India’s shift to electric vehicles rests on the ability to charge. Today, solving that has less to do with the vehicle itself than with everything it plugs into — the charger, the connector, the software, the protocols that charging networks use to talk to one another, and, most importantly, the ageing Indian electricity grid beneath all of it.

To chat about all this, we got Zohra Khan, the founder of IPEC, a Bangalore-based company that designs and manufactures EV chargers for India’s leading two- and three-wheeler OEMs, and is now moving into four-wheelers and buses. She is a third-generation entrepreneur from the MEHER Group, which has been building power quality systems and power electronic components in India for more than forty-five years — a lineage that turns out to matter a great deal when the problem you are solving is a highly-volatile Indian grid.

Me and my teammate Vignesh sat down with her to understand the parts of the EV transition nobody puts on a spec sheet: why two chargers that look identical are really running different software, why public charging has three separate ways to fail, why India is a power-surplus country with a distribution problem, and why a charger built to survive Bangalore in the monsoon might be the most exportable product IPEC makes.

IPEC is also one of three founding members of LEAF, or the Light Electric Vehicle Acceleration Forum, launched alongside Ather Energy and Hero Vida. The aim of LEAD is to standardize charging for the two- and three-wheelers that the rest of the world is watching India figure out.

Here’s 6 interesting quotes from a conversation full of them:

On why two chargers that look identical aren’t

Ask most people what differentiates one EV charger from another and they’ll point at the box — its size, its brand, the connector on the end of the cable. Zohra’s answer is that the box is the part that doesn’t differ. Strip an Ather charger and a Bajaj charger down to their power electronics and you find the same job being done the same way. The thing that actually separates them is invisible: the firmware, and the language each vehicle’s battery insists on being spoken to in.

“So honestly, they have a lot more in common than people think. At the hardware level, what you’re doing is fundamentally moving energy from an AC power grid, converting it into DC, and then charging a battery. That’s essentially what you’re doing. So the main power electronics at the core is mainly a rectifier, then you also have multiple conversion stages and protection circuits which are required to protect these electronics. This is largely the same across all OEMs. Whether it’s an Ather or an Ola or a Bajaj, the basic hardware remains common because the function is the same. Where it actually differs is on the software protocols.”

On the three separate ways public charging fails

Home charging, Zohra says, is a solved problem — the charger comes integrated with the vehicle and behaves like charging your phone overnight. Public charging is where the friction lives, and it’s worth being precise about it, because “the charging network is bad” is actually three distinct failures stacked on top of each other. Your connector might not physically fit. If it fits, the software might not agree on how to talk. And even if both line up, the grid feeding that charger might not be clean enough for it to run at all.

“The first one is the actual physical connector. Let’s say I have an Ather vehicle and I go to any public infrastructure. The first thing I need to see is, is the connector on my vehicle compatible with the connector on the charging station? So this is one distinct layer which could cause friction. This is a physical layer. The second thing is a protocol or software layer. Even if the connector matches, you usually need to also have software interoperability.”

On why standardisation is a land grab

The two-wheeler and three-wheeler segment that India leads globally still hasn’t settled on common connectors and protocols, the way four-wheelers have converged on CCS2. That sounds like a problem. Zohra frames it as an opportunity with a closing window. India is writing these standards more or less from scratch, and the firms in the room while they’re written get to shape them — which is why IPEC helped launch LEAF with Ather and Vida rather than waiting for a regulator to decide.

“And whoever joins this first, or standardizes this first — these are the people who will shape how the standards are written for the future. Every OEM now knows which side of the table they want to be on. And this is what we are leading the effort on, along with Ather and Vida. Honestly, it has been quite encouraging. We thought we’d face a lot of resistance, but OEMs have been quite welcoming, and not just OEMs — different ecosystem players like charge point operators, infrastructure players, charging connector companies, charger companies are all coming together.”

On why India is power-surplus but still short of power

The intuitive story about Indian electricity is scarcity — not enough generation. Zohra’s correction is one of the most useful frames in the conversation: at the level of generating and transmitting power, India has largely caught up. The failure is in the last mile, the neighbourhood-level distribution layer that nobody photographs for a ribbon-cutting.

“If you look at it, we’re actually a power surplus country officially. The amount of power we generate is enough to meet our needs right now. But I think what India has is a distribution problem. And that’s where the DISCOMs and a lot of these players come into the picture. This distribution layer, which is mainly the 11 kV and the 415 volt lines — that’s what actually connects the grid to your neighborhood. Right now a lot of these lines are overloaded.”

On building for the worst grid in the world as a moat

Here’s where the grid-quality problem turns into a strategy. Chargers designed for the stable grids of Europe or China assume a voltage that mostly stays put. India doesn’t offer that. IPEC’s response is to engineer for the worst case as a matter of course — wider voltage tolerance, far higher surge protection — and Zohra’s argument is that a product hardened for India isn’t just a defensive necessity. It’s the most globally competitive thing the company could possibly build.

“And that’s also an advantage, a competitive advantage that I think we would have. Because one, we’re forced to build very reliable products, but India is also a very large market, so you’re also building these products at scale. So it’ll be reliable, but you can also make them cost competitive, and then you can export these to any part of the world, because they will work anywhere. So I actually see that as an advantage. Whether it’s a Chinese charger or a German-made charger, I would say we’ll do it better and cheaper.”

On the one part India can’t make yet

For all the talk of self-reliance, Zohra is precise about where it currently ends. IPEC’s domestic value addition is already past 55% — enclosures, cables, magnetics, and increasingly the capacitors (one of its investors, Deki Electronics, has localised those). But the brain and the muscle of any power electronic device — the microcontrollers and the power semiconductors — still come from abroad, and no amount of will closes that gap quickly.

“So the hardest part, if you ask me, would be the semiconductors and microcontrollers to indigenize or localize. Because right now almost all of these are being primarily imported either from China — these are mainly US, European, Japanese manufacturers, but all their supply chains are in either China or Taiwan. So that’s just the reality of the supply chain and how it exists. But like I said, the government is pushing this, a lot of policies and incentives are in place, and we are going to see that changing in the next three to five years, but not anytime soon.”

The full conversation goes further — into how OCPP and OCPI let any charger talk to any network, who pays when an OEM changes its battery chemistry, the line between IPEC’s intellectual property and an OEM’s, whether OEMs building chargers in-house is a real threat, and why Zohra is certain IPEC will never build a scooter.

Listen to the full conversation on YouTube, Spotify, or Apple Podcasts. The full transcript of the podcast is below if you prefer to read.

Subtext by Zerodha

Zohra Khan on the business of Indian EV infra

India’s shift to electric vehicles rests on the ability to charge. Range, or rather the fear of running out of it, is still the single biggest reason EV adoption hasn’t picked up as strongly. And solving that has less to do with the vehicle itself than with everything it plugs into — the charger, the connector, the software, the protocols that charging …

Read more

9 days ago · 7 likes · Zerodha

Share

Hi folks, welcome to another episode of Who Said What? I’m your host Kashish and if you’re new here, here’s the quick context.

The idea is simple: we pick the most interesting comments from business leaders and break down what’s really going on behind them.

This builds on a newsletter we run called The Chatter by Zerodha, where we track what company managements say in their earnings calls or TV interviews. But instead of looking at those comments in isolation, this format zooms out. We track one theme across companies and over time to see how the story evolves. The goal is to connect the dots and understand the deeper shifts shaping an industry.

For today’s episode, we we wanted to know how are Indian pharma companies planning to make money off of the GLP-1 trend?

On the evening of 20 March 2026, hours before Novo Nordisk’s Indian patent on semaglutide formally expired, Natco Pharma announced that its generic vial would sell for ₹1,290 a month. The original drug, at innovator pricing, had been going for roughly ₹16,000. By the next morning, six more domestic firms had piled in. By the end of the week, thirteen companies had launched twenty-six brands. Within a month, more than fifty were on the market or in the queue.

The interesting question isn’t who launched — almost everyone did. The interesting question is what each company is actually doing: whether they’re making the drug, selling it, both, or simply renting out their distribution to someone who has already done the harder work. The same molecule, the same patent cliff, the same 250-million-strong pool of potential patients — and six very different ways to make money off all of it.

Over the last four quarters, we read through earnings calls from 27 listed Indian pharma companies — more than 140 calls in total — to map what each one is actually up to. Here’s what we found.

A note on the prize before we get there. India’s GLP-1 market — semaglutide plus Eli Lilly’s tirzepatide — was worth about ₹571 crore in 2024. By February 2026, on a rolling twelve-month basis, it had jumped to ₹1,446 crore, with the anti-obesity slice alone more than doubling. But that figure is the market the innovators built: until 21 March, semaglutide was still on patent, so almost all of it was Novo’s and Lilly’s branded money.

Three estimates of where it goes from here: Systematix sees a ₹5,000-crore incremental opportunity across India, Brazil and Canada combined over the next 12-15 months; CareEdge expects India alone to reach ₹5,000 crore by 2030; and Jefferies, more bullish, sees a market north of $1 billion in India.

Set against those projections, what the new entrants have actually booked is almost nil. Cipla’s Yurpeak — itself an innovator launch, Lilly’s tirzepatide sold under a Cipla brand — did ₹14 crore in its first month. Eris’s liraglutide, an older off-patent GLP-1, does about ₹1 crore a month. Every generic semaglutide launched only in the last week of March, so a full quarter of post-launch sales simply doesn’t exist yet. What follows is a description of strategy, not of a profit pool that has already formed.

A primer — what it is, who makes it, how it’s sold

Start with what the drug actually does, because that’s what explains the gold rush.

GLP-1 — glucagon-like peptide-1 — is a hormone your gut releases after you eat. It nudges the pancreas to put out insulin, tells the brain that you’re full, and slows down how quickly the stomach empties. Drugs like semaglutide are engineered copies of that hormone, tuned to last a week in the body instead of a few minutes. Take one injection a week and your appetite quietly shrinks, your blood sugar steadies, and the weight comes off — often 15% of body weight or more.

Now, the cast.

The two innovators. Two foreign drugmakers built the entire global GLP-1 market between them.

Novo Nordisk, a Danish company, makes semaglutide — the molecule everyone is now copying. Novo sells it under three brand names: Ozempic (an injection, for type 2 diabetes), Wegovy (the same molecule at higher doses, for obesity), and Rybelsus (an oral tablet). When people say “Ozempic” colloquially, they usually mean semaglutide in general.

Eli Lilly, an American company, makes tirzepatide — a different molecule that works on the same biological pathway but has shown stronger weight-loss results in trials. Lilly sells it under one global brand, Mounjaro. In India, Lilly has partnered with Cipla to market it under a separate brand, Yurpeak.

What “going off patent” actually means. A patent gives the original drugmaker an exclusive selling window. When it expires — the industry calls this the “loss of exclusivity”, or LOE — other companies can legally make and sell copies, called generics. Semaglutide’s Indian patent expired on 21 March 2026. Tirzepatide’s runs until at least 2030. So semaglutide is now fair game. Tirzepatide is still legally Lilly’s alone.

Pens vs vials — why the device matters. Semaglutide is an injection a patient takes once a week, for years on end. How it’s packaged changes the economics entirely.

A vial is a glass bottle of the liquid drug. To inject, the patient — or a clinic nurse — has to draw the right dose into a separate syringe. Vials are cheap to manufacture but fiddly to use: dosing errors are common, and most patients won’t self-inject from one.

A pen is a pre-loaded injector with a dial. The patient turns the dial to the prescribed dose, presses the pen against the skin, and pushes a button. It costs three to six times more to make than a vial. But it’s dramatically easier to use — and adherence, patients sticking with the therapy long enough for it to actually work, depends almost entirely on the device.

