This is the 200th edition of The Pattern. 

Across these editions, the goal has been consistent: to understand the structure behind the story. Over time, that structure becomes easier to recognise. 

The view from here 

Two hundred editions is not a milestone we’re celebrating — it’s a vantage point we’re using. When we started, Indian finance was in the middle of a boom that felt permanent. FinTech was flush. Co-lending was novel. Superapps were imminent. AI was a promise on every investor deck. Four years later, the landscape looks different — not broken but sorted. The models that deserved to survive did. The ones that needed ambiguity to function didn’t. 

This edition is not retrospective. It’s a reckoning with what actually happened and why. 

Five themes ran through nearly every edition we published over these four years. We’re naming them plainly — and saying what each one actually meant. 

THEME I — PARTNERSHIP LENDING 

The infrastructure play 

Co-lending was never about vision. It was about a problem both the sides had. Banks held cheap capital and priority sector obligations they couldn’t fill organically. NBFCs had origination depth and thin-file underwriting capability but expensive liabilities. The RBI’s co-lending framework gave the trade a structure. For two years, volumes grew and everyone benefited. 

Then came the FLDG circular of June 2023 — and it changed the calculus. First Loss Default Guarantees, which many FinTech-NBFC-bank arrangements had been using informally (sometimes at 10–15% of portfolio), were capped at 5% and required to be cash-backed. The models that depended on banks being effectively de-risked had to be restructured. Several didn’t survive the math. 

Five months later, the RBI raised risk weights on bank lending to NBFCs from 100% to 125%, increasing the cost of the entire model. The co-lending growth curve flattened noticeably in the quarters that followed. 

The lesson from co-lending is not that the model failed. It’s that it revealed something important: distribution is a commodity; credit judgment is not. The best bank-NBFC partnerships held because they were built on genuine underwriting trust. The ones held together by oversized guarantees quietly fell apart.

THEME II — FINTECH 

The regulation that selected for quality 

The Indian FinTech story of 2022–2025 is not a failure narrative. It is a selection narrative. The RBI’s August 2022 Digital Lending Guidelines were the most consequential regulatory intervention of the period — and the most clarifying. Lending had to happen on a regulated entity’s balance sheet. Pass-through models died. FinTechs that had been built on regulatory grey zones had to find licensed partners or exit. 

The credit tightening continued in October 2023, when risk weights on unsecured consumer loans were raised to 125%. The companies that had bet on personal loan growth suddenly found their funding more expensive and their credit limits tighter. 

Then came Paytm in early 2024 — and it changed the sector’s relationship with regulatory risk permanently. RBI barred Paytm Payments Bank from onboarding new customers, then directed winding-down of core services, citing persistent KYC non-compliance and governance failures. The company’s market cap fell roughly 70% in weeks. 

The sector absorbed one lesson above everything else: scale is not a shield. After Paytm, compliance became a first-order strategic asset — not a cost center. The FinTech sector didn’t shrink. It matured. 

THEME III — SUPERAPPS 

The ambition India’s infrastructure prevented 

India’s superapp moment was declared annually from 2021 to 2024. Tata Neu launched in April 2022 with the full weight of a conglomerate behind it. Jio assembled every ingredient — subscribers, commerce, media, financial services. PhonePe diversified aggressively into insurance, mutual funds, and credit. None of them became WeChat. 

They couldn’t — and the reason is structural, not executional. WeChat’s dominance was built on payment lock-in. UPI is open and interoperable. No company can build a payment moat on a public rail. Without payment lock-in, the superapp flywheel doesn’t spin in the way it does around the world. And that’s a good thing. We have multi-product apps but no superapps (yet!).  

ONDC made this permanent. By turning digital commerce into a public protocol, it removed the need for any single platform to aggregate everything. India’s policy design made a deliberate choice: interoperability over platform dominance. 

India did not fail to build a superapp. It built infrastructure that made superapps unnecessary. The question now is who benefits from that infrastructure most. The answer is increasingly: whoever owns the transaction relationship. 

