Welcome to the 111th edition of Fintech Inside. Fintech Inside provides nuance and insight to the big trends shaping financial services. It’s the fintech newsletter for people who don’t have time to read five fintech newsletters.

For the last few decades, the world moved toward globalisation. People, capital, goods, services, technology, data - all of it flowed freely across borders. The assumption underneath most of it was that this flow was basically uni-directional and basically permanent.

That assumption is under real strain right now. And I think most of the fintech commentary on stablecoins is missing why.

This week’s edition is about correspondent banking, stablecoins, economics, sovereignty and what India is missing out on.

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech.

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

The Slow Death of Correspondent Banking?

The swing back toward sovereignty

China spent two decades proving that a country can build its own internet, its own chip stack, its own AI models, largely walled off from the rest of the world’s infrastructure. That used to sound implausible. It doesn’t anymore.

The US has been making the same point from the other direction. Anthropic had to suspend access to its Fable and Mythos models on June 12 this year to comply with US Department of Commerce export controls, then restored access on July 1 once those controls were lifted. Whatever you think of the policy itself, the mechanism is the point: the US government demonstrated, in real time, that it’s willing to use export controls to decide who gets access to their frontier technology. Can’t remember the last time this happened. Technology sovereignty isn’t just a Chinese project anymore. It’s a live lever.

Data sovereignty has followed the same arc, data localisation laws and data residency requirements have gone from a niche India/EU concern a decade ago to fairly standard practice across most large economies today.

Which brings us to the layer underneath all of those: capital.

This is where things move at a much slower, much less visible pace. On the surface, there’s been a lot of noise - sovereign wealth funds getting more assertive, reserve currency diversification, tariff fights, the odd tax skirmish. Underneath, actual capital sovereignty is glacial, because money is the most globally interconnected thing we have, and the switching costs are complex and enormous.

But watch the two biggest economies play the same game from opposite ends.

China has spent over two decades building demand for the renminbi the slow way (see Edition #102): financing ports, railways and power grids across 150+ countries, mostly through RMB-denominated loans, currency swap lines with roughly 50 economies, oil settled in RMB with Saudi Arabia, its own SWIFT alternative (CIPS) now linking close to 4,800 banks across 185 countries. It isn’t subtle about the goal either, Xi Jinping has publicly called for building a “powerful currency” with global reserve status. I’ve written in Edition #105, how Ant Group is going global by following Chinese global citizens - travellers and immigrants.

The US, meanwhile, tried something more direct and blunter: a provision in last year’s tax bill, nicknamed the “revenge tax,” that would have raised US taxes on investors from countries with tax regimes it considered unfair to American companies. It was dropped within weeks after a G7 deal, so it never became law. But the fact that a G20 economy floated using access to its own capital markets as retaliatory leverage tells you where the appetite is right now, law or no law.

And then there’s the GENIUS Act, which is where this gets genuinely counterintuitive.

The GENIUS Act is US domestic legislation. It regulates who can issue a compliant dollar stablecoin inside the US and what has to back it. On paper, it has nothing to do with any other country’s monetary sovereignty. In practice, it’s arguably the single most effective de-dollarisation defence mechanism the US has built in years — because a well-regulated, dollar-backed stablecoin is exactly the kind of instrument an emerging market’s businesses and citizens reach for the moment their own currency wobbles. US Treasury Secretary Scott Bessent has said as much, plainly: stablecoins are a tool to keep the dollar dominant. A country can launch its own stablecoin, feel like it’s building sovereign digital infrastructure, and still end up more dollarised than before, because the stablecoin sitting on top of that infrastructure is backed by dollars either way.

That’s the paradox worth sitting with. Everyone is fighting for sovereignty, over technology, over data, over capital. And a loosely effective countermove any single country has found so far isn’t a wall. It’s a stablecoin.

Which is exactly the terrain India has been standing on for a while, without fully realising it.

UPI proved something the world needed to see

UPI showed that instant, real-time settlement is not just possible, it’s the right default for a mobile-first world where plastic cards were never going to matter the way they did in the US or Europe. That’s a genuine, undersold achievement, India built the payments rail the rest of the world is still catching up to.

But exporting UPI has been a slow, country-by-country slog. Bilateral agreements, one regulator at a time, one central bank relationship at a time. It’s real progress, but it hasn’t yet found the kind of traction that turns a domestic rail into global default infrastructure. And in a world where sovereignty over money is suddenly back on the table for everyone, that slow pace is starting to look like a cost, not just a rollout detail.

Which is what makes the recently launched OpenUSD worth paying close attention to right now.

First, what OpenUSD actually is

On June 30, an entity called Open Standard announced OpenUSD (ticker: OUSD), a stablecoin, backed by more than 140 businesses, which is an unusually broad coalition for anything in this space. The list is worth sitting with for a second: Visa, Stripe, Mastercard, Amex, BlackRock, BNY, Standard Chartered, Coinbase, Shopify, DoorDash, Google, Samsung. Payment networks, global banks, big tech, and crypto infrastructure, all on the same cap table.

When National Payments Corporation of India (NPCI) was formed it was promoted by a consortium of public and private sector banks in the country. The ten core promoter banks include the following entities - State Bank of India, Canara Bank, Punjab National Bank, Union Bank of India, Bank of Baroda, Bank of India, HDFC Bank, ICICI Bank, HSBC, Citibank.

Open Standard is the NPCI of USD stablecoins.

Three design choices separate OUSD from existing stablecoins like USDT and USDC:

• No fees, no volume caps: Businesses can mint and redeem Open USD at no cost, with no artificial ceiling on volume, a direct jab at the mint/redeem fees that make existing stablecoins expensive at scale.

• Partners keep the yield: Reserve earnings flow back to the partners, minus a small fee to run the thing, instead of being captured entirely by the issuer the way they are with Tether or Circle today.

• Nobody owns it: It’s run by an independent company with a board made up of its partners, not a single issuer calling the shots.

On the surface, it seems like this isn’t Stripe launching a stablecoin. It’s Stripe, Visa, Mastercard, and 140 others agreeing that no single company should own the rails everyone plans to build on. OpenUSD goes live later this year (2026).

That’s the news. Here’s why it matters more than it looks.

Stablecoins stopped being a crypto story

Zoom out from OpenUSD and look at the last eighteen months. Stripe acquired Bridge. PayPal launched PYUSD. Visa and Mastercard are wiring stablecoin settlement into their networks. Coinbase, Robinhood, and Circle keep expanding stablecoin products. This is consensus, not a coincidence.

The global stablecoin market crossed $310–320bn, more than doubling in under two years. That’s now larger than the foreign exchange reserves of over 90 countries, including developed economies like the UK and Canada. BNY’s own product chief recently projected the category could reach $1.5tn by 2030.

Five years ago, “stablecoin” meant one thing: crypto trading. Today it means cross-border payroll, treasury management, B2B settlement, remittances, and increasingly, the rails AI agents will use to transact with each other. The use case moved from speculation to infrastructure, and most of traditional finance is not noticing the shift happen in real time.

You don’t have to take anyone’s word for it. The market already voted with $300bn.

Here’s the piece of evidence I think is the most underrated in all this, and it isn’t from a central bank at all. It’s from Wall Street’s own reaction to the GENIUS Act actually passing.

A March 2026 IMF working paper did something clever: instead of asking economists what they think stablecoins will do to payments, the authors watched what investors actually did with their money in the ten trading hours around the decisive House vote on the GENIUS Act, on July 17, 2025. Comparing listed payment firms to the rest of the financial sector, and correcting for how much the market had already priced in beforehand, they estimate the vote wiped out 18% of incumbent payment firms’ market value i.e. roughly $300 billion. That’s a bigger hit than the market impact of the Durbin Amendment (debit card fee caps) or the ECB’s digital euro announcement on US firms, the two benchmark “stablecoins are disruptive” precedents people cited before this paper existed.

The part that matters more than the headline number is where the damage concentrated. Firms whose primary business is cross-border payments got hit hardest with an estimated 27% wipeout, more than double the sector average. Firms with strong existing network effects (Visa, Mastercard, the card networks) saw no significant hit at all. Neither did firms that had already engaged with crypto assets before the vote i.e. PayPal, Block, Fiserv. The market isn’t betting that stablecoins destroy payments broadly. It’s betting they gut the slow, expensive correspondent-banking layer specifically, while leaving network-effect moats and early movers largely untouched.

You can watch this prediction play out in real time. Western Union, about as pure a cross-border correspondent as exists, announced in October 2025 that it would launch its own stablecoin. Its own earnings calls tell the story of a company changing its mind in real time: in May 2025, before the Act passed, a Western Union executive was calling stablecoin economics “operationally challenging and expensive.” By July, weeks after the Act passed, the same company was “actively revising our position” with a “clear regulatory framework and guaranteed redemption value.” Same company, same underlying technology, completely different posture, the only thing that changed was the law.

Worth sitting with for a second: UPI’s dominance inside India functions a lot like Visa’s network effect inside the US. If the market’s read is right, that’s exactly the kind of moat that holds up against stablecoin competition. The exposure isn’t to UPI as a domestic system. It’s to whatever tries to move money across India’s borders the old, slow way.

Stablecoins to disrupt correspondent banking?

While the US companies and regulators are focussed on launching standards, regulations and products for stablecoins and crypto, the global institutions it manages, seem to have the opposite view.

According to the Bank for International Settlements’ 2026 Annual Economic Report, 99.4% of all fiat-backed stablecoin value is now denominated in US dollars. Every other currency combined accounts for the remaining 0.6%. That’s near-total dominance.

On June 28 2026, the BIS went further, formally declaring that stablecoins fail the basic tests of money i.e. singleness, elasticity, interoperability, and warned explicitly about “digital dollarisation,” the risk that dollar-denominated stablecoins become the default cross-border infrastructure for emerging markets, quietly weakening their monetary sovereignty in the process.

That’s a strong statement from the central bank of central banks. It reads less like commentary and more like a line being drawn.

Why this matters beyond crypto

Strip away the crypto framing for a second, because I think that’s what’s making this hard to see clearly.

This debate isn’t really about crypto. It’s about whether the next generation of internet-native businesses settles in dollars by default, simply because dollars are the easiest rail available.

Picture an AI agent paying another AI agent for compute, data, products, services etc, with no human in the loop to pick a currency. Picture a SaaS platform paying out thousands of creators or gig workers across a dozen countries in one settlement run. Picture a marketplace clearing merchants across twenty countries overnight instead of over days.

None of these care about correspondent banking relationships. None of them are going to route around a currency because a central bank prefers a different rail. They care about whichever settlement layer is easiest to plug into. Right now, that’s overwhelmingly dollar rails. Every one of these use cases is a small vote for that status quo, and there are going to be a lot more of them, a lot sooner than most central banks are planning for.

The great paradox

What’s fascinating here is that no single institution designed this outcome.

The BIS worries about financial stability. The IMF worries about monetary sovereignty. Stripe wants easier internet payments. BlackRock and BNY want to serve institutional demand for tokenised dollar exposure. The US Treasury quietly benefits from anything that extends dollar reach without it having to do anything at all.

Different institutions, different objectives. And yet all of it reinforces the same outcome: a more internet-native dollar, whether anyone intended that or not. That’s a more interesting story than “America wins.” Nobody planned this. It’s just what happens when everyone’s separate incentives point in the same direction.

Three different answers to the same question

Strip away the noise and three governments are running three different experiments on the exact same problem, how do you keep control of money as it goes digital, when the best available technology is being built somewhere else?

America’s answer is private stablecoins, regulated just enough to be trusted, precisely because a trusted dollar stablecoin is the best de-dollarisation defence the US could have built. Europe’s answer, under MiCA, is regulated private money deliberately steered toward the euro rather than the dollar — the BIS calls this a live test case for whether a “domestic currency integration” path can actually work. China’s answer is neither: an outright ban on crypto and stablecoin trading, reasserted as recently as February 2026, paired with the e-CNY as the state-backed alternative. Three different bets on the same underlying question, none of them proven yet.

Back to India and why “just export UPI” isn’t a full answer anymore

India is not the only country facing this. Indonesia, Brazil, Nigeria, Mexico, Turkey, and Vietnam are all sitting on some version of the same tension - real-time domestic rails that work brilliantly at home, and a much harder, much slower path to making that infrastructure matter beyond their own borders. India just happens to be the sharpest example, because it also built Aadhaar, DigiLocker, and the Account Aggregator framework alongside UPI, arguably the best public digital infrastructure stack anywhere.

But now, where the rest of the world’s private sector ran toward stablecoins, India, like most large emerging economies, leaned toward CBDCs instead. Different philosophy, same underlying question every one of these countries is quietly asking: how do you keep control of money as it goes digital, when the alternative infrastructure is being built by companies you don’t regulate, sitting inside a country whose regulators just showed they’re willing to switch access on and off?

UPI was built for Indian bank accounts, Indian regulators, Indian settlement rails, Indian identity infrastructure. The internet has none of those defaults, and increasingly, when a developer somewhere else needs to plug into a settlement rail, the honest, easiest answer is a dollar stablecoin, not a domestic real-time payments system, however good that system is.

Here’s where I want to correct something I nearly got wrong in an earlier draft of this piece, because it’s a genuinely important nuance, and it comes straight from the RBI rather than from me. It’s tempting to frame this as “India just needs its own rupee stablecoin.” RBI Deputy Governor T Rabi Sankar gave a keynote at BIS in December 2025 that closes that door explicitly, and it’s worth relaying accurately.

His argument: stablecoins fail the basic tests of money. They’re not fiat, they’re not “single” (there could be hundreds of competing private monies instead of one trusted one), and there’s no guarantee the issuer’s promise to redeem at par is even a legal liability. On every claimed benefit i.e. cross-border efficiency, financial inclusion, bridging crypto to the real economy, he argues India’s existing rails (UPI, RTGS, NEFT) already deliver it better, without the risks of currency substitution, capital account leakage, bank disintermediation, and a quiet transfer of seigniorage (the profit a government normally earns from issuing its own currency) to private stablecoin issuers, most of them foreign.

And crucially, he rejects the “just make it a rupee-denominated stablecoin” compromise too. In his view, a domestic-currency stablecoin still carries the currency-substitution, disintermediation and seigniorage risks, it isn’t the foreign-currency exposure that’s the core problem, it’s the private issuance itself. As he put it, the bigger threat is a stablecoin that works well.

India’s actual stated strategy has four pillars: preserve trust in the rupee and the existing payment system, safeguard monetary sovereignty, push CBDC adoption and cross-border CBDC corridors, and interlink India’s fast payment systems directly with partner countries rather than build a private alternative on top. That’s a real, coherent bet, not a policy vacuum waiting for a stablecoin to fill it.

Worth noting separately: India already has a functioning compliance apparatus for crypto activity regardless of where the stablecoin debate lands. Since 2026, exchanges and wallet providers dealing in virtual digital assets have been brought under the same anti-money-laundering law that governs banks, KYC, suspicious transaction reporting, the works. So “India has no rules for this” isn’t quite right. The AML plumbing exists. The monetary policy question is the one still being fought over.

The same open question i.e. CBDC and FPS interlinking versus some form of private digital rail, applies just as directly to the real, the rupiah, the naira, and the lira, even where those central banks haven’t stated a position as clearly as India has.

So what should emerging markets actually do?

I don’t think there’s one clean answer, and I’d be suspicious of anyone who claims there is. The BIS’s own research sketches out why this is a live, unresolved question rather than a settled one. Researchers there lay out three scenarios for where this goes: stablecoins stay a niche crypto-trading tool with limited real-world adoption; stablecoins become the de facto cross-border payment rail across emerging markets, with real damage to monetary policy and capital controls; or countries successfully build regulated domestic-currency stablecoins that interoperate with existing payment rails and CBDCs, capturing the efficiency gains without the sovereignty costs. They’re honest that the third path requires more regulatory capacity than most emerging markets currently have.

India’s bet i.e. CBDC plus interlinked fast payment systems, no private stablecoin at all, sidesteps that capacity problem entirely by not needing it. Whether that wins depends on whether CBDCs can move as fast internationally as private capital and developer tooling move by default. That’s not a question any single regulator can answer alone, and if the market’s repricing of Western Union is any signal, the companies that actually move money across borders may answer it faster than the regulators do.

The goal isn’t chasing crypto for its own sake. The goal is making sure local currencies stay relevant on rails that are increasingly internet-native by default, not government-native.

The bigger point

Every era gets the monetary infrastructure it builds for itself. The Bretton Woods era ran on correspondent banking and fixed exchange arrangements. The internet era is going to run on programmable money, and the standard-setting for that infrastructure, OpenUSD being the latest example, is happening right now, largely outside any emerging market’s control.

India has already proven that a government can build digital public infrastructure that rivals, arguably beats, anything the private sector has shipped elsewhere. That was the first payments revolution, and India won it decisively, the same way a handful of other emerging economies won their own versions of it.

The next one is a different contest. It’s no longer about building the best domestic payments system. It’s about building, or choosing, the monetary infrastructure the rest of the internet decides to use by default, especially the part that moves money across borders, which is exactly where the market’s own $300bn repricing says the pressure is landing first. UPI isn’t going anywhere at home. The real question is whether the rupee, and its equivalents elsewhere, become native to that internet on their own terms, or whether the internet just quietly transacts in someone else’s digital currency instead, the way most cross-border flows already quietly do, without anyone noticing it happened.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on Loop

Good background songs to listen as you read Fintech Inside: This weekend I was at a bar and I heard a song I hadn’t heard since college - Black Betty by Ram Jam (Spotify / Youtube). Full nostalgia trip. Enjoy! 🎶

✨ Call Outs

1. [Video] AI x FinTech: Voice, Agents & Year 1 in a Decade of Change | SummitUp by Elevation

2. [Product] Wiretap.us - use AI to scan social media for mentions of your brand

3. [Video] How to build a full body ultrasound

4. [Video] The Truth about Space Data Centers

5. [Book] What I learned about investing from Darwin

👋🏾 That’s All Folks

If you’ve made it this far - thanks! As always, you can always reach me via DM at osborne.vc/dm. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 110th edition of Fintech Inside. Fintech Inside provides nuance and insight to the big trends shaping financial services. It’s the fintech newsletter for people who don’t have time to read five fintech newsletters.

A few weeks ago I wrote about what AI can and cannot disrupt in financial services i.e. the interface layer is fair game, the system of record is not. A lot of you wrote back. The question that kept coming up: okay, so if AI can’t replace the core, what does it actually change?

Everything else. And faster than most people are prepared for.

This edition is about India’s once-in-a-generation opportunity to use AI to become the world’s services factory. Not by being cheaper. By delivering high quality outcomes, faster, than anyone else on the planet.

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

India’s once-in-a-generation opportunity

For thirty years, the global economy ran on a fairly legible division of labour.

America innovated. China manufactured. India serviced.

Each country carved out a lane and built deep competitive advantage in it. The US combined world-class universities, deep capital markets, and a culture of genuine risk-taking.

China invested in infrastructure, vocational training, and industrial clusters at a scale the world had never seen.

India became the back office of the world i.e. IT services, BPO, finance, accounting, software development, an export machine that today employs millions and generates hundreds of billions in foreign exchange.

That division is now being disrupted, simultaneously, by a single technology.

The question is not whether AI changes services. It will.

The question is whether India uses this moment to move decisively up the value chain or gets commoditised out of the very market it built.

The Wrong Lesson From China

Most people who cite China’s manufacturing success point to cheap labour.

They’re wrong. Tim Cook has been saying so since 2024.

China stopped being the world’s lowest-cost manufacturer a long time ago. Vietnamese factories are cheaper. Mexican factories are closer. And yet the world still makes things in China. Because China didn’t win on cost. It won on capability i.e. deep tooling expertise, precision engineering, supply chain coordination at scale.

The kind of industrial execution that only emerges when millions of people spend decades getting very, very good at a thing.

China became indispensable by becoming the best at manufacturing.

India needs to build the equivalent in services, and fast. And AI is the machine that makes it possible.

Our services industry is still largely selling on cost arbitrage. Indian salaries have risen to comparable global compensation standards. Remote work has globalised competition. AI is reducing the value of routine, repeatable labour.

India stopped being a low cost service provider sometime around covid, but our output quality is still lagging.

We seem to be stuck in a chalta hai (Hindi for “it’s fine”) attitude loop. Nahi chalta hai boss! (Hindi for “fine doesn’t work anymore, boss!”). We need to dramatically raise the bar on quality of services we provide domestically and internationally. Just “fine” doesn’t cut it in the AI age. That will require upskilling at a scale we’ve never seen before.

The next phase of India’s services story cannot be built on being cheaper.

It has to be built on being better. And delivering that better, faster than anyone else.

100x service provider: the leverage shift

The dominant narrative around AI in services is job displacement. AI writes the report. AI answers the query. AI files the document. Headcount falls.

There may or may not be truth to that - time will tell. But it’s the wrong frame for what India should be building toward.

The more important shift is leverage. AI doesn’t just automate tasks, it multiplies what a skilled professional can deliver in a given unit of time.

A wealth manager who serves 80 clients today may serve 200 or even 800 tomorrow. Not because she’s working harder. Because AI is handling the preparation i.e. portfolio reviews, market summaries, meeting notes, rebalancing recommendations and so on, and she’s spending her time on the judgment and relationship work that actually requires a human.

A compliance analyst who reviews 50 transaction alerts a day may review 500. Not because he’s faster at reading. Because AI has already triaged, categorised, and flagged the ones that actually need human attention.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

A loan underwriter who processes 20 applications a week may process 200. Because AI has already extracted and cross-validated the bank statements, GST filings, and income records before the file even reaches her desk.

This is not a future with fewer professionals. It is a future where the same professionals deliver dramatically more, faster, with higher accuracy, at lower marginal cost.

That is the leverage shift. And it is the foundation of India’s opportunity.

Speed in service delivery as a moat - old wine in a new bottle

In most industries, quality is the differentiator. In financial services, speed is quality.

At its core, financial services is an information-processing industry. Banks, insurers, wealth managers, asset managers, fintechs, regulators, they all spend enormous amounts of time reading documents, reviewing transactions, assessing risk, monitoring compliance, producing reports, communicating with customers. Every single one of those activities is changing fast.

Speed, powered by AI, is rapidly becoming the single most important competitive variable across every layer of financial services. And India is sitting on top of one of the largest pools of financial services talent in the world.

Let me go through the specific verticals where I think this plays out most clearly.

Wealth management. India’s wealth management industry has a coverage problem. There are tens of millions of households that have enough investable assets to warrant professional advice but not enough to justify the economics of a full-service relationship manager. AI changes that math entirely. One advisor with AI-assisted portfolio review, automated client communication, and AI-generated financial plans can cover a client base that previously required a team. The output quality goes up. The cost per client comes down. And the speed of response to market movements, to client queries, to rebalancing triggers, goes from days to minutes.

Lending. Credit underwriting in India is still painfully slow for most product segments. Fintech startups have improved this at the consumer end, but MSME and commercial lending still involves weeks of document collection, manual verification, and sequential review. AI compresses every step. Document extraction. Bank statement analysis. GST reconciliation. Bureau pull interpretation. Fraud signal detection. Each of these steps, which used to be done sequentially by different analysts, can now happen in parallel, in minutes. The underwriter’s job shifts from data collection to judgment. And judgment, applied with better information and less time pressure, produces better credit decisions.

Cross border and tax management: Indians are increasingly global citizens, working in the US, holding assets in Singapore, earning rental income in Dubai, managing an ESOP from a Cayman-incorporated startup. The tax implications of this are genuinely brutal. You need a CA in India who understands FEMA, an accountant in the US who understands FATCA, and someone who can reconcile the two without getting you into trouble in either jurisdiction. Today, that coordination is slow, expensive, and error-prone, because each professional is working in their own silo with their own tools. AI changes the coordination layer fundamentally. It can ingest income and asset information across jurisdictions, map applicable treaties, identify filing obligations across regimes, flag inconsistencies before they become compliance events, and generate jurisdiction-specific documentation for the CA or CPA to review and sign off on. The professional stays in the loop, liability and judgment remain with them, but the preparation work that used to take weeks collapses to hours. For the tens of millions of Indian professionals and entrepreneurs living globally, this is not a niche problem. It is a massive, underserved pain point.

Compliance. This is where I think India’s biggest opportunity sits, and it is underappreciated. Global financial institutions spend billions annually on KYC, AML, sanctions screening, transaction monitoring, and regulatory reporting. These are labour-intensive, rule-bound, data-heavy functions. They are exactly the functions where India already has scale — through GCCs, outsourcing firms, and consulting operations. And they are exactly the functions where AI provides the most dramatic throughput improvement. An AI-augmented compliance team does not just do the same work faster. It catches more. It files more accurately. It produces audit trails that are cleaner and more defensible. It turns compliance from a cost centre into a genuine competitive advantage for the institutions that get it right. India is positioned to be the delivery layer for this globally.

Insurance operations. Claims processing, medical record review, underwriting support — all operationally intensive, all heavily dependent on document interpretation, all changing fast. AI doesn’t replace the adjuster or the underwriter. It means the adjuster is spending time on genuinely complex claims and the underwriter is spending time on genuine risk judgment, not on extracting data from PDF attachments.

Investment research and fund operations. Portfolio monitoring, earnings transcript analysis, LP reporting, investment memo production, this is knowledge work that scales badly. A good analyst can only read so many filings. An AI-augmented analyst reads all of them and flags the ones that matter. India already has a deep bench of CFA-trained, IIM-educated analysts doing this work for global funds. AI turns each of them into a multiplier.

This is the future of knowledge work across every layer of financial services. And India has the talent base to lead it.

India’s second services wave

The first wave of India’s services exports was built on labour arbitrage. Send the work to India because the talent is cheaper. That was the pitch for IT services in the 1990s, for BPO in the 2000s, for GCCs in the 2010s. It worked. Enormously. But it was always a cost story.

The second wave is a capability story.

And it is broader than fintech. It spans every domain where India has built deep professional expertise and global client relationships i.e. legal, accounting, consulting, research, healthcare administration, hospitality, engineering services. The common thread across all of them is the same: skilled professionals, doing knowledge-intensive work, for global clients, at scale.

AI doesn’t disrupt that model. It supercharges it.

A legal services firm that drafts contracts doesn’t get replaced by AI, it uses AI to draft faster, review more thoroughly, and serve clients in more jurisdictions simultaneously. An accounting firm that handles cross-border tax compliance doesn’t get disintermediated, it uses AI to navigate multi-jurisdiction complexity that would have previously required three separate engagements. A consulting firm that produces market entry reports doesn’t get automated away, it uses AI to synthesise more data, with more rigour, in a fraction of the time.

In each case, the service is not merely software products. They are outcomes, delivered by professionals whose judgment, relationships, and domain expertise remain central, and whose leverage has been multiplied by AI.

This is what the second services wave looks like. Not AI replacing Indian professionals. Indian professionals using AI to deliver outcomes that were previously impossible at this speed, this quality, and this price point.

The companies that win this decade, in fintech, in legal, in accounting, in consulting, will be the ones that rebuild their delivery model around AI from the ground up. Not bolting AI onto existing workflows, but rethinking what the workflow looks like when AI is native to it. The pitch to global clients changes from “we are cheaper” to “we are faster, more accurate, and we get better every month.” India knows how to make a services pitch. It has been making one for thirty years. This is just a dramatically better version of it.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

The capability imperative

None of this is automatic.

China’s manufacturing throughput didn’t come from cheap labour. It came from decades of deliberate capability building and the discipline to obsess over quality until it became the default, not the aspiration.

India needs the same obsession. Not with cost. With capability building.

The winning wealth manager will be an AI wealth manager. The winning compliance officer will be an AI compliance officer. The winning CA handling cross-border tax will be one who has rebuilt their entire practice around AI, not someone who occasionally uses it to draft an email. These are professionals who understand what AI can and cannot do, and have redesigned their workflows around that understanding.

The underlying talent is here. The domain expertise is here. The client relationships are here. What’s needed now is the cultural shift, from selling hours to delivering outcomes, from competing on price to competing on quality and speed.

We don’t need to build our own large language models for this. We simply need to be the biggest users of AI models in the world. Tokenmaxxxing!

That is a harder change than adopting a new tool. But it is the change that determines whether India owns this opportunity or watches someone else build it.

The opportunity

For the past thirty years, India exported labour.

For the next thirty years, India can export leverage.

China became the world’s factory by combining people with machines and obsessing over throughput. India’s opportunity is to become the world’s services factory by combining people with AI and obsessing over speed & quality in every domain that runs the global economy.

The technology is increasingly becoming a commodity. The models will keep getting cheaper. Access will keep getting easier. What will remain scarce, genuinely scarce, is the ability to deploy that technology within a domain, with accuracy, at speed, in ways that hold up under regulatory scrutiny and institutional trust.

That is a services challenge. And India has spent thirty years building the foundations to lead it.

The question is not whether the opportunity is real.

It is whether we move fast enough to own it.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on Loop

Good background songs to listen as you read Fintech Inside: This week I’d been obsessing over Olivia Dean and her song Dive (Spotify / Youtube). She’s got a beautiful voice and songs are easy listens. They’ve helped me get through deep work sessions. I’m literally writing this edition with Ms. Dean in the background. Here’s a bonus, hour long, live jazz session of her singing.

✨ Call Outs

• [Travel] Mike Okay is my recent favourite travel vlogger. He’s currently in China.

• [Travel] Itchy Boots is my other favourite travel vlogger. She rides a two wheeler across the world. From her vlogs I got to know that China

• [Video] The Ridiculous Engineering Of Figma (ft. Figma Team)

• [Video] How Stripe Built Their New Website

• [Explainer] How a mechanical watch works

👋🏾 That’s All Folks

If you’ve made it this far - thanks! As always, you can always reach me via DM at osborne.vc/dm. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 109th edition of Fintech Inside.

I've been an investor in Indian startups for more than a decade. In that time I've watched the rupee grind from 64/$ to 83/$ - slow, predictable, almost boring. Every year, 3-4% weaker. Economists called it gradual depreciation.

The rupee hit 97 against the dollar this month. To put that in perspective: it took the rupee nine years to move from 64/$ to 83/. It took twelve months to move from 83 to 97/$.

It’s easy to worry, when the headlines call the rupee at 97/$ a crisis. The data tells a more nuanced story i.e. one where the rupee went from its most overvalued level in years to genuinely undervalued in just seventeen months. Both framings are partially true, and the distinction matters enormously if you're trying to build a business through it.

This edition is my attempt to cut through the noise. I'll walk through why the rupee fell, what the RBI has done and can and cannot do about it, and what the second and third order effects look like for startup founders. It's a long read but I think it's worth it.

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

97 and Free Fallin’: Why the Rupee is falling and what it means for startups

A disclaimer worth making: I'm an early stage investor, not an economist or currency analyst. Everything here is my own analysis and interpretation, not financial advice, not investment advice, just one practitioner trying to make sense of a complicated picture.

The rupee touched 96.9 against the dollar this month. That’s a record low.

Here’s what’s strange though, the rupee has been grinding lower for years. The 10-year data tells a clear story: 64/$ in 2015, 67 in 2016, 70 in 2019, 74 in 2020, 83 in 2023, 85 by end-2024. A slow, predictable decline averaging 3-4% a year. Economists call it gradual depreciation.

But 2025-2026 is different.

In the last twelve months, the rupee dropped from ~83/$ to ~97/$. That’s ~17%. More depreciation than the previous three years combined.

And that’s not all. The rupee weakened against most major currencies, not just the dollar. Against the Australian dollar, it fell 25%+. Against the Chinese yuan, ~20%. Against the Singapore dollar, ~15%. The only exception was the Japanese yen (3.5%) because the yen weakened alongside the rupee.

The rupee is Asia’s worst-performing currency in FY26 at 9.9%. It’s not a dollar story. It’s a rupee story.

Compare this to other emerging markets. Turkey’s lira lost 30% in 2023-24. Egypt’s pound fell 40% in 2024 alone. Nigeria’s Naira halved in 2024. These were crisis-level collapses driven by policy failures i.e. unorthodox monetary policy, capital controls, debt defaults. India’s 17% decline looks modest by comparison.

The difference is structural: India has $681bn in reserves, a Current Account Deficit (CAD) at 1.3% of GDP, and a coherent monetary authority. The rupee is weakening from a position of relative strength, not weakness. That makes this a correction, not a crisis. Either way, correction or crisis, it has implications for all of us.

My view is straightforward: It’s an acceleration and a convergence of forces that were building for years and hit all at once. The inflation differential. The trade deficit. US tariffs. The Iran conflict. An RBI that spent $53bn defending the currency and still couldn’t stop the slide.

For fintech founders and operators in India, this has direct impact on your P&L. Not in the distant future. Right now.

What Actually Happened

The structural baseline

The rupee doesn’t fall by accident. It falls because of three structural features built into India’s economy.

First, the inflation differential. Inflation measures how fast the purchasing power of a currency erodes domestically. India runs inflation at 5–6% annually. The US runs at 2–3%. That gap sounds small. Over a decade, it isn’t.

If a basket of goods costs ₹100 in India today and $1 in the US, and India inflates at 6% while the US inflates at 2%, then in ten years that basket costs ₹179 in India and $1.22 in the US. For the exchange rate to reflect the same real purchasing power, the rupee has to fall to compensate. Economists call this Purchasing Power Parity. The rupee adjusts whether policymakers want it to or not.

Signal to watch: When India’s Consumer Price Index (CPI) prints above 5% and US CPI prints below 3%, expect structural depreciation pressure. When the gap narrows i.e. India’s inflation falls, US inflation rises, the structural drift slows. The RBI’s 4% inflation target exists partly to manage this. Every time it misses upward, the rupee pays.

Second, the trade deficit. The trade deficit measures the gap between what a country imports and what it exports. India’s total imports was $979bn in FY26 and total exports was $860bn. That $119bn gap creates a dollar demand. Importers need dollars; they sell rupees. More supply of rupees, less demand and the price falls.

Every Indian importer - an oil company buying crude, a manufacturer sourcing machines, a retailer importing electronics, needs to convert rupees into dollars to pay their foreign supplier. That conversion creates selling pressure on the rupee. When imports consistently exceed exports, the selling pressure is constant. More rupees chasing fewer dollars means the rupee weakens.

Signal to watch: A narrowing deficit, driven by export growth, not import collapse, is positive for the rupee. India’s electronics export surge ($48bn) is the most important positive signal in the current data. If that trajectory continues, it could reduce the deficit over 3–5 years.

The current account deficit (CAD) is the broader version of this, it includes services and remittances alongside goods trade. India’s CAD is at 1.3% of GDP, up from 1.1% YoY, which is manageable. In 2013, when the rupee last faced a comparable crisis, CAD hit 5% of GDP. That context matters: India is running the same depreciation pressure from a structurally stronger position.