Hold that distinction in your head. It does most of the work in Bets 1 and 6.

Bet 1 — Be there on day one with a generic pen

The default play: show up on 21 March with a branded generic, fight for share through the sales force, and ride the volume.

Almost every Indian pharma company took some version of this bet — Sun, Dr Reddy’s, Lupin, Zydus, Alkem, Natco, Cipla, Glenmark, Torrent, Mankind, Eris. The interesting thing is that nobody priced the same way. Within ten days of launch, the Indian generic semaglutide market had split into three distinct price tiers — and the tier a company chose determined who its patient was.

The cheapest tier was vials. Natco opened at ₹1,290 a month for a vial; Glenmark’s GLIPIQ launched even lower, at ₹325 a week. These prices are roughly a tenth of Novo’s. The trade-off: patients can’t easily self-inject from a vial, so this tier really only works in clinics where staff can manage the dosing — or with the small fraction of patients comfortable handling a syringe.

The middle tier was affordable pens — Alkem, Zydus, Natco’s own pen device — in the ₹1,800-4,500-a-month range. Same molecule as the vial, but in the easy-to-use injector. This is where the mass-market volume will likely settle once price competition has done its work.

The premium tier was high-end pens with patient support wrapped around them — Dr Reddy’s at ₹4,200, Sun at ₹2,960-7,400 depending on dose. These companies are betting that better device design, a branded experience and patient education can justify a premium even after generics have flooded the market.

Same drug, ₹325 a week to ₹7,400 a month. Different patients, different doctors, different settings.

“Sun Pharma plans to be in market on day-one of a generic launch. We have already received the regulators’ approval for both the indication of chronic weight management as well as treatment of type 2 diabetes under the brand name Noveltreat and Sematrinity respectively.”

— Kirti Ganorkar, Managing Director, Sun Pharma | Q3 FY26

The honest part of this bet is that everyone running it knows exactly what it looks like. Crowded. Loud. Margin-thin within a few quarters.

“There’s a mad rush in the market. Every company is launching. And if at this particular time, you launch, you will be lost somewhere.”

— Rajeev Juneja, Vice Chairman & MD, Mankind Pharma | Q4 FY26

“I am not going to tell you that we are going to make a lot of money in India, I will never say that, I actually agree with you, it will be very competitive.”

— Rajeev Nannapaneni, Vice Chairman & CEO, Natco Pharma | Q2 FY26

So why run it anyway? Two reasons. The market is so big that even a thin slice is real money. And being absent on day one means being absent forever — Indian doctors tend to stick with whichever brand they wrote first. That’s why Lupin, Dr Reddy’s and Sun all set internal targets to rank top-three among generics from week one.

“Our goal is very clear to be at Top 3 player in this as generic Semaglutide and I think we’re well on track.”

— Nilesh Gupta, Managing Director, Lupin | Q4 FY26

The early post-launch read suggests Torrent has surprised everyone. The single decision that won it that share was launching the oral version alongside the injectable — something no other generic did at scale.

“Torrent held 38% share among generic players as per the Pharmatrac April data set, with 28% share in the injectable format and 100% share in the oral format.”

— Aman Mehta, Management, Torrent Pharmaceuticals | Q4 FY26

In rupee terms, even Torrent’s standout is small — about ₹17 crore of monthly sales against a quarterly revenue base of roughly ₹3,000 crore. For now, the point of Bet 1 is defensive: don’t let the competitor own the diabetes prescription pad. The money comes later.

Bet 2 — Don’t fight the generics, sell the original

Here’s the counter-bet. If the generic market is going to be a knife fight at ₹1,290 a month, why be in it at all? Instead, partner with the original drugmaker, sell their actual product at a premium, and lean on their global data and their device.

Emcure took this bet hardest. In late 2025, it signed an exclusive deal with Novo Nordisk to launch Poviztra in India. Poviztra is the same drug as Wegovy — Novo’s semaglutide for obesity — just sold under a different brand name in India by Emcure. The molecule is identical. The device is identical (Novo’s own pen). And the data behind it is the same global trial evidence Novo has been collecting for years.

“It was after a fierce competition there were at least 8 or 9 companies who were interested in partnering with Novo Nordisk and they selected Emcure. And we became the only company to launch the innovator drug in the game-changing semaglutide category.”

— Satish Mehta, MD & CEO, Emcure Pharmaceuticals | Q3 FY26

The logic is straightforward. Post-LOE, you have 17 generic semaglutide brands fighting for shelf space, and none of them have Novo’s vast trial record — the evidence on how the drug behaves in patients who also carry other conditions, what the industry calls comorbidities. So Emcure’s pitch to doctors is simple: this is the actual Wegovy, just at a more accessible price than the imported original.

“If you look at the construct of this deal, obviously, post-LOE, we anticipate a very crowded market in India. And so the ability to partner with Novo and launch this brand ahead of the competition and allowing ourselves to shape the market and have brand recall versus being lost in the crowd probably 4, 5 months down the line is something that we feel strongly will allow us to succeed.”

— Vikas Thapar, President, Corporate Development, Emcure | Q2 FY26

Cipla ran a different version of the same bet — and arguably the cleaner one. Rather than partnering with Novo on semaglutide, which is now copyable, Cipla partnered with Eli Lilly on tirzepatide, launching Yurpeak in India in late 2025. Yurpeak is Lilly’s Mounjaro, marketed in India by Cipla under a different name.

The crucial difference is the patent. Tirzepatide’s Indian patent runs until at least 2030, so Cipla isn’t fighting any generic competition at all. It’s selling a still-patented innovator drug, at innovator-adjacent prices, into a country where Mounjaro became the second-largest pharmaceutical retail product within a year of launch. It helps that tirzepatide mimics two gut hormones rather than one — GIP as well as GLP-1 — which is why its trials show even bigger weight loss.

“Tirzepatide is the first and only dual agonist of GIP and GLP-1, which makes it a very unique proposition. And while both products have been launched by the respective innovators, you can see from the uptake that Tirzepatide uptake is far ahead.”

— Achin Gupta, Global COO, Cipla | Q2 FY26

“As you see from IQVIA data, last month the molecule was clocking upwards of Rs. 130 crores a month. So, our efforts are primarily focused on that.”

— Achin Gupta, MD & Global CEO Designate, Cipla | Q3 FY26

Yurpeak’s ₹14 crore in its first month annualises to about ₹168 crore at the launch run-rate. Small against Cipla’s quarterly revenue — but it’s a healthy margin on a molecule that is still on patent, while the rest of the industry slugs it out at near-cost prices.

Both Emcure and Cipla are making the same wager: don’t try to be the cheapest. Try to be the only one allowed to sell the real thing.

Bet 3 — Differentiate the device

A third group accepted that they’d be launching a generic, but decided to win share through a better pen rather than a lower price.

Zydus went furthest here. The standard semaglutide regimen forces patients to switch devices as the dose climbs — one pen for the first four weeks at the starter dose, another at the next level, another for maintenance. Three different disposable pens over the first few months. Zydus built a single reusable pen with a 15 mg / 3 ml cartridge that handles every dose.

“We have a new formulation where it’s a very significant ease to the patient in terms of use. Also, there is a very meaningful benefit in terms of cost because today, we have in, typically for weight loss, you have to shift from the first four weeks to the next four weeks with a different dose and that means a different pen device. For us, we have a…. and we don’t need to make any changes from the initial dosing to the higher dosing.”

— Dr. Sharvil Patel, Managing Director, Zydus Lifesciences | Q3 FY26

The clever part is what Zydus did next. Instead of trying to take share alone, it licensed the same pen to Lupin and Torrent under semi-exclusive co-marketing deals. Three of the strongest sales forces in India, all pushing the same differentiated device.

“This being a highly competitive product with the multiple launches, we believe that the best strategy would be to launch with more players to create more share of voice and more impact for the new formulation with the customers.”

— Dr. Sharvil Patel, Managing Director, Zydus Lifesciences | Q4 FY26

The arithmetic is striking. Lupin’s, Torrent’s and Zydus’s generic injectable market in April 2026 — 57% of all generic semaglutide sales in India — all flow back to Zydus’s Ahmedabad production lines. The device was the moat and partnerships with others was the distribution.

“The product from Zydus is a unique pen and I think that’s differentiation in the market as well. Our patient support program is pretty solid as well. So, a so good start. I think we’re the number two company as a generic one, the number three as a product itself.”

— Nilesh Gupta, Managing Director, Lupin | Q4 FY26

Sun Pharma’s differentiation was an auto-injector — a pen designed so the needle stays hidden and the injection happens automatically when the patient holds it against the skin. Same molecule, same dose; what changes is the patient’s experience.

“The product has been received very well by the doctor community and one of the differentiating points is our auto-injector as well as the pen system which was appreciated by the doctors and patients equally well.”

— Kirti Ganorkar, Managing Director, Sun Pharma | Q4 FY26

Bet 4 — Sit out the rush

Mankind Pharma did something none of its peers did. It launched a pen, then deliberately refused to compete on price, refused to launch the cheaper vial, and started pointing its energy elsewhere.

“In this competitive market, we have launched our GLP Pen. And we are not in a hurry to launch vials. We are not in a hurry to cut down the prices.”

— Rajeev Juneja, Vice Chairman & MD, Mankind Pharma | Q4 FY26

This is the contrarian bet. The logic: when 50 companies are fighting over the same molecule at falling prices, the better business isn’t winning that fight. It’s selling all the other things patients on GLP-1 also need.

“We are focusing on GLP-1. It’s a long term and a large opportunity for any Indian company to ignore that. But we are also focusing on adjacent and supportive therapies like vitamins and minerals and protein, which is going to grow better in the time to come.”

— Sheetal Arora, CEO, Mankind Pharma | Q4 FY26

The numbers suggest the basket strategy is already working. Mankind’s anti-diabetes portfolio grew 14.4% in Q3 FY26 — about 1.9 times the market, excluding new GLP-1 launches. Chronic-care now contributes 39.3% of its domestic business. The semaglutide pen launched into an existing chronic-care engine, not into a vacuum.

“What we’re doing, we are basically working on the adjacent therapies like vitamins, minerals, protein side because ultimately, a company which would be selling the complete portfolio will be having a better advantage. So our approach is a bit different than the rest of the people actually.”

— Rajeev Juneja, Vice Chairman & MD, Mankind Pharma | Q4 FY26

There’s a longer game behind it, too. Mankind has a novel molecule, MKP10241 — a small-molecule oral obesity drug — in Phase II trials in Australia. Nothing for FY27 or FY28. But it’s the kind of asset that can turn a defensive GLP-1 launch into an offensive franchise if it works.

Bet 5 — Export the molecule

For companies with international footprints, the Indian market is the appetiser. The main course is everywhere else the innovator is still supply-constrained — Brazil, Canada, the Middle East, parts of Europe.

Dr Reddy’s is the most aggressive here, with 26 GLP-1 products in the pipeline and a plan to launch across 87 countries.

“First of all, I see that as many, many years of opportunity. Actually, we are entering a decade of GLP-1 products. Obviously, it’s going to change and evolve. The full portfolio of GLP-1 for the company is 26 products.”

— Erez Israeli, CEO, Dr. Reddy’s Laboratories | Q1 FY26

Different companies are betting on different geographies. Torrent’s whole international strategy hinges on Brazil — a market whose regulator, ANVISA, is so strict about semaglutide that pharmacies have to keep physical copies of every prescription.

“Each prescription for Semaglutide has to be issued in two copies. And with the indication on the prescription and the pharmacy is required to maintain extensive records, including a copy of each prescription.”

