The Moment That Slipped By Unnoticed
In September 2025, something historic happened in the Indian economy.
For the first time ever, Indians paid more money through digital channels — UPI, credit cards, debit card online transactions — than they withdrew as cash from ATMs.
Digital payments: ₹2.53 lakh crore in a single month.
ATM cash withdrawals: ₹2.30 lakh crore.
Digital won.
This wasn't a forecast. It wasn't an aspiration in a government white paper. It happened. And it barely registered in the mainstream financial press.
Digital crossed cash in September 2025. And nobody really talked about it.
I think that's a mistake. Because this crossover — if it holds, and the March 2026 data suggests it will — marks a genuine structural shift in how 1.4 billion people interact with money. And structural shifts in payment behaviour have a habit of reshaping entire industries, often before the people running those industries have noticed.
The Big Picture: Three Numbers to Anchor Everything
These three numbers tell you a lot. The digital payment rails are already carrying serious volume. The credit card market has enormous room to grow. And the debit card market is, for all practical purposes, a zombie.
Part One: Credit Cards — A Market Still in Its Infancy
India's credit card market grew 13.5% year on year in transaction value terms by March 2026. Monthly credit card spend is now running at ₹2.19 lakh crore.
That sounds large. But context matters. India's GDP is approximately ₹350 lakh crore annually. Credit card spend at current run rates is under 8% of GDP. In the US, credit card spend is closer to 35–40% of GDP. South Korea — arguably the world's most credit-card-intensive economy — runs at over 50%.
India is nowhere near saturation. The top 10–15% of the income distribution drives the majority of credit card spending. The next 20% of the population — urban, digitally connected, with stable income — is largely uncaptured. That's roughly 250–280 million people.
Who's Winning
| Bank | CC spend, YoY | Read |
|---|---|---|
| Federal Bank | +88% | Digital-first entry + co-brand partnerships; largest growth in the dataset |
| PNB | +41.8% | The quiet PSU surprise — watch whether it sustains |
| Kotak Mahindra | +25.8% | Consistent, higher-quality growth with better underwriting |
| SBI | +24.1% | Converting the mothership's distribution into SBI Cards customers |
Who's Not
IndusInd Bank: −23% YoY. The most alarming number in the entire dataset — and it's not being discussed widely enough. Credit card data from the RBI tends to be a leading indicator for bank-level retail stress. When cardholders pull back, either the bank is restricting credit, or customers are worried. Either way, the direction is not good. (I've written a separate note on this: IndusInd: When the Data Knew Before the Market Did.)
Amex: −5% YoY in a market growing at 13.5% is effectively losing 18 percentage points of relative ground. The premium card market is under structural pressure as domestic issuers match or exceed Amex's rewards proposition at lower fees.
Part Two: Debit Cards — The Quiet Death of a Product Category
Let me be direct: the debit card, as a payment product, is in structural decline in India. This is not alarmist. This is what the data says.
93 out of every 100 debit cards issued in India are not being used for transactions in a given month.
This is the natural result of a structural feature of Indian banking: any time a bank account is opened, a debit card is automatically issued. So 104 crore debit cards outstanding is not a measure of payment activity. It's an artifact of account-opening practices.
Why UPI Is the Reason
UPI has taken the core use case of the debit card — paying at a physical store — and made it cheaper, faster, and more widely available.
UPI has 76 crore QR codes deployed. Card PoS terminals: 117 lakh. QR codes outnumber card machines by 64 times.
The economics are decisive. A UPI QR code costs close to nothing. A card PoS terminal costs ₹15,000–20,000 upfront plus MDR fees. For the vast majority of Indian merchants — especially in tier 2, tier 3 and rural areas — the QR code wins by default. The implication for banks: the debit card portfolio, once a source of MDR fee income, is becoming a liability.
Part Three: The ATM Network — A Managed Sunset
India's ATM network is being quietly rationalised. Total ATMs were relatively stable — but off-site ATMs (petrol pumps, malls, kiosks) fell 11.8% year on year. Banks are actively decommissioning them: an off-site ATM costs roughly ₹1–2 lakh per month to operate, and when throughput falls 10–15% a year, the machine eventually runs at a loss. UPI has broken the business case.
On-site ATMs — machines inside branches — are stable. That's not panic; it's rational capital allocation. One counterintuitive data point: while withdrawal volumes are falling, the average value per withdrawal is likely rising. Cash isn't disappearing. It's becoming a deliberate, occasional choice rather than a default — mirroring the UK and Australia as digital payments matured.
Part Four: UPI — Not Just a Payments App
Most coverage of UPI focuses on volume growth. The more interesting angle is what UPI is displacing — and what it is not.
UPI has displaced: debit card PoS swipes, small ATM withdrawals, NEFT for small transfers, cash for peer-to-peer payments. UPI has not displaced: credit card spending. Credit cards provide credit — access to money you don't yet have — and the core mechanic of UPI is pulling from a bank balance. They're architecturally different. This is why the credit card market grows at double digits even as UPI expands at speed.
What This All Means for Investors
Nothing in this analysis constitutes investment advice. That said, the data suggests a few directional observations worth thinking through:
- Credit card issuers with strong digital acquisition are in a good position: a growing market, low penetration, structural tailwinds.
- Banks dependent on debit-card fee income face compression — MDR caps plus volume decline.
- ATM infrastructure businesses (white-label operators like AGS Transact, Euronet) face a structural demand headfall.
- UPI-first fintechs have volumes, but the monetisation question remains open.
- SBI Cards is worth watching given SBI's 24% YoY credit-card growth.
The Bottom Line
India's payment system is in the middle of a genuine structural transition. The September 2025 crossover is the milestone that anchors it. It's not a data point — it's a signal that the transition is done: the habits have changed, the infrastructure has changed, and the economics have changed.
The question now is: who benefits from what comes next? That's what I'll keep tracking — monthly RBI data, quarterly earnings, and the longer-term structural picture.
Explore every chart and bank-level number yourself in the interactive FY26 dashboard.
Sources: RBI ATM & Card Statistics — March 2025, September 2025, March 2026 releases; NPCI UPI product statistics. Disclaimer: personal research for educational purposes; not investment advice.