Third, the managed float. Most people assume exchange rates are either fixed (the government sets the rate) or free-floating (the market sets it). China had earlier pegged its currency to the US Dollar until 2005. But the rupee is neither. It operates under a managed float i.e. the RBI intervenes to smooth volatility without committing to a specific level.

When the rupee falls sharply, the RBI sells dollars from its reserves to buy rupees, creating artificial demand that slows the decline. When the rupee strengthens too quickly, which can hurt exporters, it buys dollars to slow the appreciation. The RBI is not fighting the market’s direction. It is managing the speed of travel.

Signal to watch: RBI reserve levels are published weekly. When reserves fall sharply, the RBI is intervening i.e. burning dollars to defend the rupee. That is unsustainable beyond a certain point. India’s reserves dropped from $705bn in September 2024 to $681bn as of 22nd May 2026. That $24bn drawdown tells you how hard the RBI fought. When reserves stabilize or rebuild, intervention pressure has eased. When they fall consistently month-on-month, the RBI is in a losing battle.

The 40-year chart tells the story. Each crisis resets the floor lower. 1991, 2008, 2013, 2018. The pattern is consistent. The direction is one way. The managed float explains why the rupee falls in steps rather than in a smooth line - the RBI absorbs shocks and releases them gradually.

The REER context adds another layer. India’s Real Effective Exchange Rate (REER), a trade-weighted measure adjusted for inflation, stood at 108 in November 2024 (peak overvaluation). By April 2026, it had fallen to 89.7. The rupee went from most overvalued in years to undervalued in 17 months. This is what a real correction looks like, not a disaster.

The REER is an important currency metric most people should look at. It adjusts the nominal exchange rate (the 97/$ number you see on your phone) for inflation differentials across all of India’s trading partners, weighted by trade volume. It tells you whether the rupee is genuinely cheap or expensive relative to fundamentals, not just relative to the dollar.

A REER above 100 means the rupee is overvalued relative to its historical average i.e. Indian goods are expensive for foreign buyers, imports are cheap, exports are uncompetitive. A REER below 100 means the rupee is undervalued i.e. Indian goods are cheap globally, exports become more competitive, imports get expensive.

This is the number that reframes everything. The 97/$ rate feels like a crisis because it’s a record nominal low. But the REER says the rupee is now cheap by historical standards which means Indian exports are more competitive, the current account pressure should ease, and the structural case for further sharp depreciation is weaker than the headlines suggest. This is what a real correction looks like.

Signal to watch: REER data is published monthly by the RBI. When REER is above 105, the rupee is expensive and depreciation pressure is building beneath the surface. When REER is below 95, the rupee is cheap and the structural correction has largely played out. Where it sits today at 89.7, suggests the worst of the adjustment may already be behind us, even if the nominal rate drifts further.

The acute shock

What changed in 2025–26 wasn’t the structure but the shock that landed on top of it.

US tariffs. The US imposed 50% tariffs on Indian goods, the sharpest trade restriction since the 1990s. That spooked foreign investors. India is reportedly in the final leg of negotiating a better trade deal.

FPI outflows followed. Foreign portfolio investors pulled ₹1.6 lakh crore (~$18bn) out of Indian equities in 2025 alone. One of the largest annual outflows on record. Every exit meant selling rupees and buying dollars.

The Iran conflict pushed oil past $110 a barrel. India imports 88% of its crude oil. Every $10 rise in crude adds ~$15bn to India’s annual import bill, which in turn gets passed onto customers. More dollar demand. More rupee pressure.

The RBI fought back. It sold $53bn in the spot market in FY26, a record. It held $103bn in negative net forward positions. It intervened daily at an estimated $1bn. It even restricted Indian banks from trading in the offshore NDF market. None of it reversed the slide.

Rate cuts removed support. The RBI cut the repo rate four times in 2025 after retail inflation dropped to 0.25%. Lower yields meant lower carry trade appeal. Foreign investors looking for yield found less reason to stay.

The result: the rupee fell from ~83/$ to ~97/$ in twelve months. A 17% depreciation against a historical baseline of 3–4%.

What the RBI does next and what it means for us

The question I hear in every conversation: “Can the RBI stop this somehow?”

Short answer: it can slow the fall. It cannot reverse the forces that drive it.

Rate hikes are the big signal to watch. The next RBI Monetary Policy Committee decision is June 5. The RBI cut rates four times in 2025. It may now pause or reverse. A 25-50bps hike would restore some carry-trade appeal and slow the rupee’s slide. But it also raises your cost of capital. If you run a lending fintech, this directly compresses your net interest margins. If you are a cross-border platform, higher rates tell you the RBI is prioritising the currency, which is good for your business.

NRI deposit schemes could bring short-term relief. The 2013 playbook: the RBI raised $34bn through FCNR deposits by offering NRIs premium dollar-denominated rates. It could do the same now. Current FCNR rates offer up to 5.2% tax-free on USD savings. A new scheme could inject $10-20bn of dollar supply into the system quickly, stabilising the rupee without rate hikes. For fintechs serving NRI remittances, that means more volume flowing through formal channels.

Reserve depletion is a wasting asset. The $53bn spent in FY26 bought time. At $110 oil, the monthly burn is $20-30bn. Reserves are $681bn and dropping. The RBI can sustain this for another quarter, maybe two. But not indefinitely. As reserves drain, the psychological buffer shrinks. Markets notice.

What the RBI might not do. It might not defend a specific INR level. It will not raise rates aggressively just for the currency. The MPC has consistently prioritised growth over rupee stability. The rupee can drift to 100/$, slowly, not suddenly. That is the managed float working as designed.

Bottom line for founders: The policy toolkit can slow the rupee’s fall and prevent a disorderly crash. It cannot change the structural drift. Assume the rupee stays in the 95-100/$ range for the next 12-18 months. Anything better is a tailwind. Plan for worse.

The other side

A reasonable person could disagree with this framing. Let me state the strongest version of that argument.

“The rupee’s fall is temporary. It stabilises when the Iran conflict de-escalates and oil prices normalise. The RBI has $681bn in reserves, plenty of firepower. India’s macro fundamentals i.e. GDP growth at ~6.5%, CAD at 1.3% of GDP, record exports and so on, are nothing like the 2013 taper tantrum when CAD hit 5%. The rupee returns to 85/$ within two years.”

I think this argument is wrong, but not unreasonable. Here is why the structural argument holds more weight.

The structural drivers predate the Iran conflict and will outlast it. The inflation differential. The trade deficit. The 86% of India’s trade still invoiced in dollars. These don’t go away when oil prices drop.

What the oil shock and tariffs did was accelerate a trajectory already in motion. The question isn’t whether the rupee stabilises. It’s where the new floor settles. ING’s May 2026 forecasts show the rupee recovering slightly to 94/$ over 12 months, a 4% improvement from current levels but still well above 85/$. BMI expects 95/$ by end-2026. MUFG and Aberdeen warn 100/$ is possible if Strait of Hormuz tensions escalate.

What would change my mind? Two things. First, if India’s export diversification continues i.e. electronics exports at $48bn, smartphones at $30bn, and reduces the structural dollar demand for imports. Second, if rupee invoicing of trade expands beyond the current 2–3% share. Both are positive trends. Both are early.

What this means

Layer 1: Consumers: A cheaper rupee hits consumers through three channels and they all compound.

1. Inflation channel: India imports 88% of its crude oil and 60% of its edible oil. Every 10% rupee depreciation adds roughly 1-1.5% to headline inflation. That flows through fuel, transport, food, and manufactured goods. The consumer feels this at the petrol pump, the grocery store, and the monthly electricity bill. Discretionary spending gets squeezed first.

2. Imported goods channel: Electronics, smartphones, gold, and machinery all get costlier. A MacBook that cost ₹1.2 lakh at 83/$ now costs ₹1.4 lakh at 97/$. Gold, a traditional savings vehicle in Indian households, becomes more expensive to buy. This shifts consumption patterns toward essentials and away from aspirational purchases.

3. Savings channel: Rupee savings held in cash or fixed deposits lose purchasing power relative to global goods. This pushes savers toward inflation hedges i.e. gold, real estate, or dollar-linked instruments. We are already seeing this. Gold demand in India rose 15% in FY26 despite record prices - as discussed in Edition #102.

Layer 2: Consumption: The chain is predictable. Higher inflation reduces real disposable income. India’s per capita income grew at roughly 9.5% nominally over the last decade, but strip out 5% average inflation and real income growth is closer to 4.5%. That 4.5% was already thin. A rupee that adds 1-1.5% to headline inflation doesn’t just slow real income growth, it cuts it nearly in half. The sectors that feel it first are travel, dining, entertainment, automobiles, and consumer durables. Discretionary spending is the first casualty of a currency correction.

This matters for fintech because consumption drives transaction volume. UPI payments, credit card spends, BNPL usage, all correlate with consumption. If consumption slows, payment volumes dip. The fee income that fintechs earn per transaction shrinks at a time when they need it most.

For lending fintechs, the picture is worse. Consumption slowdown means lower credit demand. At the same time, the RBI may raise rates to contain inflation. Higher rates mean higher EMIs, which further suppress demand. It’s a contractionary spiral.

Layer 3: Credit: A weaker rupee impacts credit markets in four ways.

1. Higher rates: If rupee depreciation pushes inflation above the RBI’s 4% target, the central bank pauses its rate-cutting cycle or reverses it. This raises the cost of funds for every NBFC and bank. As of April, 2026, Indian banks already raised lending rates by 10 basis points to 8.50%. Credit growth will slow.

2. Higher NPAs: Businesses that import raw materials or carry dollar-denominated debt, face higher costs and compressed margins. Some will default. Retail borrowers face higher EMIs and higher living costs. Delinquencies rise.

3. Tighter underwriting: In response, lenders tighten credit standards. Fewer loans get approved. The credit-to-GDP ratio, already below its trend, slips further. This hit to financial inclusion is real especially for the thin-file borrowers fintechs have been targeting.

4. Secured loan boom. One side effect: as gold prices rise (driven by both global demand and rupee weakness), gold loan providers like Muthoot and Manappuram benefit. Loan-to-value ratios improve. More households monetise their gold holdings to manage cash flow. Couple that with mutual fund SIPs growing to record levels of $3bn monthly, investors will start leveraging their MFs to take secured asset-backed loans as ell. This is one of the few credit segments that will get stronger.

Layer 4: startup ecosystem: Beyond fintech, a weaker rupee reshapes the entire startup landscape.

1. Dollar costs rise for everyone: Indian startups run on global cloud providers, AI labs, Stripe, SaaS providers and more - majority priced in dollars. A 15% rupee depreciation means a 15% jump in cloud infrastructure costs. Burn rates could go up. Runways shrink.

2. Foreign talent gets more expensive: Startups hiring senior talent from the US or Europe in dollar terms face higher INR costs. Remote-first companies that pay in USD see their wage bill expand in rupee terms. Obviously this impacts Indian domiciled startups.

3. Fundraising math changes: Foreign funds compute returns in USD. A 4% annual rupee drift means every fund needs to earn a 4% higher nominal return in INR to deliver the same USD return. This doesn’t kill the India thesis, but it tilts the playing field toward dollar-earning business models (cross-border SaaS, digital exports) and away from domestic-consumption-dependent ones (hyperlocal delivery, domestic BNPL, D2C brands with imported inputs).

4. The IPO window narrows: A weaker rupee often correlates with FPI outflows. FPIs outflows from the equity market crossed ₹2.25 lakh crore ($23.7bn) as of May, 2026, surpassing the ₹1.66 lakh crore ($17.5bn) withdrawn during all of 2025. This means lower valuations in public markets. Domestic IPOs get postponed. That delays exit timelines for early-stage investors. The liquidity crunch in the startup ecosystem lasts longer.

The Second-Order Effect

If the rupee continues toward 100/$, the RBI faces a classic “trilemma”. It cannot simultaneously manage the currency, control inflation, and keep rates low. Something breaks.

The most likely outcome: the RBI prioritises inflation control over growth. Rates go up. Credit slows. Consumption weakens. The startup funding winter, which looked like it was thawing, gets another season.

The less likely but more dangerous outcome: the RBI lets the rupee find its level while keeping rates accommodative. That risks imported inflation becoming entrenched. Inflation expectations de-anchor. The cost of fixing it later, through sharper rate hikes and deeper recession, is much higher.

What this means for founders and investors?

The rupee at 97/$ is not a single event. It is the visible surface of forces that have been building for years i.e. the inflation gap, the trade deficit, the dollar dependency, and accelerated by shocks that may or may not persist.

The founders who navigate this well will not be the ones who wait for the rupee to recover. They will be the ones who build assuming it doesn’t.

That means a few concrete things. If your cost base is in dollars, model your burn at 100/$, not97/$. The psychological buffer of “it can’t get much worse” is not a treasury strategy. If your revenue is in rupees and your customers are discretionary consumers, assume transaction volumes soften over the next two quarters as real incomes get squeezed.

The rupee correction has largely played out in REER terms. The structural drift hasn’t. Those are two different statements, and conflating them is where most analysis goes wrong. The worst of the acute currency shock may be behind us.

We might have to shift our focus on the other external shocks to the market i.e. the Iran war potentially goes on for longer which could cause oil prices to spike further, spiralling into higher inflation, lower consumption, lower credit growth and higher defaults. Oh and let’s not forget the AI causing higher expenses for enterprises which might force them to cut costs by laying off employees, which will have its own ripple effects.

We can’t seem to catch a break since 2020. The slow grind is permanent. Lage raho!

I want to hear the counterargument. If you think I’ve got this wrong, if you believe the rupee strengthens from here, I’d love to hear from you. I’ll publish the best responses in a future edition.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: Today’s title is inspired by Free Fallin’ popularised by John Mayer (Youtube / Spotify). This rendition is performed at the Nokia Theatre for his popular Where the Light Is Tour. If you haven’t watched the live performance video of that concert, I’d recommend it. Soul soothing.

Call outs

Links to non-fintech content I found super interesting:

1. [REPORT] India VC Exit Performance

2. [ARTICLE] DeepSeek’s 10 trillion USD grand strategy

3. [REPORT] SpaceX S-1 Breakdown

4. [VIDEO] The Geometry of Luck by Soleio at the Tokyo Design Forum

5. [VIDEO] How Jazz and Improvisation Works

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me via DM at osborne.vc/dm. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 108th edition of Fintech Inside.

Every pilot who has ever carried passengers has already crashed a plane.

Not on a real runway but in a simulator. Over and over, across hundreds of hours, in conditions designed to break them. Engine failure at take-off. Wind shear on approach. Total hydraulic loss at 30,000 feet. The aviation industry decided long ago that the only acceptable place to make fatal mistakes is in a room where nothing is real.

So far, the only way for the financial services industry to simulate risk management was in a spreadsheet. But the product was test in the wild.

A bank launching a new credit product runs a pilot, a few thousand real customers, real money, real consequences, and calls it testing. A fraud team discovers a new mule network pattern when the mules have already moved the money. An insurer calibrates its AI claims engine on live claims, paying the price of every wrong threshold in real time.

We carry passengers on the first flight. The question this edition explores: what changes when financial services industry can simulate risk management on synthetic users?

LLMs make that possible. But is there an opportunity for fintech startups to build?

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

Flight simulator for financial services

What world models mean for risk management in financial services and why simulation-first strategy could be a competitive advantage

When I was in college, I wanted to build a humanoid robot.

Not fix one. Build one. The whole thing. A machine that could walk, think, adapt, and exist in the world the way humans do. I was obsessed with having a never-tiring companion to do things for me, maybe I was just lazy. The problem was I was never particularly gifted at physics or mathematics, which, it turns out, are fairly important if you want to build a humanoid robot.

So here I am, a venture capital investor. But the fascination never really left.

What I did not know back in college was the hardest part of building a robot was not only the hardware. It was the training too. How do you teach a machine to navigate a world it has never seen? To react to a chair placed where no chair was before? To recover from a fall it has never taken?

The old answer was deterministic programming i.e. code every move, script every response. As processing power improved, thanks to Moore's law, and neural network approaches became commonplace, training improved, but it was still slow, brittle and tough to scale.

With the discovery of the transformer architecture, large language models can be trained on massive amounts of data, making it possible to train AI agents that can generalise to new situations.

Over the past two years, two events brought back my fascination for robots with full force:

1. The first was watching Google DeepMind’s work on world models, training robotic systems entirely inside synthetic environments, exposing them to millions of simulated scenarios, edge cases, and adversarial conditions, before real-world deployment.

   

2. The second was NVIDIA’s Omniverse and Cosmos platforms - a physics-accurate simulation infrastructure purpose-built for training autonomous systems in synthetic reality at scale.

Enter world models.

Today, Google DeepMind’s Genie 3 generates entire photorealistic, interactive environments from a text description. A robot trains inside that synthetic world - bumping into walls, grasping objects, navigating edge cases, for thousands of hours. It fails safely, learns continuously, and emerges into the real world already prepared for scenarios it has never physically encountered.

Waymo took this further. In February 2026, Waymo adopted Genie 3 to build its “Waymo World Model.” The Waymo Driver had logged 200M real autonomous miles, but billions more in simulation. The result: a self-driving car that crashed a million times in a simulator so it crashes near-zero times on real roads.

I find this approach of running simulations super relevant to financial services. Not as a metaphor but as a blueprint.

My thesis is that better risk management in financial services will come from firms that simulate before they ship. Running thousands or millions of synthetic users, transactions, and fraud scenarios through a digital twin of their financial product before it touches real customers.

The same way a robot trains in a synthetic world, a financial product can train in a synthetic economy.

What exactly is a world model?

Put simply: a world model is a digital sandbox that simulates how actual people, customers, fraudsters, borrowers, regulators, each with varying parameters of income level, locations, spending behaviour, devices, and other characteristics, will behave in specific scenarios. Simulation is running thousands of those scenarios through the sandbox to see what happens before you commit real capital.

Think of it like this: if you open a new store on a busy Mumbai street, you do not immediately stock inventory and hire staff. First, you study foot traffic. You observe competitor pricing. You understand peak hours. Then you experiment with different displays, promotions, and product placements. You simulate the economics of that store in your head, or better yet, with real data. A world model digitises that process at scale and speed. It can simulate 10,000 “stores” across 10,000 “streets” in minutes. The only difference is the sandbox runs in silicon instead of in your head.

A pattern across industries: simulate the human, not just the system

The usefulness of simulation extends far beyond finance. Consider three examples.

War games: RAND Corporation and military strategists have used simulated conflicts for decades, not to predict outcomes, but to surface second-order effects. A troop movement triggers a supply chain response that triggers a diplomatic reaction. A spreadsheet cannot capture that cascade. Only a simulation of human decision-making under stress reveals it.

Pandemic modelling: In October 2019, Johns Hopkins’ Event 201 simulated a novel coronavirus pandemic. It predicted supply chain breaks and hospital crises. When COVID-19 arrived months later, the simulation proved failure modes were knowable in advance.

Flight simulators: Every pilot trains for emergencies they have never faced. When engine failure happens at 30,000 feet, the pilot does not reason from first principles. They react from muscle memory built in a simulator.

Google’s own DeepMind pushed the frontier by famously training AlphaFold to predict protein structures and then beating the world champion at the game of Go by building AlphaGo. DeepMind achieved all this by approaching the problem as a game.

Common thread: all three simulate human behaviour under stress, not just physics or probabilities. That is what makes the LLM-powered approach different. It models people i.e. what they do, how they lie, who they blame, not just math.

Why should a financial product train in a simulator?

In aviation, no pilot carries passengers without hundreds of hours in a flight simulator. They practice engine failure, wind shear, bird strikes, emergencies most will probably never face in real life. When they do face them, they have already lived the scenario. The simulator collapses years of rare-event exposure into hours.

Today financial services operates in reverse.

A fintech launching a new BNPL product tests it on real customers with real money. A bank rolling out a new fraud detection system learns what it missed after the fraud happens. An insurer deploying an AI claims engine calibrates thresholds through live claims, paying the cost of every false positive and false negative in real time. Sure, it’ll be in “closed beta” or pilot mode to track vintage performance, but parameters change quickly when the rubber hits the road.

We test products on the equivalent of live passengers.

Simulation is not new to finance. It just simulated the wrong thing.

Banks have run simulations for decades. Monte Carlo models for portfolio risk. Value-at-Risk calculations for market exposure. Stress tests for capital adequacy. These simulate numbers, interest rate movements, default correlations, market volatility.

But, as they say, the best fiction is written in Excel.

They do not simulate people, scenarios, or behavioural patterns.

That is the limitation LLMs now address. A growing body of research uses large language models as “synthetic participants” i.e. generating behavioural responses that approximate real user decisions. A 2026 study validates LLM simulations as behavioural evidence, showing they can reproduce known economic patterns. JPMorgan Chase’s 2026 Emerging Tech Trends report explicitly calls out “AI-powered simulation” with synthetic users as a key trend i.e. synthetic personas and continuous attack simulations replacing real-world testing (Page 55 onward).

The shift is from simulating macro trends/markets to simulating users.

Applications in financial services

1. Credit underwriting

A fintech building a BNPL product faces an impossible choice: underwrite conservatively (losing customers) or loosely (accepting defaults), no middle path without data.

With a world model, a company could simulate 10,000 synthetic borrowers. Varying income levels, spending patterns, repayment behaviours, shock scenarios, job loss, medical emergency, economic downturn. Even the synthetic profiles could be auto generated by LLMs or from past data.

The simulation could reveal the default curve before the first customer checks out. The team knows exactly where risk thresholds should sit because they’ve watched 10,000 versions of their borrower behave across every stress scenario.

2. Fraud detection and mule accounts

The Indian government and RBI have flagged mule accounts as a systemic risk. The RBI Innovation Hub built MuleHunter.AI, an AI tool detecting mule accounts with 85%+ accuracy, now rolling out across 15+ banks.

But today’s fraud patterns shift tomorrow. Mule networks evolve. Social engineering tactics adapt. A static detection model chases yesterday’s crime.

A simulation environment changes this. It generates thousands of synthetic mule networks, testing detection systems against novel patterns before fraudsters deploy them. Banks stress-test their fraud models against threats that do not yet exist. The RBI Innovation Hub could open this as a shared simulation layer for all banks, a collective immune system trained on synthetic attacks.

3. Insurance claims

An insurer deploying an AI-driven claims approval engine faces a calibration problem. Approve too aggressively and fraud leaks through. Reject too aggressively and legitimate customers get denied, damaging trust.

Simulate 50,000 claim scenarios, legitimate, fraudulent, borderline and more, and the optimal threshold reveals itself. The insurer enters the market with a calibrated system, not a guessing game.

Three more places simulation bites

Credit, fraud, and insurance are the obvious ones. These three are less obvious.

4. KYC and identity verification: Synthetic identity fraud, where fraudsters blend real and fake data into convincing new identities, costs financial services millions annually. A simulation environment can generate synthetic identity profiles that push the boundaries of current KPIs. Banks can test their verification pipelines against emerging fraud strategies before real criminals deploy them.

5. Payment routing and transaction risk: Every payment network decides in milliseconds whether to approve or block a transaction. A simulation can generate millions of synthetic transaction patterns, including evolving fraud tactics, to find the optimal balance between friction and loss. Lower false declines. Faster settlements. Higher conversion. All before going live.

6. Insurance product design: Insurers set premiums based on historical loss data. But the world is not static. Climate change. New vehicle technologies. Shifting demographics. All of these change the risk equation. A simulation can model how a new product or pricing change performs under dozens of future scenarios, not just past conditions.

The best part is, an LLM/world model can literally generate scenarios and edge cases that we could not possibly imagine. Talk about dealing with the unknown unknowns!

Simulation is an illusion of accuracy which captures historical patterns without understanding the underlying causal economic drivers

The strongest opposing view deserves a fair hearing.

“Simulations are only as good as their assumptions. A synthetic user is not a real user. LLMs hallucinate. Behavioural simulations encode the biases of their training data. You might simulate 10,000 scenarios and still miss the one that breaks your product. The 2008 mortgage crisis happened because sophisticated risk models failed to capture correlated defaults. A simulation would have failed similarly.”

This is worth sitting with, especially the 2008 point, because it sounds like the kill shot. It isn’t.

The models that failed in 2008 were simulating correlated macro risks: interest rate movements, default probabilities, market volatility. They were, in effect, very sophisticated Excel sheets. What it did not simulate was human behaviour, the mortgage broker who misrepresented income, the borrower who defaulted the moment equity turned negative, the rating agency analyst who stopped asking questions. The cascade was behavioural.

That is the gap world models simulation addresses. The shift isn’t from bad models to better models. It’s from simulating markets to simulating people. 2008 doesn’t argue against this approach, it argues for it.

No simulation replaces reality entirely. A flight simulator doesn’t produce a perfect pilot. But a pilot who trained in one handles their first real emergency better than one who didn’t. A BNPL product tested against 10,000 synthetic borrowers could have a better risk profile than one tested against 100 real users in a regulatory sandbox.

The bar is “better than current practice.” But unlike the 2008 models, this approach is actually designed for the problem that broke them.

What this means for founders and regulators

India has a head start. The RBI already operates a regulatory sandbox where fintechs test products with real data. The Inter-operable Regulatory Sandbox (IoRS) spans RBI, SEBI, IRDAI, PFRDA, and IFSCA, allowing products that cross regulatory boundaries to be tested under one roof.

This is a strong foundation. But it requires a partnership between a licensed firm and uses real data at limited scale with a narrow scope.

A synthetic world model could complement it: let firms stress-test products against millions of synthetic users and scenarios before they enter the sandbox. The sandbox becomes the final exam, not the first test.

The implications flow in three directions.

For founders: A build opportunity. The startup that creates the “financial world model” i.e. a simulation platform where financial firms test products against synthetic users, becomes the AWS of risk testing. The first fintech to embed simulation into product development ships faster, fails cheaper, defends risk models to regulators with data instead of promises.

For institutions: Simulation transforms the cost of innovation. Launching a new credit product today requires more than a year of pilot testing, internal roll outs, regulatory review, and a cautious rollout. A simulation environment compresses this to weeks. Banks and NBFCs that build internal simulation capabilities out-innovate those that do not.

For regulators: The regulatory sandbox 2.0 includes a shared simulation layer. The IoRS framework could offer synthetic testing environments as a public good, letting firms prove product safety through simulation before touching real customer data. The RBI’s MuleHunter.AI shows the blueprint: build a shared tool, offer it free to all banks, improve it together.

The investment play: This is infrastructure that will be used on an ongoing basis. I don’t think simulation will be a build and sell once model. It’ll require continuous refinement.

Startups that build simulation platforms for financial services sit at the intersection of three mega trends: AI, risk management, and the digitisation of finance.

Think of it as the “AWS of risk testing”, a platform layer that every fintech, bank, and insurer will need but won’t want to build from scratch.

The business model is SaaS - priced by scenario volume, with a consumption layer for compute-heavy stress tests. Think per-simulation-run pricing at the low end, annual platform contracts for institutions running continuous testing cycles.

The moat is data: Every fraud pattern, default curve, and behavioral anomaly that passes through the simulation becomes training signal. That data feedback loop compounds in the flywheel of better simulation → more customers → more data → better simulation.

That compounds in a way that a one-time software sale doesn’t. The firms that get in early lock in a data advantage that’s genuinely hard to replicate. The cost structure is real. Inference at scale isn’t cheap, but margin expands as the models improve and the dataset deepens.

This is a business that gets cheaper to run as it gets better. That’s the kind of unit economics worth building toward.

I think AI-native startups i.e. startups that begin with the synthetic user layer and expand outward into risk management, is where I think the outsized returns live. A startup that builds a clean simulation layer, independent of any single financial product, can become the default testing environment for the industry.

What the future of risk management holds for the industry?

DeepMind proved world models work. Waymo proved they scale. JPMorgan Chase proved the largest financial institution on earth is thinking this way. The pieces are all there.

The company I’m watching for looks like this: it starts narrow, building for one vertical, one problem, probably fraud or credit underwriting, and builds the simulation layer so well that it becomes the default testing environment for that problem across the industry. It doesn’t try to be a bank or a risk engine. It sells picks and shovels: run your product through our synthetic world before you ship it to real customers.

The wedge is regulatory. The first firm that walks into a SEBI or RBI sandbox conversation with a simulation report - “here are 50,000 synthetic scenarios we stress-tested before touching a single customer”, could potentially change the conversation with regulators permanently. That becomes the new standard of care. Every firm that follows has to meet it.

Within five years, launching a financial product without simulating it against synthetic users will seem as reckless as flying a plane without a simulator.

The question is which founders build the infrastructure, and which institutions adopt it first.

If you could simulate 10,000 versions of your customer before your next launch and know exactly where the defaults, fraud, and failures would come from, would you still launch the same way you do today?

I want to hear why I’m wrong. Simulation has limits. But so does testing with real money and real customers. Which ones are you willing to live with?

Would love to hear your views.

If you’re a founder working on any of the use cases above, credit underwriting, fraud detection, KYC infrastructure, or regulatory tech, I’m actively looking at this space. Let’s talk.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: Never thought I’d get hooked onto AI music, but this youtube channel (Shanti Instrumentals) is on loop for me playing in the background all day as I work. It’s calming, it helps me focus and I just can’t believe it’s made with AI. Blows my mind.

Call outs

1. [VIDEO] What does the $ actually stand for? (I never really thought to question where the dollar sign comes from, this video explains the fascinating history).

2. [VIDEO] The World’s First Smart Home (built in the 19th century by William Armstrong in Cragside, UK).

3. [ARTICLE] The Trap India Faces Isn’t Just Geopolitical Anymore (related to the INR/USD rate, but goes deeper).

4. [ARTICLE] The Mathematical Reason Most People Never “Make It”

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me via DM at osborne.vc/dm. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 107th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

Every quarter, I sit down with every earnings call transcript and analyst presentation from the companies that run India’s financial system: the banks, the NBFCs, the insurers, the brokers, the fintechs. I read them as a practitioner trying to understand what is actually happening inside the financial services industry.

What are the best operators in the country seeing that the market hasn’t priced in yet? Where are the structural shifts hiding inside the quarterly numbers? What does the behaviour of 115M Bajaj Finance customers base tell us about where consumer credit is going? What does Groww’s ₹900 CAC tell us about the cost of acquiring India’s next investor?

This quarter, I covered 23 companies. 160 slides. 100+ insights.

The Q3 FY26 India Financial Services Insights Report is now out.

View the full report now

Why I write this report

Fintech Inside started as a newsletter to make sense of India’s financial services industry, not for traders, but for founders, operators, and investors who are building inside it.

This report is the most concentrated expression of that mission.

Every quarter, I synthesise what India’s most important financial institutions are telling their shareholders and translate it into signal for those of us who are building companies, deploying capital, or advising founders in this space.

The goal is simple: give you a map of where the white spaces are, where the blind spots are, and where the real opportunities are forming, before they become consensus.

What’s in this quarter’s report

Here are five things that stood out to me from Q3 FY26:

1. Every product is now a deposit funnel

India’s banks have quietly re-architected their entire product stack. Credit cards, auto loans, home loans, even payment devices, every product is being reframed as a deposit acquisition and relationship retention tool. NIM compression is structural. The race is no longer for loan market share. It’s for low-cost liability primacy. Founders building in lending infrastructure, payments, or banking SaaS need to understand this shift. It changes what banks will pay for.

2. AI has crossed from pilot to P&L

This is the quarter where the language changed. Nobody is “testing AI” anymore. Bajaj Finance has 800+ autonomous agents deployed across 24 business areas. AU Small Finance auto-resolves 60% of fraud alerts. Angel One’s engineers ship 37% AI-generated code. Pine Labs generates 20% of new product code with AI. These are operational metrics disclosed under analyst scrutiny. The cost-per-unit-of-output curve is bending in real time.

3. The unsecured credit cycle is behind us

Every major lender is simultaneously quality-upgrading or shrinking their unsecured book. Capital is rotating into gold loans, SME, business banking, and secured vehicle finance. When all the best operators make the same credit call at the same time, that’s a structural signal, not a coincidence.

4. India’s mutual fund industry has crossed a threshold

Monthly SIP inflows hit an all-time high of ₹31,000 crore in December 2025. The SIP asset base is now ₹17tn ($183bn), over 20% of industry AUM. The industry has crossed the threshold where redemption risk is permanently cushioned by perpetual monthly inflows. For anyone building wealth management, distribution, or savings products, the compounding dynamics have fundamentally changed.

5. The market has drawn a hard line on fintech

Paytm’s redemption arc is complete. Groww continues to outperform. Pine Labs and MobiKwik are bleeding below Nifty. The message from public markets is unambiguous: path-to-profitability is the only valuation currency. GMV storytelling is dead. If you’re raising fintech capital in 2025, the only question that matters is how many months to contribution positive.

Get the full report

The full 160-slide report, covering 23 companies across banking, NBFCs, insurance, wealth management, and fintech, is available to view now.

View the full report now

As always, this report is for educational and informational purposes only. It does not constitute investment advice. The author may hold positions in companies mentioned.

Preparing this report took a lot of time and effort. If this was useful to you, forward it to one founder or investor who should be reading it. Better yet, provide your feedback and subscribe to the newsletter.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: Ease into the weekend with a breezy track: Following the Sun by Super-Hi and Neeka (Spotify / Youtube). Good background music to listen to while you’re reading through this report.

✨ Call outs

1. [Newsletter] The folks at Popvax are bringing generative AI to cancer research and treatment. This post follows the story of Yash Bindal’s cancer journey and how Popvax is helping him find a cure specific to Yash and his cancer type.

2. [Photos] The team at NASA released stunning images clicked by the Artemis II astronauts. This collection is special and maybe its worth watching this video to understand the context behind the images. It’ll help you appreciate the images more.

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me via DM at osborne.vc/dm. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 106th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

Every quarter, I spend a couple of weeks reviewing investor presentations and analyst call transcripts of public financial companies to understand how user behaviour is changing, where the industry is headed and what these large institutions are doing to navigate the future. I published a report covering the period Q1 FY26. Would you want to see the report for Q3 FY26? Let me know, will publish it.

When reviewing the data for Q3 FY26, I started noticing a trend - default rates at record lows, provision buffers growing, lending mix increasingly conservative, unsecured credit in free fall and more. Almost as if… Indian financial institutions forgot how to take risk.

But then the West Asia conflict happened. And suddenly it seems like this risk aversion might be what shields the Indian economy from economic shocks.

The irony is uncomfortable. In normal times, I'd criticise these numbers, institutions sitting on record capital buffers, retreating to gold loans, starving MSMEs of credit. But with oil at $100 and supply chains fracturing across the Strait of Hormuz, the most overcautious financial system in decades might also be the most resilient one. That tension is what this edition is about.