— Sanjay Gupta, Executive Director, International Business, Torrent | Q1 FY26

“Roughly the market is $1 billion for Semaglutide. You can divide it into the injectable at 75% and oral at roughly 25%. Currently, the Ozempic market is declining very fast and Wegovy is growing very fast. There are no generic launches in Brazil as of today.”

— Management, Torrent Pharmaceuticals | Q4 FY26

Biocon is going direct in Europe — selling its own generic liraglutide, an older GLP-1 drug, in the Netherlands rather than through a partner — and out-licensing semaglutide to Ajanta Pharma for 26 countries across Africa, the Middle East and Central Asia.

“We launched generic Liraglutide for diabetes and obesity in the Netherlands as our first ‘direct-to-market’ GLP-1 in the EU. We also signed an out-licensing agreement with Ajanta Pharma to market our vertically integrated drug product, Semaglutide, in 26 countries across Africa, Middle East and Central Asia.”

— Kiran Mazumdar Shaw, Executive Chairperson, Biocon Group | Q3 FY26

Biocon’s cost edge here is structural. The company spent years building fermentation-based capacity to make its peptides — growing them in living cells — rather than taking the more common route of stitching them together through synthetic chemistry. At scale, fermentation brings the cost of the active ingredient down from roughly $200 a gram to between $20 and $50 — a 4-10x advantage rivals can’t easily replicate. In a generic market that will eventually be priced like a commodity, that’s the difference between making money and not. In the September 2025 quarter, Biocon’s generics business grew 24% year-on-year to ₹774 crore, with margins on its biosimilars — near-copies of complex biological drugs — expanding to 28% at the operating-profit (EBITDA) level.

Lupin is leaning on South Africa and Brazil. Emcure picked up Quebec through a Dr Reddy’s partnership. Zydus is taking its differentiated pen to 20-plus markets.

The wrinkle is that this bet has been slower than anyone expected. Canada — meant to be the first big market after the patent expired — stalled for most of the year, with Health Canada clearing no generic GLP-1 at all. Dr Reddy’s, after a Notice of Non-Compliance in October 2025, finally won approval — the first one through. For most of the year, the picture looked like this:

“Of course, the review cycle is long drawn, especially in markets such as Canada, where we have not seen a single generic GLP being approved, including liraglutide, which has not been approved by Health Canada.”

— Siddharth Mittal, CEO & MD, Biocon | Q3 FY26

The export bet has a cleaner economic logic than the domestic one. Innovator pricing abroad is still high, generic competition is thin, and the ₹5,000-crore incremental opportunity Systematix talks about is, in the analysts’ own framing, largely an export number — concentrated in far fewer hands than the domestic chaos suggests.

Bet 6 — Sell the shovels

In any gold rush, some people decide not to dig. They sell the shovels.

In Indian GLP-1, the shovel-sellers split into two distinct camps. One is making the pens for everyone else launching this year. The other is making the chemical raw materials for the next generation of drugs, the ones nobody has launched yet. Both businesses are quietly more profitable than selling the generic itself.

Sub-part A — Make the pens for everyone else

Semaglutide is a hard drug to manufacture. It’s a 31-amino-acid lipidated peptide — a chemically intricate chain that most Indian small-molecule drugmakers can’t produce consistently at industrial scale. And even the companies that can make the active ingredient often can’t fill it into pens themselves, because cartridge fill-finish — the sterile step of loading the drug into the device — needs specialised, isolator-based sterile lines. There are perhaps three or four such facilities in the whole country.

So a separate set of companies decided not to launch generics at all. They decided to make the pens for everyone who is.

OneSource Specialty Pharma is the dominant player. It’s what the industry calls a CDMO — a contract development and manufacturing organisation, a company that builds a drug end-to-end for others to sell under their own labels. And the pitch is exactly that end-to-end-ness: develop the formulation, fill the cartridge, assemble the pen, serialise the box.

“Our biggest value proposition to our customers is end-to-end we supply. Not only we do development for them, we will do commercial manufacturing for fill finish for cartridge, pack it, put it in the box, serialize it. So the box which a patient opens at home is the box coming out of one source site.”

— Neeraj Sharma, CEO & MD, OneSource Specialty Pharma | Q2 FY26

The scale is real. OneSource has signed up more than twenty customers — and has stopped taking new ones. Capacity is being scaled from 40 million units in FY26 to 220 million in FY27. Its customer base holds roughly two-thirds of the Indian generic semaglutide market by value. Several customers have paid capacity-reservation fees up front; others are on take-or-pay contracts, which means they owe OneSource for the reserved capacity whether or not they actually take the product.

“And today as we speak, our customer base total as on end of April holds almost two-thirds of the Indian market share by value of the generic market… We are the first and the only CDMO partner for the first three generic Semaglutide approvals in all the highly regulated markets, US and Canada.”

— Neeraj Sharma, CEO & MD, OneSource Specialty Pharma | Q4 FY26

Gland Pharma is running the same play with cartridges — capacity expanded from 40 million to 140 million units a year, eight GLP-1 contracts signed, six or seven more in the pipeline.

“In the GLP-1 space, we have made significant progress with eight contracts already signed and an additional 6-7 expected to be signed soon. Our current cartridge capacity now stands at 140 million units. Our approach remains disciplined and value-focused, positioning this as a strong mid-to-long-term opportunity with meaningful upside.”

— Srinivas Sadu, Executive Chairman, Gland Pharma | Q4 FY26

The neat thing about this bet is that these companies don’t care who wins the Indian launch war. Whoever ends up selling the pen — they made it.

Sub-part B — Make the building blocks for the next wave

Go one layer deeper. The fill-finish CDMOs need cartridges, the cartridges need an active ingredient, and that active ingredient is a peptide that has to be synthesised from raw chemical pieces. That’s where the second group of shovel-sellers sits.

A quick definition. A peptide is a chain of amino acids — essentially a small protein. Semaglutide itself is a 31-amino-acid peptide. To make a peptide drug, you take individual amino acids, chemically “protect” them so they don’t react in the wrong places, link them together in the right order, and then strip the protections off at the right time. The whole thing get assembled like Lego into the final drug.

Divi’s Laboratories made the most interesting choice here. It flatly refused to make generic GLP-1s. Instead, it sells the Lego pieces — fragments and amino acids — to the original innovators like Novo Nordisk and Eli Lilly, who use them to build their own peptide drugs.

“Divi has strategically decided that we will not look at generic part of peptide synthesis. We are right now fully occupied with the amount of CS projects we have. So, we do not want to venture into that mode.”

— Dr. Kiran S. Divi, Whole-Time Director & CEO, Divi’s Laboratories | Q2 FY26

The economics are very different from Sub-part A. Divi’s makes its own protected amino acids in-house — dull to describe, but it gives the company a grip on cost and on impurity that pure peptide assemblers don’t have. It has earmarked roughly ₹700-800 crore per project for three new dedicated peptide units. Custom synthesis — what it calls its innovator-facing business — already accounts for 56% of Divi’s revenue, and that segment grew 23% year-on-year in Q2 FY26.

“The one unique thing about Divi’s is, we manufacture our own protected amino acids, which gives us both in natural and unnatural, which gives us an edge over everyone, because we control our quantities, we control our cost, we control our impurity profile.”

— Dr. Kiran S. Divi, Whole-Time Director & CEO, Divi’s Laboratories | Q2 FY26

The trade-off is clean. The branded generics are scrapping over a fragmented Indian retail market. Divi’s is selling shovels to the prospectors who still own the gold mine.

Neuland is running a similar bet, but framing it bigger. The global peptide CDMO market, in its telling, has gone from a $1-1.5 billion category to $5-6 billion in just a few years — and most of that explosion is downstream of GLP-1’s success, which has made innovators want to develop the next generation of peptide drugs.

“What used to be a $1 or $1.5 billion market space is now a $5 or $6 billion market space and it is growing very rapidly because more and more drugs are coming out into the market, which are peptides.”

— Saharsh Davuluri, Vice-Chairman & MD, Neuland Laboratories | Q1 FY26

“It’s not just about the weight loss drugs that have been commercialized, but it’s also about the next gens, not just from the large companies, but other companies as well. So there is a plethora of development candidates, which are peptides.”

— Saharsh Davuluri, MD & CEO, Neuland Laboratories | Q4 FY26

Laurus Labs, Syngene, Piramal Pharma and Shilpa Medicare are all running variations of this bet. Different scales, different positioning, but the same thesis: don’t bet on which generic semaglutide wins in India — bet on the entire peptide modality, globally, over the next decade.

A coda — the diagnostics tag-along

There’s one more bet, and it isn’t pharma at all. If millions of new patients go on a powerful weekly injection, somebody has to test their bloodwork. Diagnostic labs see the GLP-1 boom as a downstream opportunity.

Metropolis Healthcare is the most explicit about it:

“On the GLP-1 front, the receptor agonist market in India is currently valued at USD 110 million in 2024 and is projected to grow to USD 500 million by 2030. This shift is likely to drive higher demand for regular monitoring of glucose, lipids, liver, kidney, and cardiac parameters.”

— Ameera Shah, Chairperson, Metropolis Healthcare | Q1 FY26

Dr Lal PathLabs is less convinced — at least so far:

“Patient volume growth is partly a factor of the improved collection and lab network we have put in place. It is definitely not driven by GLP-1.”

— Shankha Banerjee, CEO, Dr Lal PathLabs | Q4 FY26

A bet that’s real but small. The kind of thing that shows up in the numbers two or three years from now, not next quarter.

A footnote on what’s not a bet — the next-generation Indian molecules

Beneath the semaglutide land grab, three Indian companies are quietly building their own proprietary metabolic drugs — ones that could end up mattering more than any generic launch.

Sun Pharma has GL0034 (also called Utreglutide), an in-house GLP-1 molecule now in global Phase II trials for both Type 2 diabetes and MASH, a form of fatty liver disease. Mankind has MKP10241, a small-molecule oral obesity drug, in Phase II in Australia. Lupin is developing its own oral semaglutide in-house and has partnered with China’s Gan & Lee on Bofanglutide, a novel fortnightly GLP-1 agonist.

None of these reach the market for two to four years. None move FY27 earnings. But they’re the only assets in the listed Indian pharma universe that could turn GLP-1 from a defensive franchise into an offensive one.

TL;DR

Semaglutide came off patent in India on 21 March 2026, and within a week thirteen companies had launched twenty-six generic brands. But the synchronised launch hides six different answers to the same question — how do you actually make money off this molecule? Most went for the generic pen and split into price tiers; Eris reframed it as a metabolic drug for “thin-fat” India; Emcure and Cipla skipped the generic fight and sold the innovators’ own products; Zydus won by licensing a better pen to its rivals; Mankind sat out the price war; Dr Reddy’s, Torrent and Biocon chased exports; and OneSource, Gland, Divi’s and Neuland sold the shovels. The point isn’t that everyone launched — it’s that almost no one chose to compete head-on with the company next door. Combined listed-player revenue is still just ₹100-130 crore a month. The real money is two years out, and most of it will be made outside India.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

Hi folks, welcome to another episode of Who Said What? I’m Kashish and if you’re new here, here’s the quick context.

The idea is simple: we pick the most interesting, sometimes spicy, comments from business leaders and fund managers and break down what’s really going on behind them.

This builds on a newsletter we run called , where we track what company managements say in their earnings calls or TV interviews. But instead of looking at those comments in isolation, this format zooms out. We track one theme across companies and over time to see how the story evolves. The goal is to connect the dots and understand the deeper shifts shaping an industry.

For today’s episode, we cover how AI is changing the rules of Indian IT services industry. This edition couldn’t have happened without the help of my colleague, and in-house IT sector expert, Pranav Manie, so a huge shoutout to him.

