution after reward redemption, operating costs and credit performance, not the advertised earn rate.

Reward breakage is an economic assumption

Breakage is the proportion of earned rewards that is not ultimately redeemed. Higher breakage lowers the issuer’s realised reward cost relative to the headline promise. The briefing described historical breakage around 30–50%, with pressure as customers became better informed and redemption became easier.

That is a significant sensitivity. Suppose an illustrative programme promises benefits with a nominal cost of 2% of spending. If 40% is not redeemed, realised cost before other effects would be 1.2%; if only 20% is not redeemed, it becomes 1.6%. The 40-basis-point difference can be material relative to the source’s interchange ranges. This calculation is explanatory, not an observed programme result.

Issuers can respond through changes in eligibility, caps, redemption value, annual fees or targeted benefits. Each response can affect customer engagement. A programme that reduces cost aggressively may lose the spending behaviour on which its revenue model depends.

Funding and credit costs can overwhelm volume growth

A growing card book does not guarantee expanding profitability. The cost of funds, portfolio seasoning and credit losses determine how much of the additional revenue reaches earnings. Rapid acquisition can initially make performance look strong before newer vintages have passed through a full loss cycle.

The source argued that near-term net interest margins were biased lower over the following 12–18 months. That remains the original historical call. Issuer-level economics are excluded from this report.

Credit analysis should separate origination vintages, customer segments and product changes. A fall in reported losses can reflect slower growth or tighter underwriting rather than a permanent improvement in risk. Conversely, a rise can reflect seasoning rather than a sudden deterioration in every cohort.

The briefing proposed mild and severe stress cases with credit-cost increases of approximately 100 and 300–400 basis points, respectively. These were scenario inputs, not predictions. Their usefulness lies in testing whether the issuer’s earnings and capital can absorb deterioration while continuing to fund and service the portfolio.

Figure 02 · Qualitative comparison

Four metrics that should never be substituted

Four metrics that should never be substituted

Measure	What it describes	What it cannot establish alone
Cards or active accounts	What it describesCustomer access and engagement	What it cannot establish aloneSpending intensity or profitable credit
Transaction volume and value	What it describesPayment activity	What it cannot establish aloneBorrower balances or lender earnings
Revolving balances	What it describesCredit outstanding	What it cannot establish aloneAffordability, losses or net margin
Net contribution	What it describesRevenue after defined costs	What it cannot establish aloneCustomer welfare without outcome evidence

Premiumisation is a strategy, not a free margin

The historical thesis favoured premiumisation because higher-income customers can generate larger discretionary spending, lower default rates and broader relationship value. International travel, affluent consumption and wealth relationships can reinforce that proposition.

However, competition for the same customers raises acquisition costs and benefit expectations. A premium portfolio is not automatically profitable if its rewards, lounge access, service and marketing absorb the revenue advantage. The correct metric is lifetime contribution after retention costs and actual redemption, not the prestige of the card design.

The source contained institution-specific profitability and portfolio estimates that were not separately disclosed or verified. They should not be treated as established facts merely because the institution is now anonymised. The sector argument can be assessed using its general unit-economics framework without asserting a particular bank’s private profitability.

Real-time credit changes the competitive boundary

The source viewed real-time payments as dominant in transaction count while expecting cards to retain a role in larger discretionary purchases and credit. It also discussed credit lines accessible through real-time-payment interfaces as a potential disruptive scenario.

The economic issue is whether those products combine broad acceptance with competitive funding, underwriting and customer engagement. A new rail can make access easier, but lenders still need to price credit losses, acquisition and servicing. Moving the interface does not eliminate credit risk.

Merchant economics also matter. A product can gain acceptance by lowering merchant charges while reducing issuer revenue per transaction. Higher volume may offset that loss, but only if it is incremental rather than a migration of spending the issuer would otherwise have captured more profitably.

The historical briefing quoted differing interchange economics across conventional card transactions and card-linked real-time payments. Those rule-sensitive figures should be read at the source’s historical cut-off, not used as current merchant-pricing instructions. The durable analytical task is to separate incremental spending from cannibalised spending and to model the net economics of each.

Replace a conflicting forecast with a testable model

The original report alternately labelled 180–200 million cards and 150–170 million cards as its FY 30 base case. The conflicting forecasts remain in the original source record and have been withdrawn from this current public article. Choosing the more attractive number would create false precision.

A reproducible forecast starts with active accounts, annual activation and attrition, spending by customer cohort, the share carried as interest-bearing balances, average repayment and vintage losses. The receivables forecast must reconcile to the spending and repayment assumptions. Customer growth cannot be counted independently in both penetration and spending-per-card assumptions.

The model should publish sensitivity ranges only after the starting observations reconcile. It should not assign probability weights merely to make a forecast table look complete.

Figure 03 · Conceptual framework

Premiumisation creates competing pressures

How the parts connect

1. APremium positioningcommercial upsideHigher spending
2. BPremium positioningcustomer expectationsCompetitive benefits
3. CPremium positioningcompetitive responseAcquisition pressure
4. DHigher spendingincomeCohort economics
5. ECompetitive benefitsexpenseCohort economics
6. FAcquisition pressureinvestmentCohort economics
7. GCohort economicsevidenceSustainable franchise

Expansion beyond major cities needs a different model

The briefing identified lower penetration outside leading metropolitan markets as a growth opportunity. Yet a large population of potential users is not the same as an economically serviceable market. Income visibility, bureau history, distribution, ticket size and collection economics all influence the achievable return.

Digital acquisition can lower some costs, but it can also attract customers with limited repayment history or multiple unsecured products. The issuer needs underwriting that matches the segment rather than simply extending a metropolitan scorecard. Portfolio concentration across lenders and products matters when household obligations are dispersed.

The investment case should therefore separate customer acquisition from profitable activation and sustained use. Cards issued, cards active, spending per active card, revolving balances and losses are distinct measures. Growth in one can coexist with weakness in another.

A monitoring framework for the sector thesis

Fraud and credit loss should also be separated. Fraud can arise from account takeover, social engineering or unauthorised transactions even when the customer could repay. Credit loss arises when authorised borrowing is not recovered. The source discussed low historical fraud rates in the domestic market alongside growing social-engineering risks, but its issuer-specific estimates were not a substitute for comparable disclosed data.

A more complete performance review would consider false declines as well as fraud prevented. Excessively restrictive rules can reduce legitimate spending and weaken the customer relationship, particularly in a premium portfolio. The objective is not simply the lowest recorded fraud rate; it is a controlled balance of loss prevention, customer experience and contribution.

Capital consumption belongs in the same model. Higher required capital can lower the return on equity of a card portfolio even if its accounting margin is unchanged. A regulatory change should therefore be translated into capital and funding consequences before it is used to explain a sector valuation. This report preserves the historical source’s direction of concern without treating its dated rule references as current instructions.

Variables that test the sector thesis

The historical report’s strongest argument is that cards can remain economically relevant even in a market dominated by real-time payments. Its weaker point is forecast discipline: inconsistent central projections and unverified institution-specific estimates cannot be treated as a clean sector model. A stronger investment assessment preserves the premiumisation thesis, tests its net economics and keeps the uncertainty in the original numbers explicit.

Source record and update scope

• Reserve Bank of India — UPI volume grew 22.2% in Q2 FY2026–27; separate from credit balances. Observation/publication: 2026-10-07. Verified primary source; observed date stated.

• NPCI — RuPay credit cards may be linked to UPI for permitted merchant payments; exclusions apply. Observation/publication: retrieved 2026-10-08. Verified primary source; observed date stated.