This edition is best read with your favourite cuppa. Let’s get into it!

Happy new financial year to those who celebrate!

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

India’s financial institutions forgot how to take risk

It might be what shields India in the short term

Like most of you, I’ve been glued to the news from West Asia. The Iran conflict isn’t abstract for me, I have friends, family and founders across countries in the region, and the first few days were spent checking in on people.

But once the immediate personal concerns settled, the fintech investor in me kicked in. I was following the politics from a distance, but the economic impact? That’s my lane for curiosity. And the question that kept nagging me was simple: How can this war impact us in the Indian fintech ecosystem?

So I did what any venture capitalist would do - I put on my 2-weeks-only armchair economist hat and went deep. I pulled Q3 FY26 investor presentations from every major Indian bank and NBFC. I cross-referenced them with RBI’s long-term asset classification data going back to 1997. I read through CareEdge’s latest credit and deposit research.

I watched Neelkanth Mishra’s sobering analysis on the Strait of Hormuz (he’s the Chief Economist at Axis Bank and one of the sharpest macro thinkers in the country). I’ve listened to every Odd Lots podcast’s Iran coverage, interviewing experts going deep on topics across fertilisers, ammunitions, oil, logistics, insurance and more.

What I found was interesting. The Indian financial system is the strongest its been in decades, by every metric. But the more I dug into the data and the narratives, a different story emerged. A story about a financial system that has become so obsessed with safety that it may have forgotten what it’s supposed to do i.e. take risk.

And the irony of it all? That very same excessive caution, the caution I’d normally criticise, might be the thing that protects us from what’s happening in West Asia.

Share

The case that banks are playing it too safe

Bad loans have practically vanished. Banks make money by lending - ~85% of banking income is interest income. I’ve done a whole deep dive in Edition 65 covering the revenue split of banks in India.

Sometimes borrowers don’t pay back. When a borrower stops paying for 90+ days, that loan becomes a “Non-Performing Asset” or NPA, a bad loan sitting on the books. The percentage of loans that go bad is the Gross NPA ratio, think of it as the bad loan (default) rate.

In 2018, India’s system-wide Gross NPA (GNPA) ratio peaked at 11.2%. For government-owned banks (Public Sector Banks), it was a catastrophic 14.6%. For context, anything above 4% is considered really bad. Above 7-8% is a crisis. India was deep in crisis territory.

Today? The system-wide GNPA has collapsed to 2.2%, the lowest since 2008. Public sector banks have gone from 14.6% to 2.6%. Private banks sit at 1.8%.

The NPA chart from RBI data tells this story visually, three decades of bank defaults, and the current numbers are at the absolute floor.

A 2.2% default rate sounds great. It means 97.8% of all loans are being paid back on time - maybe some delinquencies (delays), maybe debt collections platforms are showing better collections efficiency.

The Net NPA numbers are even more extreme. Net NPA strips out the provisions banks have already set aside for bad loans, it’s the actual residual risk on the books. Here’s what Dec, 2025 (Q3 FY26) looks like for India’s top banks:

Net NPAs of 0.3% to 0.4%. Even the NBFCs (Non-Banking Financial Companies, lenders that operate outside the traditional banking structure) are showing extreme discipline: Bajaj Finance at 0.44% net NPA, Sundaram Finance at 1.06%, HDB Financial at 1.25%.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

It seems like India’s financial system has become risk averse - too terrified of making a bad loan.

A quick caveat on the percentages. The last time system-wide GNPA was at 2.2% was in 2008 and 2010-11. But back then, the total loan book was roughly ₹20 lakh crore ($460bn at 43.3 USDINR) . Today it's nearly ₹198 lakh crore ($2tn at 94.0 USDINR), 10x larger in INR terms. So 2.2% today translates to roughly ₹4.4 lakh crore in absolute bad loans, compared to ₹68,000 crore in 2008-09.

The percentage is at a historic low, but the rupee value of bad loans is still substantial. What's different this time is the provision buffer sitting underneath, banks have set aside 75-92% of those bad loans in cash reserves. The absolute exposure is bigger, but the net risk after provisions is the lowest it's ever been.

Provision buffers built for a crisis that already ended

The Provision Coverage Ratio (PCR) measures how much cash a bank has set aside to cover bad loans. If a bank has ₹100 in bad loans and has set aside ₹75, that’s a 75% PCR. Think of it as an insurance reserve.

During the 2018 IL&FS crisis, PCR was 50-60%. Banks were caught underprepared. Fair enough, they learned a painful lesson. But the crisis ended eight years ago. Bad loans have fallen by 80%. And yet, provision coverage keeps climbing.

Major banks are running PCR of 75-83%. SBI’s coverage, including certain recovery accounts, is 92.4%. For every ₹100 of bad loans on SBI’s books, loans that already represent just 1.57% of the portfolio, ₹92 is covered.

Where the lending is actually going

The clearest evidence of safety-first thinking isn’t in the NPA data. It’s in what banks are choosing to lend for.

Following RBI’s tightening measures, unsecured personal loan growth has collapsed from 21.1% YoY in FY24 to 7.0% YoY in December 2025. These are loans without collateral, the kind that fund small business expansion, consumer purchases, aspirational spending. Banks have slammed the brakes.

Meanwhile, loans against gold surged 127% YoY. The fastest-growing loan category in Indian banking is one backed by physical gold sitting in a vault. Muthoot Finance, the largest gold loan NBFC, has delivered 40% stock returns this year riding this boom.

Personal loans now make up roughly 35% of all bank credit, up from 19% a decade ago. Industry lending has fallen from 45% to 23%. And even within the personal loan category, banks are retreating from anything unsecured.

Capital buffers that are almost absurdly thick

Banks need to maintain a minimum capital ratio, CRAR (Capital to Risk-Weighted Assets Ratio), essentially the cash cushion relative to the risks they take. India’s regulatory minimum is 9%.

Today, all 78 scheduled commercial banks are above 10%. Zero below 9%. Zero in the danger zone. Kotak sits at 23.3%, more than 2.5x the minimum. SBI Card at ~27%. Even SBI, with the country’s largest loan book, has a comfortable 14%.

When every single bank is over-capitalised and simultaneously posting the lowest default rates in history, the capital is sitting idle when it could/should be funding growth.

Why too little risk is actually not great

This is worth pausing on. There’s a temptation to look at near-zero defaults and record capital buffers and think “great, the system is safe.” But risk-taking is not a bug in financial systems. It’s the entire point.

Banks exist to intermediate risk, to take money from depositors and deploy it toward productive activity that grows the economy. When they stop doing that effectively, the consequences ripple outward.

1. For borrowers, the credit squeeze is real. The small business owner who needs ₹20 lakhs to buy equipment but doesn’t own gold or property is increasingly out of luck. The first-generation entrepreneur without a credit history can’t get a working capital line. The consumer who needs a personal loan to bridge a medical emergency faces fewer options. When unsecured lending growth drops from 21% to 7%, that’s millions of Indians who could have been borrowers, aren’t. MSME credit growth has decelerated from 20.6% to 14.1%. These are businesses that employ 110M+ people.

2. For fintechs, the partnership model is under pressure. The last decade of Indian fintech was built on a thesis: technology companies would originate, underwrite, or distribute credit on behalf of banks and NBFCs. Co-lending, digital lending partnerships, FLDG arrangements, all of these depend on banks and NBFCs being willing to take risk on the underlying borrower. If a bank doesn’t want unsecured exposure, your sleek digital lending app doesn’t solve their problem.

3. For the broader economy, credit growth drives GDP growth. India needs 13-15% credit growth to sustain 7%+ real GDP growth. Credit is currently growing at 12.2%, which is adequate. But the composition matters, the growth is increasingly coming from safe, secured, low-risk-weight categories. The sectors that need credit most are getting proportionally less. When banks only lend where they can’t possibly lose money, capital flows to those who already have capital. That’s not financial deepening.

4. For innovation, excessive caution kills experimentation. Cash flow-based lending, invoice financing, supply chain credit - these models have higher default rates than gold loans. That’s by design. They serve riskier segments at higher yields, and when done well, the portfolio economics work. But in a system where regulators are raising risk weights on unsecured lending and banks are competing on who can have the lowest NPA, there’s no institutional appetite to underwrite models that might show 4-5% defaults even if they’re highly profitable.

The result: the very lending innovations India needs to reach the next 200M borrowers are being strangled in their infancy.

How did we get here? trauma and regulation

The overcorrection is a product of two forces.

1. The NPA PTSD: The 2015-2018 NPA crisis was brutal. Banks, especially public sector ones, were loaded with bad loans from reckless infrastructure lending. The cleanup through RBI’s asset quality reviews, and forced recapitalisation, was long and painful.

2. The RBI’s regulatory tightening: The bigger driver is regulatory. The RBI systematically tightened the screws over the past two years. Increased risk weights on unsecured lending (Nov 2023), tighter norms for NBFC lending, particularly around co-lending and digital partnerships, stricter asset classification rules, Enhanced KYC and compliance requirements that increase the cost of onboarding new borrowers, particularly in underserved segments.

The RBI isn’t wrong to be cautious. They’re the ones who had to orchestrate the 2018 cleanup. But regulation has a cost. When you make unsecured lending expensive, you make it harder for the borrower who doesn’t own gold or property to get a loan. When you tighten NBFC norms, you constrain the institutions designed to fill gaps traditional banks can’t.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

The deposit war is forcing conservatism

There’s a structural squeeze too. India’s Credit-to-Deposit ratio has hit a historic peak of 82.7%. Banks are lending faster than they’re collecting deposits. Credit is growing at 12.2%. Deposits at just 10.2%.

The CASA ratio, the share of lower cost current and savings account deposits, has dropped to ~38%. Retail investors have found better options: mutual funds, equities, higher-yielding fixed deposits. Banks are fighting a brutal war for deposits, which keeps funding costs sticky even as the RBI cuts rates.

The spread (net interest margin) between what banks earn on loans (9.04%) and what they pay on deposits (6.64%) has narrowed by 43 basis points to just 2.40%. When margins compress, you get more selective about the loans you make. You don’t take chances on borderline credits when every basis point of spread counts.

What the market is paying for all this safety

The Price-to-Book (PB) ratio tells you how much investors are paying for each rupee of a bank’s net assets. A PB of 1.0x means the market values the bank at exactly its book value (assets minus liabilities). Above 1x means investors believe the bank will generate returns above its cost of capital. Below 1x usually signals distrust.

The Bank Nifty currently trades at about 2.2x PB. Over five years, this has ranged from 1.6x (COVID panic, March 2020) to 3.2x (2023 bull run). We’re in the middle of the range i.e. not cheap, not euphoric, but trending downwards.

HDFC Bank at 2.1x is notable. This bank historically commanded 3-4x book because of its perceived governance premium. Between merger integration, NIM compression, and a recent governance episode, the market has taken roughly 50% off the valuation premium.

PNB and Bank of Baroda at 0.9x book. Sub-book-value with 15%+ ROE and net NPAs under 0.5%. Five years ago these banks were genuinely distressed, sub-book made sense. Today, the balance sheets are clean and the market still doesn’t fully trust them. Either the deepest value in Indian equities, or the market correctly pricing in the risk that PSU bank management will eventually find a way to squander the turnaround.

The market is paying 5x book for Bajaj Finance and 0.9x for PNB. Both have ROEs between 15-19%. The gap isn’t about asset quality anymore, it’s about trust in management, growth trajectory, and the ability to deploy capital productively.

The standout story: Muthoot Finance at 4.2x PB and 40% returns, built entirely on the gold loan boom. The lender who holds your gold in a vault and gives you 70% of its value is outperforming everyone. What does that tell you about this system’s appetite for risk?

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

Enter the storm

So here’s where I landed after two weeks of spreadsheets: Indian banks are playing it too safe. Default rates too low. Provision buffers too fat. Lending mix too conservative. Unsecured credit in free fall. Gold loans booming because nobody wants to take a chance. MSMEs being rationed out. Fintechs losing their bank partnership foundation.

And then the West Asia conflict changed the calculus.

The Iran conflict put the Strait of Hormuz in play. And suddenly, all that excessive caution started looking less like a problem and more like preparation.

The math of the Strait (according to Mr. Neelkanth Mishra)

The Strait of Hormuz, the narrow waterway between Iran and Oman, carries about 20% of the world’s crude oil. That’s roughly 21M barrels per day, or about 40% of all globally traded oil. Additionally, 2.5% of global gas and 50% of the world’s sulfuric acid (a base ingredient for virtually all chemical manufacturing) flows through it.

A sustained blockade cuts off 7% of global energy supply. Unlike a pandemic, where demand drops to offset the shock, an energy supply cut has no natural substitute. A 7% energy shortfall translates roughly to a 7% contraction in world GDP.

What $100 oil does to India

If oil stays at $100/barrel for a sustained period, India could face three simultaneous shocks:

1. The spending power shock: India’s trade deficit expands. That’s approximately 2.1% of GDP being pulled out of domestic purchasing power. Money that would have funded consumer spending and business investment goes to paying for more expensive imports.

2. The currency shock: To bridge the resulting balance of payments gap, the rupee would need to depreciate. A weaker rupee makes imports more expensive, creating a feedback loop.

3. The inflation shock: Even with partial pass-through of fuel prices, consumer inflation takes a direct hit. That’s before the cascading effect on food (fertiliser disruptions, China has already halted exports), construction, auto parts, and chemicals.

The second-order effects are worse than the headline. Global supply chains carry only 2-3 weeks of inventory. India holds roughly 25-30M barrels of oil reserves. China, which anticipated disruption, holds 100 days’ worth. Mishra’s assessment: the global economy cannot sustain a disruption lasting longer than 4 weeks.

The path from oil to banking stress

Lesser oil → higher transport and input costs → higher consumer prices → RBI pauses rate cuts or tightens → borrowing costs stay elevated → consumers and small businesses under pressure → loan defaults rise → banks tighten lending further → credit growth slows → economic growth slows → more defaults.

That’s the inflation-consumption-default spiral. And it specifically hits the segments already showing early cracks i.e. MSMEs, small businesses, unsecured borrowers. Exactly the borrowers banks had already pulled back from.

The irony: too safe becomes just safe enough

Everything I flagged earlier i.e. the excessive caution, the hoarded provisions, the retreat to gold loans, the stifled risk-taking - suddenly looks like exactly the right positioning for a world where oil is at $100+ and supply chains are breaking.

The last time India faced a comparable external shock was the 2013 “taper tantrum” combined with oil volatility. The banking system was structurally fragile:

Even in a severe scenario where bad loans double from 2.2% to 4.4%, that’s still below 2018 peak levels. Credit costs could rise 100-200 basis points and the system would remain solvent, well-capitalised, and capable of lending.

Those “excessive” provisions? They’re the buffer that absorbs the first wave of defaults without triggering a capital crisis. That “too-conservative” lending mix? It means the loan book is skewed toward secured assets - gold, property, vehicles, that hold value in a downturn. The collapsed unsecured growth? It means less exposure to the borrower segment that defaults first when inflation spikes.

The system I was ready to criticise for not taking enough risk turns out to be well-positioned for an external shock nobody saw coming.

Is that luck? Is it the RBI playing a longer game than any of us realised? Or is it the happy accident of institutional trauma - the hangover from 2018 producing caution that accidentally prepared the system for 2026?

Probably a bit of all three.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

One question I’m wrestling with, and I don’t have a clean answer.

In normal times, the Indian banking system is probably too cautious. But these are not normal times. And the question of “how much risk should banks take?” looks very different when there’s a conflict in West Asia and supply chain disruptions are shutting down factories.

The tension between these two truths is the defining question for Indian banking right now. Too little risk starves the economy of growth capital. Too much risk leaves it vulnerable to shocks. The system is currently planted firmly on the cautious end. The conflict validates that positioning in the short term.

But when the conflict ends we’ll need to have a serious conversation about whether this level of caution is sustainable for a country that needs 7-8% GDP growth.

How should founders plan for this period?

1. Credit is flowing but narrowing. Credit growth is projected at 13-14.5% for FY27. That sounds healthy. But secured lending is booming while unsecured is being rationed. If you have hard assets to pledge, you’re in a borrower’s market. If you don’t, options are shrinking.

2. Build working capital buffers. If you’re in anything MSME-adjacent, logistics, transport, manufacturing, anything tied to input costs, assume your working capital cycle lengthens by 30-60 days. The founders who survive geopolitical shocks are the ones with the most cash, not the best product.

3. Partner risk is real. If your fintech relies on third-party program managers, payment aggregators, or API partners, their compliance failures become your problem. Audit your partners. The regulatory appetite for enforcement is clearly rising.

4. Watch your customers’ customers. If you serve small businesses, their end consumers cut spending first when inflation bites. Model your revenue with a 10-15% consumption sensitivity.

What this means for fintech startups in India

1. The system’s caution is your white space. Banks have retreated from unsecured lending. Growth collapsed from 21% to 7%. But demand for credit hasn’t disappeared, it’s just unserved. Fintechs that can underwrite the “unbankable” borrower more intelligently, using cash flow data, GST analytics, supply chain signals, are filling a gap that’s wider than it’s been in years. Build the underwriting intelligence and banks will partner with you.

2. Novel collateral types are the unlock. If the system wants collateral, give it collateral, just not the traditional kind. Warehouse receipt financing, invoice discounting, equipment-backed lending, receivables factoring. Every fintech that can create a trusted collateral layer for borrowers who don’t own gold or property is building a wedge into the biggest gap in Indian financial services.

3. The deposit war is a fintech opportunity. Banks are desperate for deposits. CASA ratios are falling. Anyone who can help banks acquire or retain depositors, better savings products, embedded deposit features, yield-optimised offerings, is solving a ₹240 lakh crore ($2.5tn) problem.

4. Risk and collections infrastructure will see a demand spike. Even a mild uptick in defaults will trigger a buying cycle for early warning systems, collections automation, and restructuring workflows. Build for the stress cycle now, sell into the demand when it arrives.

5. Cross-border and trade finance is the sleeper. Supply chain disruptions from West Asia will force Indian businesses to rethink sourcing, hedging, and trade finance. Currency volatility is rising. Cross-border payment efficiency matters more. This is a problem that just became urgent.

6. Compliance-as-product wins. The Fino and HDFC episodes share a thread: the cost of compliance failure is going up. KYC on API partners, suitability checks on complex products, GST traceability through payment chains. Fintechs that build compliance as a core feature rather than a checkbox will win bank partnerships and avoid existential regulatory risk.

Indian banks are not taking enough risk. The data is clear: historical low, near-zero net defaults, record provision coverage, collapsing unsecured lending, booming gold loans. The system has overcorrected from 2018, aided by an RBI that learned the hard way what happens when banks run wild.

In normal times, I’d argue this is a problem. MSMEs are being rationed. Fintechs are losing their partnership foundation. Innovation is being throttled. Capital is sitting in provisions instead of funding growth.

But these are not normal times. There’s a conflict in West Asia. Oil is at $100. Supply chains are breaking. And the Indian banking system, almost by accident, is the most fortified it’s been in three decades.

The fortress was built for the last conflict. It might just save us in this one.

The real question comes after. When the crisis passes, will the system find the courage to take risk again? Will the RBI loosen the regulatory grip? Will banks start funding the economy’s potential, not just its safest bets?

That’s the conversation we’ll need to have. But not today. Today, we’re grateful for the fortress.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: I’m forever fascinated by movies about space and space travel, so had to watch Project Hail Mary this past week. The movie had the song Winds of Change by the Scorpions (Spotify / Youtube) which I hadn’t heard in a while. Good nostalgia. Which is your favourite movie about space?

✨ Call outs

1. [Product Release] Union Square Ventures released its venture capital agents. Meet the Agents at USV: Arthur, Ellie, Sally, and Friends.

2. [Report] 2026 Market Overview by Hamilton Lane

3. [Post] The IMF Part 2 - How the fund actually works

4. [Report] India’s Data Centre Capacity to Grow 4x to ~4 GW by 2030

5. [TV Show] Shrinking - my latest favourite show, available on Apple TV.

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me via DM at osborne.vc/dm. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 105th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

This week’s edition is a bit different. It’s a single, long-form deep dive into a company I’ve been researching for a while - one that I think every fintech founder, investor, and policymaker in India should understand deeply but almost nobody in our ecosystem talks about.

Ant Group.

Alibaba’s financial arm. Parent of Alipay. The world’s largest fintech company by virtually every metric - users, transaction volume, revenue.

I spent the last few weeks going through as much information on Ant as is available in English - equity research, academic papers on its regulatory crackdown, Ant International’s sustainability reports, and a stack of Chinese media commentary (translated by Kimi, I don’t speak Mandarin obvs).

What I found was a story that’s part masterclass in ecosystem building, part cautionary tale about regulatory hubris, and part geopolitical chess move that has direct implications for India’s own fintech ambitions.

Fair warning: this is a longer read than my usual long reads. There was so much more but I had to cut it out in the final edit. I think it’s worth it though. If you only read one Fintech Inside edition this year to understand where global fintech is actually headed, not where X/LinkedIn thinks it’s headed, make it this one.

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

The Invisible Juggernaut: How Ant Group is Quietly Rewiring Global Finance

If you ask an investor or founder in our circles to name the world’s most important fintech company, you’ll probably hear Razorpay, or Stripe, or PayPal, or maybe Visa. All of these companies are important and innovating a lot in the payments and broader fintech sector.

They’d be missing out on one very important company.

I’ve long held the belief that one of the most consequential fintech company in the world is one that barely registers in our discourse: Alibaba’s Ant Group. While the world has been debating stablecoin regulations, credit card interchange fees, agentic commerce, cross border payments and more, Ant Group has built a financial ecosystem of genuinely staggering scale - 1.8bn monthly active users, 150M+ global merchants, and over $21trillion in annual transaction volume.

Ant Group generated an estimated $26-27bn in revenue in 2025, holds 23,550+ patents (including 10,000+ in blockchain alone - the world’s largest portfolio), and 59% of its 24,700 employees are in tech roles. For context, Visa processed about $15tn in total volume in 2024. Ant is already larger. And most of us couldn’t tell a single thing about its business model.

But here’s why it’s important to know the story of Ant Group. It is no longer just a Chinese domestic payments champion. It has become the tip of the spear in a much larger structural shift: the slow, deliberate construction of an alternative global financial architecture that doesn’t run through New York, doesn’t depend on SWIFT, and doesn’t need Visa or Mastercard.

And the strategy powering this expansion? It isn’t a government mandate or a Belt-and-Road loan package. It’s the Chinese diaspora. Tourists, students, e-commerce exporters - ordinary people carrying their financial habits across borders.

I think there’s a masterclass in here for India. Let’s get into it!

The Holy Grail - a financial walled garden

To understand what Ant Group is today, we need to understand why it was created. And the origin story is surprisingly familiar to anyone who’s watched Indian fintech evolve.

In 2004, Jack Ma had a trust problem. Chinese consumers wanted to shop on Taobao (Alibaba’s consumer marketplace), but buyers didn’t trust sellers, and sellers didn’t trust buyers. There was no established online payment infrastructure, no credit card penetration to speak of, and no culture of transacting with strangers on the internet.

Ma’s solution was elegantly simple: Alipay, an escrow service. The buyer sends money to Alipay. Alipay holds it. The seller ships the goods. Once the buyer confirms receipt, Alipay releases the funds.

But Alipay didn’t stop at payments. And this is where the story gets really interesting.

2013: The Yu’e Bao moment. Ant launched Yu’e Bao, literally “Leftover Treasure”, which allowed users to invest their idle Alipay wallet balance into a money market fund. The minimum investment? 1 yuan. That’s roughly ₹10 in 2013. Within six months, 49M users had parked their money in it, accumulating $36.5bn in assets. The fund was offering 4-6% yields at a time when China’s state-run banks were paying 0.35% on savings deposits.

By 2018, Yu’e Bao had swelled to RMB 1.7tn ($247bn) in AUM with 600M users, the largest money market fund in the world, bigger than any single fund in the US. (It’s since been capped by regulators and has plateaued around RMB 700bn, but that initial explosion is what matters.)

From there, Ant kept layering services on top of its payments base:

• Huabei (a virtual credit card / BNPL product with 500M+ users and $140bn+ in annual volume)

• Jiebei (cash loans serving 200M+ users)

• Ant Fortune (wealth management hosting 5,000+ mutual funds from 150+ asset managers, with AI-driven robo-advisory and a Vanguard joint venture for mass-market investing)

• Ant Insurance (partnering with 100+ insurers to offer 2,500+ products, processing 7.25M health claims in 2024 alone with 96% AI automation), and

• MYbank (a fully cloud-native digital bank that underwrites SME loans using satellite imagery and AI, with a “3-1-0” model i.e. 3 minutes to apply, 1 second to approve, 0 human intervention). MYbank has served over 50M small businesses, 80% of whom are first-time borrowers who’d never had a bank loan before.

The revenue mix tells the story of just how deep the ecosystem runs: roughly 50% from digital payments, 23% from credit products, 12% from wealth management, 12% from international operations, and the rest from insurance and tech services.

By the time Ant was ready to go public in 2020, it had become a financial walled garden - once a user entered the ecosystem for one service, the gravitational pull of convenience kept them there for everything else. Payments, credit, savings, insurance, investments - all in one app. The Holy Grail for any financial company.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

The $315bn IPO that never was

In late 2020, Ant Group was preparing for a dual listing in Shanghai and Hong Kong. The target valuation: ~$315bn. It would have been the largest IPO in human history. The order books were oversubscribed. The listing was two days away.

Then Chinese regulators pulled the plug.

The narrative often reduces this to a single cause: a speech Jack Ma gave in October 2020 at the Bund Summit in Shanghai, where he called China’s banking regulations “a seniors’ club” and accused Chinese banks of having a “pawnshop mentality.”

But while the speech could have been the spark, Chinese media reported the reason as being systemic financial risk that regulators had been watching for years.

Through Huabei and Jiebei, Ant was originating enormous volumes of consumer loans. But it was keeping almost none of the risk on its own balance sheet. Ant was funding just about 2% of the loans with its own capital and offloading the rest through an Asset-Backed Security (ABS) loophole.

The mechanics were straightforward but the scale was breathtaking.

If a wave of defaults hit, Ant would walk away with its fees while state-backed banks would absorb the systemic damage. Chinese regulators saw this and acted, even if the timing and manner of intervention were heavy-handed by any standard.

The aftermath was painful. Ant was forced to restructure into a financial holding company subject to bank-like capital requirements. The rectification demands were sweeping: establish a licensed consumer finance entity with proper capitalisation, cease regulatory arbitrage in lending, enhance consumer protection, restrict cross-selling, cap Yu’e Bao’s size, and most dramatically - restructure corporate governance so that the board had a majority of independent directors.

Jack ceded control entirely: his voting rights were slashed from 53.4% to just 6.2%, and ten major shareholders agreed to vote independently. Hangzhou’s city government took a 10% stake in the consumer finance subsidiary. Ant’s valuation cratered from $313bn to roughly $79bn in a 2023 share buyback. And in July 2023, regulators imposed a RMB 7.12bn ($1bn) fine - one of the largest penalties ever levied on a Chinese internet company - formally closing the chapter on the crackdown.

Then came the rebound

By 2024, Ant had adapted. Net profit grew 61% YoY to an estimated $5.3bn on roughly $25bn in revenue. The company’s valuation recovered to roughly $103bn by 2025. The business model pivot was real: 98% of consumer loans are now funded by partner banks (versus the pre-crackdown model where Ant was originating on its own balance sheet), and the company has shifted decisively towards B2B tech infrastructure - its distributed database OceanBase now serves 2,000+ enterprise customers, and its blockchain, AI, and cloud services are growing 40%+ annually.

Most importantly, Ant International, the company’s Singapore-headquartered global arm, generated nearly $3bn in revenue in 2024 (up 35% YoY), grew another 20-25% in 2025 to an estimated $3.7bn (roughly 10% of group revenue), and has now produced eight consecutive profitable quarters. It has its own independent board and is being prepared for a potential standalone spinoff and IPO.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

The diaspora strategy or, how to build global rails

This is where the Ant Group story shifts from an interesting Chinese tech narrative into something with direct geopolitical implications. And it’s the part I find most relevant for India. I’ve briefly written about this in Edition 102 - Gold at $5,000, Edition 91 - Trade, Tariffs and Finance and other editions.

For decades, cross-border payments have been controlled by a US-centric architecture: SWIFT for banking messaging, Visa and Mastercard for card rails, correspondent banking for settlement. These systems work, but they’re slow, expensive, and, critically, controlled by institutions that sit under American regulatory jurisdiction. The US has demonstrated repeatedly (Iran sanctions, Russia’s SWIFT disconnection) that it’s willing to weaponise this infrastructure.

Ant Group’s international strategy doesn’t directly challenge this system with rhetoric or government backing. It bypasses it. And the bypass mechanism is simple: follow the Chinese diaspora.

Ant International runs on a three-pillar structure, each pillar targeting a different layer of global financial flows:

Pillar 1: Alipay+ (the consumer gateway). This is the cross-border payment and digital solutions platform, connecting local e-wallets globally so consumers can “pay like a local” anywhere. More on this below, because it’s the most strategically important piece.

Pillar 2: Antom (the merchant infrastructure). A one-stop acquiring and payment orchestration platform for global merchants. Through a single integration, a merchant can accept 250+ payment methods from across Asia and beyond i.e. bank cards, digital wallets, online banking, BNPL. Antom also offers an AI-powered “Payment Copilot” that automates payment interface integration, drastically reducing the technical barrier for merchants going cross-border. Think of Antom as Ant’s version of Stripe but built for Asia-first commerce.

Pillar 3: WorldFirst (the B2B engine). Acquired from the UK in 2019 for $700M, WorldFirst provides multi-currency “World Account” services for SMEs and cross-border sellers. A small exporter can manage up to 42 currencies in one account, connecting with markets in 200+ countries, collecting payments, converting currencies, making supplier payments, and accessing supply chain financing. Over 38,000 merchants used WorldFirst for cross-border e-commerce payments in 2025.

Beyond these three pillars, Ant also runs a digital bank in Singapore, Alipay wallets for Hong Kong and Macau, and has recently launched Bettr, a lending platform piloting in Hong Kong and Thailand.

The international operation follows it’s domestic strategy - acquire and grow via digital payments but soon diversify into a broader financial services business.

The tourist as a Trojan Horse. In 2025, roughly 87% of Chinese tourists abroad used Alipay as their preferred payment method. Ant went to merchants in Japan, South Korea, Southeast Asia, Europe, and the Middle East with a simple value proposition: “Want Chinese tourist spending? Accept Alipay.” Today, over 150M global merchants do. Ant’s Smart Tourism initiative alone powered payments for 18M Chinese travellers across Europe and Asia. No Visa acquiring agreement needed. No Mastercard interchange. Just a QR code.

The B2B export engine. Through its 2019 acquisition of UK-based WorldFirst and the launch of Antom (its merchant payment and digitisation services arm), Ant built a cross-border collection and disbursement platform for global merchants and SMEs. A small factory in Vietnam selling to Chinese buyers? Ant handles the payment, the FX, the settlement, across 42 currencies in a unified ledger.

Cross-border payment volume grew 25% in 2025, reaching over $350bn. And the partnerships keep expanding: a 2025 Shopify integration enabled 250,000 merchants to accept Alipay from Chinese shoppers worldwide, and collaborations with Uber (85+ countries), Starbucks (12,000 stores), AirAsia MOVE, and Booking.com are embedding Alipay into the travel and commerce workflows that Chinese consumers already use.

And then came the masterstroke: Alipay+

Launched in 2020, Alipay+ is a cross-border payment gateway but not for Alipay users alone. It connects local e-wallets from around the world into Ant’s global merchant network. Here’s how it works in practice: a Filipino tourist travels to Japan. She doesn’t have Alipay. She doesn’t need it. She opens GCash (the Philippines’ leading e-wallet and partly owned by Ant Group), scans a PayPay QR code at a Japanese convenience store, and the transaction is routed, converted, and settled via the Alipay+ backend. The merchant never knows the difference. The tourist pays in her home wallet. Ant clips the transaction fee.

As of mid-2025, Alipay+ connects 40 partner wallets and banking apps to merchants in over 100 markets, with 1.8bn connected user accounts. The scale of the individual partners is what makes this network so formidable:

• PayPay in Japan (60M users, 3 million+ merchants, owned by Softbank an early Alibaba investor),

• GCash in the Philippines (80M users),

• Kakao Pay in South Korea (45M users),

• Touch ‘n Go in Malaysia (20M users),

• TrueMoney in Thailand (20M users),

• ZaloPay in Vietnam (15M users),

• plus OCBC Digital in Singapore, Vipps in Norway, Bluecode in Germany and Austria, and Tinaba in Italy.

  That’s a long list of integrations across Asia and Europe.

Alipay+ also partnered with national QR code operators PayPay in Japan, DuitNow in Malaysia, SGQR in Singapore, ZeroPay in South Korea, KHQR in Cambodia, NepalPay QR in Nepal and many more, so that partner wallet users can pay at virtually any merchant in these countries without the merchant needing to do anything special. Over 90% of Alipay+ merchants are MSMEs. In the first half of 2025 alone, 6.5M consumers used Alipay+ cross-border payments for the first time, and intra-Asia transactions by partner wallets grew 32% YoY.

The regional strategy is also differentiated, which is worth noting.

In Southeast Asia, Ant goes deep by investing directly in local fintechs (Ascend Money in Thailand, GCash in the Philippines, earlier Paytm in India), providing wallet technology infrastructure, and integrating across e-commerce and financial services.

In Japan and South Korea, they primarily serve inbound tourism through merchant QR acceptance.

In the West Asia, they’re building out teams in Dubai and Riyadh and introducing Antom for local payment digitisation.

In Europe and North America, they mostly serve outbound Chinese tourism and cross-border trade.

Each market gets a tailored approach, deep cultivation in emerging markets, technology export and network connection in developed ones.

I want to pause and let that sink in. Ant has, effectively, built a parallel Visa network for Asia, except it runs on QR codes and smartphones instead of plastic cards, it’s interoperable across dozens of local wallets, and it completely bypasses the traditional card schemes.

And the way they built it was through a deliberate invest-and-integrate strategy that started in 2015-2017: Ant invested $857M+ in Paytm (India), $200M in Kakao Pay (South Korea), strategic stakes in GCash/Mynt (Philippines), Ascend Money (Thailand), Touch ‘n Go (Malaysia), TrueMoney (Thailand), and ZaloPay (Vietnam). They even tried to acquire MoneyGram for $880M in 2017, but were blocked by US regulators on national security grounds.