For nearly three decades, the Indian IT services pitch was simple to draw on a whiteboard. Take work that costs $150 an hour in New York, route it to engineers who cost a fraction of that in Bengaluru, keep the spread, and scale by adding more engineers. The model was so durable that the listed Indian IT industry — TCS, Infosys, Wipro, HCL, Tech Mahindra and the next tier — grew into a $250-billion-plus export machine on the back of it.

FY26 was the year that loop got publicly disowned by the people running the companies. Not because clients stopped buying services — bookings, by and large, held up — but because management itself began naming the model’s mortality on earnings calls. Some called it deflation. Some called it accretive. Some called it disruption they were inflicting on themselves before someone else did it to them. The vocabulary varied. The shape underneath did not. What follows is FY26 in the words of the operators, across five fractures in the old way of doing business.

[1] Revenue and headcount come apart

The first fracture is the load-bearing one. For three decades, headcount was a leading indicator for revenue at every Indian IT major. In FY26, the link visibly broke. Here is HCL’s CEO C. Vijayakumar, in plain numbers:

“If you see our revenue in the last couple of years, we have grown 4%-5% and our headcount has not grown. So, that gives you a sense there is some non-linearity playing out. Even this quarter, in revenue growth and headcount, there is at least 1.5% or 1% difference.”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q2 FY26

Two years of growth without bodies. The word the industry has started using is “non-linearity” — a clean piece of jargon that translates to: the old equation, where one more engineer billed one more set of hours, no longer governs how revenue is made. Mphasis says the same thing in its own language:

“This naturally means that there is certain amount of de-linkage between revenue growth and headcount growth, which, again, we’ve been seeing for the last few quarters.”

— Nitin Rakesh, CEO, Mphasis | Q2 FY26

The most uncomfortable version of the same point came earlier in HCL’s year, when Vijayakumar was asked what happens to the people whose work is now being automated:

“Of course, we have had a good amount of people released due to the productivity improvements. Now, not all of them are readily redeployable because the requirements for some of the entry level or lower end skills are being addressed through Automation and other elements.”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q1 FY26

“Released” is the operative word. So is “not readily redeployable.” The Indian IT majors employ a combined two million people. Even small percentage shifts in redeployability translate to large absolute numbers of careers that have to find a new direction.

Infosys is the outlier — but the way it’s framed is the more interesting part. CEO Salil Parekh confirmed Infosys had added 13,000 people through the first three quarters and would keep adding:

“For the year, we have added ~13,000 net headcount for the first three quarters. My sense is, we will continue to add headcount as we go through. And it sort of comes back a little bit to an earlier discussion we were having, which is there is a macro element and there is an AI element.”

— Salil Parekh, CEO & MD, Infosys | Q3 FY26

Read carefully, that’s not a defence of the old model — it’s a hedge that the macro is still hungrier for engineers than AI is yet capable of replacing them. The decoupling is happening at HCL and Mphasis already. At Infosys it is, for the moment, being absorbed.

The cleanest data point on the decoupling, though, comes from a company that didn’t appear in this set of calls. Coforge — a mid-cap that has had the strongest growth print in the listed Indian IT universe this year — grew revenue roughly 30% in FY26 while its employee cost base grew only around 20%. The gap between the two lines is the entire thesis of this section, expressed in a single income statement. Coforge management hasn’t framed this on a call the way HCL’s Vijayakumar has, but the numbers describe the same phenomenon: the unit of revenue is detaching from the unit of labour.

[2] The margin paradox — same phenomenon, two stories

The strangest thing about FY26 is that the same productivity gain shows up on one CFO’s slide as a discount the company is forced to give clients, and on another’s as a price premium the company is finally able to charge. The cleanest illustration of the dichotomy is that Infosys’s own CFO articulated both views on the same earnings call.

On the accretive side, Jayesh Sanghrajka pointed out that Infosys’s pricing has actually firmed up — and credited AI for it:

“In terms of pricing, I think pricing environment for us has remained stable. On the contrary, actually, most of our growth this year has been pricing-led because the volumes have been softer. And that in a way corroborates with the fact that the AI revenue are coming at a better pricing.”

— Jayesh Sanghrajka, CFO, Infosys | Q4 FY26

And then, a few questions later on the same call, on the deflationary side:

“Market is competitive. As I said, the competitive intensity in the market has gone up and the productivity will get passed back to the client largely.”

— Jayesh Sanghrajka, CFO, Infosys | Q4 FY26

Both statements are true. New AI-led work commands premium pricing because clients are still figuring out what it’s worth. Existing services, where the productivity savings are quantifiable and the competition can match the offer, give the savings back. The phrase “AI is accretive” and “AI is deflationary” describe the same underlying economics applied to two different parts of the book.

HCL has taken the most explicit deflationary view in the industry — to the point of pre-announcing the revenue it expects to lose:

“I think we are being very transparent. We are telling the clients, if you allow us to use AI Force and use all the recipes that we’ve created, we will showcase to you the optimization that is possible. And it will mean some reduction in revenue for us. And we are okay with that.”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q1 FY26

By the end of the year, Vijayakumar had put numbers to it. A $100-million deal in the old shape, he said, was now closing for closer to $80 million:

“I mean, $100 million deal would be much lesser today - maybe 80 million, just on a rough ballpark. So, deal TCV is flat. But technically, it does require at least 25%, 30% more effort to convert and get to the same number.”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q4 FY26

And, taking a longer view, he sized the structural drag on HCL’s own portfolio:

“If we look at the industry today and categorize it, 40% of the industry runs the risk of being disrupted by AI and can shrink 3% to 5% CAGR for a few years and can eventually be 25% of the enterprise spend... The 3% to 5% deflation that I mentioned in the AI disrupted services, based on the mix of services that we have, it would translate to 2% to 3% for our portfolio.”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q4 FY26

To make room for that pivot, HCL ran a restructuring program through the year. The cost showed up plainly on the Q4 P&L: a reported operating margin of 16.5%, against an underlying 17.7% — a 120-basis-point drag the company chose to absorb in a single year.

Mphasis sits in the accretive camp, and Nitin Rakesh has been the most forceful about why:

“The good news though is that because we have this approach, we are at least not playing the pricing game alone when it comes to winning business. We are playing the Savings-Led Transformation game, and while we win business, we don’t have to sacrifice profitability of those deals because we are using this as a leverage.”

— Nitin Rakesh, CEO, Mphasis | Q2 FY26

TCS, predictably more measured, made the same point on revenue productivity but flagged the timing wrinkle — early-stage AI delivery still carries investment costs that distort margin comparisons:

“On the AI and data part, the revenue productivity is definitely much better than the TCS average or the traditional business, both at onsite and offshore. Margins, I will not call out because there would be investments which would be temporary or in the initial phase, so it wouldn’t be like-to-like for comparison.”

— Samir Seksaria, CFO, TCS | Q4 FY26

The two camps are not actually contradicting each other. They are describing different ends of the same portfolio: new AI work prices well because it’s scarce and unmeasured; old work that has been re-priced under AI savings clauses gives margin back. The companies whose CFOs say “accretive” are the ones whose mix is tilted, today, toward the new end. The ones who say “deflation” are sizing the headwind from the old end. FY27 will tell us which dominates as the new work scales and the old work gets re-papered.

There is, however, a related move that two of the larger players have begun making — and it is the most concrete signal yet that they are willing to defend pricing with the one lever services firms rarely pull: walking away from deals. HCL’s Vijayakumar disclosed in Q4 that the company had voluntarily declined to chase a meaningful share of available pipeline:

“We have lost some deals which are voluntary losses. We have walked away from some deals which will not make sense and that would have easily contributed at least $1 billion more to this number. It’s only prudent to be a little bit more careful about this...”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q4 FY26

A billion dollars of foregone TCV is a serious admission for an industry that has measured its quarterly press releases in TCV growth for two decades. Tech Mahindra has been making the same call, and saying so explicitly:

“We have stayed extremely disciplined in large deals. So we have stayed disciplined so that large deals don’t end up creating a problem for us in the future, or doing large deals that don’t make business sense. Clients don’t want us to do deals that don’t make sense either, right?”

— Mohit Joshi, MD & CEO, Tech Mahindra | Q4 FY26

Tech Mahindra’s CFO Rohit Anand confirmed the same posture, in CFO language:

“We’re extremely conscious on what margins and the risk profile we sign it up... [we’ve] been very selective not just on the two large deals that we’ve announced but even on the deals that we’ve been ramping up for the last 3-4 quarters. Our as-sold margins on each of these deals from a portfolio perspective are accretive.”

— Rohit Anand, CFO, Tech Mahindra | Q4 FY26

Notably absent from this chorus is Infosys, which spoke of “financial discipline” and a “margin protection programme” through FY26 but did not, on its Q4 call, name any deals it had walked away from. The question this raises is not whether HCL and Tech Mahindra are bluffing — both have given up real TCV to make the point — but whether either of them can keep doing it through FY27. Walking away from a billion dollars of deals is a luxury in a year where reported growth was already supported by macro tailwinds and currency. If FY27 starts with softer demand and the competitive set begins matching whatever pricing HCL refused, the discipline will be tested in a way it wasn’t this year. Investors should watch quarterly TCV disclosures with that filter on: is the deflation in deal sizes coming because AI is shrinking the work, or because HCL and Tech Mahindra are choosing not to take it? Through FY26, both stories have been true. They may not stay true together.

[3] The pricing reframe

If the labor-hours model is in retreat, what replaces it? The honest answer from FY26 is: no one knows yet, but everyone is auditioning a candidate. The sharpest reframe came from Tech Mahindra’s CEO Mohit Joshi — the most transparent any large-cap Indian IT services chief executive has been about the pricing logic of the AI era:

“The way of thinking about running an AP function for our client, for instance, in the age of AI, is thinking about the work overall in terms of the number of service tokens that you will deliver to a client. So a service token in the context of an AP could be a sub-process of AP that you need to deliver for a telco. And as the combination of human labour and digital labour changes over time, and as the pricing for digital labour changes over time, the result is very transparent to the client.”

— Mohit Joshi, MD & CEO, Tech Mahindra | Q4 FY26

The “token” framing is borrowed, deliberately, from how large language models are priced. A unit of work, abstracted from how it gets done. Whether the work is performed by a human in Pune, an agent on a GPU, or some blend of both, the client buys outcomes by the token. Notice what it strips away: the offshoring rate-card. There is no longer a “billing rate” in the traditional sense.

Mphasis has taken a different route to the same destination — selling the savings, not the hours:

“Clients are also asking us to not just show me on a PPT or tell me but actually show me in a live sandbox environment in many cases. So, think of this as ‘RFPs are turning into hackathons’, and that’s their yardstick of who can deliver on what they’re promising versus not. So, it has in a way become a lot more about the ability to showcase through execution.”

— Nitin Rakesh, CEO, Mphasis | Q2 FY26

That sentence is worth pausing on. RFPs — request for proposals — have been the industry’s procurement language for thirty years. A multi-hundred-page document goes out, vendors respond with a multi-hundred-page document back, the decision is made on the document. If RFPs are being replaced by live sandbox demonstrations, the selling motion itself has changed: showing the working AI agent in the room beats describing it. The salesforce that wins is the one with engineering in the field, not slideware in the back office.

Infosys, more cautiously, hinted that pricing is in transition without committing to where it lands:

“Over a longer period of time, on the back of AI, etc., we may expect some part of newer pricing models emerging. It could be outcome-based pricing model. It could be pod-based or studio-based pricing model, etc. So there are various new pricing models that are emerging as we speak. I do not think over the next year or so the entire model is going to change.”