THEME IV — PAYMENTS AS DISTRIBUTION 

When the transaction layer became the sales layer 

The most consequential shift of this period happened sans fanfare, without a product launch or regulatory circular. Payment platforms — with hundreds of millions of users who trusted them for everyday transactions — discovered that distributing financial products to those users cost almost nothing. Banks and insurers, who had spent decades and enormous capital building agent networks and branch distribution, found themselves competing with a push notification. 

PhonePe became one of India’s largest mutual fund distributors by volume. CRED turned credit card bill payments into a financial marketplace for prime borrowers. And across the stack, UPI extended from debit into credit — first through RuPay credit cards on UPI rails, then through pre-approved credit lines drawn directly on UPI. This converted the payment layer into a credit distribution layer. 

When a payment app’s cost to acquire a financial product customer is near zero and a bank’s is ₹2,000–3,000, the distribution moat shifts. That asymmetry does not resolve in the bank’s favour over time. 

The Paytm episode showed how fragile this integration is when the payment foundation is disrupted — which is precisely why protecting the payment relationship is now the most important strategic priority for anyone in this space. 

THEME V — ARTIFICIAL INTELLIGENCE 

The honest accounting 

Between 2023 and 2025, almost every bank and NBFC announced an AI strategy. Fewer than a quarter deployed AI at a scale that changed their operating economics. The gap between announcement and deployment is the most important thing to understand about AI in Indian finance right now.

What was deployed: collections voice bots, document processing for KYC and loan origination, fraud detection at network level, customer service automation. What remained in pilots: GenAI for customer-facing financial advice, autonomous credit decisions, real-time personalised banking. 

The pattern is consistent — AI scales where the task is narrow, the output is verifiable, and human oversight is easy to maintain. It stalls where decisions require accountability to a regulator or an explanation to a borrower. 

The genuine structural development was more discreet: the Account Aggregator framework maturing. When AA-native data feeds ML underwriting models, credit decisions improve for thin-file borrowers in ways bureau-only models cannot replicate. This is where the real AI story in Indian credit is being written — by lenders who don’t make many announcements. 

The AI gap in Indian banking is not primarily a technology gap. It is a data infrastructure gap, a talent retention gap, and a decision-making culture gap. Institutions that don’t solve the first three will not benefit meaningfully from the fourth — regardless of how many AI partnerships they announce.  

The Meta-Pattern 

These five themes are not independent. Co-lending created distribution scale for NBFCs. FinTechs that survived used that distribution to build regulated lending businesses. Superapps couldn’t form because India’s infrastructure was deliberately interoperable. Payment platforms filled the distribution gap that superapps couldn’t. And AI is now being layered onto the distribution, underwriting, and servicing infrastructure that all of this built. 

The four years from 2022 to 2026 did not produce a revolution in Indian finance. They produced a consolidation — of regulatory posture, of business models, of who the credible players are. The institutions that understand this are better positioned for what comes next than the ones still chasing the disruption narrative. 

Across all five themes, the same force ran underneath: Indian finance became more regulated, more interoperable, and more concentrated in the hands of players who had earned regulatory trust. That is not a loss. Markets that mature through regulatory consolidation tend to be more stable, more competitive, and more trustworthy for the people they serve. Two hundred editions in — that’s what the pattern shows. 

From the archives 

We usually end with a reading list from the week.

For this edition, we went back instead. Five pieces from across our work — not just The Pattern, but everything we’ve published — that best capture how we think about lending, risk, and credit infrastructure. These are the pieces we keep coming back to.  

1. How AA reduces fraud and improves conversions: A FinBox case study  

2. What fraud can tell us about FinTech  

3. State of Digital Lending 2026: A FinBox report 

4. FinBox’s guide to partnership lending  

5. Why AI killing the loan application form is good for lending 

Thank you for reading. If you liked this edition, forward it to your friends, peers, and colleagues. You can also connect with me on X here and follow FinBox on LinkedIn to get the latest updates.    

See you next week.   

Cheers,   
Mayank