Each investment came with technology transfer, Ant’s Wallet Tech platform helped these local wallets evolve from simple payment tools into super-apps.

I had a whole section on Alipay becoming a Super App in China, but it made this edition too long (5,000+ words). I deleted it, but would you want to read about that too? Write to me.

The tech moat — AI, NFC, and Blockchain

Ant Group’s R&D spending in 2024 was RMB 23bn ($3.26bn), an 11% increase YoY. About 40% goes to AI and machine learning, 20% to blockchain, 15% to their distributed database (OceanBase), 15% to security and biometrics, and 10% to cloud infrastructure. They hold over 23,550 patents globally (96% are invention patents), with the world’s largest blockchain patent portfolio at 10,000+.

AI-powered commerce. In February 2026, Alipay announced that its “AI Pay” feature had surpassed 120M weekly transactions. The use cases are expanding fast: AI agents embedded in apps and mini-programs, voice-activated payments through smart glasses, and most ambitiously, the Agentic Commerce Trust Protocol launched in January 2026 with Alibaba’s Qwen AI app and Taobao Instant Commerce. Users simply chat with an AI, with payment handled seamlessly in-agent.

Ant has also built its own foundation models Ling-Plus and Ling-Lite, with 100 bn+ parameters, targeting finance and healthcare applications. Their AI financial assistant Zhi Xiaobao serves 70M monthly active users, and AI-powered fraud detection has reduced fraudulent activity by 83% compared to legacy systems.

Hardware innovation. “Alipay Tap!” an NFC-based contactless payment solution launched in 2024, recently crossed 100M daily transactions (as of January 2026).

The experience has been specifically designed for accessibility - embraced by younger users, older adults, and people with visual impairments.

Blockchain and tokenisation. Through AntChain, Ant processes 100M+ daily data uploads with consensus performance exceeding 100,000 TPS. They’re cutting cross-border remittance times by 45% by bypassing correspondent banks entirely. And they’ve launched the Jovay Network (an Ethereum Layer 2, hybrid ZK + TEE architecture, targeting 100,000 TPS) to help banks issue digital bonds and tokenise real-world assets.

Ant also built OceanBase, a distributed database that handles 61M+ queries per second (it holds the TPC-C world record) and is now used by 2,000+ enterprise customers outside of Ant, including banks and telecoms. OceanBase is to Ant what AWS was to Amazon, an internal infrastructure product that’s becoming a standalone business.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

The Ant playbook - what founders can take away

The four points above are mostly about policy and ecosystem design. But I want to go deeper into what Indian and global fintech founders can specifically learn from how Ant was built - the operational and strategic decisions that turned an escrow button into a $100bn ecosystem.

These are the lessons I’d want in front of me if I were starting a fintech company today. I’ve talked about these in several past editions, and you’re probably tired of listening to me like a broken record.

1. Solve the trust problem first. The financial product comes second.

Alipay wasn’t created because someone wanted to build a payments company. It was created because buyers and sellers on Taobao didn’t trust each other.

I see too many Indian fintech founders start with the financial product - “we’re building a savings product” or “we’re doing cross-border remittances” or “we’re making tax filings easier”. In my evaluation and diligence, I’m always asking founders what trust deficit they’re solving and how.

The best fintech companies in India have followed the Ant pattern, often without realising it. Zerodha solved the trust problem of “is my broker ripping me off with hidden charges?”. Slice solved it for young Indians who couldn’t get a credit card from a bank.

Start with the trust gap. The product will reveal itself.

2. Build the data flywheel before you build the lending product.

Ant didn’t start with lending. They collected payments data for nearly a decade before launching Huabei and Jiebei in 2015.

The lesson isn’t “collect lots of data.” It’s that the best lending businesses are built on top of proprietary distribution and data advantages that take years to accumulate. In India, the companies best positioned to become the next Ant-like lenders aren’t necessarily the ones with NBFC licenses today, they’ll be the ones sitting on the richest transaction data.

The question is whether they’ll have the patience and the regulatory room to make that transition.

3. Monetise through the stack, not the transaction.

Ant’s revenue breakdown is instructive. The payment itself is a low-margin, high-volume anchor. The real business is everything that sits on top of it.

I think about this a lot in the Indian context. UPI made the payment layer essentially free. A lot of founders look at that and see a problem, “how do I monetise if payments are free?”. But Ant’s playbook shows that free payments are a feature, not a bug. They’re the distribution channel. The question is what high-margin service you can attach to the payment flow that the customer actually needs.

The merchant doesn’t need to leave the ecosystem. Every layer is a monetisation opportunity. Indian fintechs that obsess over payment MDR are missing the forest for the trees.

4. International expansion follows people, not pitches.

This is perhaps the most actionable learning for Indian founders thinking about global markets. Ant went global by following three natural flows: Chinese tourists spending abroad, Chinese exporters selling cross-border, and Chinese tech companies expanding internationally. Every international product maps directly to one of these flows.

The Indian equivalent of these flows exists, and it’s arguably larger: 32 million NRIs sending $135+bn in remittances annually, Indian IT services companies operating in every major market, Indian students at universities worldwide, and rapidly growing outbound tourism.

5. Regulatory survival is a design principle, not an afterthought.

Ant’s story includes what is probably the most dramatic regulatory intervention in fintech history - something we, in India are all to familiar with. Ant survived. Not just survived, thrived. Profits surged 61% the year after the crackdown ended.

The lesson for Indian founders: build your company assuming the regulatory environment will change significantly during your lifetime. Don’t optimise for today’s rules, build structural flexibility into your capital structure, and your partnerships from day one.

6. AI is infrastructure, not a feature.

This is the lesson most relevant to right now. Ant is spending on core AI products.

Indian fintech founders building today should be asking: “If AI were not an add-on but the core operating system of my business, what would my cost structure look like? What markets could I serve that are currently uneconomical?”.

The bottom line

While the West continues to view global finance through the lens of plastic cards and SWIFT codes, Ant Group is already processing trillions through QR codes, AI agents, and blockchain rails. The infrastructure of the future isn’t being debated in Washington or Brussels. It’s being built in Hangzhou and Singapore.

For Indian builders and policymakers, the question isn’t whether to pay attention to Ant Group. It’s whether we can learn fast enough to build India’s version of this story, using our own extraordinary assets (UPI, India Stack, the diaspora, the trade flows), before the window closes.

I’m optimistic we can. We need to move faster. And we need to think bigger. We need to ask ourselves “how do I build the financial operating system for the next billion users?”

That’s the question Ant answered for China. It’s India’s turn to answer it for the world.

Which of the founder lessons resonates most with what you’re building? I’d love to hear your thoughts - hit reply or find me on Twitter/LinkedIn.

If you enjoyed this deep dive, share it with a friend who’s interested in fintech, geopolitics, or just wants to understand what the world’s most important company you’ve never heard of is actually doing. And as always, thank you for reading.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: Dracula by Tame Impala (Spotify / Youtube) has been on loop lately. Classic Tame Impala song - good beats to groove to. Let the party never end!

✨ Call outs

1. [Video] Petrodollars: Important post given the current environment. Understanding petrodollars is understanding geopolitics. Worth your time.

2. [Report] Pandora’s Bog: the global energy shock of 2026 by JP Morgan. According to the report, a ~20% disruption to global energy supply has triggered a broad commodity shock, driving inflation, pressuring corporate margins, and increasing downside risk for global equities.

3. [Project] This is a Claude Cowork skill built by Ameya Pimpalgaonkar. It’s a “…plugin that decodes any technology or market signal into 2nd and 3rd order business insights — using layer-by-layer question sequencing that takes you from noise to structured understanding.”

4. [Report] 220-slide report by Persistence Capital in India highlighting key trends and insights from 8 sectors of companies in the public markets. Worth a read through for the signal.

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me via DM at osborne.vc/dm. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 104th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

Last week, I vibe-coded an entire core banking system in a single prompt. Role-based logins, general ledger, deposit and loan accounting, transaction views - the works.

A year ago, this would’ve taken me weeks. Today, one prompt. That’s how fast coding agents have improved.

This speed is exactly what’s driving the panic. Anthropic launched a legal plugin for Claude in February and wiped roughly $285bn off the legal tech sector in a single day.

Software stocks are cratering. The “SaaS is dead” discourse is deafening. And honestly? If you only look at the interface layer, the panic makes sense.

But you come to Fintech Inside for the nuance, not the noise. So this edition goes deeper. I explore what AI can and cannot actually disrupt across the banking stack - from the interface layer (fair game) to the controls layer (harder than it looks) to the core system of record (not happening anytime soon).

And I lay out how I think fintech founders should think about building and selling to financial institutions in an AI-accelerated world.

Three-time champions, first-ever title defence, and 255 on the board in a final — India didn’t just beat New Zealand, they made it look unfair. What a tournament. Go India! 🇮🇳

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

Vibecoding a Bank - What AI can and can't disrupt in financial services

I vibecoded an entire core banking system using Kimi K2.5. In one shot. One prompt. You can play with it at cbs.osborne.vc. I open-sourced both the prompt and the code if you’d like to make it your own.

Try the Core Banking Demo

It has role-based logins (admin, teller, customer service, operations, credit officer, finance), a general ledger, deposit and loan accounting screens, transaction views - the full interface of what a core banking system looks like.

It looks like a modern core banking system. But it doesn’t work like one.

I built the interface layer in one shot. The middle layer i.e. controls, compliance, fraud, AML, would take months of domain expertise. The core layer i.e. the actual ledger, the system of record, would take years.

That gap between what you can vibe code in an afternoon and what a bank actually needs? That’s the entire thesis of this edition.

I’ve built 21 projects (and counting) with AI in the last year and a half. Some would say I’ve got an addiction, and they may or may not be right.

A dilution visualiser for founders. A news digest I use every week for the TWIF Asia newsletter. A CLI-based CRM for VCs. A smart ingredient scanner. A pastry chef dashboard that helps chefs brainstorm dessert recipes using AI. A text to speech converter for my read-later list. A flappy birds clone for the tech ecosystem called FounderUdd, which remains one of my finest contributions to society. You can find a list of some of my projects here.

I’ve experimented and used all the big lab coding tools i.e. Cursor, Claude, Gemini, Codex, Kimi. Some projects took a weekend, some took weeks of late nights and started-from-scratch frustration. But the throughput was extraordinary. Ideas I’d been sitting on for years were suddenly shippable in days.

And the tools got dramatically better along the way - in just one year! My first project, the dilution simulator, was rough - prompt ChatGPT, copy paste code, debug manually, wrestle with errors I barely understood. By the time I built Lokta (try it here), a full DocSend alternative with AI-powered document chat, the experience was unrecognisable.

Claude Code in Cursor was architecting components, running security audits, handling authentication flows, catching vulnerabilities I wouldn’t have spotted myself. The progression from “AI helps me write code” to “AI helps me build software“ happened in about 12 months.

Subscribe now

But the more I built, the more something became clear. What I was generating across all 21 projects was interface. Dashboards. Tools. Workflows. Things that present information, collect input, guide users through a flow.

I really tried building a system of record. The first thing I really wanted to build was an entire operating system for my venture/investing activities. It was a complex requirement and I’d constantly face challenges. I deleted everything and started from scratch - thrice. Maybe I’m expecting too much from AI too early in its journey. Or maybe I’m just incompetent.

Not one was a system of record. Not one maintained financial state. Not one had to reconcile a ledger, survive an audit, or report to a regulator.

That distinction is the most important one in the current AI “SaaSpocalypse” discourse, especially for those of us in financial services.

Because it’s one thing to vibe code a dashboard. It’s another to vibe code a balance sheet.

The “SaaSpocalypse” and what everyone’s arguing about

If you've been anywhere near tech Twitter lately, you've heard it all. SaaS is dead. AI agents will replace every enterprise tool. Seat-based pricing is finished. VCs won't touch a pure SaaS pitch anymore.

The trigger was Anthropic’s recent product blitz i.e. Claude Code, Cowork, Claude for Excel/Powerpoint, Claude for legal review, Claude for Equity Research, all landing in quick succession. After Anthropic launched a legal plugin on Feb 3, Thomson Reuters dropped 16%, RELX fell 14%, LegalZoom sank nearly 20%, wiping roughly $285bn from the legal tech sector in a single day. The “SaaS-pocalypse” had its poster child.

The discourse has split into camps, but any sane person would agree the future is more nuanced than “SaaS is dead.”

Some say AI will eat everything. HSBC’s CIO Academy team published a note calling this the “AI fear trade” - the market selling first, asking questions later. Their key insight: the two concurrent fears (AI will disrupt all sectors and the return on $650bn in AI capex is uncertain) are contradictory.

Share

Sidu Ponnappa, founder of RealFast AI, makes a sharper structural argument: it’s not that SaaS pricing is under pressure, it’s that the economics of software as an asset is breaking. If building an HRMS takes an AI agent a weekend instead of a team six months, the output isn’t an asset. It’s inventory. You don’t amortise inventory. You liquidate it.

The view I concur with: it depends entirely on what layer of the stack you're building on. On a recent episode of No Priors, Elad Gil put it well - “people are projecting five-person startup behaviour onto the Fortune 100. A technical team can vibe code a CRM over the weekend. Before that, they just used a spreadsheet. Does that mean Bank of America is going to displace their core systems with something hacked together in a day? Probably not”.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

The slop problem with vibecoding

There’s a serious code quality problem with vibecoding that gets ignored.

When AI generates enormous amounts of code and nobody’s reading it closely, nobody deeply understands the codebase, and there’s no systematic review - you get fragility.

That’s a slop problem. Not slop from random weekend projects, but vibe coding slop in actual production codebases, created by every engineer taking the path of least resistance. Generally, I’ve found when vibecoding that incremental changes ends up introducing slop.

Now extrapolate that to a core banking system where billions of transactions are being processed every minute, 24x7, 365 days a year. Without a break. Processing that much volume requires code ultra-streamlined code. Any aspect of the system that causes even millisecond delays in a transaction, compounds at the scale and speed of UPI as an example.

A single reconciliation error can cascade into a regulatory event. Where audit trails need to be perfect. Where the code isn’t just running a feature, it’s maintaining legal state.

And it’s not just code quality. There’s a change management problem. If you ask the engineer annoyed about paying $10/seat for Jira whether they’d want to handle change management at a bank involving aligning hundreds of people on a new workflow, managing security, maintaining the system indefinitely, the answer is probably not. Enterprise software isn’t just code. It’s organisational infrastructure.

The tools are getting better, fast. My own experience proves it. AI coding agents now run security scans, catch dependency vulnerabilities, generate test suites, and architect with less intervention than a year ago.

Token costs have collapsed, according to MIT Research, for frontier models, the cost to achieve a given benchmark performance has fallen 5x-10x per year, with top-tier models seeing declines up to ~30x per year. The trajectory is clearly towards more capable, more auditable code generation.

But “getting better” and “ready for core banking” are very different bars. The gap between 4th best provider and 3rd best is wider than the gap between 4th and 200th.

That’s the distinction that matters for financial services.

Banking Is not software. It’s a system of record.

Most software is a system of interaction. It manages tasks, workflows, communication. If your project management tool goes down, you lose context. Annoying, but recoverable.

Banks are systems of record. They maintain who owns what, who owes what, what capital exists, what liabilities are outstanding, and what must be reported to regulators. A bank’s core system isn’t just software, it is the legal memory of the institution.

If your CRM fails, you reschedule a demo. If a consumer order doesn’t get placed, you ask them to do it again. If your ledger fails, someone’s deposit is wrong. That’s not a UI bug. That’s a liability event.

Most AI-optimists looking to “disrupt banking” are looking at the interface. They should be looking at the plumbing too.

The banking stack has three layers

To understand what AI can and cannot change, it helps to break the banking stack into three layers.

The core layer i.e. the ledger. Core banking system, general ledger, deposits and loans accounting, settlement engines, regulatory reporting. This layer defines state. It records every transaction, reconciles every money movement, backs every reported number.

Leave a comment

Replacing it is not refactoring. It’s organ transplant surgery.

Core systems accumulate years of transaction history, reconciliation logic, regulator-reviewed processes, and audit trails nobody fully understands but everyone depends on. Migration means perfect balance replication, zero reconciliation gaps, regulator sign-off, operational retraining, and vendor re-integration.

That’s why core replacements take years. Not because engineers can’t write code but because institutions can’t afford to be wrong. AI can generate CRUD apps in minutes. It cannot generate and migrate 15 years of reconciled state.

Using Sidu’s framing: the core banking ledger sits firmly above the line. The difficulty isn’t in the code, it’s in the understanding the code encodes.

And replacing a core banking system isn’t just a technical migration. It’s retraining thousands of people. It’s renegotiating vendor integrations. It’s convincing a regulator that your new system is at least as reliable as the old one. No amount of faster code generation compresses that timeline.

The middle layer i.e. controls and orchestration. AML/KYC engines, fraud detection, risk scoring, treasury workflows, regulatory reporting pipelines, payment routing, reconciliation tools. This is where control lives.

Some of this can absolutely be AI-augmented. Fraud models will improve. Document review will accelerate. I’m seeing this in my portfolio already.

But this layer must remain explainable, auditable, deterministic at decision boundaries, and defensible in front of regulators. False-positives and false negatives are both problematic. If an AI-vibecoded AML engine fails, suspicious transactions pass through, accounts freeze incorrectly, regulators intervene. This isn’t feature risk. This is institutional risk.

This is the “code slop” risk that matters most. Every line of unexplained, unreviewed code in a regulated system is an audit finding waiting to happen.

Nihar Bobba at Better Tomorrow Ventures wrote a great piece about “last mile defensibility”. AI companies survive when they own the coordination between intelligence and real-world systems: regulatory endpoints, audit requirements, liability. The middle layer of banking is exactly this. You can enhance it, but you cannot trivialise it.

Just because AI can write code doesn’t mean it will be trusted. History shows the cheapest solution doesn’t always win. The trusted one does.

The interface layer i.e. where AI actually moves fast. Mobile apps, admin dashboards, internal tools, analytics views, customer onboarding, support automation. Systems of interaction. They present state, collect input, guide users.

This layer can absolutely be vibecoded faster. I’ve done it 21 times. Fintech startups will ship interfaces in days and replace front-ends cheaply.

But surface speed does not equal core replacement. You can modernise the interface all you want. The ledger still remembers everything.

Using Sidu’s “asset vs. inventory” framing: if what you’ve built is primarily an interface, you don’t have a defensible product. You’re below the line. Anyone with Cursor and a weekend can approximate it. But if you’re building something that touches financial state, regulatory compliance, or settlement - that’s still an asset. Possibly more of an asset than ever.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

Why systems of record are so sticky

I ask myself this a lot when evaluating enterprise SaaS startups: if AI is so powerful, why can’t we just rebuild banking from scratch?

Because systems of record accumulate things code generators cannot replicate. You cannot just prompt:

1. Data gravity. Years of transactions, exceptions, manual adjustments, and edge cases resolved by people who’ve long since left the company.

2. Regulatory memory. Tests passed. Models approved. Protocols approved. Reports accepted by specific regulators who know the institution’s history.

3. Integration surface area. Payment rails, card networks, clearing houses, partner APIs, each connected through layers of contracts, certifications, and operational agreements.

4. Operational muscle memory. Teams trained on specific workflows. Knowledge that lives in people’s heads, not in documentation.

5. Embedded risk performance. Underwriting data. Fraud loss ratios. Capital buffers calibrated over cycles.

Consider a mid-sized bank with a commercial loan that was originated in 2016, restructured once during COVID under RBI's resolution framework in 2020, had its provisioning norms changed by a regulatory circular in 2022, was upgraded back to standard after 12 months of satisfactory performance, and now sits in a pool that's being evaluated for securitisation. How do you prompt Claude Code to add this to your CBS?

Replacing a system of record means revalidating all of that simultaneously. Boards fear it. Regulators scrutinise it. Operations teams dread it. Now consider that you have quarterly financials releases and every month you have to show growth, margin protection and profit.

This maps to what Nihar describes as “barriers that compound.” Every successful audit strengthens credibility. Every edge case handled adds institutional knowledge. Every year of compliance history deepens dependency. These aren’t features. They’re accumulated trust and trust doesn’t have a vibe-code shortcut.

Competitive advantage theory didn’t just disappear. The production side of software is getting cheaper, yes. But distribution, trust, and institutional embedding - those are as hard as ever.

What AI compresses and what it can’t

I’m not making a Luddite case against AI in fintech. Far from it.

AI will compress feature moats, internal tooling costs, and speed to launch. Fintech teams will prototype faster, build systems cheaper, improve analytics, enhance underwriting, and reduce compliance effort. I’m living it - As a builder and as an investor watching portfolio companies ship faster than ever.

But AI doesn’t eliminate the need for a ledger. And it doesn’t make financial correctness probabilistic. In banking, you’re either right or you’re wrong. There is no “mostly correct” deposit balance.

As interfaces accelerate, the core becomes more important. Faster surface activity increases stress on reconciliation, settlement, and compliance. The faster your cars go, the better your brakes need to be.

Agentic AI won’t kill enterprise software, it will reset its business model from effort-based billing to outcome-based pricing. But the platforms orchestrating AI and non-AI components remain essential. AI becomes the brain; software remains the body. They’re complementary, not substitutional.

The proof is already here, at scale

None of this is theoretical. The most sophisticated fintech companies are already shipping AI-written code into production.

Stripe’s internal coding agents, “minions,” now produce over 1,300 merged pull requests weekly, completely AI-written, human-reviewed. Each minion operates unattended: picks up a task from Slack, writes code, runs tests, submits a PR. But Stripe built these in-house because their codebase is hundreds of millions of lines of mostly Ruby with proprietary libraries and compliance constraints that off-the-shelf agents couldn’t handle. The guardrails matter more than the model.

Ramp implemented over a million lines of AI-suggested code, with 50% weekly active usage across engineering and an 80% reduction in incident investigation time through AI-powered internal tooling.

Brex’s engineering lead described it as a mindset shift: engineers now try three different approaches in the time it used to take for one. The output is cleaner architecture not just faster code.

And Kailash Nadh, CTO of Zerodha, one of India’s largest brokerages, wrote what I think is the most honest assessment from a financial services builder. His piece argues that software development as done for decades is over - compressing weeks into hours, not through vibe coding, but through disciplined use of LLM agents. This is someone maintaining systems millions of retail investors depend on daily.

Harshil and team at Razorpay have always been at the bleeding edge of implementing AI for their systems and products. Last year, Rahul Chari at PhonePe said they don’t use AI for their core systems at all. I’m assuming things have changed for him since then.

The question is no longer whether financial firms will adopt AI coding. They already have. The question is: what layer of the stack can it actually touch and make a difference?

The gap between “can we ship this?” and “should we?”

AI coding agents will keep getting better, faster and cheaper. The progression from my dilution simulator to my DocSend alternative will look quaint in two years. We’ll see agents that maintain long-running context across codebases, flag compliance issues in real-time, and generate audit-ready documentation alongside code. The code slop problem? Every major lab will optimise for accuracy. I expect it gets substantially addressed.

Token costs will keep collapsing. The bottleneck shifts from “can we produce this code?” to “should we ship this code?” - a question requiring human judgment, domain expertise, and institutional context. Exactly what matters most in financial services.

But even as coding tools approach production-grade quality, the non-code barriers in banking remain enormous. You still need a licence, regulator relationships, years of underwriting data, and the operational muscle memory of teams who’ve handled a thousand edge cases.

Technology cycles are compressing decade-long shifts into a year or two. But institutional change, the kind that governs banking, doesn’t compress that way. Technology cycles and institutional cycles operate on fundamentally different clocks. That gap is where the most interesting fintech opportunities of the next decade will live.

What this means for fintech founders

If you’re building an interface-layer fintech i.e. a better dashboard, a prettier app, a faster onboarding flow, your moat just compressed. Defensibility has to come from distribution, data, or regulatory positioning. Not the interface.

If you’re building control-layer infrastructure i.e. compliance tooling, risk engines, fraud systems, AI enhances your product but doesn’t erase it. You need domain expertise, regulatory awareness, and battle-tested trust. Good place to be building.

If you’re replacing a system of record i.e. a new core banking platform, a next-gen ledger, your challenge isn’t code generation. It’s trust migration. And trust migrates slowly, painfully, and expensively.

The question isn’t what vertical you’re in. It’s what product choices you’re making. Are you a tool for professionals, where the user retains liability? That’s vulnerable to AI model creep. Or are you owning outcomes, absorbing liability, coordinating with real-world regulatory systems? That’s structurally defensible.

The best defence is to build a bundle i.e. multi-product surface area that becomes default infrastructure. Cross-sell into the institution. Become embedded. Make switching costs real. In a world where code is cheap, distribution and entrenchment are what compound.

I wrote in my last edition #103 about three structural levers for startups: attack the margin profile, compress time to money, and deliver a Delta 4 experience. AI supercharges all three especially the second.

If AI collapses inference and analysis costs to near-zero, the competitive edge shifts to delivery speed. The fintech that underwrites in seconds, not days. The compliance platform that files across jurisdictions in hours, not weeks. Or even if you’re building SaaS for financial firms, the speed of your response time can dictate if you win that customer.

Every layer of AI compression feeds directly into shrinking the gap between “customer has a need” and “customer has money in hand.”

I believe the winners in AI-era fintech will be services businesses that own the outcome, absorb the liability, and get paid when value is delivered not SaaS companies selling seats. This is where Nihar’s “long last mile” thesis and my “compress time to money” thesis converge. The companies that own the outcome and deliver it fastest will be extraordinarily hard to displace. They’re not selling software. They’re selling certainty, delivered instantly.

The next decade of fintech will not be won by better dashboards. It will be won by companies that own a piece of the system of record, control financial state, accumulate proprietary transaction data, and embed deeply into capital flows.

Because state compounds. Interfaces change.

You can vibe code an app. You cannot vibe code institutional memory. And financial services, at its core, is institutional memory wrapped in software.

What layer of the stack are you building on? How are you thinking about defensibility in an AI-accelerated world? Hit reply — let’s talk.

1-min Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: This week I’ve been listening to Mountain Sound by Of Monsters and Men (Spotify / Youtube) on loop. It’s one of those songs that makes you want to drive somewhere with no destination. Perfect energy for a week spent thinking about systems that don’t move and code that moves too fast.

✨ Call outs

1. [REPORT] Quick Commerce in India: The Money Machine That Has Not Made Money Yet by Karan Sehgal and Yash Panditrao of 3F Capital

2. [BLOG] Situation Awareness - on AGI and the fund investing in the future

3. [REPORT] The Middle Class Math is not Mathing by AngelOne

4. [NOTES] Thoughts From the Road by Henry McVey, Head of Global Macro

   & Asset Allocation, CIO of KKR’s Balance Sheet

5. [VIDEO] “I Visited Most Restricted Area of T20 World Cup” - a video on the tech that goes into streaming the T20 World Cup

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me via DM at osborne.vc/dm. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 103rd edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

This week's piece comes from a conversation I've been having on repeat. Seed-stage founders building different products, across different sectors, all asking the same thing: how do we compete against entrenched players? Decided to write down what I usually tell them.

The answer matters more now than it did three years ago. The incumbents founders are competing against today aren't just banks with legacy tech. They're well-funded fintechs with real distribution and deep data moats.

I believe there are three structural levers: 1. Attack the margin profile, 2. Compress time to money, 3. Deliver a “Delta 4” experience.

The breakout companies almost always do one of three things differently. Often two.

If you're building in fintech right now, or thinking about it, this is the lens I'd use to pressure-test whether your startup has a real edge.

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

Three ways a startup can compete against incumbents

If I had a dollar for every time a seed-stage founder asked me “How do we compete against incumbents?” I’d have enough to start my own venture fund!

Over the past year, I’ve had this conversation with several founders. Different sectors. Different products. Same anxiety. And honestly, I understand the anxiety. When you’re at the seed stage, the incumbents feel enormous. They have the customers, the data, the regulatory relationships, the brand. You have a pitch deck and a Notion doc.

Over the past decade, competition have become tougher. Earlier, the question used to be about banks. Founders would ask how to compete with SBI or HDFC or ICICI, and I’d walk them through the typical playbook: banks are slow, their tech is old, their customer experience is terrible, there are gaps everywhere.

That conversation has changed. Now, increasingly, it’s: “How do we compete with the new fintech incumbents?”

PhonePe. Razorpay. Zerodha. CRED. Groww.

That second question is harder. These companies don’t have the same weaknesses as banks. They’re tech-first. Well-funded. Distribution-heavy. Data-rich. They’ve already crossed the trust threshold with the customer segment you’re probably targeting.

Last month, a founder building in the SME lending space put it simply: “The banks don’t scare me. PhonePe scares me.”

So how does a seed-stage startup, with limited capital, no brand, and no distribution, realistically compete and win against an incumbent and a new incumbent?

I’ve been thinking about this for a while, and I keep coming back to the same framework. I think there are only three structural ways a startup can compete against an incumbent:

1. Attack the margin profile

2. Compress time to money

3. Deliver a “Delta 4” experience

I want to be precise about the word “structural” here. I’m not talking about having a better sales team, a sharper brand, or a more charismatic founder. Those things matter, but they’re tactical advantages. They can be copied or hired away.

Structural advantages are rooted in how your business is built. They’re hard to replicate because matching them would require the incumbent to change something foundational about how they operate.

And these structural advantages come from spending time with users. I’ve written about extensively in Edition #65 on Unique Insights.

Everything else is tactics. Let’s walk through each one.

1. Attack the Margin Profile

Let’s start from first principles.

An incumbent’s pricing reflects their cost base, their distribution model, their risk appetite, and their shareholder return expectations. These aren’t strategic choices that a CEO can reverse in a board meeting.

Unless you’re Sergey Brin and you own majority of the voting class shares. Then, yes, you can decide unilaterally to spend several hundred billion dollars to save your company from getting pummelled in a new technological transition.

But for the rest of us mere mortals, the cost base is a structural constraint baked into how the business runs. The branch network costs what it costs. The relationship managers need salaries. The investors expect a certain ROE.

This means every incumbent has a price floor: the minimum they need to charge to keep the business viable.

If a bank needs 600 bps on a lending product to cover origination, servicing, credit losses, and return expectations, and you can build a business that works at 300 bps, you don’t need to be better at anything else. You don’t need better marketing, a slicker app, or a bigger brand. You just need to exist at that price point, and a certain segment of the market will find you.

This is what I mean by structural asymmetry. Your cost structure makes possible a price point that the incumbent literally cannot sustain without breaking their P&L. They’d have to fire their RMs, close their branches, and renegotiate their investor returns to match you. That takes years, , if it happens at all. But more importantly, it requires risk-taking leadership.

Zerodha, from back in 2010’s, is the cleanest Indian example.

Flat ₹20 brokerage. No relationship manager model. No branch network. Bootstrapped from day one, which meant no investor pressure to juice margins for a fundraise. By 2024, Zerodha had over 7.8M active clients, roughly 15% of all active demat accounts in India. Full-service brokers like ICICI Direct or HDFC Securities, with their heavy cost structures and RM models, couldn’t match that pricing without gutting their existing revenue streams.

Nithin Kamath didn’t out-feature them. He built a fundamentally different cost structure and let the pricing speak for itself and still became the most profitable stock broker. Today, their operating stack is base expectation. You’ll rarely find a stock broker veer too far away from this model.

Wise (formerly TransferWise) did this globally for cross-border payments.

Banks price international transfers at opaque spreads of 3–5% (PayPal is/was at 10%), bundled into the exchange rate so you don’t even see the fee. The opacity is the business model. Wise offered the real mid-market FX rate plus a transparent fee, typically under 1%.

By 2024, Wise was processing over $35bn in cross-border volume per quarter, growing ~20% YoY. Lower margin per transaction, but higher trust and higher volume. The transparency itself became a competitive moat because once a customer sees the real rate, the bank’s opaque pricing feels like a scam.

They changed the unit economics of the entire category.

One thing to flag for founders: margin attack only works when your cost structure is fundamentally different. I see a lot of startups that think they’re doing margin attack by offering the product for free hoping to raise prices later, but they’re actually just spending capital to subsidise lower prices. That’s not the same thing. If your free/low pricing depends on your next funding round, you don’t have a structural advantage. You have a runway problem.

The 2021 vintage taught us this clearly. Dozens of fintechs offered zero-fee or heavily subsidised products, grew fast, and then had to raise prices when funding dried up in 2022-23. Customers churned. The incumbents are still standing.

Real margin attack means you can charge less and still make money because of how your business is built.

2. Compress Time to Money

In fintech, trust doesn’t build when users download an app. It doesn’t build when users browse the interface or read the onboarding screens.

Trust builds when money moves.

I’ve been thinking about this a lot lately because of how many pitches I see where the founder obsesses over “engagement” and “DAUs” without talking about the speed of the first transaction. A user who downloads your app and browses for two weeks is not an engaged user. They’re a user who hasn’t decided if they trust you yet.

Here’s how I define it: Time to money = the time between user signup and first completed transaction.

Why does this matter so much? Because in financial services, the first transaction is the trust event. That’s the moment the user takes a real risk on you. They put their money (or their data, or their credit) in your hands and see what happens. Everything before that is browsing. Everything after that is a relationship.

The shorter this window, the faster habit forms, the faster data accrues, and the faster lifetime value compounds. But there’s something deeper going on too. The first transaction creates a psychological commitment. Once a user has moved money through your app/platform and had a good experience, the switching cost goes up dramatically, even if a competitor offers something marginally better.

In my opinion, time to money is the new CAC. You can spend ₹500 acquiring a user, but if they don’t transact for three days, you’ve effectively lost that spend. The user will forget about you, get distracted by a competitor, or simply lose motivation. Whoever captures the first transaction often captures the relationship.

In the early days, Jar was founded with this thesis.

Their initial premise was to get a user to start saving within 60 seconds. The whole app and product experience was designed to enable that. And it worked. Before Jar, the user was expected to do homework and know which mutual fund to invest in - growth funds, large/small/multi cap funds, what’s the expense ratio, what’s the exit load and so on. That’s a lot of jargon-filled, cognitive load. With Jar, users invested in digital gold - an instrument they know and understand.

Square in the US understood this better than anyone.

Before Square, a small merchant wanting to accept card payments dealt with terminal purchases, paperwork, underwriting, and multi-week waiting periods. The process was designed for large merchants and grudgingly adapted for small ones. Square collapsed all of that. Sign up, plug in the reader, accept your first payment on the same day. By 2023, Square’s seller ecosystem was processing over $200 billion in annualised GPV.

Think about what Square actually did: they didn’t make card acceptance cheaper (their take rate was actually higher than traditional acquiring). They made it faster. The speed of activation was the entire product.