— Jayesh Sanghrajka, CFO, Infosys | Q1 FY26

“Pod-based” and “studio-based” are worth unpacking briefly. A pod is a small dedicated team — sometimes ten people, sometimes three plus an agent stack — billed as a unit rather than per head. A studio is a longer-running engagement priced on capacity and outcomes. Both share a common property: the client never sees an hourly rate.

[4] From labour to platforms and IP

If the work is no longer priced by the hour, where is the leverage? Every CEO in the brief converged on roughly the same answer: stop being a pure services firm, start owning intellectual property and platforms that get embedded into client environments. HCL’s Vijayakumar said it most directly:

“We believe this industry will have to evolve from being a pure labor-based service provider to people plus IP and platform-based service provider. When you have the platform as a third-party platform, there is very little leverage, very little stickiness that we can build. And we can really deliver very good quality vertical IP solutions, which can be replicated across customers.”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q2 FY26

The argument is structural. A consultancy that uses someone else’s platform — Microsoft’s, Salesforce’s, ServiceNow’s, OpenAI’s — is renting leverage. A consultancy that builds its own platform, even a narrow vertical one, captures the leverage. HCL’s AI Force is the lead exhibit; by the end of FY26, the company reported $155 million in quarterly Advanced AI revenue, up from a $100-million annual milestone just two quarters earlier.

Wipro, in the most explicit organizational signal of the year, set up a separate business unit around the same thesis:

“As intelligence becomes industrialized and widely accessible, we are making a deliberate strategic pivot to stay ahead. We have launched a dedicated AI-native business and platforms unit to expand beyond a services-only model to a services-as-a-software approach. This unit will operate with dedicated leadership, focused investments and a distinct operating model to accelerate enterprise-grade agentic AI solutions.”

— Srini Pallia, CEO & MD, Wipro | Q4 FY26

“Services-as-a-software” is the phrase to watch. It is the inverse of “software-as-a-service” — instead of subscription software that requires services around it, the service itself is delivered as software that runs continuously. The CEO described the resulting structure as a “dual engine”: traditional services on one side, AI-native platforms on the other.

Tech Mahindra is making the same move under a different label — rebranding entire service lines rather than spinning up a new unit:

“We are repurposing our application development and maintenance services to agentic development and modernization services and it’s not just a name change, right. It’s not just a name change because it’s about how are we driving value to the customers. Now, how are we bringing that experience in the agentic form to our customers for building application agentic development is something that we are very, very focused on...”

— Atul Soneja, COO, Tech Mahindra | Q4 FY26

TCS has stated the destination in the largest terms:

“With AI, TCS aspires to be the World’s largest AI-led Tech Services company. This aspiration is powered by capitalizing on AI-led renewals, vendor consolidation and cost optimization deals resulting in market share gains; using new-age services and adjacencies that enable enterprises to ‘Get ready for AI’; becoming a full-stack AI services player — Infrastructure to Intelligence — thereby delivering maximum ROI to clients on their AI investments; and building new revenue streams such as building AI infrastructure.”

— K. Krithivasan, CEO & MD, TCS | Q4 FY26

The most provocative version of the labour-to-IP thesis, though, came from Infosys’s co-founder and chairman Nandan Nilekani, in a line that pointed at something even further out — the threat to the SaaS industry the IT services companies have spent twenty years serving:

“As AI becomes a bigger part of the spend, the balance of advantage is moving towards ‘build’ rather than ‘buy’. If you see some of the concerns about what will happen to SaaS companies and all that, it is because of this, that building applications has become so simple that very often you may just build, or you may replace something that you have, which you bought, with something to be built.”

— Nandan Nilekani, Chairman, Infosys | Q3 FY26

If Nilekani is right, the next chapter is not just IT services repricing labour. It is enterprise software being eaten by the same wave — and the consultancies that ride it being the ones that own the new application stack, not the old one.

[5] Where the new money is

The closing fracture is the most underappreciated one in the brief. The growth pocket nobody had on their bingo card eighteen months ago has turned out to be the building, running, and feeding of AI itself — what HCL calls “Day-1 services”:

“The real acceleration in what we are seeing is not necessarily in deploying AI within enterprises, but really ‘Day-1’ services, which are foundational for enabling AI, like a lot of work in our engineering services. Like I mentioned about custom silicon for edge inferencing, it is a big area with a lot of companies across multiple industry verticals. This is not restricted to semiconductor industry.”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q3 FY26

Edge inferencing is the act of running AI models on the device — a car, a sensor, a factory floor controller — rather than in a centralized data center. Custom silicon is the chip designed to do that efficiently. Both used to be niche semiconductor specialties. In HCL’s telling, they have become mainstream IT services work because every industrial company building AI capability now needs both. By Q4, HCL had a concrete deal to point to:

“A global technology major selected HCLTech for another AI Factory program worth over $100 million. The HCLTech solution will fast-track the client’s requirements of building and operating next-generation AI data centers to support cutting-edge AI workloads using the latest GPU technologies. A global semiconductor major selected HCLTech’s AI engineering services to support ASIC development across multiple advanced node chips, strengthening its position in Physical AI.”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q4 FY26

TCS made the most striking infrastructure announcement of the year — a move into the data centre business itself:

“The promise we see in our HyperVault Business — which has made significant progress this quarter on its journey to build out 1 GW of capacity. This includes winning customer commitments, land parcel finalizations and partnering agreements.”

— K. Krithivasan, CEO & MD, TCS | Q4 FY26

One gigawatt of data centre capacity, for context, is a serious number. Indian data center capacity in early 2026 totals roughly 1.4 GW. TCS, a services company, has signalled it intends to build something approaching the size of the entire existing Indian colocation industry — and intends to sell that capacity to AI workloads.

The more interesting move underneath HyperVault, though, is what TCS is positioning around it. The company’s stated aspiration of being a “full-stack AI services player — Infrastructure to Intelligence” reads, at first, like investor-day boilerplate. Read against the HyperVault build-out, it describes a sequence no other Indian IT firm is set up to deliver: an AI lab buys compute from TCS, then turns to TCS again for GPU operations, then cloud management, then the model fine-tuning and red-teaming work that Wipro and Tech Mahindra are doing on a project basis. Each successive layer is more services-intensive and more margin-rich than the layer below it. Owning the data centre at the bottom of the stack gives TCS a contractual reason to be in the conversation at every layer above it. That is structurally accretive in a way that running someone else’s GPUs cannot be — and, for now, no other Indian IT firm has the balance sheet, the customer book, and the infrastructure muscle to attempt the full stack the way TCS is.

Wipro and Tech Mahindra are working a different end of the same picture — getting paid to run models for the model-makers themselves:

“In my first example, a leading global technology company has engaged Wipro to help run and improve its frontier AI models. Wipro will manage the end-to-end operation of these AI models from training, governance and evaluation to domain-specific validation. In fact, this engagement will be done through a specialized global delivery platform.”

— Srini Pallia, CEO & MD, Wipro | Q4 FY26

“When we look at our BPS services, for example, close to one-tenth of our current business in BPS is working with technology players, high-tech players creating their AI models, continuously fine-tuning them and managing them over a period of time. And more and more work we are doing on the BPS side is actually towards that now.”

— Atul Soneja, COO, Tech Mahindra | Q4 FY26

For a layer of context here: BPS — Business Process Services — used to be the lowest-margin, most automatable layer of the Indian IT stack. The fact that one-tenth of Tech Mahindra’s BPS book is now training and fine-tuning AI models for hyperscalers is a sentence that would have read as nonsense three years ago. The hyperscalers, ironically, have become the buyers that need the most armies of humans — to label data, evaluate outputs, and red-team models.

HCL has put the cleanest framework on what comes next — a three-bucket split of its own portfolio:

“We will see differential growth rates in all the three different categories: AI disrupted, AI amplified, and AI native or Advanced AI services. We really look forward to growing our AI-native services in the 25% to 30% range. And that will truly be the validation of how we are evolving as a company.”

— C. Vijayakumar, CEO & MD, HCL Technologies | Q4 FY26

Three buckets, three growth rates. AI disrupted shrinks. AI amplified grows in line with the business. AI native compounds at 25-30%. The whole question for FY27 is whether the third bucket gets big enough, fast enough, to outrun the drag from the first.

What it adds up to

The Indian IT services industry spent FY26 publicly admitting that its core business is changing shape — and putting numbers, language, and organisational decisions behind that admission. The vocabulary fractured into deflation and accretion, tokens and pods, disrupted and native — but the underlying movement was a single one. The companies that used to sell engineering hours by the thousand are now trying to sell platforms by the licence, savings by the contract, and infrastructure by the gigawatt. None of them have completed the pivot. All of them have started it. FY27 is the year the math becomes legible — when the new revenue mix is large enough to either validate the bet or expose the gap.

What to Watch

• HCL’s 2-3% portfolio deflation forecast: Vijayakumar has put a number on the headwind; track whether reported FY27 revenue mix bears it out, and whether AI-native services growth (targeted at 25-30%) is fast enough to offset it.

• TCS HyperVault execution: 1 GW is a number large enough to reshape the Indian data centre market. Watch for customer commitments converting to live capacity, partnership announcements with GPU vendors, and the first revenue disclosure.

• Tech Mahindra’s “service tokens” pricing: No client has yet been named signing a token-priced contract. The first concrete deal at a stated token rate will be the signal that the reframe is real.

• Infosys headcount additions in FY27: Salil Parekh hedged “we will continue to add” — but the gap between Infosys’s hiring posture and HCL/Mphasis’s de-linkage narrows with every quarter of productivity gain. Watch the net adds in Q1 and Q2.

• Whether any IP/platform business crosses 10% of group revenue: HCL’s Advanced AI sits around 4-5% on a $13-billion run-rate; Infosys put its AI Hexagon at 5.5% in Q3. The 10% line is when “we’re pivoting” becomes “we have pivoted.”

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

The idea is simple: we pick the most interesting, sometimes spicy, comments from business leaders and fund managers and break down what’s really going on behind them.

We’ve been tweaking the format a bit, and this is a new version that I’m trying right now. Our previous episode was the first time we experimented with this, and people seemed to have liked it, so we’ll continue..

This builds on a newsletter we run called The Chatter, where we track what company managements say in their earnings calls or TV interviews. But instead of looking at those comments in isolation, this format zooms out. We track one theme across companies and over time to see how the story evolves. The goal is to connect the dots and understand the deeper shifts shaping an industry.

For today’s episode, I wanted to understand what’s up with E-waste in India and how have the rules surrounding it evolved. Here’s what we found.

This was originally published on 26th April, 2026 under The Chatter by Zerodha.

Every year, India throws away the equivalent of a small mountain of electronics. In FY 2024–25, the Central Pollution Control Board counted 14 lakh tonnes of e-waste — discarded phones, televisions, refrigerators, cables, batteries — generated across the country. That’s up from 7 lakh tonnes in 2017–18. Almost double in seven years. By total volume, India is now the third-largest e-waste generator in the world, behind only China and the United States.

The rules to handle this have technically existed since 2012. But having rules and enforcing them have been, for most of that time, entirely different propositions. The story of India’s e-waste framework is the story of a government that kept tightening its grip — through mandatory targets, customs enforcement, a digital certificate trading platform, and now a minimum floor price contested in court — until the industry, reluctantly, started moving. What follows is that story, told through the people building and paying for the new system.

The Paper Tiger (2012–2016)

The first iteration of India’s e-waste rules arrived in 2012 and operated on an optimistic principle: that producers of electronics would voluntarily take responsibility for what happened to their products at end of life.

It did not work.