Affirm did this for consumer credit at checkout.

No branch visit. No “we’ll get back to you in 5-7 business days.” Credit approval in seconds, at the point of purchase. The transaction happens instantly. Value is experienced instantly. By the time a traditional lender processes the application, Affirm has already captured the customer and the merchant relationship.

In India, I’d argue Slice (now merged with North East Small Finance Bank) compressed time to money for a generation of first-time credit users. The traditional credit card application (documentation, income proof, 2-3 week approval cycles) got collapsed into minutes. That speed mattered because it built trust before the user had a chance to drop off or second-guess.

CRED did this too, their credit card bill payment experience at launch was instant. Today it’s the standard.

Look at what quick commerce has done in India. Zepto, Blinkit, and Swiggy Instamart compressed delivery from "2-3 days" to "10 minutes," and in doing so, they've permanently rewired consumer expectations. A generation of users now considers 30 minutes slow.

That expectation bleeds into every other category, including financial services. If your fintech product has a 48-hour activation cycle, you're competing against a user whose last commercial experience was a 10-minute delivery, where even the payment (UPI) was instant.

You must be thinking, it’s very hard to compress time to money in financial services - KYC process, onboarding, data sharing and more. It’s a hassle. But we’ve seen so many examples over the years of founders figuring out adjacent products just to keep the user engaged and trust the platform, then get the user to go through the core process. This comes from studying user behaviour and developing insights constantly.

For founders thinking about this lever: map out every step between “user signs up” and “user completes first transaction.” Every form field, every verification step, every waiting period. Each one is a point where you lose people. The companies that win on time to money are ruthless about eliminating steps that don’t directly serve trust or compliance.

3. Deliver a “Delta 4” Experience

This concept is attributed to Kunal Shah. The idea: switching happens when the experience gap between the old way and the new way is so wide that going back feels irrational.

4x better. 10x better. Visibly, unmistakably different.

Most product improvements are incremental. A slightly faster load time. A cleaner settings page. A new feature that 15% of users might try. These improvements matter, but they don’t change behaviour. Users will stick with what they know unless the gap is wide enough to justify the effort of switching. And in financial services, where trust is hard-won and switching costs are real (primary bank accounts, credit lines, investment portfolios), the gap has to be very wide.

I think about Delta 4 slightly differently when evaluating pitches. I look for a specific signal: does the user’s emotional state change? Not just their workflow, but their actual feeling about the category.

Here’s what I mean. If your product makes someone feel smart for the first time in a category that previously made them feel stupid, that’s Delta 4. If your product makes someone feel in control in a category that previously made them feel anxious, that’s Delta 4. The experience gap has to be wide enough that the user can articulate it to a friend without you prompting them. “You have to try this” is the Delta 4 test. If users say it spontaneously, you’re there.

Stripe fits here.

If you were a developer in 2012 trying to integrate payment processing, you were dealing with acquiring banks and gateway APIs that looked like they were designed in 1998. Weeks of back-and-forth. Technical documentation that assumed you had a payments team. Stripe offered seven lines of code, clean documentation, transparent pricing, instant sandbox testing.

The experience gap was wide enough that switching felt inevitable. By 2023, Stripe was processing over $1tn in total payment volume. Developers, who had never been the target customer for payments infrastructure, became the most valuable distribution channel in the category.

Paytm’s Soundbox did this for offline merchant payments in India.

Before the Soundbox, a merchant accepting a digital payment had to stop what they were doing, pick up their phone, unlock it, check the notification, confirm the amount, then go back to serving the next customer. For a busy kirana store or a street food stall during the lunch rush, this was painful.

The Soundbox made payment confirmation audible. “Paytm pe aapko 500 rupaye mile.” The merchant hears it and keeps working. The customer hears it too, which eliminated disputes on the spot. The shift was sensory, not just digital.

By late 2024, Paytm had deployed over 11M soundboxes. And here’s the Delta 4 tell: ask any merchant who has one whether they’d go back to checking their phone. They look at you like you’ve lost your mind. The old way (unlock, scroll, squint, confirm) feels absurd once you’ve experienced the new way (just listen). That’s the one-way door. The entire offline merchant payments category in India reorganised around this single device.

One thing I see founders get wrong: they confuse “good UX” with Delta 4. A cleaner interface or a faster loading time doesn’t qualify. You can polish a product to perfection and still be only 20% better than the incumbent. Delta 4 is a discontinuity. The user cannot imagine going back to the old way. If your users would go back to the incumbent after a month if you disappeared, you haven’t achieved Delta 4.

One lever is an advantage. Two is a moat.

Breakout companies almost always attack at least two of these levers simultaneously. Like a 1-2 in boxing: the first lever creates the opening, the second does the damage. And I think this is the most important part of the framework, because it explains why some startups build durable advantages while others get competed away.

Margin + Speed: Zerodha didn’t just offer cheaper brokerage. The account opening was faster. The execution was faster. The entire experience was built around removing friction alongside cost. An incumbent trying to respond had to simultaneously lower prices and overhaul their tech stack. Most couldn’t do both.

Speed + Delta 4: Affirm at checkout is both instant (time compression) and emotionally different from traditional credit applications (Delta 4). A bank trying to match Affirm can’t just build a faster approval process. They also need to change how the credit decision is surfaced, how the offer is presented, and how the user feels at the moment of purchase.

Margin + Delta 4: Wise is both cheaper and more transparent. The experience of seeing the real exchange rate, after years of hidden bank spreads, creates a trust gap that’s hard for incumbents to close. A bank can’t match Wise’s pricing without revealing how much they were previously overcharging, which creates a brand problem.

This is why two levers together are so powerful. Matching one lever is hard. Matching two at the same time requires the incumbent to change two foundational things about their business simultaneously. Most organisations can barely change one.

If you’re building right now and you can only clearly articulate one of these three advantages, think about the another vector you can add to your offering. Ask yourself: what’s the second lever? If you’re cheap, how fast is your activation? If you’re fast, how wide is the experience gap? If the experience is transformative, can you also be structurally cheaper?

So What Does This Mean If You’re Building Today?

If you’re a founder reading this, the question to ask yourself is: Which structural constraint am I changing?

Be specific. “We have better UX” is not a structural answer. “We can underwrite this segment at 200 bps lower because of [specific data advantage]” is. “Our onboarding gets users to first transaction in 90 seconds versus the incumbent’s 3 days” is. “First-time users in our category say they’d never go back to the old way” is.

If the answer to “which constraint are you changing?” is none of the above, you’re probably competing on better UX. Better UX matters. But incumbents have more money, more time, and more distribution to play that game. You won’t out-brand HDFC or out-market PhonePe. You need a structural edge they can’t easily copy.

I could be wrong about the exact framing here. Maybe there’s a fourth or fifth lever I’m missing. But when I look at the breakout fintech stories, in India and globally, they almost always fit into one of these three buckets.

If you see it differently, I’d love to hear it. What structural advantages have you seen startups exploit that don’t fit this framework?

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: Judge me all you want but, this week I’ve been hooked on Golden (Youtube / Spotify) from the excellent Netflix movie K-Pop Demon Hunters. There’s also this video by Vanity Fair interviewing the creators of the song where they dissect what went into making the song. It’s really good.

✨ Call outs

1. [Interview] Varun Mayya’s interview of Demis Hassabis is really good, partly because Varun’s questions were really good. The responses were top notch.

2. [Article] Sidu Ponnappa, founder of RealFast.AI has something to say about the whole “SaaS is dead” debate. And he makes a really strong point. There is no product. Read all about it.

3. [Video] Picking off from my last week’s edition on Gold, here’s a video of rare access to the gold vaults at the Bank of England.

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 102nd edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

Gold is at $5,000 and behaving in ways it never has before. Central banks are buying at all-time highs. The real yield relationship is broken. India is sitting on ₹380 lakh crore ($4.5tn) worth of gold, more than our GDP.

The price volatility over the last few weeks told us more about the state of the global financial system than any earnings call or policy speech this year.

In this edition, I’m breaking down what seems to be happening with gold, why the old rules stopped working, what it means for India’s $4.5tn gold hoard, and what we should be watching over the next 12-18 months.

This is a long, data dense one - grab a cuppa with my song recommendation in the background.

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

Gold at $5,000: What Broke, What It Means, and What to Watch

Quick disclaimer: I'm a private markets investor trying to make sense of macro, markets, and the global economic machine. The world order is far too complex for my brain to comprehend in its entirety. I'm sure I've gotten things wrong here. If you spot errors in the analysis please drop me an email or DM.

I asked a jeweller about gold recently. He said, the weight surprises you first. Holding a standard 10 gram gold bar feels less like holding metal and more like holding frozen time. “It feels final,” he says. In early February 2026, as gold traded around $4,960 per ounce, more than double its price from mid-2023, that physical weight became impossible to ignore.

But the price isn’t the story. The story is what happened in late Jan, 2026.

Gold dropped 9% in a single day. Not over a week. Not during a gradual selloff. In one trading session, gold erased over $600 per ounce, wiping out roughly $600 billion in value. Silver crashed even harder - 26% in the same session. These aren’t normal market corrections. These are “something in the system is off” moves.

For Indian fintech founders, finance executives, and investors, what’s happening with gold matters. Gold stopped following its own rules. And when a 5,000-year-old asset starts behaving in ways it never has before, it’s worth paying attention.

The rules that stopped working

Gold has always been simple to understand. It pays no interest. It generates no cash flow. Prices increase over time because of demand/supply movement. But gold has always been used as currency because it’s not too rare or too common.

Moreover, gold’s not used in industry as much as other metals - 90% of gold is used as jewellery or financial asset. So if the economy tanks, the value of gold doesn’t get affected, since industry doesn’t use it as much.

So when government bonds offer attractive returns after accounting for inflation, what economists call “real yields”, investors avoid gold. Why hold a metal that does nothing when you can earn 5%+ safely from a treasury bond?

This logic has broadly governed gold pricing for decades. When real yields rise, gold falls. When real yields fall, gold rises.

But something broke in 2024 and 2025.

Gold’s rally began when real yields were low, which made sense. But then gold kept climbing even as real yields in the US rose toward 2%. That shouldn’t have happened. Multiple market analysts noted that gold’s “long-standing inverse relationship with real yields seems to have broken down.” In 2025, gold posted its sharpest annual gain since 1979, averaging $3,431 per ounce, up 44% from 2024’s average of $2,386, despite real yields staying near multi-decade highs.

The traditional thermometer broke. And nobody is quite sure why.Three forces changing the game

Three Forces Changing the Game

The most common explanation points to central banks. Since 2022, emerging market central banks have increased their gold purchases fivefold. China’s central bank bought gold every single month for over two consecutive years, even as prices hit record highs. They weren’t trying to buy low and sell high. They were building something else.

To understand what, you need to understand what happened in February 2022.

The Russia precedent: when digital became “revocable”

After Russia invaded Ukraine, Western governments froze approximately $300 billion in Russian central bank reserves. These weren’t physical bills in a Moscow vault. These were digital holdings, numbers on screens in banks in London, New York, and Frankfurt. With a few keystrokes and some coordinated phone calls, Russia’s foreign reserves became inaccessible.

Central banks in China, Turkey, Poland, India, and dozens of other countries watched this and learned a harsh lesson: if your national wealth is a number on a screen in someone else’s country, it isn’t truly yours. Digital reserves, they realised, are essentially “revocable permissions” granted by Western governments.

This wasn’t theoretical anymore. It was demonstrated capability. Add to it, the fact that there are several wars ongoing, governments being toppled and more, countries being acquired, tariff threats and more. Pick your poison - this is a very volatile time for the global economy.

So central banks started buying gold. Not as an investment. As insurance against a world where your reserves could be frozen if you end up on the wrong side of a geopolitical environment.

Central banks as “institutional preppers”

According to World Gold Council data, central banks purchased 250 tonnes of gold in Q1 2025 alone - the highest quarterly purchase ever recorded. This buying continued into 2026 even as gold prices surged past $5,000. If you’re trying to maximise returns, you don’t buy an asset at all-time highs. But if you’re building a parallel financial infrastructure that no one can freeze, sanction, or turn off remotely, price becomes secondary to quantity.

Gold seems to have reassumed its historical role as the “final settlement asset” - the one asset that requires no intermediary, no clearinghouse, no permission from a foreign government to use. Every other major asset, from bank deposits to Treasury bonds to corporate stocks, represents someone else’s promise to pay. Gold just exists.

The geopolitical fragmentation accelerator

Add AI dominance to this mix and you have the world reorganising into competing blocs. The US and its allies on one side. China and its partners on another. A large group of countries trying to stay neutral in between. As this fragmentation accelerates, the question every central bank faces is: whose currency do we trust for the next 50 years?

The dollar still dominates with roughly 59% of global foreign exchange reserves are held in US dollars. But that dominance is eroding. The euro’s share is stable. The renminbi’s share has doubled from 2% to nearly 5% in five years. Gold’s share has been creeping up.

When central banks fear their reserves can be frozen, as the Russia precedent demonstrated, the traditional calculation around gold’s “opportunity cost” completely breaks down. Gold’s zero yield stops mattering when the alternative might be total confiscation.

This is why gold kept rising despite higher interest rates. The risk being priced shifted from “will inflation erode my wealth?” to “will the system even let me access my reserves?”

What this means for India: The $4.5 trillion question

India sits at the centre of this story in a unique way.

India holds the largest private stock of gold in the world. Between households, temples, and family vaults, Indians collectively own an estimated 28,000 tonnes. At current prices, this hoard is worth approximately $4.5 trillion, larger than India’s entire projected GDP of $4 trillion for FY2026.

Now, this isn't comparing apples to apples. GDP is annual economic output (a flow). Gold holdings are accumulated savings (a stock). But that's precisely the point. Generations of Indian families saved in gold, and that gold just doubled in value over two years.

This became a problem for the Indian government that issued its Sovereign Gold Bonds (SGB). Price increases meant, the Indian government’s liability increased dramatically i.e. average +79% from its issue price to ₹1.2tn ($13.3bn). With only 17 tonnes of gold redeemed out of 147 tonnes issued since 2015, the government stopped issuing new SGB tranches since Feb 2024. This bond was supposed to give the government easy money, with low (<3% annual interest cost) liability, but it quickly ballooned to become a $13bn liability (as of 2024) because the SGB value was pegged to the value of gold.

Other than stopping new SGB issuances, the government also recently added tax caveats to the earlier tax-free redemptions. As of the 2026 budget speech, only primary subscribers holding until full 8-year maturity retain tax-free capital gains status. Secondary market purchasers face 12.5% long-term capital gains tax without indexation, even if held to maturity.

And the retail investor entered this party too. As of January 2026, gold ETF inflows jumped 106% MoM to a record ₹24,040cr (~$2.65bn), slightly higher than equity mutual fund inflows of ₹24,029cr (~$2.65bn). Read that again. Gold ETFs beating equity mutual funds. That's a sentiment shift.

The wealth effect that doesn’t quite work

For Indian households that accumulated gold over decades at prices ranging from $300 to $1,500 per ounce, this represents a massive wealth gain. Every ₹1,000 per gram increase in gold prices adds nearly ₹1 lakh to average household wealth through mark-to-market gains.

But what’s strange is Indians aren’t selling. Unlike stock market rallies that often trigger consumption booms, gold price increases in India primarily expand credit capacity rather than spending. Gold in India isn’t just wealth, it’s insurance, inheritance, social security, and cultural identity rolled into one physical asset.

Loans against gold in the personal loan segment, grew 128% YoY and 86% FY YoY growth, reaching ₹3.83 lakh crore ($42.3bn). Borrowers are just able to borrow much more, even with limits on loan to value (LTV).

The banking sector is generally suffering from a problem of lower net interest margins (NIM), driven partially by the RBI’s clamp down on unsecured lending in 2023, which is why this gold price increase was an ideal scenario for banks to grow gold backed loans. The problem could be, if gold’s price swings the other way. Reductions in gold prices, could impact the loan-to-value of the gold backed loans, potentially causing issues for lenders.

This creates a liquidity multiplier throughout rural and semi-urban India. But it comes at a cost.

The import trap

India imports roughly 87% of its gold consumption. When gold prices rise, the import bill rises accordingly, assuming volumes stay constant, which they largely have. Indian demand for gold is remarkably price-inelastic. Weddings still need jewellery. Festivals still drive purchases. Cultural practices don’t adjust quickly to price signals.

The arithmetic is brutal: a 10% rise in global gold prices increases India's gold import bill by roughly 10%. All else equal, this can widen India's current account deficit by approximately 15 basis points of GDP. Higher gold prices mean more pressure on the rupee and more imported inflation.

This puts the Reserve Bank of India in an impossible position. High gold prices support household wealth and credit expansion (good for financial inclusion and growth). But they also strain the rupee and fuel inflation (bad for macroeconomic stability). The RBI can either defend the rupee by raising interest rates, or accept currency depreciation and support growth by keeping rates low. It can't do both.

The obvious question is: why not mobilise the $4.5tn already sitting in Indian households? The Gold Monetisation Scheme, launched in 2015 to let households deposit physical gold with banks and earn interest, has mobilised less than 30 tonnes (as of Nov, 2024) in over a decade. That's a rounding error on 28,000 tonnes of private holdings. Doesn’t seem like Indians want to melt down family jewellery, and face making charges (8-20% of value) that make the economics unattractive. Until someone solves for trust and incentive design, India's gold stays locked in family vaults, enormous latent value.

The currency layer: Dollar, Renminbi, and the fog ahead

Gold’s rise is a symptom of currency system stress. To understand where gold might be going, let’s understand that context.

The dollar dominated global finance since Bretton Woods in 1944 (for a detailed history, read Edition #91 - Trade and Tariffs), not because it’s inherently superior, but through network effects. Oil is priced in dollars. International trade is invoiced in dollars. Financial institutions settle through dollar-based systems like SWIFT. Today, the dollar is used in 88% of all foreign exchange transactions. If you’re a Thai company buying German machinery, you’ll probably pay in dollars even though neither country uses dollars domestically.

China’s challenge: slow but methodical

China isn’t being subtle about this anymore. In early 2026, Xi Jinping publicly called for building a “powerful currency” that achieves global reserve status, his most direct statement yet on challenging dollar dominance. Beijing has been laying the groundwork for over a decade: currency swap agreements with ~50 economies, RMB inclusion in the IMF’s Special Drawing Rights basket in 2016, oil trade settlements with Saudi Arabia in RMB, and the expansion of CIPS (Cross-Border Interbank Payment System) as an alternative to SWIFT, now linking ~4,800 banks across 185 countries. Through 2024-25, RMB’s share of global payments rose from around 2.0% to 4.7%.

But the real ground work is in lending. According to AidData’s research, China has extended $2.2tn in loans and grants across 200 countries between 2000 and 2023, making it the world’s largest official creditor, outspending the US on a more than two-to-one basis and the World Bank by nearly $50 billion in 2023 alone. This isn’t charity. It’s infrastructure for influence. When you finance ports, railways, and power grids across 150+ countries with RMB-denominated loans, you create structural demand for your currency. The playbook isn’t new. The World Bank and IMF built dollar dependence across the developing world through the same mechanism for decades (read more, in Edition #86 - HDFC’s History). China is simply running the same play with different branding it seems. Here’s an insightful post on what will need to happen for China to have it’s way with being the reserve currency.

Despite all this, China faces fundamental barriers: capital controls that prevent deep liquid markets, an opaque and politically influenced legal system that creates a trust deficit, and convertibility restrictions that limit the RMB’s utility. As one analyst put it, Beijing’s immediate goal isn’t to replace the dollar overnight, it’s to build a “strategic counterweight that limits US leverage in a fracturing financial order.” And after watching Russia’s reserves get frozen, who trusts anyone’s digital currency system?

The Stablecoin Wild Card

Here’s where it gets interesting. There’s approximately $216 billion in stablecoins currently in circulation, with 95% pegged to the US dollar. These stablecoins hold reserves in US Treasury bonds_, essentially dollar dominance 2.0, spreading dollar usage through blockchain rails rather than traditional banking infrastructure.

But the same infrastructure that enables dollar-pegged stablecoins can enable RMB-pegged or gold-backed stablecoins. If China launched a gold-backed digital yuan that could be held on any smartphone globally, suddenly the 85% of the world’s population that doesn’t live in “the West” would have an alternative to the dollar in their pocket.

China seems to be tightening its rules on stablecoins lately though.

The Unanswerable Question

Is dollar dominance eroding slowly over 20-30 years, or are we watching the early stages of rapid fragmentation into currency blocs? The slow scenario: the dollar’s share declines from 59% to 45% over a generation, volatility is manageable. The rapid scenario: a crisis causes sudden loss of confidence, trade fractures into blocs i.e. US/West (dollar), China/aligned (RMB), neutrals (maybe something else), and volatility is severe.

Gold’s behaviour suggests markets are pricing something closer to the second scenario. When gold rises despite high real yields and then crashes 9% in a single day before recovering, that’s not normal price discovery. That’s fear trading and forced liquidation.

Not sure if anyone knows which scenario plays out. That’s the honest answer.

What to watch

Three macro signals worth tracking: gold and equities falling simultaneously (forced liquidation, historically precedes funding winters) or credit spreads widening sharply (global credit tightening).

For India specifically: watch rupee volatility (sharp moves of 3-5% in a month signal stress), sustained FII outflows (three consecutive months above $5-7bn is a pattern shift), gold loan NPAs (a leading indicator of household stress despite on-paper wealth gains), and any sudden tightening of gold import duties back toward 15%+ (the current account deficit turning critical).

The only honest conclusion

This is confusing because multiple things are true simultaneously. Indian households are wealthier AND the economy is more vulnerable. Gold is a rational asset AND it signals systemic stress. There’s massive TAM unlocked AND a funding winter is likely. India benefits from geopolitical neutrality AND suffers from trade fragmentation.

These aren’t contradictions. They’re different facets of the same complex situation. And nobody has clean answers because this is the first time in 80 years the reserve currency system faces a serious challenge. Gold’s behaviour is unprecedented. India’s position is unique. The gold price crash showed that even professional traders don’t fully understand the dynamics.

For founders, investors, and finance executives, the answer isn’t “buy gold” or “avoid gold.” The answer is: operate with higher uncertainty buffers. Extend financial runways beyond what feels comfortable. Monitor macro signals, not just your internal metrics. Build flexibility into everything, fundraising timelines, hiring plans, treasury management, product roadmaps.

Gold at $5,000 isn’t the story. Gold behaving in ways it never has before, that’s the story. The rules that governed the global financial system for 80 years are under stress. Maybe they’ll hold. Maybe they’ll break. Maybe they’ll transform into something new.

There are just too many moving parts in our complex economic machine with controls for different parts in each passengers hand.

The weight of that gold bar isn’t just the weight of the metal. It’s the weight of uncertainty about the system itself.

What are you seeing in your corner of the market? I’d love to hear from you. DM me here.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: Neat new band discovered this week: King Bee and the Stingers and their song Don’t Move so Fast (Spotify / Youtube). It’s a beautiful, soul, funk, blues, rock song - think raw harmonica, powerhouse vocals, and a rhythm section that won't let you sit still. Slow down, pay attention, don't move so fast. Felt right for this edition. :)

✨ Call outs

1. [AI Benchmark] N OF 1: What happens when you give an AI model a stock trading account and tell it to go wild and make you money? This benchmark is a great experiment in that direction. Fascinating to see how the models think, trade and win/lose money. h/t to reader and friend - Manan Agarwal.

2. [Report] Crisil and Oister report on alternate investment fund benchmarks. Good analysis and charts on the state of AIFs in India.

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 101st edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

"What's left to innovate in financial services?" An investor asked me this recently, and I've heard variations of this question dozens of times over the past year. The mood around Indian fintech has shifted, funding is slower, regulation is tighter, and many believe the best opportunities are behind us.

I disagree. Strongly.

In today's One Big Thought, I'm laying out exactly what I want builders to work on next. These aren't vague themes, they're specific problems worth dedicating years of work to.

If you've been wondering whether fintech still has room for growth or room for more startups, this post is for you.

As usual, there's also a song recommendation at the end.

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. Send me a DM here.

🤔 One Big Thought

From access to outcomes: Where Indian fintech goes next | Requests for Startups

In Edition #83, back in May 2024, I wrote that Indian fintech ecosystem reached base camp. Not the summit. Over the past year, I’ve been asked the same question by founders, operators, and investors: “Are you still excited about fintech in India?”. One investor I spoke to literally asked me “what’s left to innovate in financial service?”. If I had a dollar for every time I was asked questions like these…

Jokes aside, my answer is yes - I’m still excited about fintech in India. But for very different reasons than five years ago. Take this post with a pinch of salt from someone who’s biased and optimistic about India and financial services.

The mood has changed

If you spend time in the ecosystem today, the tone is noticeably different.

Five years ago, fintech was the consensus favourite. Every fund wanted exposure. Every founder pitch had a fintech angle. Growth covered mistakes. Today, funding is slower, regulation is tighter, unit economics matter again, and distribution is expensive.

Many consumer fintechs look similar. Moats are thinner than expected. The optimist has become a skeptic.

Many people read this as a signal that fintech is “done.” I read it differently.

Every industry goes through this phase, when hype fades and reality sets in. The tourists leave. The builders stay. Financial services is not supposed to be easy. It deals with trust, risk, regulation, and people’s money. Hard sectors produce durable companies.

That’s why I’m more excited now than I was at the peak of the hype cycle. Not because the problems are solved, but because the real problems are now visible. And visible problems have buildable solutions.

What the last 15 years actually achieved

Let’s start with clarity. India didn’t just produce fintech startups — it built financial infrastructure at national scale.

Over 15 years, India executed a buildout most countries never achieved. We all know this story all too well. First came identity infrastructure with Aadhaar, which created universal, low-cost identity verification. eKYC reduced onboarding friction dramatically. Then banking access expanded through Jan Dhan accounts, bringing basic accounts to millions. Next, UPI changed retail payments behaviour at national scale - instant settlement became normal and QR-based acceptance exploded. Finally, Account Aggregator is enabling consented financial data sharing and digital document layers matured.

Institutions like the RBI, NPCI and SEBI pushed public digital infrastructure in a way few countries have managed. This matters because most countries build fintech companies first and fix rails later. India built rails first.

If you think of Indian fintech as a startup, the last 15 years were the Idea → MVP → PMF phase. In Edition #83, I argued that Indian fintech has achieved Product Market Fit as a sector and entered its Series A stage. That view hasn’t changed.

But PMF at the rail layer or the access layer does not mean PMF at the outcome layer. And that distinction is critical.

Access improved. Experience did not.

We solved access. As the fintech ecosystem, we’ve made it super simple for users to get onboarded. The past 15 years has been about making better onboarding systems. It’s been about expanding the market. It’s been about improving transparency of information.

All of this was critical as it changed user behaviour. We don’t see people standing in lines to pay their electricity bills anymore. We don’t have to send heaps of paperwork to open a stock broking account. We don’t have to go through weeks of process to get approval for a credit card. We don’t have to know the bank branch manager to open a bank account. These experiences are the rotary dial telephones of financial services. GenZ and younger generations will have no idea we used to live like this.

Despite all of this behavioural shift, experience in using financial services is still broken - not in a cosmetic sense, but in a structural sense. Except for maybe the experience with payments and the stock markets, every other financial product usage experience continues to lag the onboarding experience. And it’s hard to get the experience right because it’s a mix of products and services. Getting financial products right is only half the battle. The real breakdown happens in delivering a 10/10 service.

Credit is still opaque. Users don’t understand true pricing, repayment stress is poorly managed, and collections remain aggressive in parts of the market. Users still struggle to get approved for products beyond just a small ticket short term loan.

Insurance is still distrusted. Claim rejection fear remains high, policy wording is unreadable, mis-selling continues, and post-sale service is weak. Trust breaks at claim time, not purchase time.

Wealth products are still intimidating. There are too many choices, too much jargon, risk is poorly explained, advice is conflicted, and long-term planning is rare.

MSME finance is still painful. Documentation burden is high, underwriting is rigid, working capital cycles are misunderstood by lenders, and cash-flow based lending is underdeveloped.

Support is fragmented. Users must coordinate between bank, lender, app, and insurer. They escalate through bots and navigate multiple grievance systems.

In several of my previous posts, I’ve talked about how the % penetration of digital financial products to total financial products, across sectors, except for payments and stock broking, continue to be at the same levels from 10 years back.

Digital distribution improved. Financial clarity did not. I believe as with most things, Indians have a different consumption pattern than global users, but financial services in the country was imported as is (thanks to help from the IMF), without much customisation to mould around the behaviour or patterns of us Indians. That’s created a gap. That gap is where the next wave will be built.

We’ve seen green shoots of products being built around Indian behaviour patterns, but we’re yet to see it be widespread. This is what excites me. True localisation of financial products and services in India.

Why this cycle is healthier

This cycle is slower, more regulated, and less forgiving. That’s a feature, not a bug.

Three structural shifts stand out. First, regulation is now a design consideration. Earlier, fintech innovation often ran ahead of regulation, but now frameworks are more defined. Digital lending norms are clearer, data consent frameworks exist, payment rules are tighter, and compliance expectations are explicit. This raises entry cost, but it also raises trust. Serious, long term builders prefer this environment.

Second, capital is more selective. Capital is no longer chasing top-of-funnel growth. Investors now ask: Where is the margin? Where is the repeat usage? What changes user behaviour? What survives regulation? This filters shallow models.

Third, founder conviction in new fintech theses has dramatically improved over the past two years. There are more second-time founders, more regulatory literacy, and more distribution realism. This is a better founder base than the 2016–2021 wave, IMO.

Requests for startups: what I hope to see builders to work on next

So with that, instead of listing broad themes and being hand wavy about it, here are specific unsolved problems in my opinion in Indian fintech worth dedicating years of work to.

Many of these will start as point solutions. But the ambition should be to build full suites and platforms that compound into category-defining, multi-billion-dollar outcomes.

I must mention, that some of these ideas/requests might not get VC funded or might not be large valuation businesses or might already have a number of companies building, and that’s okay. We still need builders to build in financial services and I’m on a crusade to get more talent and the best talent to build.

1. Powering Indian commerce

The problem is simple: Indian businesses don’t fail because they lack software. They fail because they lack financial control.

Cashflow visibility is weak, credit access is episodic, and finance tools are fragmented. Most small businesses fail due to cash timing, not profitability.

What should you build? Start with a small business cashflow intelligence engine that ingests bank, GST, and invoice data to create forward cashflow projections, working capital alerts, and lender-ready reports. Sell it to MSMEs, lenders, NBFCs, and accountant networks.

Or build vertical-focused finance for restaurants, logistics, healthcare, creators, or exporters. Each vertical has unique cash cycles and needs that generic tools can’t address.

Or build an AI-native CFO suite that answer real questions like: Can I afford this hire? This expansion? This loan?

The ultimate vision is a true small business financial operating system that combines payments, credit, forecasting, and reconciliation in one place.

Why this can be big: The winner becomes the default financial control plane for Indian commerce. Payments are the wedge. The suite is the outcome.

2. Innovating access & experience

The problem is that access improved faster than understanding.

Users still misjudge risk, mistrust insurance, churn from wealth products, and make short-term financial decisions.

What should you build? Create a financial product explainers layer, a neutral tool that translates loans, insurance, and investments into plain language with scenario modeling. Build “what you actually pay” calculators, claim probability models, fee transparency scores, and product comparison with risk flags. Think of it as financial nutrition labels.

Or build household CFO tools that aren’t just budgeting apps but decision engines. Create loan prepayment optimisers, insurance coverage gap finders, tax-aware investment planners, and risk exposure dashboards. Position it as a financial control center.

Or develop AI-native wealth tools built for Indian behaviour, not western assumptions. Make them language-native and multilingual, working through both text and voice.

Or design insurance products around specific life moments not policy distribution, but claim execution and life event support.

Or embed financial literacy directly into product flows so understanding is built into the transaction itself.

Why this can be big: The company that owns financial decision-making, not just transactions, owns trust. And trust compounds.

3. Modernising financial infrastructure

The thing is that regulation and scale have made financial operations complex again.

Compliance, reconciliation, underwriting, and claims remain painful and manual for most financial institutions.

What should you build? Create compliance infrastructure for financial institutions with API-based compliance checks, audit logs, regulatory reporting generators, lending rule engines, and consent lifecycle managers. This is fintech infra for financial builders.

Or build claims success platforms, not for policy distribution, but for claim execution. Guide claim filing, detect missing documents, pre-audit submissions, escalate correctly, and benchmark claim timelines. Users don’t want policies. They want payouts.

Or reinvent the collections stack. Collections is still blunt-force. Build behaviour-based repayment nudges, hardship detection, restructure workflows, borrower communication intelligence, and regulator-compliant scripts. Better recovery plus less harassment equals win-win.

Or develop AI-driven underwriting engines, identity and consent lifecycle management tools, or advisor productivity software.

Why this can be big: These are not flashy businesses. They are critical infrastructure with high switching costs, deep moats, and large ACVs.

4. Reimagining business finance & global flows

The problem is that Indian businesses are increasingly global, but financial infrastructure is still local.

Cross-border finance is slow, expensive, and fragmented.

What should you build? Create account aggregator native products. Examples include dynamic credit limits, risk-based insurance pricing, savings automation, portfolio rebalancing, and SME underwriting engines.

Or build SME-first cross-border payments infrastructure, digital trade finance platforms, merchant-of-record infrastructure, or cross-border business banking.

Or create fintech + SaaS hybrids. The strongest new models will combine financial rails with workflow software. Examples include treasury plus accounting, payroll plus credit, invoicing plus lending, or tax plus cashflow.

Why this can be big: The next generation of Indian companies will be born global. Whoever becomes their financial backbone wins a durable, expanding market.

A crucial caveat on ambition

Most large fintech companies did not start as full suites. They started as a narrow workflow, a painful wedge, or a single insight. But they expanded relentlessly.

Start focused. Earn trust. Own a workflow. Then build horizontally and vertically. Ambition matters as much as execution.

What will not work easily

Worth stating clearly: some categories are structurally hard today.

Undifferentiated digital lending, cashback-led acquisition, copycat neobanks, feature-only investment apps, and pure distribution plays without margin control are all uphill battles. Distribution without depth is expensive now.

The next phase: from reach to depth

The first wave of Indian fintech expanded reach. The next wave must improve depth.

We need to move from access to outcomes, from signup to success, from transaction to resilience, from product to decision, from distribution to trust, and from data to decisions.

India has already built the rails. Now we need builders willing to do the slower, harder work of building intelligence, clarity, and control on top of them.

We’re still climbing

Base camp meant rails built, PMF achieved, regulators engaged, and talent attracted. The climb ahead is different - less sprint, more altitude, more discipline, better gear.

For many, base camp was enough. And that’s fine. For builders, the summit still matters.

Welcome aboard. We’re still climbing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: Black by Pearl Jam (Youtube/Spotify). One of the greatest rock ballads ever written. Eddie Vedder’s vocals are haunting, the guitar work is beautiful, and it’s been stuck in my head all week. Pearl Jam’s probably my favourite band of all time, and I keep coming back thanks to Eddie’s vocals and energy.

✨ Call outs

1. [Report] State of the AI markets by a16z: Stunning report by a16z, gives an analytical view of the AI market with a slight bias towards a16z’s portfolio. Good read nonetheless.

2. [Video] Interview of India’s Chief Economic Advisor: Hour long interview with some good questions but even better responses. Never heard the CEA give a podcast interview before.

3. [Game] Crystal ball trading: Ever wished to trade stocks if you knew the front page of tomorrow’s newspaper? This is a solid simulation game to see how well you’d trade stocks and bonds with a starting value $1M. Try it!

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 100th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

A century of editions! When I started this newsletter back in 2020, I didn’t think I’d make it past ten, let alone a hundred. What began as a way to organise my own thinking about fintech has turned into something I genuinely look forward to writing. Thank you for reading, for sharing, for the feedback, and for sticking around through the long posts and the occasional rant. Here’s to the next hundred!

This week’s edition is a little different. The One Big Thought isn’t about a market trend or a company. It’s about something I experienced firsthand. I attended RBI’s open house, Finquiry, in Mumbai last week, and what I saw changed how I think about the regulator. I’m sharing it because I think more founders need to be in that room.

In Top Three this week: Wint Wealth’s Series B signals that online bond platforms have survived the regulatory gauntlet and are now scaling, FIU-IND’s updated AML guidelines are turning crypto platforms into full-fledged regulated financial institutions, and PhonePe got SEBI clearance for its IPO with a target listing in mid-2026.

Thank you for supporting me and sticking around. Enjoy another satisfying week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. I’m at os@osborne.vc

🤔 One Big Thought

Inside RBI’s Finquiry: What I Didn’t Expect From India’s Most Powerful Regulator

Last week, I attended RBI’s open house i.e. Finquiry, in Mumbai, partly because I was curious and partly because I’ve always believed fintech innovation and regulation need to work together, not against each other.

I’ve always wanted to contribute to policy making in some way, and I think there needs to be more collaboration between fintech builders and the regulator. So when I saw the open house announced, I figured I’d go see for myself who the people behind India’s most consequential regulator actually are and how they operate.

I went in with no expectations, just curiosity. I left with more respect for what the RBI does and how open they are to learn from the fintech ecosystem. I’m sharing this experience so more founders, operators, and investors show up next time, because we really should.

Walking into RBI felt like entering a temple

Walking into the RBI building felt surreal to me - like entering a religious place. It had a quiet seriousness. It’s a historical building, beautiful circular atrium in the centre, banking counters around this atrium, stunning large columns supporting the roof! So much history in that building. Wish I got to see more of it. Alas, I was handed my entry pass and then I was guided up to the second floor conference room.

I entered the conference room a little late (parking in Mumbai!), just as Mr. Suvendu Pati, the Chief General Manager (CGM) of RBI’s Fintech department, was wrapping up his opening remarks. The topic for the day was Consumer Protection in Financial Services, and the format was fairly straightforward - an opening address by Mr. Pati, followed by remarks from (I missed his name because I was late, but I think..) someone in the Ombudsman department, then an address by the CEO of FACE (the SRO) on new consumer protection standards they’ve introduced, and finally an open Q&A with both online and in-person participants.

There were about 6–8 people from RBI in the room and roughly 30–35 people in person from the fintech ecosystem - founders, product managers, professors, legal teams from banks. The session was scheduled from 3 to 5pm, but it went well past 6pm because the RBI team kept encouraging participants to ask questions and then answering them.

What surprised me about RBI

I didn’t know what to expect going in, but I was pleasantly surprised on multiple fronts.

Firstly, the RBI fintech team, especially Mr. Pati, is sharp, very sharp. What stood out immediately was their command over detail. They quoted regulation names, notification dates, clause numbers, and even international precedents from global regulators, all without checking any notes. Yes, it’s their job to know this stuff, but I’ve seen senior lawyers pause mid-conversation to look up legal references. This was a completely different level of preparation, and no question was treated as “too dumb” to answer.

What made this feel like a real public hearing rather than a “good to have” event was that anyone could raise their hand and anyone could ask whatever they wanted - directly to the CGM of RBI’s Fintech Department. Most answers came directly from the CGM himself, and he didn’t dodge the tough questions. His attitude throughout was what surprised me most - not defensive, not bureaucratic, not dismissive. The tone was essentially: “Help us understand what’s happening on the ground so we can improve this together.”

Their stated objective was clear: address concerns where regulations seem too rigid, and understand what can be done to improve user protection. That kind of openness from a regulator is genuinely rare.

As a side note, I loved that they served the classic government office snacks - coconut water, puff pastry, jeera biscuits, dry fruits, and a chutney sandwich. Standard Mumbai fare. Small detail, nice touch.

One moment that stayed with me

At one point during the Q&A, a founder kept pushing back on the CGM’s explanation of a particular regulation/process. The founder repeated the same point multiple times, refused to accept the CGMs reasoning, and ended up taking a lot of time while everyone in the room grew increasingly uncomfortable. Some folks tried to intervene and move things along.

But the CGM stopped them. He said something to the effect of: “Let’s not encroach on someone else’s time. Let him finish.”

Then he calmly addressed the founder’s concern point by point, continuing until the founder himself had nothing more to add.

I don’t know many senior government officials or frankly, many people in positions of power anywhere, who would handle a situation like that with such patience. He’s the CGM - he had every reason to shut that person and move on. He chose not to. That moment really stayed with me.

The conversation after that changed my mind

After the formal session ended, I stuck around and got talking with a few members of RBI’s fintech team. We discussed increasing the participation from the startup ecosystem, and I mentioned something I strongly believe: senior fintech talent had left the ecosystem.

Founders and CXOs who built the first wave of Indian fintech are now moving into AI, healthcare, and SaaS. One big reason, in my view, is the perception of regulatory overhang or the sense that building in fintech means constantly fighting an uphill battle against regulators and banks.

Now, I brought this up because it’s something I’ve been wanting to do something about for a while. And honestly, I expected a polite dismissal, or at best a “that’s not really our problem” response. Most government folks would typically brush this away and put the onus back on you.

Instead, the RBI team leaned in. They asked questions. They wanted data on talent movement. They asked what RBI could do differently to attract builders back to the ecosystem.

That genuinely surprised me. This is completely not what I expected from the regulators.

It’s easy to hate on things when we don’t know the people behind the organisations. From the outside, RBI looks opaque and intimidating, like a distant authority that can shut down your “little fintech project” overnight if you end up in their crosshairs.

But I didn’t feel that when I met the team. They care a lot about equity in regulations, about consumer protection, and about the spirit of the law rather than just its letter. They seemed open and collaborative, genuinely trying to balance policy with innovation to improve outcomes for everyone.

The biggest letdown: our side

Here’s the uncomfortable part though: the quality of questions from the ecosystem was quite poor.

After the initial consumer protection discussion faded, the Q&A quickly turned into product pitching, “I’m building X, Y regulation is getting in the way” requests, company-specific lobbying, and feature-level complaints. After a point, every startup wanted to position their product or solution as the best or one that should become the gold standard in India. There was very little systems-level thinking happening in the room.

The CGM was gracious about it. He’d mention that founders can send an email to the team with specific product details, but then he’d still take the time to talk through RBI’s thought process around regulations for their specific sector. Very few people, though, asked about the things that actually matter to the whole ecosystem: industry-wide grievance frameworks, fraud liability standards, dispute resolution infrastructure, interoperability challenges, data responsibility, or how to build long-term consumer trust.

We have an opportunity to engage directly with the regulator on issues that matter to everyone, and we largely wasted it on narrow, self-interested questions.

If founders want better regulation, we need to show up with better questions, prepared questions backed by evidence and framed around industry-wide perspectives, not feature requests for our specific products.

So how can we do better next time? Founders - please come prepared with regulations you want to discuss rather than vague complaints about regulations impeding growth. Founders could coordinate before these open houses to identify the top three to five systemic issues affecting multiple players, so the conversation is focused and productive. You could compile some data so questions are grounded in evidence rather than anecdote. The fintech community needs to treat these sessions as what they actually are, a chance to shape policy, not a pitch meeting.

My request to founders

As the fintech ecosystem, we need to do better and actually participate in these open houses. Tell the regulator what you’re doing, engage with them directly. Even if there’s no specific regulation for your product yet, even if you think regulation should exist for your category but doesn’t - you should be in the room. No topic seemed off-topic for them at this forum.

No one had preferential access at Finquiry. Anyone could attend and anyone could speak. That openness exists. We’re just not using it the best we should.

So my one request for 2026 is simple: engage more with RBI. Tell them about your products, your solutions, your ideas. Keep an eye on their social channels for future dates and topics. Show up, and ask better questions.

As an investor, I walked out of Finquiry feeling more optimistic about fintech in India than I have in a while. Not because regulation is becoming easier or because I expect a wave of relaxations. But because my perception about the regulator changed after the open house.

The regulator is clearly willing to have an open channel with the fintech ecosystem. Yes, this was my first time interfacing with a regulator directly like that and maybe things aren’t as simple/optimistic as I’m making it out to be. But there seems to be clear intent from the regulator to engage. I’ll take that.

Unfortunately, at that forum, I didn’t see strong participation from our side. So I’ll end with a plea to founders: please attend and actively participate in these open houses.

Fintech in India won’t grow by avoiding RBI/SEBI. It’ll grow by engaging with them.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

3️⃣ Fintech Top Three

1. Wint Wealth, an online corporate bond distribution startup, raised INR 250 crore ($27M) in Series B funding.

   • Impressive growth: WintWealth was among the first startups to launch this corporate bond distribution product. Facilitating ₹8,000cr ($871M). in total bond investments with 300K registered users but most importantly, with zero defaults across its portfolio is very impressive. Good on them for what seems like clear focus on customer trust and success. This is still a small fraction of the $76bn in net new AUM of the mutual fund industry, or the $310bn in net new deposits with banks in 2025. This industry has a long way to go from here, but strong start.

   • Category creation to formalisation and growth: Most of these online bond platforms launched in 2020/2021, when there was a need for the product but virtually no regulations per se. This was an institutional product with broken experiences. There was some form of real estate fractionalisation as an investment product, but again fraught with poor experiences and no regulation. In 2022, SEBI came along with its Online Bond Platform regulation. When a new regulation comes along, there’s usually a fear that it’ll effectively kill the business model or make it unviable for an early stage startup (no matter how good one’s intentions). Besides, regulations almost always favour large incumbents - thankfully this sector/product didn’t have any incumbents. The BluSmart implosion also could have impacted

     growth of passive investment products by these platforms. But this regulatory impact being a net positive is a rarity. The founders and investors in this space correctly decided to continue building, obviously with clear focus on consumer benefit.

   • Consumer behaviour shift for good? This is not an apples for apples comparison but a decent proxy for consumer intent, I guess. As of Dec, 2025, 18% of total mutual fund industry AuM is in passive funds, growing 31% YoY, up from 12% of total AuM in 2021. Over the past 12 quarters, there have been an average of 12 new fund offers launched each quarter in the passive space. NFOs were probably a big contributor for the capital growth in passives. 68% of all retail investors in the industry are invested in at least one passive mutual fund. But we’re in sort of a bull market, one where capital in the markets is aplenty - across all asset classes. We may never know how long this bull run will last. But then again, if we avoid the near term view and opt for the long term view, there’s no doubt that passives will continue to grow as percentage of total AuM. Will digital bonds continue to factor in that growth?

     

2. India’s Financial Intelligence Unit tightened the screws on crypto & virtual digital asset platforms (again)

   • From “benign neglect” to full-spectrum regulation: India’s relationship with crypto has been a slow march from ambiguity to containment.

     • 2018: RBI’s banking ban (later struck down by the Supreme Court in 2020).

     • 2022: 30% tax + 1% TDS policy signalling without a licensing regime.

     • 2023: VDA service providers formally brought under PMLA and FIU-IND reporting obligations.

     • 2026: The latest AML/CFT guidelines make it explicit - crypto is no longer an “edge case,” it’s a first-class financial risk category.

     This update isn’t a philosophical shift. It’s an operational one. The government had already decided crypto would live inside India’s financial surveillance perimeter. Now it’s making sure platforms actually behave like regulated financial institutions.

   • What actually updated (and why it matters): Three things stand out in the updated regulations to me.

     • Full fledged registration: VDA platforms now go through a real registration and review process with FIU-IND, including in-person walkthroughs of AML systems, travel rule tooling, sanctions screening, and incident response. This isn’t checkbox compliance anymore.
       Implication: fly-by-night exchanges, offshore entities serving Indians quietly, and “we’ll fix compliance later” startups are effectively dead.

     • The travel rule + wallet scrutiny
       Transfers to and from unhosted wallets, interactions with foreign VASPs, and use of mixers / privacy tokens are explicitly flagged as higher-risk. Platforms are expected to collect originator–beneficiary data, apply enhanced due diligence, and justify how they monitor cross-chain flows.
       Implication: expect more blocked withdrawals, friction for DeFi-native users, and tighter on/off-ramp controls.

     • Governance is being regulated, not just transactions: The rules go deep into: appointing a designated director and principal officer, board-level AML oversight, periodic enterprise-wide risk assessments, independent audits, staff training and escalation protocols. This is basically RBI-style compliance architecture being copy-pasted into crypto. Implication: fixed costs go up materially. Only scaled or well-capitalised players survive.

   • Other notable changes (that didn’t make the headline list): Identity verification is now important (live selfie, liveness checks, geo-tagging, device/IP fingerprinting, stronger bank verification). High-risk users must go through periodic re-KYC and risk reclassification. STRs are being actively analysed to build fraud typologies, not just filed and forgotten. And offshore platforms are explicitly in scope if they serve Indian users. Cybersecurity is now a regulatory input, not just an IT problem. None of this changes the product surface dramatically for users. But it materially changes the cost structure, legal risk, and operational maturity required to play this game.

   • Where the opportunity could be (and it’s not just exchanges): At face value, this regulation looks like pure friction. Structurally, it creates a new SaaS market: compliance tooling for crypto — travel rule systems, wallet risk scoring, STR automation, cross-chain monitoring, audit trails. Think “Signzy for crypto,” but chain-aware and regulator-friendly. Longer term, this also quietly lays the groundwork for regulated on-chain payments and stablecoins. If and when RBI formalises a private stablecoin or on-chain settlement regime, only players already inside this AML perimeter will qualify. Net effect: less Coinbase, more “AWS for regulated crypto,” and a small set of too-compliant-to-fail platforms that survive the regulatory grind.

3. PhonePe obtained regulatory clearance for Initial Public Offering (IPO) targeting mid-2026 listing with 100% offer for sale.

   • PhonePe’s retail payments scale dwarfs most content and consumer platforms: PhonePe’s scale is staggering with over 650M+ registered users (2x the population of USA) and ~301M monthly active users as of late 2025, and approaching 11M monthly active merchants. It processes hundreds of millions of transactions daily, with annualised TPV in the multiple ₹125–₹150tn+ ($1.5tn+) and consistent leadership in UPI volume at ~45–48% of the ecosystem. That places PhonePe’s throughput on par with or ahead of major global payments companies in sheer payments volume, even if the core UPI rails themselves don’t directly generate revenue. On the financials side, total income grew 33% YoY in FY25 to ₹7,630cr ($833M). Revenue from payment services as a % of total revenue declined from 97.7% in FY23 to 91.3% in FY25 and further to 86.9% in H1FY26.

   • Revenue diversification is underway but payments still dominate: PhonePe’s monetisation narrative is shifting. Historically, the company made almost all of its income indirectly through payments services and incentives tied to UPI and wallet usage (which itself has near-zero merchant discount revenue). Revenue from payment services as a % of total revenue declined from 98% in FY23 to 91% in FY25 and further to 87% in H1FY26. Revenue from Lending and Insurance Distribution more than doubled to ₹4,530cr ($495M) in H1FY26. With its immense distribution, attaching ancillary financial products, ads and other products is basically playing a conversion game with high margin products. I suspect it’ll be easier for PhonePe to reduce its payments income share in the future. Besides, a company of this size in any other country would have had multiple billions in payments revenue.

   • The UPI market-cap problem regulation is alive… but effectively deferred: PhonePe’s dominance has raised regulatory eyebrows for years. The National Payments Corporation of India (NPCI) proposed a 30% UPI volume cap on any single third-party app to reduce concentration risk. But enforcement has been delayed repeatedly, most recently pushed out to the end of 2026, with little clarity on if or how it will be applied. In practice, this cap has become a talking point more than an active constraint. That regulatory laxity is not a guaranteed long-term “free pass,” but seem to be unlikely to derail PhonePe’s near-term trajectory.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here’s an earworm: Supersonic by Oasis (Youtube / Spotify). This one's been on loop for me lately - loud, confident, and weirdly motivating. Good positive energy for powering through a long week.

✨ Call outs

1. NSE Passive Funds dashboard: didn’t know this existed, but the NSE has a beautiful dashboard on the passive mutual fund industry. It’s a neat visualisation of the growth of passives. Just wish it had the ability to select time period as a filter.

2. India in Charts: My friends Rahul and Aaryaman have been on a tear lately - banger after banger with their recent posts. This post - India in Charts, is one such banger. Painfully sourced charts on India that make you stop and think.

👋🏾 That’s all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 99th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets. Almost at 100 editions!

This is a special edition. I've spent the past few weeks diving deep into something that's been nagging at me since earnings season wrapped up.

India's financial services companies just reported Q1 FY26 results, and beneath the usual metrics lies a story most people were missing. While everyone focuses on individual company performance, I kept seeing the same pattern repeat across 23 different institutions - from mega banks to insurers to listed fintech companies.

Credit stress is forcing uncomfortable decisions. Technology is delivering real productivity gains. But most importantly, companies are making radically different strategic bets in response to identical market conditions.

The implications for where Indian household savings flow next, and which business models survive the current cycle, are enormous.

In this edition, I break down what I found across all 23 companies and why this strategic split might be an important trend shaping India's financial sector.

Thank you for supporting me and sticking around. Enjoy another great week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. I’m at os@osborne.vc

🤔 One Big Thought

Trends and Insights Report - Q1 FY26: The Great Divide

Why India's financial giants are choosing opposite strategies driven by credit stress

India's financial services sector just revealed a defining moment. After analysing the Q1 FY26 performance of 23 largest financial companies across sectors, one pattern dominates: institutions are making radically different strategic bets in response to similar market pressures.

The catalyst? Widespread credit deterioration that's forcing uncomfortable choices.

Banks and NBFCs are reporting customers with 3+ active loans becoming the norm, not the exception. Commercial vehicle financing faces structural headwinds. Even home loans show stress amid brutal pricing competition. The numbers tell the story - Kotak Bank's PAT fell 7% YoY while net interest margins compressed 37 basis points.

This credit stress is creating two distinct responses:

The Defensive Play: Established institutions are pulling back. Multiple banks are reducing unsecured portfolios, tightening underwriting, and prioritising margin preservation. The message is clear - protect what you have.

The Aggressive Expansion: Growth-stage companies are doing the opposite. PB Fintech's management stated they're "clearly not optimising for profits right now" while targeting ₹1 lakh crore ($11bn) premiums annually by 2030. Paytm achieved profitability but immediately signalled continued aggressive investment in merchant payments and international expansion.

Meanwhile, technology deployment is accelerating across both camps. ICICI Lombard cut headcount by 22% while boosting productivity 78% through digital consolidation. SBI deployed GenAI across 10+ use cases, processing $4bn in loans during Q1 via YONO. Bajaj Finance is quick to redeploy 150 employees to build out the firm’s AI strategy. At the same time Paytm wants to become an AI first company for all customer facing products.

Oddly, other than the above mentioned firms, there’s no mention of a concrete AI strategy or roll out from any of the other firms. It’s mid-2025!

The implications extend beyond individual companies. This strategic divergence is reshaping capital allocation across India's financial system, with profound consequences for where household savings flow next.

I've documented the complete analysis across all 23 companies, across 82 slides, including the specific strategies each is pursuing, management statements of bullishness or bearishness and what this means for customers, investors and general market watchers.

Read the full 82-slide report

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here's an earworm: This week I’d been binging on my third favourite band. This song casually crept in except it wasn’t from the same band. This song is She’s Not Gone Yet but She’s Leaving by The Fratellis (Youtube / Spotify). The sound and vocals are exactly like my third fav band - grunge-y, high pitched and fun. Can you guess which band I’m referring to?

✨ Call outs

1. [REPORT] SVB released its State of the Markets report for H1 2025 and it’s a data-packed read. It gives a very detailed review of the current state of the VC ecosystem and startup financing trends. Definitely worth your time.

2. [VIDEO] I’ve been learning a lot about nuclear energy lately so learning about the design choice for the massive “exhaust” from the nuclear facility seemed like a nice thread to pull on. Not knowing what to expect, the video surprised me in its simplicity, puns and general wonder for the unknown using a slice of pizza as a metaphor.

3. [VIDEO] Another weirdly fascinating video the algorithm sent my way, was this new “light-field” display concept video. It breaks down a new kind of display technology for the AR glasses future being built out of Switzerland. The technology is complex and still in prototype stage but the implications could be massive if miniaturised and if energy requirements are met.

👋🏾 That's all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 98th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets. Just two editions away from the big 100!

The unwritten rules of Indian finance are being rewritten in real-time. Across insurance, lending, and payments, we're seeing established models challenged and new precedents being set that will define the market for years to come.

In our Top Three this week, we dissect these pivotal moments:

1. The Insurance Gambit: Private equity firm WestBridge is creating a fascinating regulatory puzzle, owning major stakes in two potentially competing insurers. We explore the conflict, the "Berkshire model" for PE in India, and why the insurance sector is suddenly so hot.

2. The "Vanilla" Gold Rush: Tech VCs are falling in love with "boring" but profitable NBFCs. The $230M acquisition of Infinity Fincorp is the latest example of a massive trend where stable returns are beating high-tech hype.

3. The End of Free UPI? ICICI Bank made a bold move, introducing a transaction fee for UPI aggregators. This small fee could signal a tectonic shift towards a sustainable UPI model, finally answering the question of who pays for the world's best payments infrastructure.

These aren't just headlines; they are signs of a maturing market grappling with its own success.

If you want to know where the market is headed, follow the smart money. Here's the inside track.

Thank you for supporting me and sticking around. Enjoy another great week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. I’m at os@osborne.vc

3️⃣ Fintech Top Three

1️⃣ PE Firms Are Building Insurance Portfolios: How PE Money Is Reshaping India's Coverage Game

WestBridge, the PE firm, partnered with Neelesh Garg, ex-CEO of TATA AIG General Insurance, to form Kiwi General Insurance - a digital first insurance company. In July, 2025, Kiwi received R1 approval from IRDAI to set up a general insurance company in India.

1. WestBridge joins other PE firms: WestBridge joined the league of Fairfax, Warburg Pincus, Premji Invest, TPG Capital, TrueNorth, Carlyle and other PE firms in owning stakes in Indian insurance companies. WestBridge will own 60% stake in Kiwi General Insurance, along with its existing 40% stake in Star Health Insurance.

2. Doesn't the ownership create a conflict? Banks have always owned large stakes in general, life and health insurance companies, so there's no issue per se. But this is the first time the regulator is faced with a private holder owning such stakes. With WestBridge owning a meaningful stake in both Kiwi and Star Health, the regulator is in a fix of its own making. Kiwi could launch health insurance products, but will directly compete with Star Health. There already seems to be talk of IRDAI potentially curbing Kiwi from launching retail health products. But there's almost a year long journey to go from the R1 phase of approval to R3 and final "go live". Will be interesting to see how IRDAI tackles this, because this will set precedent.

3. Health insurance is key for new insurers: The motor insurance market seems to have plateaued though and any growth seems to be a factor of price increase. Growth in the market is coming from health insurance segment with some factor of price increase. So health is clearly important for any insurer. However, claims ratios are increasing for the insurers and could impact profitability.

4. Why would PE invest in insurance? We all know the now famous Berkshire model of investing in insurance businesses and using the "premium float" to invest in other businesses. Today, global PE firms like Apollo Global and KKR have tweaked that model in buying out insurance companies - Athene and Global Atlantic, and using that insurance premium float, to lever up and launch private credit products, creating a "financial perpetual machine". I'm not going to get into the pros and cons of that model, but it's SUPER interesting to read up on, if you're interested in it. In India though, all insurance companies have to follow the IRDAI (Investment) Regulations, 2016 (page 57 onward in English). This regulation basically outlays what an insurer can do and cannot do with the customer funds and shareholder funds. There cannot be any deviation in adhering to this regulation. General insurers in India can invest in approved securities like government bonds, AAA-rated corporate debt, infrastructure instruments, and select equities and mutual funds, with IRDAI-defined exposure limits and diversification norms. Insurers can invest in public markets, but only in certain, stable stocks. So the universe of investable assets is very limited. But the Indian insurance sector is hot and owning an insurance asset is high growth at best and stable growth at minimum. This means, the return profiles are relatively low risk to hedge a portfolio for decent returns.

5. Insurance sector is heating up: The IRDAI is not too quick in doling out insurance license approvals though. GoDigit and Acko received general insurance license approvals in 2016/2017. Then Kshema General Insurance and Narayana Health Insurance received approvals in 2022 and 2024 respectively. That's it. Only in 2025 did it approve Vauleattics, Digit's reinsurance business, and potentially Reliance Jio and Allianz' reinsurance JV. There have been several approvals for insurance brokers in the recent past. Even on the regulatory front, the IRDAI and Ministry of Finance have introduced reforms to entice more participants in the insurance industry including removing the FDI ownership cap of 75%, composite broker license, flexible pricing for motor insurance, PE participation in insurance including as a promoter and much more.

6. Can IRDAI do more?: Over the past decade, it surely seemed like IRDAI, India's insurance regulator, has been slow to implement reforms in India's severely under-penetrated insurance market. And while that's in some ways true, IRDAI, has taken a risk-off approach to letting participants operate in this market. Startup founders are not allowed, VC firms are not allowed and many other restrictions. There are pros and cons to allowing it, but at India's population scale, we need to be more risk-on and get more people to participate and increase India's insurance participation.

2️⃣ Hottest new fintech in town

Infinity Fincorp, a small business finance company, raised $70M+ in 2025. Partners Group acquired a 75% stake in Infinity Fincorp for $230M.

Takeaways:

1. The master of one: Infinity Fincorp isn't really a fintech, but a vanilla financial services company. Founded in 2016, it received its NBFC license in 2017 and has just one product - loans against property. It's grown with that one product to 120 branches across eight states (3 states in south contribute 75% of AUM) and 50,000+ small and micro business customers across agriculture, trading and manufacturing. Since 2016, Infinity disbursed more than 2000cr ($230M) in loans, 42% of disbursals was just in the last two years.

2. Why is Infinity hot these days? Going by its recent fund raise activity, Infinity announced raising $35M in Jan, 2025 and then soon after, announced raising another $40M in April, 2025. And now, as of June, 2025, it announced the majority stake sale to Partners Group. I'm not able to verify, but I think, Infinity Fincorp was founded by and majority owned by TrueNorth, the PE Fund, or a fund advised by TrueNorth. Seems like this would have been a good return profile for them over the nine years since founding. This is PE working well.

3. Strong financial growth: Infinity seems to be growing at breakneck pace, for a single product NBFC. In FY25 (ended Mar, 2025), it did INR 240cr ($28M) in revenue, that's 67% more than the INR 144cr ($16M) it did in FY24 (ended Mar, 2024). It's net profit (PAT) for FY25 was INR 44cr ($5M), compared to INR 26cr ($3M) it did in FY24. It's gross NPA (DPD90+) stood at 1.94% at at Dec, 2024 and net NPA was 1.42%. This may seem really strong loss performance given the small business customer base, but for a secured product, it's higher than expected while nothing that cannot be absorbed by its net interest margins. All that being said, Infinity grew its AuM to a strong 1,083cr ($124M) as of Dec, 2024. I don't have much to go on (don't know book value, don't know enterprise value etc.), and while price/revenue multiple is not the right metric, its 10-11x revenue multiple seems high. Infinity is as vanilla as it gets in terms of business model, and yes, growth is better than most at 67% YoY, but even then, valuation seems rich for such a straightforward lending business. Great outcome for the business and the early investors.

4. Vanilla NBFCs are hot among VCs in 2025: Jungle Ventures and Beams Fintech are not the only VCs playing this game though. Every other large VC in India is buying stakes in traditional NBFCs. Elevation, PeakXV, Stellaris Ventures, Saison Capital and others seem to be hedging their portfolio with these investments. Elevation invested in Mahaveer Finance's $24M round, Stellaris and Saison Capital invested in Techfino's $8M round, PeakXV invested in SarvaGram's $67M round. All being secured/small business finance in some way. Haven't done the math on these businesses, but I'm assuming a similar higher revenue multiple across transactions and reflects the premium investors are willing to pay for predictable, secured lending businesses in an uncertain market.

3️⃣ To be free or not to be, that is the UPI question

ICICI Bank introduced a 2-4 basis points transaction fee for payment aggregators using its Unified Payments Interface (UPI) services.

1. Long time reader? If you've been a long-time reader of Fintech Inside, you'd have read my in-depth analysis of UPI in Edition #66 in 2022, its infrastructure, costs and my take. Back then, I said that 1) UPI should not be free, 2) it should be charged at a flat 10 basis points per transaction, and 3) that most likely the government will not let a fee be introduced. The jury is still out, but i'm claiming I was right about 2 out of 3 counts. Surprised, that it took only 3 years and even the government is discussing introducing a "sustainable model" for UPI.

2. Tables are turning: ICICI Bank made a bold move. It introduced a 2-4 basis points transaction fee for payment aggregators (PA) using its UPI rails. This is the first time a large private bank has openly charged aggregators for UPI access. Surprisingly the fee structure is what I would have imagined HDFC would have designed. If the PA uses an ICICI escrow account, then the fee is 2 basis points, else, it's 4 basis points. And if the PA directly settles between the user and the merchant, to the merchants ICICI account, then its free. Fascinating fee structure, favouring using ICICI's banking services.

3. Why does this matter? Since inception, UPI was “free” for merchants and aggregators. Banks absorbed the infrastructure cost. Axis and YES Bank charged PA's some portion of the cost. Then, UPI volumes exploded, especially on the peer to peer side, causing serious strain on the infrastructure. ICICI’s fee is small, just 2-4 bps, but it sets a precedent.

4. What happens next? Payment aggregators will face a tough choice. Absorb the fee, or pass it on to merchants. Some may try to negotiate. Others may look for workarounds. But the era of “free” UPI for business is likely over. The bigger question: if UPI economics change, what happens to the fintech business models built on free infrastructure?

5. RBI Governor chimes in: At an event, the RBI Governor said "This (UPI) is an important infrastructure. The government has taken a view it should be available free and the government is subsidising it. And I would say it has borne good fruits,” Malhotra said. “The important thing is that the UPI, or any other payment system for that matter, is accessible, cheap, secure, and sustainable…and it will be sustainable only if someone bears the costs. So as long as it’s the government or someone else — that’s not so important — the important thing is that costs of any service should be paid, whether collectively or by the user." Quote from Indian Express. His comment and the ICICI UPI fee introduction were completely independent of each other, I think.

6. Caveat: This is early days. The regulator and government may step in. NPCI might issue new guidelines. But one thing is clear—UPI’s economics are changing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here's an earworm: Portugal, the Man, the band famous for the song Feel it Still, has this lighthearted song about refusing to be bogged down by anxiety or the pressures of adulthood. Enjoy What, Me Worry? (Youtube/Spotify). Ideal for the times, I guess.

✨ Call outs

1. [Report] HDFC Ergo’s Annual Report FY25 was a good read about the general insurance industry in general. Health insurance outpacing motor, 32% customers coming direct surpassing brokers, and more such insights packed in the report. However, I must point out that HDFC Ergo is terrible at chart making.

2. [Report] NSE’s Market Pulse report. NSE publishes a solid 300-page monthly report with key stats on India’s public markets and market participation. Good read with data.

3. [Video] How every car brand got their name. Hilarious video on all the major car companies of the world. Most are just the last name of the founders e.g. Ferrari, Lamborghini but some others are quite fascinating.

👋🏾 That's all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 97th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets. So close to the 100th Edition!

While the world economy navigates headwinds and uncertainty, India's financial ecosystem seems to be defying gravity. The evidence from this week is overwhelming, and it points to a market that is not just stable, but aggressively building the future.

In our Top Three this week, I explore three pillars of this strength:

1. The Rule-Breaker: Slice did what many thought was impossible, launching a physical bank branch and becoming the youngest founder-led bank in India. It's a story of grit, pivots, and finally taking on the incumbents on their own turf.

2. The Thriving Market: Forget the global IPO slump. I analysed data from 274 IPO’s over the past 5 years. The insights show Indian IPOs have delivered returns of up to 436% since 2020, creating a vibrant public market that's the envy of the world.

3. The Resilient Foundation: I summarise the key findings from the RBI's Financial Stability Report, which confirms that India's banks are the strongest they've been in a decade, insulating our economy from global shocks.

From a single startup's audacious move to the confidence of an entire nation's capital markets, the message is clear: India is the bright spot. Here’s the deep dive.

Thank you for supporting me and sticking around. Enjoy another great week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. I’m at os@osborne.vc

3️⃣ Fintech Top Three

1️⃣ Slice enters the big leagues as it opens its first branch

slice launched an UPI-based credit card and opened India's first UPI-powered physical bank branch and ATM in Bengaluru.

Takeaways:

1. Historical moment for Indian fintech: I barely saw any excitement about this on social media but this launch is historical in many ways, for slice and the broader ecosystem. For slice, this is its first branch as a small finance bank. This is the first branch to be powered by UPI. This is the beginning of a change in consumer banking in India. For the broader ecosystem, this is India's first new consumer bank launched in India post covid. This is the first time a fintech startup, by itself and not as a consortium, was approved to acquire a small finance bank in India. Rajan Bajaj is the youngest founder a bank in India today. None of these are a small feat by any measure.