“In India, the EPR policy first came into picture in 2012 where it was voluntarily compliance. Not taken very seriously, right, so no producer essentially wanted to do anything.”

— Nitin Gupta, Founder and CEO, Attero Recycling | Aug 2025

Extended Producer Responsibility is the idea that the company which profits from selling a product should bear responsibility for what happens to it when the consumer is done. As a principle, it’s intuitive. As a regulatory mechanism, it requires enforcement. The 2012 rules provided the principle without the enforcement.

Four years later, the government upgraded from voluntary to mandatory.

“And then the government actually made it mandatory compliance in 2016. Still nobody did anything...”

— Nitin Gupta, Founder and CEO, Attero Recycling | Aug 2025

The 2016 rules set escalating collection targets — 10% of historical sales in the first year, stepping up to 70% by 2022–23. But the rules had a structural gap: verification. Producers self-reported their compliance. Penalties were modest. The informal sector — hundreds of thousands of waste pickers, kabadiwalas, and crude dismantling units operating across Indian cities — continued to handle the overwhelming majority of e-waste, exactly as it always had.

Enforcement finds its teeth (2016–2022)

The breakthrough came not from environmental enforcement but from trade logistics.

“...then the checking of EPR compliance was handed over to customs. So as an OEM, you could not import a product or a subassembly or a component without EPR compliance. That’s when a lot of OEMs started taking it seriously...”

— Nitin Gupta, Founder and CEO, Attero Recycling | Aug 2025

India’s electronics industry runs on imported components. A company that couldn’t clear customs couldn’t sell. Tying EPR compliance to import clearance was the enforcement lever that actually worked — because the cost of non-compliance was no longer an abstract penalty. It was a supply chain shutdown.

The rules continued tightening. In 2019, the scope of products covered under EPR expanded from 21 categories to 30. The E-Waste Management Rules 2022 — notified in November 2022, effective April 1, 2023 — expanded that further to over 100 product categories, adding tablets, GPS devices, modems, air purifiers, medical equipment, solar PV panels and modules, and laboratory instruments, among others.

“So back in 2019 it was 30 products covered under EPR, then they made it 100 products. Then they introduced a platform for trading of these EPR certificates as well and now they’ve introduced the minimum floor pricing right which is in contention in the courts because some OEMs have opposed it.”

— Nitin Gupta, Founder and CEO, Attero Recycling | Aug 2025

The trajectory is worth pausing on: from 21 to 30 to 100+ products; from voluntary to mandatory to customs-enforced; from self-reporting to a digital certificate trading market with a government-set price floor. Each step was a tightening of the system.

What the rules actually ask for?

Before getting into how industry has responded, it’s worth explaining the mechanics — because EPR certificates are doing a lot of work in this story and the con cept is less intuitive than it sounds.

An EPR certificate is proof that a certain quantity of e-waste has been collected and scientifically processed by an authorized recycler. Under the E-Waste Management Rules 2022, every producer — any company that manufactures or imports electronics — must meet annual recycling targets based on what it sold in prior years. The current target is 60% of historical sales volume, stepping up to 70% by FY 2025–26 and 80% by FY 2027–28.

A producer can meet its target one of two ways: run its own collection and recycling program, or buy EPR certificates from authorized recyclers who have already done the processing. This certificate-trading mechanism transforms a compliance obligation into a market. CPCB launched its EPR Electronic Trading and Settlement Platform (EPRETP) to host it. Certificates are denominated in kilograms, priced by category, and valid for two years.

As of 2024–25, EPR certificate prices for e-waste range from roughly ₹10 to ₹50 per kilogram, with a minimum floor price of ₹22/kg established by government amendment in March 2024.

The floor price is the live controversy in the system. The logic: without a price floor, certificates could collapse toward zero, making it economically irrational for formal recyclers to invest in infrastructure. The counterargument, advanced by several of the country’s largest electronics brands, is that the floor distorts the market and constitutes an unfair financial burden.

In April 2025, a group of major OEMs — Daikin, Samsung, LG, Havells, and Voltas among them — challenged the floor pricing mechanism in the Delhi High Court. The case is ongoing. It captures the central tension in the system: the government is trying to make formal recycling economically viable; the brands whose products create the waste are pushing back on what that costs them.

What compliance looks like in practice?

For appliance and consumer electronics brands, the shift from theoretical compliance to actual accounting showed up in earnings calls from around FY 2023–24 — roughly when CPCB moved from accepting broad industry data to issuing specific, company-level targets.

“You would be aware that the government had brought out the E-waste regulation particularly consumer durables a few years back but at that time there was no clarity on how that should be treated. Now the government / CPCB has come out with a lot of clarity and we all the industry players have submitted all sort of data so they have come out with the specific number which every company needs to comply with.”

— Rajiv Goel, Executive Director, Havells India | Q3 FY24

This is the moment the rules became real for companies. Until CPCB issued company-specific numbers, the obligation existed but the liability was undefined. Once defined, it had to be provisioned. And the quantum is not small — nor is it stable.

Bajaj Electricals, whose lighting and appliance business generates significant e-waste obligation under the Consumer Electrical and Electronics Equipment category, gave a clean accounting of where it stands.

“So EPR for this year is about Rs. 9.5 crores and last year also was similar. Going forward next year it will be a charge of about Rs. 18 crores.”

— E.C. Prasad, CFO, Bajaj Electricals | Q4 FY25

The cost is essentially doubling. And this isn’t an anomaly — it’s arithmetic. EPR targets step up every two years; the compliance cost scales accordingly. For Havells, a company with broader product categories and a larger sales base, EPR has graduated into a named line item alongside commodity inflation and energy efficiency regulation changes.

“I think it is consumed, and we are hoping for better contribution margins. It’s not comparable as against last year because EPR liabilities have also increased in the current year, which obviously when the season comes, when we are in a position to pass on that price to the market, we’ll have to do that. But right now, because of low season, we’ve also been restrained to do so.”

— Anil Rai Gupta, Chairman and MD, Havells India | Q2 FY26

“We remain optimistic about a gradual recovery in demand. However, we are cognizant of the industry headwinds such as cost increases on account of commodity inflation, BEE changes, e-waste, etc. We are in the process of taking calibrated price hikes and enhancing operational efficiency.”

— Anil Rai Gupta, Chairman and MD, Havells India | Q3 FY26

The ‘BEE changes’ reference is to Bureau of Energy Efficiency star rating upgrades for appliances, which impose their own compliance costs. Companies are now managing a convergence of regulatory cost pressures — and e-waste is clearly one of them. The costs will eventually move into product prices. The question is timing.

Amara Raja’s ₹700 Crore bet

The producer-side response of absorbing costs, buying EPR certificates and timing price increases is one approach to the new compliance reality. Amara Raja Energy & Mobility chose a different one.

Battery companies operate under a parallel framework to the E-Waste Rules: the Battery Waste Management Rules 2022, notified in August 2022 and effective from February 2023. The structure mirrors the e-waste system — EPR registration, collection targets, certificate trading — but the targets are steeper. For automotive and EV batteries, the requirement is 90% collection and recycling by FY 2026–27, with minimum recycled content mandates in new batteries from FY 2027–28 onwards.

Amara Raja is one of India’s two largest lead-acid battery manufacturers. For a company whose entire business runs on lead — buying it, forming it into batteries, selling those batteries, and now bearing regulatory responsibility for what happens when they die — in-house recycling isn’t just a compliance option. It’s a raw material strategy.

“The refining operations... are expected to commence the commercial production in the month of September or October. And then we’ll be using that lead coming from that factory for our purposes. And the battery breaking operations are expected to commence the production, maybe about four, five months after the refining operations are completed.”

— Y. Delli Babu, CFO, Amara Raja Energy & Mobility | Q1 FY25

Lead-acid battery recycling runs in two stages: battery breaking — shredding spent batteries to separate lead plates, plastic casings, and acid — and lead refining, which smelts the recovered material into usable metal. Amara Raja’s facility, called ARCSPL, was being built in phases. Refining came first; battery breaking would follow. This sequencing matters because a refinery without its own breaking operation still buys pre-processed lead from the market. Only once both stages are integrated does the company control the full loop.

The investment rationale is direct.

“...once we do the entire 1,50,000 tons, that is what is being envisaged in that particular location, that should give us close to about 30% of our overall requirement from our own internal recycling sources. So that’s the objective of that investment.”

— Y. Delli Babu, CFO, Amara Raja Energy & Mobility | Q1 FY25

Phase 1 is approximately ₹500 crore; the board approved an additional ₹200 crore for Phase 2. At full scale, the plant would supply 30% of Amara Raja’s lead from recycled material — insulating that portion of its raw material from London Metal Exchange (LME) price volatility and from the margins lost when sending scrap to outside smelters.

By Q3 FY25, the refining phase had commenced commercial operations.

“Then on the recycling plant in this quarter, we have commenced the 1st Phase of commercial operations of the refining capacity of 50,000 tons, we have started the commercial operations in the current quarter, and it will continue to operate at a full capacity for the next quarter as well. And then we are expecting to start the smelting operations, that is the battery breaking operations, sometime towards the end of Q1 of next financial year.”

— Y. Delli Babu, CFO, Amara Raja Energy & Mobility | Q3 FY25

Then came the harder math.

“Amara Raja is running a battery collection program over a period of years. And today, we are able to collect 75% to 80% of the batteries sold. As per the BWMR regulations from the current financial year, we have an obligation to collect 90% of the batteries sold 3 years ago. This percentage was 70% till last financial year, and we met our obligations.”

— Swajitha Rapeti, Head, Corporate Finance, Amara Raja Energy & Mobility | Q2 FY26

A 10–15% shortfall in collection means buying EPR certificates. And buying EPR certificates means provisioning for a cost that is uncertain today and likely to rise.

“We are also buying our EPR credit wherever there is a shortfall, and we anticipate the demand for this EPR certificate may go up in future. Hence, on a prudent basis, we have provided for EPR credit cost in our books based on our revised estimates on scrap collection program. Going forward, as we ramp up our collections, provided the scrap recycling price is competitive with the LME rates, these provisions may come down or even may not be required.”

— Swajitha Rapeti, Head, Corporate Finance, Amara Raja Energy & Mobility | Q2 FY26

The clause ‘provided the scrap recycling price is competitive with the LME rates’ contains the whole economics of formal battery recycling in one sentence — and it’s worth unpacking, because it explains why regulation alone can’t fix the supply problem.

The LME — London Metal Exchange — is the global price benchmark for industrial metals. Lead trades on the LME the same way oil trades on Brent: it sets the reference price that every buyer and seller in the world uses to negotiate. When Amara Raja buys lead as a raw material, it pays something close to LME spot. When it processes scrap through ARCSPL, the value of the lead it recovers is also measured against LME. So the economics of recycling are simple: if your cost to process scrap into usable lead is lower than the LME price, recycling makes sense. If it’s higher, you’re better off just buying virgin lead at market price.

The problem is what this does to scrap collection. A formal recycler — paying GST, running compliant acid disposal, employing workers with safety equipment, maintaining regulatory certifications — carries significant overhead. That overhead limits how much it can afford to pay for used batteries at the collection point and still turn a profit after processing. The informal sector carries almost none of these costs. No GST. No compliant disposal. Often no safety equipment at all. Their lighter cost structure means they can offer battery sellers — auto repair shops, retailers, consumers — a higher price for scrap than formal recyclers can while still making money.

This is the outbidding problem. Batteries flow to whoever pays more. In most parts of India, that is still the informal operator. Amara Raja’s collection program falls short because the scrap it needs is being outbid at the source. The gap then has to be closed with EPR certificate purchases — which is a financial cost, not a recycling solution.