2. slice is not new to financial services: Founded in 2016, slice is no stranger to pivots, regulatory action, funding crunch and more. That company has dealt with many blows over the years since founding and still managed to emerge and run a stellar business, finding opportunity where few ventured when it was unpopular. Until RBI regulation in 2022, slice was the 2nd largest "issuer" of BNPL cards in India. It had an NBFC license since 2018/2019 and always operated with that license along with off-book loan facilitation as well. Over the years, it raised $340M+ in equity and debt. Other than Blume Ventures, it does not have a single "top tier" Indian venture fund as an investor.

3. Bank license opens a new whole world for slice: A banking license is coveted, largely because of two reasons a) it's not easily attainable, not even Sachin Bansal or Vijay Shekhar Sharma are able to get one. and more importantly b) you can launch a whole host of financial products without much dependence on others for capital or product integrations. Being able to capture customer deposits is a big moat that literally no other fintech startup can do. This means slice now competes with incumbent banks, not just fintech startups. Consumer deposits are the lowest cost of capital, allowing the lender to offer competitive interest rate loans. Slice will take some time to get there, but it's a huge moat, compared to others. If early product launches are any indicator, Slice could capture an increasingly younger user base away from incumbent banks. It's "not-your-father's-bank" website, app and branches are opinionated and unique.

4. Unfortunately not much is written about the company's financial performance and Rajan has been a bit media shy. But Slice was last known to have ~$100M in revenue (FY23) and close to profitable.

I'm very excited for this new phase for fintech in India which will involve more fin than tech, finally capable of truly taking on incumbent institutions.

2️⃣ India IPO performance shows India’s capital markets is thriving, how long will it last?

I recently listened to the All In Podcast where Thomas Laffont of Coatue Management was talking about US public market performance. He said the companies that went public (excluding SPACs) in US in 2021, were trading 40% down, on average, within one year of going public and currently 50% down, on average. That's terrible performance and it's probably what's driving poor investor confidence in the US public markets.

So I evaluated data on IPO performance in India. Of the 274 Indian companies that went public in the last five years (2020-2024) they've returned, on average, 69-100%. 2020 was an outlier year, with companies returning 436% till date! That's the complete opposite performance of the US, as noted by Thomas Laffont.

Takeaways:

• Number of companies going public is increasing: Since 2020, the number of companies going public grow 5.8x from 16 in 2020 to 93 in 2024 - the most we've seen in over a decade! The market response to companies going public is driving more companies' confidence to go public - diametrically opposite to that of the US markets. Public markets already have a strong pipeline of companies expected to go public over the next couple of years - driven by strong business turnaround by private companies and investor appetite to subscribe to these businesses.

• Success rates: of the companies that went public in each year, 2/3rds are still trading above issue price. This is a very strong performance indicator. 52% delivered positive returns on both listing day and long-term. Success rates for each year are:

  • 2020 - 94%

  • 2021 - 67%

  • 2022 - 67%

  • 2023 - 67%

  • 2024 - 58%

• Average and median returns: Depending on the vintage year, average and median return performance till Jun, 2025 is off the charts.

  • 2020 - avg is 436%, median is 117% (NIFTY50 - 136%)

  • 2021 - avg is 98%, median is 58% (NIFTY50 - 62%)

  • 2022 - avg is 96%, median is 34% (NIFTY50 - 57%)

  • 2023 - avg is 58%, median is 35% (NIFTY50 - 30%)

  • 2024 - avg. is 25%, median is 11%. (NIFTY50 - 4%)

• I'm no public markets expert, so it's hard to pin point what's driving this increasing investor confidence: these IPOs were probably well priced at issue and listing, signalled by strong investor demand and oversubscribed IPOs. There could also be the fact that there's been general strong performance in India's benchmarks. Indian corporate performance is also the strongest we've ever seen - growing 7.2% YoY in FY25 among listed, non-financial services companies, compared to 4.7% YoY in FY24. Lastly, general frenzy in the markets, driven by general price appreciation or the fact that margin trading facilities have grown from less than $1bn in 2020 to $10bn today!

• There's a lot happening in India's public markets that's diametrically opposite to global markets. And with a strong, healthy pipeline of companies expected to go public over the next two years, we can only expect it to get even more exciting. But with consumer spending reducing, while deposits growing, who knows till when this IPO party will last.

  ---

3️⃣ RBI Financial Stability Report - India continues to be the bright spot in the global economies

RBI released it's semi-annual Financial Stability Report. The report is full of deep research, data and insights on India's financial stability. It's a lot to read and analyse, so allow me summarise the most important points for you.

Takeaways:

• Global markets are not doing so well: The global economic outlook has weakened, with major international agencies (IMF, OECD, World Bank) downgrading growth forecasts. This is primarily driven by a cocktail of global issues including trade tensions and elevated geopolitical risks. These external shocks unfortunately can exacerbate existing intrinsic economic vulnerabilities. Soaring public debt, stretched asset valuations and stalled disinflation are increasing risks of global financial instability.

• India is doing relatively better: India remains the fastest-growing major economy, with real GDP projected to grow at a steady 6.5% in FY26. Growth is underpinned by strong domestic demand, insulating the economy from the worst of the global headwinds. Domestic inflation steadily declined, reaching a six-year low of 2.8% in May 2025. The Current Account Deficit (CAD) is eminently manageable at 0.6% of GDP, supported by strong services exports. Financial markets have shown resilience supported by RBI's liquidity measures.

• India's financial system is the most resilient it's ever been:

  • Scheduled Commercial Banks (SCBs) - Strongest in Over a Decade: The Capital to Risk-Weighted Assets Ratio (CRAR) for SCBs reached a record high of 17.3% in March 2025. Asset quality has shown remarkable improvement. The Gross NPA (GNPA) ratio fell to 2.3%, and the Net NPA (NNPA) ratio stood at just 0.5%. Banks continue to report strong earnings and profitability, although a slight moderation in Net Interest Margin (NIM) is noted due to rising deposit costs. Even under a severe stress scenario, the system-level CRAR is projected to remain at 14.2%, well above the regulatory minimum of 9%. Importantly, no single bank is projected to breach the minimum capital requirement.

  • Non-Banking Financial Companies (NBFCs) - Healthy but with Pockets of Vulnerability: The NBFC sector remains healthy with strong capital buffers and robust interest margins. Overall credit growth has accelerated to 20.7%. Delinquency levels have improved, though some stress is visible in the microfinance portfolio. Under a high-risk stress scenario, 15 middle-layer NBFCs might breach the minimum CRAR of 15%, highlighting a specific area for monitoring.

• Systemic Risk, Interconnectedness, and Key Risks: The RBI surveyed folks from the financial ecosystem on systemic risks. They seem to have high confidence in the Indian financial system's stability but decreasing confidence in the global system (as mentioned above). The RBI believes bi-directional exposures among financial entities will continue to grow. While this fosters efficiency, it's also a potential for risk transmission. The inter-bank market is dominated by PSBs. A critical RBI finding is that the system is resilient to contagion. According to the RBI's stress test, a hypothetical failure of the largest bank, NBFC, or HFC would not cause a domino effect leading to the failure of any other bank.

• Regulatory developments and outlook: RBI's key priority is fraud prevention. Introducing the .bank.in domain and rules for telecom service providers to curb digital payment fraud are initiatives in that direction. The RBI also issued updated directions for Digital Lending, Project Finance, and revised the Liquidity Coverage Ratio (LCR) framework to account for the speed of digital withdrawals. There's a proactive stance taken by RBI to shield the Indian economy from domestic and global shocks.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here's an earworm: Remember Seven Nation Army by the brother-sister duo White Stripes? Happened to listen to this beautiful soul/jazz rendition of the song while in Goa. This is "Remasterisée" version, by Ben L'Oncle Soul (Spotify / Youtube).

✨ Call outs

1. [Report] The Rainmaker Group’s Rain Gauge Private Markets report. The report is the best report I’ve come across with data on private matured startups and their performance comparisons with public market comps. I’m sure you’ll learn something new from that report. If you have more time, do check out their public markets report too.

👋🏾 That's all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 96th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

This week I’m bringing back the Fintech Top Three section. It’s always been a popular section and I’m glad to be sharing my hot takes…errr… insights with you.

Busy week for payments across India and Southeast Asia. This week I cover insights and takeaways from the biggest news in fintech.

1. India saw Pine Labs file its draft IPO prospectus with the regulator to raise $300M.

2. Singapore’s MAS announced forming SPaN, a payments body to oversee the country’s payment systems

3. Major stablecoin and digital asset activity in South Korea and Thailand.

Thank you for supporting me and sticking around. Enjoy another great week in fintech!

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. I’m at os@osborne.vc

3️⃣ Fintech Top Three

1️⃣ Pine Labs filed to go public in India

Pine Labs is the latest payments company, after Paytm and Mobikwik, to file to go public in india. Pine Labs is aiming to raise INR 2,600cr ($304M) via fresh issue and ~15M shares via offer for sale. Investors including PeakXV, Actis, Paypal, Mastercard and others will be selling their shares. PayU and Razorpay are reportedly also preparing to go public.

Takeaways:

1. Unique proposition to stand out: I found it interesting that Pine Labs is positioning itself as a digital affordability enabler and a closed loop gift card processor, in addition to the vanilla payments business. It's important to show its uniqueness in a hyper-competitive Indian payments market with potentially better valuation multiple outcomes. The pure POS terminal biz would have not been kind from a market cap perspective - the company will be utilising 10% of the IPO proceeds to purchase 275K POS devices. With this positioning, it's clearly carving out a niche for itself. It's affordability solutions GTV is roughly 45% of total payments GTV (not including pre-paid cards volume) as of Dec, 2024.

2. Good performance numbers: Pine Labs processed $88Bn in total transaction value for the nine months ended Dec, 2024, growing 68% YoY, from its network of 1.7M checkout devices with 916K merchants. It's revenue from operations grew 33% YoY during the same period to $196M i.e. 23 basis points in revenue over transaction value. It's revenue from ex-India operations was ~14%, but it's associated costs were ~19-20%. Pine Labs has 475M pre-paid/rewards cards issued, and earned $56M in revenue from $4.5Bn in GTV at 1.26% revenue margin (compared to 0.17% from payments). This is probably what's driving the 72% contribution margin.

3. Hot, competitive domestic payments market: While Pine Labs are carved a nice niche for itself, the domestic payments market is competitive. Online, Pine Labs competes with Razorpay, PayU and others, Offline it competes with Paytm, Razorpay and Banks, on UPI, it competes with PhonePe, Google Pay and others. It's a lot of attack surfaces. Razorpay always targeted building a consumer-facing brand so merchants demanded for Razorpay over others. That strategy seemed to work with its $293M revenue in FY24 (ended Mar, 24). PayU is used by large enterprises and other merchants for its online payments and infrastructure solutions - generating $675M in revenue in FY25 (ended Mar, 25). Not to forget PhonePe that generated $592M in revenue in FY24.

Exciting period for India's payments market!

2️⃣ Government entities want to be at the centre of the payments revolution

National Payments Corporation of India (NPCI) reported INR 1,500cr ($175M) net profit, growing 42% YoY, and INR 3,000cr ($351M) revenue in FY25. Monetary Authority of Singapore (MAS) and Association of Banks in Singapore (ABS) incorporated Singapore Payments Network (SPaN), a not-for-profit company, to oversee governance of national payment schemes.

Takeaways:

1. Most well run and profitable government entity? At $175M net profit in FY25 (ended Mar, 2025) within 10 years since launch of UPI, NPCI is probably the most well run government "startup". NPCI is innovative and quick to launch newer products to keep UPI and other products relevant. It's the central, non-regulatory body that powers and steers India's hyper competitive payments market, described above.

2. Should NPCI go public? At $175M net profit a year, that's a lot of profit for a non-profit organisation. As Pravin, founder of Dhan and Madhu, founder of M2P, noted NPCI should probably go public. Madhu mentions that other similar associations started as non-profits, converted to for-profits and are now public payments institutions. Those associations are Visa and Mastercard. At today's revenues and growth, NPCI is already a $5-7bn valuation company.

3. Singapore going the India way for retail payments? Singapore is doing an India by incorporating SPaN - the Singapore Payments Network. It's the exact same thing that India did in 2008 by incorporating NPCI. SPaN is also being incorporated as a non-profit with the MAS and seven Domestic Systematically Important Banks (D-SIB) as initial shareholders. The SPaN will initially onboard the other stakeholders and then takeover administration of Singapore's existing payment systems from the current administrators. If all goes well, SPaN is expected to live by end-2026.

3️⃣ South Korea wants to compete with Thailand as the capital for Southeast Asian digital assets activity

KB Kookmin Bank, South Korea's largest bank, filed trademark applications for KBKRW and KRWST stablecoins pegged to the Korean won, signalling growing institutional interest in blockchain-based finance in Asia. KakaoPay, a South Korean payments company, filed 18 trademarks related to a potential Korean won-backed stablecoin, causing its stock price to surge over 200% in a month. Thailand's Securities and Exchange Commission (SEC) sought public input on proposed changes to criteria for selecting digital assets offered on local exchanges, aiming to align regulations with evolving digital asset landscape while preserving investor protection.

Takeaways:

1. Active regulatory environment: Almost all regulators across Singapore, Vietnam, Thailand, South Korea, Philippines and Malaysia are currently working on some form of regulation or is in active public consultation on the topic. And these aren't just CBDC regulations, these are regulations for digital asset trading or stablecoin issuance and activity.

2. Banks are actively participating too: It seems that the regulatory clarity with South Korea's draft Digital Asset Basic Act introduction as of early, Jun, 2025 is driving the rush for stablecoin issuance among banks as well. Aside from the trademark applications, Banks are rushing (surprisingly) to be able to compete with large payments companies including the likes of KakaoPay and others.

3. Thailand is actively competing: Thailand’s "boom" isn’t hype; it’s policy-engineered. A cascade of permissive rules (stablecoins, tokenised bonds, tax holidays) hit at the same time that major banks rolled out custody and payment rails. That alignment, plus a politically endorsed vision of becoming a regional digital-asset hub, has flipped the market from experimental to execution mode in under a year. But I'm not going to claim I fully understand what's happening in Thailand.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here's an earworm: Been obsessed with the F1 movie soundtrack and this song OMG! is a banger (Spotify / Youtube). Solid track to get you in the mood for the week. Have a happy July!

✨ Call outs

1. [Video] What ails Indian manufacturing by Amit Varma and Ajay Shah’s Everything is Everything podcast. This is a great deep dive into the history of Indian manufacturing and what are some issues faced by manufacturers.

2. [Video] Behind the scenes of Notion Design team. This interview/video gives some good insight into the hundreds of prototypes and design choices that goes into making the simplistic functional frontend we see on Notion.

3. [Article] Valuation Multiples by Michael Mauboussin of Morgan Stanley. This is a research report that examines valuation multiples (P/E and EV/EBITDA), explaining how they fail to capture key value drivers like return on invested capital and why they've become less informative due to the rise of intangible investments that are expensed rather than capitalised.

👋🏾 That's all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 95th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

We talk about financial services a lot around here - that's what you're here for. But there's one sector in financial services I've never covered: the one I actually work in. Venture capital. And honestly, we're living through the most dramatic transformation this industry has ever seen.

I love studying the history of VC. The traditional model was beautifully simple: raise a fund, invest in startups, hopefully exit within ~10 years, return capital to investors. Rinse and repeat. But that’s slowly changing over the past few years.

Top VCs today aren't just early-stage investors anymore. They're becoming multi-faceted investment platforms capable of supporting companies with equity, debt, or whatever instrument makes sense, from inception through IPO and beyond. One is even building its own GPU infra!

As someone working in this space daily, I wanted to break down what's happening and why it matters - especially for founders navigating this new landscape.

Thank you for supporting me and sticking around. Enjoy another great week in fintech!

Subscribe now

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. I’m at os@osborne.vc

🤔 One Big Thought

The Great VC Reinvention: Why Traditional Funds Are Becoming Financial Platforms

We talk a lot about fintech around here. But today, we're covering the industry that funds it all: venture capital. And right now, the VC world is undergoing the most historic transformation since its inception. The old playbook—raise a fund, invest in startups, exit in 10 years, repeat—is being torn up by the very firms that perfected it. This is especially true for the larger venture funds.

Venture capital is no longer a cottage industry. Venture funds today are seemingly flexible, diversified investment platforms capable of supporting companies with equity, debt or other instruments, from inception through IPO and beyond. It has evolved from a niche, artisanal craft into a high-finance business.

The sector has grown from $53bn in global investment during the 2008 crisis to a record $640+bn in 2021, representing a fundamental shift in how innovation capital is deployed worldwide. And the top firms are reinventing themselves to deploy this capital at an unprecedented scale.

Why the shift? Part of the reason is scale and opportunity. The tech industry now supports companies that stay private longer and require larger capital infusions. Meanwhile, novel asset classes (like crypto or secondary shares) and founder needs (like liquidity and wealth management) have opened avenues for VCs to provide more than just equity funding.

The transformation addresses critical limitations of traditional VC structures: forced exits based on fund timelines rather than company performance, inability to maintain board relationships post-IPO, and exclusion of individual investors from private markets.

If you speak to active investors in financial assets, they'll tell you that each of their strategies have a ceiling on the total capital they can deploy for a particular strategy. For example, a long short strategy may have an upper limit of ~$50M, a quant strategy may have a ceiling of $100M, a mutual fund scheme could deploy maybe $150-200M etc. etc. Those aren't real numbers, but should give you a sense of what I mean by the ceiling.

This ceiling depends on depth in the market, risk-reward ratio and much more. the traditional model of VC, seems to have reached that ceiling and it now has to evolve or face the risk of being irrelevant. I think that ceiling gets breached for a fund size of $300-500M and total AuM of $2-3bn. Market depth matters too and some markets might not be able to absorb more capital.

20 years ago, a 99th percentile outcome in VC meant, a $1bn valuation. Today, 99th percentile outcome is a $10bn exit valuation. Further, capital requirements of companies has grown. Lastly, LP's have more capital than ever to deploy. All of this warrants a business model shift for the venture capital industry.

The result is an industry in flux... finding itself.

As a student of this industry, I wanted to highlight the strategies/structures employed by some of these funds as they usher us into the future. To me, these structures are fascinating, interesting and thought-provoking. Let's dive in.

(Standard disclaimer applies - I'm not an expert on this topic, and my analyses are based on publicly available information.)

a16z's switch to a Registered Investment Advisor

Andreessen Horowitz (a16z) was the first among venture funds (to my knowledge) to transform its business model by becoming a Registered Investment Advisor (RIA) in 2019. This wasn't just a legal tweak; it was a strategic overhaul that unlocked immense flexibility for the $45bn+ firm.

How it works:

• The structure is nothing fancy - the license to operate switches from a fund to a registered investment advisor.

• Seems like, a16z will continue to raise capital into dedicated funds for specific themes like biotech and crypto.

Pros

• Flexibility in assets held: the biggest benefit of an RIA is the ability to trade in private equity, crypto tokens, public equity, secondary markets and other alternative assets

• Flexibility in holding tenor: the RIA is not time bound in its trades and can buy or sell as needed. a VC fund had a fixed timeline of 10+2 years.

• Can still have dedicated fund structures: Specialised fund operations span multiple sectors with the 2024 fundraise of $7.2bn across Growth Fund ($3.75bn), Infrastructure Fund ($1.25bn), Apps Fund ($1bn), Games Fund ($600M), and American Dynamism Fund ($600M). Each fund operates with dedicated investment teams while leveraging shared platform services across all portfolio companies.

• Portfolio optimisations: a16z can structure its investments more tax-efficiently across different asset classes. It can also potentially win investment opportunities in securities, asset classes or jurisdictions that other funds cannot.

Cons

• High compliance costs: Reportedly, this structure comes with ~8x higher costs.

• Potential operational drag: Employees have to go through background checks and register their investment holdings. They’ll also be more restricted in what they can trade individually and how they talk about financial performance.

• Conflicts of interest: When you're investing across public/private/crypto simultaneously, whose interests come first?

• Mission drift risk: Are they still great at finding the next Google, or are they just becoming a generic asset manager?

The interesting thing is that a16z launched The Oxygen infrastructure platform. It's a major competitive differentiator, providing AI startups access to over 20,000 Nvidia H100 GPUs through hundreds of millions in hardware investment. The program offers below-market pricing, flexible short-term access, and equity-based arrangements, enabling startups to compete with Big Tech on AI model development without $10M+ GPU commitments. This sort of program can only be built by a firm with massive AuM. Imagine any other fund building their own AWS!

Other developments include fundraising for a $20bn AI mega fund targeting growth-stage AI investments, launch of the a16z Perennial wealth management division with $822 million AUM, and active advocacy for modernised SEC crypto custody rules.

Sequoia's evergreen fund

Sequoia Capital's "The Sequoia Fund" seems to be the most radical departure from traditional VC structures. It announced it was "breaking with tradition, abandoning the traditional fund structure" with set 10-year cycles. Instead, Sequoia created a "singular, permanent structure" called The Sequoia Fund – essentially an evergreen fund that can hold investments indefinitely. That fund has grown to $20bn in assets as of 2024.

How it works:

• Sequoia also became a Registered Investment Advisor (RIA) where one master fund serves as the sole LP for multiple sub-funds (seed, venture, growth, crypto).

• When portfolio companies go public, shares flow into the master fund instead of being distributed to LPs, enabling indefinite holding periods and continued board participation.

• Fee structures differentiate between sub-funds and the master fund. Sub-funds maintain Sequoia's premium 2% management fees and 30% carried interest, while The Sequoia Fund charges less than 1% management fees with performance fees tied to very high hurdle rates. This creates a hybrid model where traditional VC economics apply to private investments while public holdings operate under lower-cost structure.

• LPs get to choose redemptions in semi-annual windows. LPs maintain accounts with The Sequoia Fund and can request allocations to specific sub-funds, providing some control over investment strategy mix.

Pros:

• Simpler product for limited partners: LPs benefit from simplified public equity management and potential for higher long-term returns through extended holding periods, with Sequoia's internal analysis showing $8+bn in additional returns from holding shares just 12 months longer over the past 15 years.

• Redemption flexibility for LPs: LPs can choose to stay invested or redeem semi-annually.

• Longer partnership with founders: Portfolio companies benefit from continued strategic support post-IPO and reduced pressure for premature exits, while Sequoia gains permanent capital base eliminating fundraising cycles.

• Not limited by fund or market cycles: With the Sequoia Fund, it can immediately redeploy gains instead of waiting 2-3 years to raise the next fund. With this, Sequoia can support portfolio companies through multiple economic cycles.

Cons:

• Structure works for funds with a large public market exposure: Sequoia's portfolio companies have an aggregate 20% market cap of the NASDAQ! Sequoia has $85bn in total AUM, and $20Bn in the evergreen fund - that's 1/4th in publics. This is liquid holding that Sequoia could use to manage redemptions. Most other funds might not have that much access to the public markets.

• "Bank runs": In a very remote chance where majority LPs want to redeem/exit their investment, Sequoia may have to dip too far into its public market pools which could cause a strain on the rest of the private market pools as well.

General Catalyst's multi-product strategy

General Catalyst has evolved into a comprehensive "investment and transformation company" through two major initiatives: GC Wealth - a wealth management platform; and the Customer Value Fund - a revenue-based financing model. With $32bn in total assets under management, these diversification efforts represent Hemant Taneja's "Beyond Venture" strategy to create multiple revenue streams while deepening ecosystem relationships.

Strategy:

1. Transformation company with capital: Nothing in its core funds structure seems to have changed, but GC is rethinking what being a venture capital firm is. Launching wealth management, revenue based finance and doing buyout deals (operational companies) is not very VC-like. But GC doesn't seem to care and I like that.

2. Wealth management: GC Wealth launched in 2024 with $2.3bn in AuM and 20+ employees. GC Wealth is led by Dave Breslin, former First Republic Private Wealth head who previously grew AUM from $60bn to $290bn. The platform targets founders, entrepreneurs, and high-potential individuals within GC's "Famiglia" community, providing personalised wealth management with access to private investment opportunities including GC funds. The wealth management model is intended to create powerful integration synergies with traditional VC operations. GC Wealth introduced ~40 new founders to GC's venture ecosystem, with some receiving funding, while GC's venture portfolio provides wealth management clients. The platform allocated $25M of client assets to GC's $8bn Fund XII.

3. Revenue based finance: The Customer Value Fund operates through revenue-based financing mechanics, providing non-dilutive capital secured by company recurring revenue streams. The most recent major investments include Grammarly ($1bn, the largest CVF investment) and FINOM (€92.3M). The structure allows companies to fund sales and marketing costs while paying back a capped percentage of revenue generated from funded customer acquisition activities.

4. Multiple revenue streams: Fee structures across the diversified platform create multiple revenue streams: traditional wealth management fees (likely 1-2% AUM), revenue-based returns from CVF capped at fixed percentages of customer acquisition outcomes, and standard VC fees of 2% management and 20% carried interest on $32+bn AUM. This diversification reduces cyclical dependence on traditional VC fundraising and exits.

Pros:

• Suite of products to offer: GC now has a whole suite of products to offer founders, LPs and the ecosystem. It's not bound by just VC equity partnerships. It could easily offer revenue based finance to a startup where it's not an equity investor. Similar strategy for other products too.

• Strategic positioning advantages include first-mover advantage in integrated wealth management among major VCs.

• Powerful network effects through the GC Famiglia community

• Unique end-to-end entrepreneur lifecycle support from startup funding to personal wealth preservation

Cons:

• The approach risks focus dilution across multiple business lines and potential conflicts between VC and wealth management activities.

• Fiduciary conflicts: Given there are so many fiduciary responsibilities, could there be any conflicts of interest? How do you serve someone's personal wealth interests when it could conflict with your portfolio company interests?

Coatue's CTEK Fund to democratise access

Coatue's CTEK Fund structure is the most interesting structure in my opinion - largely because I've not seen anything like it before. Claude tells me there are some specialised funds in US that use a similar structure though. With Sequoia, a16z and General Catalyst, the whole fund was restructured, but in Coatue's case, the main fund is the same, it's just that they've launched a new vehicle which gives them the flexibilities that the other mentioned funds enjoy.

Strategy:

• This Coatue CTEK fund is a significant democratisation of institutional-quality technology investing, targeting qualified individual investors through an innovative tender offer fund structure. With $1bn in initial anchor capital from Jeff Bezos and Michael Dell's family offices, the fund reduces minimum investments from typical $5M Coatue requirements to $50,000 for broader accessibility.

• The fund structure operates as a statutory trust registered as a non-diversified, closed-end management investment company. Three share classes (S, D, and I) offer different fee structures and minimum investments, with Class I requiring $1M minimums while Classes S and D require only $50,000. The product is being distributed via UBS US Wealth Management.

• Investment strategy focuses on 80% minimum allocation to technology and innovation companies across AI, enterprise software, next-gen infrastructure, fintech, and climate technology.

• Liquidity mechanisms differentiate CTEK from traditional private funds through quarterly redemptions up to 5% of net assets, monthly purchase opportunities, and NAV-based pricing. This structure provides enhanced liquidity compared to traditional 10-12 year lockups while maintaining exposure to private market returns.

• Fee structures include 1.25% annual management fees and 12.5% performance fees above a 5% annual hurdle rate with high water mark protection. Distribution fees vary by share class from 0% (Class I) to 0.85% (Class S).

Pros:

• This structure allows Coatue to deploy capital across early to growth to public markets and across sectors. In some ways, this will operate like The Sequoia Fund.

• Massive TAM: Helps Coatue move from institutional-only to retail limited partners and potentially opens up trillions in AuM opportunity.

Cons:

• Retail expectation mismatch: Individual investors may expect mutual fund-like liquidity in private markets. Individual investors also may not understand the risk that comes with investing in instruments like this.

• Regulatory scrutiny: Retail investor protection rules are much stricter.

• Operational nightmare: Managing thousands of $50K investors vs. a few $50M institutions is completely different

Competitive dynamics

The venture capital industry is maturing and morphing. The lines between venture capital, growth equity, private equity, and asset management are increasingly blurred.

The emergence of these innovative fund structures reflects broader industry trends toward permanent capital, enhanced liquidity, and democratized access. PitchBook projects evergreen funds will grow to approximately $500bn by 2029, with institutional evergreen funds increasing from $130bn to $300bn, representing fundamental shifts in how private capital operates.

Competitive advantages for innovative structures include continuous fundraising eliminating discrete cycles, flexible exit timing without forced asset sales, expanded investor bases accessing high-net-worth individuals and retail channels, and operational efficiency through single evergreen structures versus multiple sequential funds.

However, operational complexity requires sophisticated infrastructure with advanced technology stacks, dedicated personnel, third-party partnerships, and risk management systems.

Early trends and limited track record data suggests comparable risk-adjusted returns to traditional structures, though most evergreen funds lack long-term performance history.

Risk factors include higher operational costs (estimated 2-3x traditional fund administration), regulatory complexity across jurisdictions, and market concentration among larger GPs with operational capabilities.

Performance measurement challenges emerge from NAV dependency, benchmark comparison difficulties, and time horizon differences requiring new evaluation methodologies.

For founders, the mega-fund era might mean an even greater focus on later-stage or large fund raise requirements, “sure bet” companies (to deploy those billions), potentially making it harder for smaller startups to get attention. There’s also the question of returns: can venture funds manage $20B and still deliver venture-like performance? Or does VC inevitably become more conservative as it scales up? These are debates now playing out in real time.

What seems clear is that venture capital is no longer a cottage industry. It’s becoming a high-finance business, complete with complex structures and broad offerings. The classic model of a VC firm solely raising a $200M fund to invest in Series A rounds is being upended by firms like Sequoia, GC, Lightspeed, and a16z leading the charge in redefining what a venture firm can be. As more firms follow suit, the ecosystem will likely become more founder-centric and competitive – and perhaps more resilient, with diverse ways to weather market cycles.

Venture’s Ongoing Metamorphosis

These changes are opinionated bets on how venture firms can remain at the cutting edge. It’s an analytical response to a shifting market: adapt structure and strategy to capture more value and provide more value, or risk irrelevance. For general tech and finance observers, this evolution is fascinating (and a bit reminiscent of how investment banks or private equity firms diversified in the past). For entrepreneurs and LPs, it means more options – and also more factors to consider when picking a VC partner.

It’s also a statement about the maturing of the VC industry: top firms like a16z are now capable of absorbing PE-like sums of money. If the venture asset class used to be boutique, it’s increasingly looking institutional in scale. As Reuters noted, if Andreessen Horowitz pulls off $20B, it would be surpassed in size only by SoftBank’s vehicles, and it highlights ongoing debate about “how scalable” venture capital is while maintaining strong returns.

One thing is certain: venture capital’s metamorphosis is underway, and it’s making the industry more dynamic than ever. Don’t be surprised if tomorrow’s top “venture” firm looks as much like a multifaceted Wall Street financial institution as it does an old-school Sand Hill Road partnership. The business of funding innovation is innovating itself – and we’re seeing the first glimpses of what the VC firm of the future might look like.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here's an earworm: God Went Crazy by Teddy Swims (Youtube / Spotify). Smooth listening.

✨ Call outs

1. Rahul Sanghi of the Tiger Feathers published his latest and greatest post chronicling the company behind institutions like Bombay Canteen, O Pedro, Papas and more. I haven’t finished the piece yet (classic Rahul), but it’s a beautiful read so far (also classic Rahul). And I’d recommend you read it too.

2. If you’re an RHCP fan, you’d appreciate this one. Have you heard the song Can’t Stop? That intro riff sounds so simple and inane, right? This video takes you through the details of that simple riff explaining how complicated it really is and why even expert guitarists find it tough to replicate. The genius that is John Frusciante!

👋🏾 That's all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 94th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

We're witnessing something unprecedented: the financialisation of content and the entertainment-ification of finance. The psychological barriers to both investing and real money gaming have collapsed simultaneously, thanks to UPI making tiny transactions feel like sending a WhatsApp message.

Here's what's wild: India's real money gaming industry now has 200M+ transacting users while our stock markets have 110M unique investors. That's right - more Indians are actively trading on Dream11 than on the NSE.

I'm fascinated by how the same technological enabler - UPI micropayments - is creating two massive parallel economies with completely different regulatory treatment, risk profiles, and social perceptions.

This week, I dive deep into India's micropayment revolution and how it's enabling businesses to earn massive revenue with new business models. The data tells a story that most people aren't paying attention to, and the implications for where Indian household savings flow next are enormous.

As usual, there’s also a beautiful song recommendation at the end, if you’d like to listen to a song in the background while you read this. Do share your recommendations with me too :)

Thank you for supporting me and sticking around. Enjoy another great week in fintech!

Subscribe now

Considering angel investing? I get a bunch of fintech founders reaching out to me for investors. I’d be happy to put you in touch. I’m at os@osborne.vc

🤔 One Big Thought

The Big Promise in Tiny Payments

How Tiny Transactions Are Building Massive Markets

In April 2025, monthly SIP inflows touched INR 26,632cr ($3.1bn). Not even five years ago, that number was INR 8,376cr ($1.1bn). At the same time, India's real money gaming market, valued at $3.7bn in 2024, is projected to reach $9bn by 2029. Not only that, digital product startups seem to have found product market fit and growing revenues like never before.

What do these seemingly unrelated sectors have in common?

They're all being powered by Indians' growing aspiration for convenience and higher household income made possible by UPI.

The hero of this story is UPI, which has fundamentally altered what's financially possible. When transferring INR 10, INR 100 or INR 1,000 becomes as effortless as sending a text message, entirely new behaviours emerge.

This post is about some of those new behaviours, powered by the digital super highway of UPI and it's suite of sub-products.

AD: Speaking of making things simple, today's post is sponsored by Simplehuman

Simplehuman is a browser extension that makes using Gmail a blazing fast. Save time when using Gmail with keyboard shortcuts, follow up reminders, 1 click unsubscribe, and command bar.

Simplehuman is built by my fellow fintech friend Phalgun for folks like you and me to free our time for more productive purposes.

Use code FINTECHINSIDE for a 15% discount on yearly or lifetime plans.

Install Simplehuman

Stocks, SIP's and IPO

The Indian security brokerage market is on a clear growth trajectory. The stock broking industry's revenues was around INR 23,500cr ($2.7bn) in the first half of FY25 which was expected to decline to around INR 20,500cr ($2.4bn) in the second half. That's ~$5.1bn revenue in FY25.

Behind these impressive numbers is a profound democratisation of market access:

• NSE registered unique investors reached 110M by Jan 2025, up from 91M a year earlier

• The derivatives segments grew 19x in active traders to ~9.6M in 2024 from ~0.5M in 2019

• About half of the new investors in 2024 participated in derivatives

• The share of mutual funds in Indian household savings grew from 7.6% in FY21 to 8.4% in FY23

What's driving this surge? The friction of entry has collapsed. When I first started investing in stocks over a decade ago, the process was cumbersome - requiring significant paperwork, wet signatures, minimum investment amounts, high fees and complicated payment mechanisms. Today it's a 5 minute onboarding process - thanks to digitisation, Digio and UPI. Costs of onboarding and friction of participation have gone down dramatically.

The UPI-powered IPO application process tells a similar story. Back in 2019, one time mandate for UPI was launched and applied to the IPO process. This meant, you could apply for an IPO using UPI to set up a one time mandate, and your funds would be auto debited from your account if you get allocation at IPO. Don't even ask about the IPO application process pre-UPI.

Today UPI is among the most preferred method to subscribe to an IPO and is used in 10-20% of the total IPO value. The retail portion of most IPO's is around that much, if I'm not wrong.

The INR 5 lakh transaction limit for IPO applications via UPI, compared to the INR 1 lakh limit per regular transaction, is high enough to accommodate most retail investors, removing what was once a significant barrier to participating in primary markets.

This "SIP-ification" of savings is becoming a structural feature of our economy. Platforms like Groww, Zerodha, and others have leveraged UPI to make investing accessible to demographics that were previously excluded - not just the affluent in metro cities, but also middle-income individuals from Tier 2 and 3 cities. Uttar Pradesh crossing the 1 crore registered investor mark is testament to this geographic expansion. For the uninformed, an SIP is a Systematic Investment Plan, which allows the user to decide to invest a certain amount each month in a certain mutual fund's scheme.

Today, I watch as friends casually set up INR500 ($6) monthly SIPs through a few taps, with the money automatically moving from their bank account via UPI autopay.

I remember having a conversation with a senior mutual fund executive who told me that they had to work hard to ensure costs were lower than current levels to launch the latest INR 100 ($1.2) SIP. It was all made possible because of these advancements.

The mutual fund industry is getting $3bn a month in guaranteed capital flows and it probably doesn't know what to do with it. Part of this monthly payment flows are made possible by UPI Autopay.

From Investing to Entertainment

On the other end of the spectrum, India's Real Money Gaming and Fantasy Sports (RMG/FS) sector has exploded on the back of the same micropayment infrastructure, but with a fundamentally different value proposition.

The Indian gaming market, valued at $4.4bn in 2024, is projected to reach $7.6bn by 2028, with a user base of 600M. Within this ecosystem:

• Fantasy sports reported revenues of INR 9,100cr ($1bn) in FY24, engaging a massive user base of 220-230M

• Companies like Gameskraft saw a 30% YoY revenue increase to INR 3,475cr in FY24

• MPL's revenue grew by 22% to INR 1,068cr, while PokerBaazi saw a 55% surge to INR 415cr

RMG leverages the same UPI payments infrastructure for entertainment spending. A typical user makes dozens of small deposits ranging from INR 10 to INR100 for game entries, with each action requiring minimal thought due to the frictionless payment process. The hope for users with real money gaming is that a user wins a game with a large potential cash reward.

The sector's explosive growth hasn't come without challenges. The 28% GST on deposits implemented in October 2023 has significantly altered unit economics. Fantasy sports revenues were projected to decline by 10% in FY25 to INR 8,200cr due to these tax changes. Companies are adapting by adjusting their commission, focusing on segments with higher ARPU like poker and rummy, or diversifying toward casual gaming.

Investing and Fantasy Sports - an Indian dichotomy

Despite the 28% GST addition, revenues are expected to decline just 10% YoY. On similar lines, SEBI (India's securities regulator), introduced a "True to Label" regulation in July, 2024 that was intended to make broking costs fair and standardised for all users. This was supposed to impact the industry as well, but CARE Ratings projected just a 4% decline in revenues.

The reason I'm talking about both these markets is because users are participating in both stock trading and RMG/FS for a windfall gain. The risks are similar, outcomes are similar, yet the real money gaming industry has over ~200M transacting users, vs. the stock trading industry has 110M unique investors i.e. half as many users as the real money gaming industry! Revenue of the two industries will soon catch up too, if not already there.

I keep thinking why that is, especially when the stock markets are regulated, while the RMG/FS industry is not. My current thesis is that users are familiar with RMG/FS - they know who a Virat Kohli is and they watch an IPL match, so, it's easier for them to understand what assets to trade in - everyone thinks they know better. This is not the case with assets in the stock markets.

Secondly, unlike the stock markets open only during a certain trading window, RMG/FS is available 24x7. That allows users massive flexibility in trading.

I could be wrong about these and I'm open to learn if you have differing views.

Beyond Financial Services: Powered by UPI Autopay

This micropayment revolution extends beyond just finance and RMG/FS. Content platforms, commerce, savings and others are increasingly adopting sachet pricing models - letting users pay small amounts for individual articles or limited access rather than full subscriptions.

The magic here though is the Autopay feature. UPI Autopay is a subscription payment platform that's getting platforms to sign up users for recurring payments. Autopay seems to be working because of low transaction values and trust in the UPI platform.

On looking at the data, I was shocked to see that there are ~1.2bn UPI Autopay mandate executions a month. That's ~4-5 transactions per user per month. Unfortunately, success rates are an abysmal 18% - so ~210M Autopay transactions are successful. That's a LOT of money left on the table.

Jar has daily/weekly/monthly UPI Autopay for savings in digital gold. It continues to grow revenues significantly.

• Digital Content Monetization: Micropayments have cracked the code for monetizing digital content. Paying ₹5 for a news article, ₹99 for a monthly OTT plan (Netflix adopted UPI AutoPay in 2021), or tipping a creator a few rupees is now frictionless.

Astrotalk, an online astrology platform, leverages the same infrastructure to enable consultations priced as low as INR 10 per minute. Their revenue skyrocketed by 130% YoY to INR 651cr ($76M) in FY24, demonstrating how micropayments can transform even traditional services into scalable digital businesses.

FRND is a social discovery app for "Bharat" uses virtual coins bought via UPI for interactions, grossing INR 74cr ($9M) in FY24 revenue. Dream11 with 200M+ users, it saw revenues of INR 6,384cr (~$800M) in FY23 and turned profitable. Probo an opinion trading saw its revenue skyrocket 32x to INR 86cr ($11M) in FY23 and an astounding INR 459cr ($54M) in FY24, with INR 92cr ($11M) in net profit. Zupee a casual gaming startup doubled its revenue to INR 832cr in FY23 and became profitable in FY24 at INR 146cr ($17M) with INR 1,123cr ($131M) in revenue.

Industry reports even predict Indians will spend more on online gaming than on films by 2025.

Kuku FM, an audio content platform, Pratilipi, another content platform and many others like them listed in the tweet below, are breaking the revenue chasm. From my conversations with friends in the industry, I'm told some of these apps employ dark patterns and/or have low MAUs. Unfortunately, I don't know enough about these platforms to do a deep dive, but you get the general gist.

The common thread? When the friction of small transactions disappears, entirely new business models become viable.

The Power of Financial Accessibility

The micropayment and autopay revolution represents more than just a technological shift - it's a fundamental democratisation of financial access. Whether channeled toward wealth building through SIPs or content and entertainment, the ability to move small sums of money effortlessly is reshaping behaviours and business models alike.

UPI is a massive case study - build the rails and people will use it, unlocking economic activity. If only the India stack also included our roads.

What sectors do you think are being and will be transformed next by the micropayment revolution? Drop your thoughts in the comments or reach out directly.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here's an earworm: Nostalgia crept in this week. Re-listening to this classic song by Sean Paul We Be Burnin’ (Youtube / Spotify). Enjoy!

✨ Call outs

1. Incredible data put together by Cole Rotman, investor at DST Global. The data shows people and firms that invested at Series A in companies that are currently valued >$5bn. Just four investors in that list who invested at Series A in 3 companies currently valued at $5bn+. I wrote a blog post about it too.

2. Insightful conversation with Jared Cohen of Goldman Sachs six months ago, on the the impact of AI on geopolitics. He was fairly prescient when he said that Saudi Arabia/the Middle East will be the ideal place for the US to build its data centers. Recently, the US had a big show in the ME and made billions of dollars in commitments toward building data centers in the region. FYI Goldman Sachs has an under rated youtube channel.

3. Stumbled upon Taylor Bell’s Youtube Channel, where in this specific episode she breaks down Erehwon’s business model. Erehwon is an ultra luxury supermarket and is fairly unique. Taylor’s videos are generally accessible with her funny style but deep insights on business models.

4. MobLand is my latest show obsession (available on Hotstar in India). As the name goes, MobLand is about modern day gangs and their rivalry, but made by Guy Ritchie and stars Tom Hardy, Pierce Brosnan, Helen Mirren and more. Excellent show!

👋🏾 That's all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 93rd edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

Just a year ago, the idea of machines autonomously creating valuable output, let alone transacting, felt distant. Yet, AI is rapidly evolving from chatbots analyzing information to 'agents' actually doing things. Imagine an AI agent handling that complex orchestration autonomously.

But this leap raises a fundamental question: how do these agents pay for the services they book or the goods they order? This need has sparked a wave of activity from the biggest names in payments – Visa, Mastercard, Stripe, PayPal, Razorpay – all launching initiatives around Agentic Payments.

Is this the dawn of truly autonomous commerce, or just sophisticated new plumbing for existing rails? This week, I explore the emerging infrastructure, the disruption to business models, and the real potential behind the buzz.

As usual, there’s also a beautiful song recommendation at the end, if you’d like to listen to a song in the background while you read this. Do share your recommendations with me too :)

Thank you for supporting me and sticking around. Enjoy another great week in fintech!

Subscribe now

🤔 One Big Thought

Agentic payments | The AI agents are coming… for your wallet

In the 1970s-1980s, expecting flying cars by 2020s was considered science fiction. Today there are car makers Alef and Xpeng who are actually building flying cars. Every generation keeps redefining what "science fiction" means to them and that pushes our collective imaginations and ambitions.

We're living through a period of exploring our definition of science fiction - thanks to GenAI. We're reimagining what it means to create, consume and thrive. We're reimagining business models and revenue models. We're reimagining our role, not to get too philosophical.

Even a short 12 months ago, we would have never imagined machines creating things for us autonomously. Today, 25-30% of all code at Google and Microsoft is writing entirely by AI. I write all my code using AI.

These AI "agents," as they're increasingly called, promise a future where tasks from simple to complex, from managing household supplies to orchestrating sophisticated business processes – are all handled autonomously.

If you're 25+, you'll relate to this example: think about the challenge many of us face managing our elderly parents' healthcare needs – ensuring they eat and stay healthy, tracking doctor appointments, ensuring follow-ups, ordering prescription refills, and handling payments for consultations or pharmacy bills.

Today, this involves constant calendar reminders, phone calls to clinics, coordinating transport, trips to the pharmacy, and manually paying bills via UPI or cards after receiving them. It's time-consuming, stressful, and prone to missed appointments or delayed refills. If only I had an EA just to manage my parents needs and their health.

Now, imagine telling/prompting your AI agent: "Manage Mom's cardiology care with Dr. Sharma at Apollo. Ensure she has her monthly check-up scheduled, remind her one day before via a call, book an uber cab if needed, book pathology tests from Orange Health based on prescription, track her vitals from her smartwatch, ensure her prescriptions for Metformin and Lipitor are refilled 3 days before they run out via Welcome Medical by calling them, and handle payments using my linked ICICI credit card, staying within a INR 10,000 monthly budget for consultations, tests and meds. Alert me if any vital reading is concerning or if Dr. Sharma recommends a change in treatment". *wake up, Osborne!* Ah! what a dream! Snap back to reality.

This AI Agent doesn't just remind you; it acts for your. It proactively schedules, coordinates transport, monitors vitals, checks pharmacy inventory, places refill orders, and – crucially – initiates payments for the consultation fee or the pharmacy bill or the pathology tests, using delegated, secure credentials within the set budget. It operates autonomously based on rules and real-time needs. This shift from human-initiated to agent-initiated commerce requires a new kind of payment infrastructure – Agentic Payments.

Why Now? The Shift to Agent-Centric Commerce

Claude's Model Context Protocol framework of building the "Type-C - universal charging point" equivalent of the Agentic world, seems to be igniting a lot of this discussion and experimentation.

We're still in the very early days of building a true, universal, autonomous agent. There are so many considerations of user interfaces, data structures, systems, proprietary access and other aspects of the internet we've come to know and love that we'll need to get right before we have such universal agents.

The underlying agentic framework is maturing rapidly though. Most large tech companies are experimenting with their vision for such agents, but that's not the topic for today's discussion, so will skip it. However, our current systems are showing their limitations.

Today's APIs and payment flows were largely built with human developers and end-users in mind. They aren't optimised for AI agents communicating and executing complex, multi-step tasks autonomously. These agents require programmatic, secure, and context-aware access to financial rails.

For a long time I've held the view that the payments form factor will evolve based on the evolution of our points of consumption. We've seen it evolve from coins to cash to plastic cards to now our mobile phones. Our payment flows will continue to evolve as our points of consumption evolve. My initial hypothesis was that the payments flow for live commerce was broken, the payments flow for renewing subscriptions on our TV is broken. That thesis was only partially wrong - seems I got the direction right, but not the use case.

In an agentic future, our point of consumption is evolving yet again. And that's creating a need for the payments form factor to evolve once again. Except this time, along with the form factor evolving, our business model is evolving as well.

The Business Model Switch: Pricing in an Agentic World

The most significant catalyst for agentic payments is the fundamental disruption occurring in software business models - especially the traditional SaaS per-seat pricing model becoming increasingly untenable.

Why the shift? SaaS sprawl creates unused 'shelfware', data silos across different tools reduce overall efficiency, and the significant compute power needed for sophisticated AI features puts pressure on margins. Most critically, if AI agents start performing tasks previously done by humans, the very notion of charging "per seat" or "per user" seems to break down.

This necessitates a move towards pricing models that better align cost with the actual value delivered. We're seeing the rise of alternatives moving away from the traditional seat-based approach:

• Usage-based: Pay per API call, compute time.

• Outcome-based: Pay when a specific result is achieved.

• AI FTE-Based: Pay based on the equivalent human full-time employees the AI replaces.

• Unit-Based: Pay per unit of output created.

However, implementing these new models isn't simple. Measuring outcomes accurately, establishing clear attribution (did the AI really complete the task?), and providing budget predictability for customers are major challenges.

This is where Agentic Payments seems to fit in – they provide the underlying infrastructure needed to measure, track, and settle value based on usage or outcomes, programmatically and reliably.

Enter the Players: The Payments Giants Respond

Recognizing this paradigm shift, the major payment networks and processors have started unveiling their strategies - all within the last 6 months:

• Visa's Intelligent Commerce: Chat-based shopping. Visa wants to enable agents to "find and buy" securely. Focuses on "AI-Ready Cards" using tokenization for secure agent transactions, backed by their fraud prevention expertise and partnerships.

• Mastercard Agent Pay: Introduced "Agentic Tokens," extending existing tokenization products. Collaborations with Microsoft and IBM target both consumer and B2B use cases, aiming for seamless payment integration into conversational AI platforms.

• Stripe Agent Toolkit: Offers developers an SDK to integrate Stripe functionality (including Issuing for virtual cards) into agent workflows. Stripe has gone the extra mile and solved for "usage-based pricing" by including tools for metered billing based on agent usage.

• Coinbase Agentkit: Giving AI agents crypto wallets for on-chain interactions, enabling stablecoin payments and monetization, designed to work with any AI framework or wallet.

• PayPal Agent Toolkit: Provides an SDK connecting AI frameworks (using the MCP standard) to PayPal's APIs for tasks like invoicing, subscriptions, and payments, enabling agents to manage commerce workflows.

• Razorpay MCP Server: An early move from India, establishing a server based on MCP to standardize how agents could potentially interact with its payment gateway APIs in the future.

There are several more platforms, including early stage startups like Prava and others addressing agentic payments in their own way.

Several common threads run through the above initiatives:

1. Tokenization as Infrastructure: Securely representing payment credentials for machine use is the foundational layer.

2. Security & Trust: Agent verification, user controls (limits, permissions), and robust fraud detection are critical. The "Know-Your-Agent" (KYA) challenge remains a significant hurdle – establishing the identity, authority, and legal standing of an autonomous agent.

3. Developer Enablement: Providing developer SDKs and clear documentation is the MO most of these platforms are going for to build adoption by those building the agents.

4. Ecosystem Collaboration: Given these are early experiments - a sort "throw them at the wall and see what sticks" approach, all are going for partnerships and collaboration across AI labs, frameworks, cloud providers, and payment networks.

5. The Billing Infrastructure Gap: While payment execution is being addressed, the infrastructure to easily support complex outcome-based billing driven by AI agents is still nascent.

Like telecom, there are only two models to charge the user - prepaid or postpaid. Depending on the agentic payment, users will be required to add amounts to their prepaid wallet or virtual card for use, else the user will be billed and charged at the end of the period. It seems like we may default to prepaid model to build trust initially, and then graduate to postpaid, once trust is built and the user knows what to expect. This has been our experience in India with digital payments too.

The amazing thing is that with agentic payments, new financial products and services might spawn as well, each with their own problems to solve. These include:

1. Outcome Insurance: Guaranteeing agent performance or insuring against errors.

2. Agent Credit Systems: Financing agent actions in a postpaid model.

3. Cross-Agent Settlement: Infrastructure for agents to transact with each other.

4. Value Attribution: Reliably proving which agent actions led to specific outcomes.

5. Crypto's potential rebound? Blockchain's programmable money and instant settlement capabilities seem tailor-made for agent-to-agent transactions. Unlike traditional rails that assume human oversight, crypto smart contracts can autonomously execute complex conditional payments between agents—imagine agents creating escrow contracts or even forming economic alliances without human intermediation.

The Future: Beyond Just Plumbing?

Is Agentic Payments just a feature? Looking at the current offerings, it's easy to see them primarily as enhanced plumbing. They leverage existing tech like tokenization and provide SDKs for what feels like automating existing payment flows, just triggered by AI instead of a human click.

Stripe’s focus on metered billing seems the most immediately practical application for the evolving SaaS landscape.

From an investment opportunity lens, maybe I'm being unimaginative, but it's hard to build conviction in the India opportunity today. If agentic payments is just a feature, as displayed by the major payment giants, then distribution will win and these larger payments companies have solved for distribution and more importantly trust.

Secondly, the AI coding agent use case, the most sophisticated agent framework today has achieved only about 30-35% autonomy! That's it. Most other use cases, similar to the one I mentioned above, are far more complex, with far more "agents" to coordinate with. Any product manager will tell you that the more "agents" you have coordinate with, the more "hops" your agent has to do, increasing the failure rate of your orchestration agent. All this to say that I don't know if autonomous agents are 6 months away or 6 years away. With the pace of AI advancement these days, it's hard to say, but I'm currently leaning toward the longer timeline. And the longer timeline is never a good thing for venture capital allocators with a tight fund timeline.

On the flipside, this "feature" view might be too narrow. The fundamental difference isn't necessarily how the payment moves, but who or what initiates it and based on what context. These new payments infrastructure enable:

• An outcome-based economy: Agentic payments along with metered billing provide the mechanism to reliably measure and charge for value delivered by AI, making sophisticated and complex outcome-based pricing feasible.

• Autonomous Commerce at Scale: Moving beyond suggestions to automated execution requires payment rails built for machine interaction and control.

• Sophisticated B2B Automation: Enabling agents to handle complex financial operations like procurement or treasury management requires programmable, secure money movement.

• Crypto's time to shine:

The real innovation gap – and perhaps the space where new, focused companies could emerge – might not be in just executing the payment transaction itself, but in the essential surrounding layers:

• Agent Identity & Authorization (KYA): Building the trusted systems to manage agent consent, permissions and liability.

• Agent-Native Billing Platforms: Creating flexible engines designed explicitly for complex outcome/usage-based pricing driven by AI performance.

• Cross-Platform Orchestration: Enabling seamless interaction and settlement between different agents and payment systems.

Conclusion: The Dawn of Agentic Finance

Incumbents like Visa, Mastercard, Stripe, Razorpay and PayPal have a significant advantage in rolling out these initial toolkits because of their distribution. The underlying shift towards programmatic, agent-driven value exchange however is more than just incremental. It feels foundational.

Think of it like the early internet protocols – simple in concept, but they enabled entirely new ways of communicating and transacting. Agentic Payments could be the financial protocol layer for the coming age of autonomous systems.

While the specific tools launched today might seem like features, building the trusted, secure, and interoperable system for this new agent economy is the real long-term opportunity and challenge.

The question isn't if agentic commerce will grow, but whether its financial backbone will be merely an extension of today's networks, or if it will necessitate new, specialised infrastructure players to address the unique complexities of a world where machines manage money.

Which pricing model do you think will dominate the agent economy—usage-based or outcome-based? And if you're building an AI agent product today, would you develop your own payment integration or partner with an existing provider? Drop your thoughts in the comments or reach out directly.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here's an earworm: This week Glorious by Macklemore and Skylar Grey (Youtube / Spotify) got me bobbing my head to their beats and lyrics. Give it a listen!

✨ Call outs

1. Apollo Global’s 40-slide presentation on how US consumers and businesses are responding to the tariffs. Spoiler: Recession by US Summer.

2. Matthew Prince, Cloudflare CEO, pitching Cloudflare at Techcrunch Disrupt back in 2010. Technically, Cloudflare is a very complex business, it abstracts away a lot of the tooling, piping and more of the internet for us. But his presentation doesn’t get into any of the technical details, he instead constantly repeats that Cloudflare can be used by clicking just one button. That’s his entire pitch. Love it!

👋🏾 That's all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.

Welcome to the 92nd edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

Been thinking a lot lately about the plumbing of global finance. We live in this incredibly connected world – instant messages, global e-commerce, video calls across continents. Yet, sending *money* across borders often feels like wading through sludge, stuck on infrastructure built for a pre-internet era. Why the disconnect?

It’s this tangled mess – the global spaghetti of correspondent banks, SWIFT messages, and opaque fees – that holds back businesses and individuals. So, how do we fix it?

This week's edition explores a company stepping up with a potential infrastructure upgrade: the Circle Payments Network (CPN). They're aiming to build new, internet-native rails using stablecoins. Could this finally untangle the knots? Let's dig in.

As usual, there’s also a beautiful song recommendation at the end, if you’d like to listen to a song in the background while you read this. Do share your recommendations with me too :)

Thank you for supporting me and sticking around. Enjoy another great week in fintech!

Subscribe now

🤔 One Big Thought

Untangling the Global Spaghetti: Can Circle Fix Cross-Border Payments?

Over the past month, I've had two very diametrically opposite experiences with cross border payments from India.

I was in Sri Lanka recently and made my first International UPI payment. It felt natural/normal - pulled out my phone, scanned the QR code, entered my UPI pin, 15 seconds later the payment was successful and I walked out of the store with my stuff. The problem is that QR acceptance was abysmal in Sri Lanka and banks charge 3% fees for all digital payments, apart from the forex fees. This led me to the insight that for International UPI to work, you need other countries to have a wide spread QR acceptance network - but this is topic for another day.

On the other hand, I wanted to make a transfer to GIFT city for an investment in US. I struggled for 3-4 weeks. First I walked up to my bank branch and told them I wanted to make a transfer to GIFT city, the teller thought I wanted to buy a GIFT card! No jokes! Unfortunately, I had to resort to calling anybody and everybody I knew at my bank, signed physical documents, sent in a physical cheque (which by the time it got delivered to the bank, the forex rate changed, so I had to send it again). But finally, after all of that running around, I managed to get the transfer done.

The pain is real. Anyone want to join my GIFT city remittance survivors support group?

In today's digital world, our global economy has never been more interconnected. Every day, trillions of dollars flow across borders, supporting international trade, supply chains, remittances, and digital commerce.

Businesses source materials from one continent, sell to customers on another, and collaborate with teams scattered across time zones. E-commerce has made global marketplaces accessible from our phones, and communication is instant, virtually free.

There in lies the paradox: while we can instantly stream movies, video chat with colleagues across the world, or order products from thousands of kilometers away, one critical aspect of our global economy remains stubbornly anchored in the past: the infrastructure that moves our money around the world.

Moving money across borders remains frustratingly slow, expensive, and opaque. It's like having a high-speed train network forced to run on decades-old, narrow gauge tracks.

The Broken State of Cross-Border Payments

I know I'm not alone in experiencing the pain of sending money across borders. Anyone who's dealt with international wires knows the pain points all too well.

Despite technological advances in nearly every industry, cross-border payment systems continue to rely on frameworks established in the pre-internet era. The correspondent banking system that underpins international wire transfers emerged in the 1970s, designed for a world that bears little resemblance to our digital-first, global economy.

The consequences of this outdated infrastructure are felt by businesses and individuals worldwide:

• The Plumbing: At the heart of the old system lies "Correspondent Banking". Banks hold accounts with each other in different countries – Nostro ("our" money held by them) and Vostro ("your" money held by us). To actually move money, they send instructions back and forth using SWIFT, a secure messaging system. SWIFT standardises the message (codes, language etc.), transfers the message, audits the messages, but doesn't move the money itself. A typical cross-border payment relies on and passes through 3-5 financial institutions, each adding their own fees and potential delays.

• Banker's Hours: Global commerce operates 24/7, but the banking system largely doesn't. Time zone differences and bank holidays create frustrating bottlenecks in the SWIFT message chains and settlement processes.

• Eye-Watering Fees: All those hops cost money. Sending money internationally is expensive. The global average cost of sending $200 was 6.65% and doesn't include forex conversion rates. This is painful for small businesses and individuals sending remittances home.

• Settlement Limbo: Forget real-time, because of the messaging delays and reliance on Nostro/Vostro reconciliation, traditional cross-border payments can take days, sometimes even longer, to actually settle. This uncertainty wreaks havoc on cash flow planning for businesses and leaves individuals waiting anxiously for critical funds.

• The Black Box Effect: All of the above lead to the eternal question - where is the money? Often, neither the sender nor the receiver have clear visibility into the payment's journey through the correspondent banking maze. Good luck getting anyone in the banking system to

As per RBI, "South Asia continues to offer the lowest remittance transaction costs worldwide, with an average of 5.8 percent for sending $200 (Migration and Development Brief 40, June 2024). This rate represents 80 basis point reduction from the global average of 6.4 percent in the same period. However, it still exceeds the Sustainable Development Goals (SDG) target of a three percent ceiling, underscoring a critical area for improvement."

Think about an importer waiting days for a payment to clear, potentially missing supplier discounts or delaying production. Or imagine a freelancer losing a significant chunk of their earnings to FX fees and wire charges. Or a family reliant on remittances seeing the value eroded before it even reaches them. This isn't just friction; it's a tax on global economic progress.

The Innovation Landscape: Startups (and Incumbents) Chip Away

Naturally, where there's friction, there's innovation. The fintech sector has recognised these problems, leading to a wave of startups developing solutions to improve aspects of cross-border payments. They can be categorised broadly into three kinds:

• Forex Specialists: Companies focused on providing better FX rates and faster transfers for specific corridors (e.g., Wise, Remitly, Niyo).

• Blockchain & Crypto Solutions: Using cryptocurrencies or stablecoins for faster, cheaper settlement, though often facing regulatory hurdles and volatility concerns.

• B2B Payment Platforms: Building rails specifically for business payments, streamlining invoicing and reconciliation.

Globally, startups have begun to chip away at the market share of incumbents and they're starting to take notice.

Back in March 2020, Visa announced it was piloting settling transactions directly in USDC over the Ethereum blockchain, partnering with Crypto.com and using Anchorage Bank as a custodian.

In 2023, Paypal announced the launch of its stablecoin PYUSD. This past week, it announced offering a 3.7% yield on PYUSD holdings and a rewards program, attempting to make its stablecoin more attractive to end users. Just last week, it was reported that the total stablecoin transaction value crossed past Visa's but Visa believes that's not true if you remove bots and other use cases of stabelcoin transactions. It even released a dashboard to make sense of stablecoin data.

Stripe is not far behind either.

As mentioned above, UPI has its own ambitions for international payments. I've covered it in some detail in the previous edition #91. Alipay is another major payments institution and the world is sleeping on its global expansion network. We all are disregarding just how widespread Alipay acceptance is across Asia, Africa and Europe for over 900M Chinese users of Alipay.

In that same previous edition of Fintech Inside, I also covered Nexus Network that the central banks of Singapore, Malaysia, India and others in SEA are building. Now, a few digital asset providers in US have come together to build the Lynq network.

This is a major signal – incumbents acknowledging the need to improve cross border payment rails and potentially use stablecoin as a vehicle to address it.

Circle Payments Network: A New Coordination Layer?

Enter Circle Payments Network (CPN). Circle, the second largest issuer of USD-backed stablecoin after Tether, announced the launch of CPN to serve as a new protocol layer in a comprehensive, open, and internet-based settlement system, with regulated payment stablecoins like USDC and EURC at its core.

Instead of just building another payment app or focusing on a single niche, CPN aims to be a foundational coordination layer connecting existing financial institutions in a new way. Conceived as a network, CPN enables banks, payment service providers, and digital asset-native financial institutions to connect directly with each other through a trusted counterparty framework.

Circle wants you to think of CPN less as a direct payment rail and more as an intelligent, compliant switchboard for global money movement, built for the internet age, leveraging stablecoins. It seems to want to get rid of the entire correspondent banking system we've all come to know and hate.

Here’s what makes it interesting, IMO:

• Stablecoin Core: CPN leverages regulated, fiat-backed stablecoins like USDC (USD-backed) and EURC (EUR-backed) as the primary settlement asset. This offers the speed and global nature of digital assets combined with the stability required for institutional finance.

• Compliance by design: From the ground up, CPN seems to be designed with compliance and governance at its core. CPN wans to upgrading stablecoins within regulatory frameworks, including robust AML/CFT standards and KYC/KYB checks for participating institutions.

• Near-Instant Settlement: By reducing intermediaries and using blockchain rails for the value transfer leg, CPN facilitates near real-time settlement across borders, 24/7/365.

• Lower Costs: Fewer intermediaries inherently mean lower transaction costs compared to the multi-hop correspondent banking system involving Nostro/Vostro accounts and SWIFT fees.

• Programmability: Built on blockchain principles, CPN opens the door to programmable payments and smart contract integration, enabling new use cases beyond simple transfers (think automated escrow, conditional payments, etc.).

Importantly, similar to Visa, CPN does not seem to want to move funds directly; rather, it wants to serve as a marketplace of financial institutions and acts as a coordination protocol that orchestrates global money movement and the seamless exchange of information.

How CPN Works

Imagine a US business needing to pay a supplier in Mexico:

1. The US business initiates the payment via their bank or payment provider (the Originating Financial Institution - OFI).

2. The OFI, connected to CPN, converts the USD to USDC (or uses existing USDC) - crypto on-ramp.

3. CPN facilitates discovery and connection with a suitable partner institution in Mexico (the Beneficiary Financial Institution - BFI). This involves checks on compliance, capabilities, and FX rates if needed.

4. The USDC is transferred almost instantly over supported blockchain networks to the BFI.

5. The BFI receives the USDC, converts it to Mexican Pesos, if required (crypto off-ramp), and credits the supplier's local account.

6. The key is that CPN orchestrates this flow, ensuring compliance data is shared appropriately (like FATF Travel Rule info), FX can be handled efficiently (potentially via integrated venues), and settlement occurs rapidly between trusted, vetted institutions, potentially reducing the need for multiple SWIFT messages and Nostro/Vostro hops.

This model benefits:

• Businesses: Faster supplier payments, quicker receipt of international sales, improved treasury management, reduced working capital requirements, and simplified operations.

• Individuals: More value reaching recipients and less uncertainty about when funds will arrive. Potential for low-cost micropayments.

• Financial Institutions: Access to modern, efficient, global payment rails; new product opportunities; reduced operational complexity for cross-border transactions.

The Future Outlook: Towards the Finternet?

CPN is an interesting solution to modernising out global financial networks and making it truly internet-native. CPN seems to be the practical implementation of the "Finternet" – a concept explored by folks at the Bank for International Settlements (BIS). The Finternet envisions multiple, interconnected financial ecosystems, much like the internet itself, designed to be user-centric, interoperable, and leveraging technologies like tokenization and unified ledgers.

I've long held the belief that nothing will shake Visa or Mastercard from its duopoly position. They're just so well rooted into the banking system. The wave of nationalistic protectionism globally is one reason that stronghold could be weakened, but CPN makes for a compelling reason as well.

Even SWIFT is so engrained in the process for cross border payments among banking systems, that it seems tough to even think that CPN will kill SWIFT. Banks today are incentivised to keep things status quo, they lose all incentive and control by working with CPN. However, Standard Chartered, Deutsche Bank and others are pilot partners with CPN on this network. Let's see how it works.

Further, CPN works in countries where there are existing stablecoin/crypto regulations. Most countries globally are cautiously adoption stablecoins and introducing regulations, with the big exception of India. India has stifled growth in the market by introducing prohibitive tax and requiring approvals from the Financial Intelligence Unit to operate a crypto/stablecoin platform.

Growing adoption for CPN is not going to be easy - there're so many moving parts, entities and systems to integrate. It's a gargantuan undertaking, IMO.

It's perhaps more accurate to see CPN as an alternative set of rails designed for the digital asset age, potentially complementing or competing with traditional systems depending on the use case. It could significantly disrupt traditional B2B cross-border settlement and remittance corridors currently dominated by the old correspondent banking model.

The journey requires network adoption, continued regulatory clarity, and seamless integration. But CPN, and the broader move towards stablecoin-based settlement and the Finternet vision, present a compelling picture of a future where moving value globally is as easy as sending an email. It’s definitely one to watch.

What do you think? Is CPN the missing piece, or just another layer in a complex puzzle? Let me know your thoughts.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here's an earworm: Discovered only recently that Billie Eilish’s brother Finneas has his own solo tracks. This one - For Cryin’ out Loud (Youtube / Spotify) makes for easy listening and you’ll very quickly find yourself singing the lyrics during the day. The trumpets and saxophone are added benefits.

✨ Call outs

1. Video interview of Fred Wilson and Brad Burnham, founders of Union Square Ventures, reflecting on the last 20 years of their journey with USV. Excellent excellent interview of the two legends - no fluff, just straight insights, strategy and grit.

2. The Studio on Apple TV (trailer). I love Seth Rogen and he’s really knocked it out of the park with the Studio. Funny, easy watch but visibly made with a lot of love for the movies and music.

3. Video on Why Going To Space Costs So Much. In the process of explaining why rockets need to be reusable, this video goes into a lot of detail about the business model of space - the revenue, costs, margins, logistics and so much more.

👋🏾 That's all Folks

If you’ve made it this far - thanks! As always, you can always reach me at os@osborne.vc. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.