Amara Raja’s own recycling plant partially changes the equation. When you process your own scrap, you capture the full value of recovery without paying a middleman. You also have more control over what you pay to collect — because you’re not relying on an external recycler’s pricing. The question, as Delli Babu frames it, is whether the plant’s recovery ratios are good enough to make in-house processing cheaper than buying lead at LME prices. That’s the real test still ahead.

That said, by Q3 FY26, the recycling plant was contributing to margins — modestly, but measurably.

“Our lead recycling plant led to a margin accretion of around 0.6% at EBITDA level during the quarter. At a LAB level, we are able to sustain the operating margins of about 12% despite cost pressures at raw material levels.”

— Swajitha Rapeti, Head, Corporate Finance, Amara Raja Energy & Mobility | Q3 FY26

The battery-breaking phase — when Amara Raja begins shredding its own scrap rather than buying pre-processed material — was expected in Q4 FY26.

“The battery breaking is going to start from Q4. And as Swajitha has articulated earlier, the refining operations are providing that additional margin comfort at this point of time. But we hope after the battery breaking gets into full shape, I think we should see some mitigation of the lead cost that we are currently incurring. But of course, recycling operations are always kind of lower margin business. So we hope with the technology, what we have put in place, our recovery ratios will be better, and then we’ll be able to improve our overall operating margins for the lead acid business.”

— Y. Delli Babu, CFO, Amara Raja Energy & Mobility | Q3 FY26

The last sentence is an important acknowledgment. Recycling is structurally a lower-margin business than manufacturing. The economics work through volume, through recovery ratios — how much usable metal you can extract from each kilogram of scrap — and through the avoided cost of buying virgin material at market prices. This is a margin improvement story, not a margin transformation.

The informal sector refuses to die

For all the progress on the formal side — certificate trading, customs enforcement, company-level EPR targets — the informal sector remains the dominant physical reality.

“...four years ago 99% of e-waste was being recycled in the informal sector and 1% in the formal sector. Today that number is 75% in the informal sector and 25% in the formal sector. That shift is happening...”

— Nitin Gupta, Founder and CEO, Attero Recycling | Nov 2024

Moving from 1% to 25% formal in four years is real progress. It is also a reminder of how far the system has to go. Studies have found that 76% of workers in informal e-waste units suffer from respiratory ailments. An estimated 4,00,000 to 5,00,000 children between ages 10 and 15 are engaged in e-waste handling across India. They work with lead, mercury, cadmium, and arsenic, typically with no protective equipment, in conditions that cause permanent health consequences.

The informal sector persists not because of ignorance of the rules but because of a structural economic advantage. Informal operators pay cash, carry no GST overhead, have no compliance costs, and can outbid formal recyclers for scrap at the street level. The 18% GST on e-waste and metal byproducts falls only on registered businesses — giving the informal sector an embedded cost advantage before any other factor enters the calculation. The minimum EPR floor price of ₹22/kg was partly designed to help formal recyclers compete; informal operators aren’t bound by it.

“Indian policy today is good from a regulation standpoint on EPR angle which is basically shifting the industry from informal to formal, but there has to be a lot of emphasis on capacity creation. There’s a lot of emphasis on sustainability and critical minerals aspect and probably a lot more financing is required for the sector to be able to develop the capacity and the speed that the capacity needs to develop.”

— Nitin Gupta, Founder and CEO, Attero Recycling | Nov 2024

The ask from the formal recycling industry isn’t a change in tax rates. It’s a structural fix to make the market traceable.

“Majority of the supply chain is shifting from informal to formal, there has to be some sort of GST rationalization that is required here and not because of rates. In the metal scrap industry or e-waste or lithium battery, the government should introduce a reverse charge mechanism. We are not saying change the tax rate of stuff but bring in RCM for better transparency in the sector.”

— Nitin Gupta, Founder and CEO, Attero Recycling | Nov 2024

A note on what Reverse Charge Mechanism would actually do here — because it sounds technical but the logic is simple.

Start with how GST normally works. When a kabadiwala sells scrap batteries to a recycler, the kabadiwala is supposed to collect 18% GST from the buyer and remit it to the government. But the kabadiwala is almost certainly not GST-registered. He’s a street-level collector working in cash — there are hundreds of thousands like him across India. Enforcing GST registration at that level is practically impossible. So in practice, the transaction happens with no tax collected, no receipt issued, and no record of it anywhere. The e-waste just moves and disappears into the informal system.

Reverse Charge Mechanism flips who is responsible. Under RCM, the buyer — in this case, the registered formal recycler — pays the GST directly to the government, instead of collecting it from the seller. The kabadiwala doesn’t need to be GST-registered. The formal recycler, who is already in the system, simply includes the purchase in their GST return and pays the tax on it.

Here’s what that changes: every time Attero or Amara Raja buys scrap from any source — registered or not — the purchase becomes a line in their GST filing. CPCB can then see, at any point, exactly how much material is flowing through the formal recycling system, where it’s coming from, and at what volumes. Today, none of that is visible. The informal supply chain leaves no paper trail.

Crucially, the total amount of tax paid doesn’t change — RCM is not a rate hike. The economic cost of the transaction stays the same. What changes is that it becomes legible. This matters enormously for EPR compliance, because right now the government cannot reliably verify whether the recycling claimed on EPR certificates actually happened. RCM would attach a paper trail to every purchase at the point where the informal supply chain meets the formal one — which is the only point where it’s practical to enforce. That the formal sector has been asking for this for years, while the government hasn’t acted, says something about how low administrative traceability sits on the priority list.

The urban mine

India’s e-waste rules were designed primarily as an environmental compliance framework. They’re increasingly becoming something else: a resource security mechanism.

India has 100% import dependency on several critical minerals — lithium, cobalt, nickel — that underpin the EV and electronics supply chains it is simultaneously trying to build. China dominates global rare earth supply and, as of 2025, has imposed export restrictions on several of them. Every tonne of e-waste processed formally is a tonne that partially offsets that dependency.

“The government of India is also considering a significant sort of incentive mechanism for incentivizing the rare earth industry in India to ensure that India becomes self-reliant on the rare earth material supply chain and is not dependent on China. So from that perspective, we will definitely benefit — as the industry will benefit — from the current geopolitical shifts that China has announced.”

— Nitin Gupta, CEO, Attero Recycling | Q1 FY26

A printed circuit board contains roughly 0.02–0.20% gold, 0.01–0.45% silver, and trace amounts of palladium and platinum. The value embedded in India’s 14 million tonnes of annual e-waste is estimated at approximately USD 6 billion — most of which is currently being captured by informal operators through crude acid-bath processes, or exported as raw scrap to processors elsewhere.

The formal sector’s response has been to build digital supply chains capable of competing with informal networks on transparency and traceability, if not always on price.

“Over the last few quarters we have very successfully transitioned our entire supply chain to a complete digital supply chain based on an AI-based pricing engine. Our supply basically is through Metal Mundy which is an online aggregator collection network — completely digital, completely transparent and traceable.”

— Nitin Gupta, CEO, Attero Recycling | Q1 FY26

Attero claims over 30% market share in the EPR certificate market — more than three times its nearest competitor. That concentration reflects how new and fragile the formal sector still is. A market where one player holds 30%+ of volumes is not a mature market. It is one in early formation, which is exactly what makes the next few years consequential.

“Completely aligned — in fact, in the EPR business, we are a leader in the country today. Our market share is more than 30%, the next best player is less than one-third of our size, and there is significant benefit that we are deriving from the government of India’s EPR policy.”

— Nitin Gupta, CEO, Attero Recycling | Q1 FY26

India’s e-waste rules have traveled a long distance since 2012 — from voluntary to mandatory, from self-reported to customs-enforced, from no market to a digital certificate trading platform with a government-set floor price now being challenged in court. The formal sector’s share of physical e-waste processing has moved from 1% to 25% in four years. Companies that once treated EPR as a footnote now provision for it as a growing line item.

But the numbers also show how much distance remains. Three-quarters of India’s e-waste still flows through an informal system that rules alone cannot reach — not because enforcement has failed, but because informality has a structural economic advantage that compliance mandates haven’t closed. The formal sector needs more than rules. It needs GST architecture, financing, and the infrastructure to compete on price.

And the next wave is already in the pipeline. Solar PV panels are now covered under the E-Waste Management Rules 2022, but India’s standalone solar waste management framework remains unfinished. By 2030, projections suggest India will generate approximately 340 kilotonnes of solar panel waste, concentrated in five states. EV batteries — governed by the Battery Waste Management Rules — will test the recycling infrastructure that companies like Amara Raja are just beginning to build. The rules that found their teeth over the last decade will need to grow a new set.

What to watch?

• Delhi High Court ruling on EPR floor pricing:
  If the court strikes down the minimum price mechanism, formal recyclers lose their margin buffer and the economics of certificate trading shift. The outcome will tell you whether the government can hold the line on making formal recycling financially viable.

• Amara Raja’s battery-breaking operations (Q4 FY26):
  Full vertical integration of its ₹700 crore recycling plant comes online this quarter. The recovery ratios it achieves — how much usable lead it can extract per kilogram of scrap — will determine whether formal recycling in India can compete on economics, not just compliance.

• GST Reverse Charge Mechanism for e-waste scrap:
  Attero has been advocating for this for multiple quarters. Any indication from the Finance Ministry or GST Council signals whether the government is willing to address the structural cost disadvantage of the formal sector.

• Solar Waste Management Rules finalization:
  MNRE and CPCB have the roadmap; standalone solar waste rules haven’t arrived. When they do, EPR obligations will land simultaneously on Adani Green, Tata Power Solar, Waaree Energies, and Premier Energies — companies whose panels are already aging in the field.

• India’s National Critical Minerals Mission linkage to e-waste:
  Whether the government ties urban mining incentives to formal e-waste processing will determine if the sector gets the financing it needs to scale fast enough to matter for the rare earth supply chain.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

You (and maybe your portfolio too) are probably tired of hearing this, but the age of AI agents has been a thorn on Indian IT’s side. Its business model is getting stale each passing day. But what we often don’t understand is what this means for the day-to-day nitty-gritties of the industry.

Now, there is a possibility that Indian IT might just adapt to the new paradigm. We’ve briefly covered some semblance of this possibility in past Daily Brief stories. And usually, we can’t say for sure what such a possibility even entails. After all, what does adaptation for Indian IT look like? What are the forms of inertia they will have to overcome to successfully change themselves? And even if they do adapt, will they be able to defend their business from the new threats that AI will inevitably give birth to?

To answer these questions, we had a nuanced conversation with Ameya Pimpalgaonkar, a storied professional with two decades of experience in the global IT industry, working across giants like IBM, Accenture and Infosys. He has donned many hats, from software developer to co-founder and CTO. He’s also a prolific technology investor in both public and private markets, and is very well-known on X for his investing takes. We’ve quoted his work in the past in our coverage of Indian IT on The Daily Brief. I highly recommend reading his newsletter.

Personally, I learnt a lot. Most mainstream conversations about Indian IT are either a black-box, or are too one-size-fits all. Ameya’s answers cut through both of those problems succinctly. The idea was not to stick to a highly doomer or highly optimistic narrative, but weave a story that pinpoints the opportunities, the threats, and understanding the depth of how much AI can commoditize any possible future moat.

Some choice quotes that I think are worth highlighting:

On why the business model shift, while real, is still slow:

“First, the contractual challenges. All service level agreements are still tied to response time, not to a business outcome. If you go to Jira, or any ticketing tool like ServiceNow, you would clearly see the service catalog — the SLAs as per the category of an incident — and the SLA translates into how many hours that incident should be resolved in. When that time elapses, you get the escalation emails automatically. But inherently, this is still tied to response time. It is not tied to the quality of the fix being delivered.

There is no connection to the business outcome defined yet. And what this means is you have to renegotiate a contract that is already a legal document — an already time-tested relationship with the customer. How would you even define a business outcome? That’s the biggest challenge right now.”

And it's slow because, to begin with, defining the new scope of work itself is a challenge:

“In my 20 years of career, what I’ve seen is that whatever scope of work you do in the first year — or even the first six months — with a customer engagement, it always balloons four to five times compared to the initial SOW. So what you end up delivering is, I think, four to five times more than what was defined in the SOW.

Since most enterprise customers can’t define what success and closure looks like in the early phases of SOW definition — that’s what I mean by client readiness. So the transition is definitely real, but it’ll happen deal by deal. It won’t happen as a wave.”

On how data annotation is emerging as a new (if boring) revenue stream for Indian IT, and why it’s well-poised to take advantage of it:

“Any data that’s anonymized and pseudonymized is perfectly suited to be transformed. This is where data labeling richness comes in. A surgeon looks at an X-ray and says, okay, there’s a nodule in the chest — that line by the surgeon is an annotation labeling that data. If you just pass that X-ray to an AI, the AI will say this is this, this is that — but it’s missing the surgeon’s label, the surgeon’s context for interpreting what it means. That data labeling, in my view, is going to be one of the quickest revenue streams opening up for many of these companies.

What it means in practice is: you take models with pre-trained weights, bring in domain experts from legal, geography, STEM, and so on, and have them engage. You must have experienced this — whenever you use ChatGPT and ask a strategic question, sometimes it generates two outputs and asks which one you prefer.

That’s exactly ChatGPT doing reinforcement learning to identify your preferences. Now think about doing that at an LLM level, at an enterprise level. India has the workforce and the muscle to do this. It’s not glamorous, but it’s real revenue.”

On why organizational structure matters more than ever in AI, and why smaller, agile firms might be better poised to win in AI:

“And if you look beyond the top 4, 5, 6 IT companies to smaller companies that have always run leaner structures, they’re much better positioned to take advantage of this.

One reason is they don’t have the overhead. There’s no risk of firing large numbers of people and creating chaos in the market. Instead, they can just increase the productivity of existing employees — and I know for a fact that’s happening. For smaller companies, it’s much easier to adopt AI tools.

These companies are starting to convert their product features into markdown files. Every product backlog, every PI planning exercise — the output is a markdown file today, which feeds directly into whatever coding copilot you’re using. And once you have everything defined in a markdown file, building a feature becomes so much easier.”

On accountability in the age of agents, and where you'll need a human still (with a very interesting personal tale):

“Sales cycles are not technology-driven. They’re driven by intangibles — delivery quality, what happens after delivery. And in the age of AI, the importance of audit trail is something we can’t even begin to describe. I was in a call yesterday with a company where they discussed an incident — they had an army of agents, and three of them didn’t agree with what the rest concluded. The system just froze.

The product had no way to determine a resolution. Who has the final say? Who says “okay, this is it”? Even the orchestrator agent couldn’t come out of that state. And it froze even though the product had a built-in rule — if agents disagree, do this, do that. Agents were self-aware, and it still happened.

We’re still discovering new aspects of working with these tools. You need a human coming in to handle the dirty work. But do you have that human if you’ve already let people go?”

You can listen to the full conversation on YouTube, Spotify, or Apple Podcasts.

And if you like reading more than listening, here’s the transcript of the podcast, shared earlier on Subtext by Zerodha.

Subtext by Zerodha

Ameya P on how Indian IT can flip the AI script

Hi folks, Pranav here…

Read more

2 months ago · 8 likes · Zerodha

Share

The idea is simple: we pick the most interesting, sometimes spicy, comments from business leaders and fund managers and break down what’s really going on behind them.

We’ve been tweaking the format a bit, and this is a new version I’ll be trying over the next couple of weeks. If you like it, let me know in the comments. If you don’t… go easy on me.

This builds on a newsletter we run called , where we track what company managements say in their earnings calls. But instead of looking at those comments in isolation, this format zooms out. We track one theme across companies and over time to see how the story evolves. The goal is to connect the dots and understand the deeper shifts shaping an industry.

For today’s episode, I wanted to answer one question: has the microfinance credit cycle turned? To get there, I went through 144 concalls from 18 Indian companies over 8 quarters. Here’s what we found.

This was originally published on 12th April, 2026 under .

The Microfinance Credit Cycle

From “not expecting any known risks” to “the issues are clearly behind the industry” — eighteen companies, eight quarters, and a credit cycle that rewrote the rules of Indian microfinance, told entirely through the words of the people who lived it.

The Goldilocks Trap

In March 2024, the Indian microfinance industry was in a celebratory mood. Asset quality had recovered from the pandemic. Growth was strong. Guidance was ambitious. The sector had disbursed record volumes and the worst of COVID-era stress was firmly in the rearview.

Satin Creditcare’s HP Singh called FY24 a “momentous chapter” in the company’s 33-year journey. Muthoot Microfin’s CEO saw a clear path to 4.5% ROA. CreditAccess Grameen, the country’s largest pure-play microfinance NBFC, set its sights on numbers that would have seemed outlandish just two years earlier.

“Assuming a stable operating environment, we look forward to achieving loan portfolio growth of 23% to 24% in FY25. We are anticipating credit cost of 2.2% to 2.4%... Overall, we aim to achieve ROA of 5.4% to 5.5% and ROE of 23.0% to 23.5% in FY25.”

— Udaya Kumar Hebbar, Managing Director, CreditAccess Grameen | Q4 FY24

Bandhan Bank, which had spent three painful years cleaning up pandemic-era stress in its microfinance book — the “Emerging Entrepreneurs Business” in Bandhan parlance — declared the legacy behind them. Its SMA-0 pool had halved. Slippages were trending down. The technical write-off of the old book was done.

SMA-0 pool is the set of loan accounts that are overdue by up to 30 days, indicating the earliest stage of repayment stress.

Slippages are loans that move from standard (performing) to NPA status during a period.

“So I think as Rajeev said, we believe that decisively, we are out of the pandemic problem. In fact, all the parameters, as you see, you can clearly see that practically the slippages has come down significantly. Recovery rates have improved. DPD pool has dried down.”

— Ratan Kesh, Executive Director and COO, Bandhan Bank | Q4 FY24

When analysts asked CreditAccess’s managing director what risks he saw on the horizon, his response carried the confidence of a man who had navigated COVID, demonetization, and the Andhra Pradesh crisis — and come out the other side each time.

“We are not expecting any known risks actually. Known risk if you see the last 10 years or 15 years, the risks which have come to this industry are unknown. Whether it is COVID or demonetization, the second COVID, all are unknown. The industry has not lost because of any known risks.”

— Udaya Kumar Hebbar, Managing Director, CreditAccess Grameen | Q4 FY24

It was a reasonable statement grounded in history. Every previous microfinance crisis — Andhra Pradesh in 2010, demonetization in 2016, COVID in 2020 — had been triggered by an external shock that no one could have modelled for. The sector’s risk management frameworks were built around this assumption: the threats come from outside, and when they pass, business returns to normal.

Not everyone agreed. A few voices struck a different note that quarter. Kotak Mahindra Bank’s new CEO Ashok Vaswani looked at the same data and saw a warning sign rather than an all-clear.

“If you look at what is happening in the industry and look at loss rates and stuff like that post the COVID provisions, it has really been a Goldilocks period for the last 2-3 years at least since COVID. Now, all of us know that credits go through cycles. At this point in the cycle, to go out and get aggressive on credit, I don’t think it is a smart thing to do.”

— Ashok Vaswani, MD & CEO, Kotak Mahindra Bank | Q4 FY24

HDFC Bank was quietly building buffers while the sun was shining — creating what it called “countercyclical provisions,” essentially setting money aside during good times for the bad times it expected would eventually come.

“We have, as a part of prudent risk management, created a countercyclical provision, which is a provision in good times.”

— Sashidhar Jagdishan, MD & CEO, HDFC Bank | Q4 FY24

IndusInd Bank, one of the largest MFI lenders through its subsidiary BFIL, was even more specific about where it saw cracks. Its CEO Sumant Kathpalia said the bank had spotted stress building in Punjab eleven months before the rest of the industry acknowledged it.

“We saw Punjab coming 11 months ago. We saw the waves of Punjab coming and even some parts of Odisha and Bihar also coming and some parts of UP also. We started exiting those portfolios at that point.”

— Sumant Kathpalia, MD & CEO, IndusInd Bank | Q4 FY24

Whether the early detection translated into early protection would become one of the more revealing questions of the cycle.

• • •

Something Breaks

The first tremors arrived in Q1 FY25 — the April-June quarter of 2024. A brutal heatwave across northern India made fieldwork difficult for loan officers accustomed to collecting repayments door-to-door. General elections in April and May disrupted collection schedules — in several states, political activists ran “Karj Mukti Andolans,” loan forgiveness campaigns that encouraged borrowers to stop repaying.

“Majority of that activity has stopped. It was almost linked with the election activity that was going on during that period in Q1. So now we don’t have too much of activism.”

— Sadaf Sayeed, CEO, Muthoot Microfin | Q1 FY25

These were plausible, temporary explanations. Heat waves pass. Elections end. But underneath the seasonal noise, something structural was breaking. At IDFC First Bank, the JLG collection efficiency — the percentage of current-bucket borrowers who pay on time — slipped from 99.7% to 99.2%. Half a percentage point sounds trivial. In microfinance, where the entire model depends on near-perfect repayment discipline, it was a red alert.

“Normally, pre the floods, our credit loss in the JLG book used to be around 1.6%. Now our estimate of credit cost for this year on the joint liability group, because we are 60% concentrated in Tamil Nadu, it is expected around 5% or a little short of it.”

— V. Vaidyanathan, MD & CEO, IDFC First Bank | Q1 FY25

A brief detour is useful here to understand what was actually going wrong. Microfinance in India works primarily through the Joint Liability Group model — small groups of borrowers, typically 5-10 women, who collectively guarantee each other’s loans. The model’s genius is that it replaces traditional collateral with social pressure: if one borrower defaults, the group faces consequences. It works remarkably well in normal times, keeping credit costs at 1.5-2.5% despite lending to some of the most economically vulnerable households in the country.

But the model has a critical weakness. It doesn’t control how many groups — or how many lenders — a single borrower joins. And since 2022, something had quietly shifted. Credit bureau data showed that borrowers were taking on loans from more and more institutions simultaneously. Ticket sizes were creeping up. Total indebtedness per borrower was rising fast.

Fusion Finance, a mid-sized NBFC-MFI, laid this out in stark numbers.

“I would like to draw your attention to Slide #8 on our customer leverage. As you can see, our outstanding per customer is mostly below INR 40,000. But 33% of these customers have outstanding greater than 1 lakh across micro finance loans. This increased from 23% in March to 32% March ’24.”

— Devesh Sachdev, MD & CEO, Fusion Finance | Q1 FY25

In twelve months, the share of borrowers carrying more than Rs 1 lakh in microfinance debt had jumped by 10 percentage points. And these were just the loans visible on credit bureaus. Equitas SFB’s veteran founder P.N. Vasudevan — who had been in microfinance for nearly two decades — identified another layer that was harder to track.

“What is not really understood or clear is also there’s some kind of a parallel lending happening to the same set of borrowers by so called app loan companies.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q3 FY25

Spandana Sphoorty’s data quantified the damage from overleveraging: borrowers with five or more lenders were just 12% of the company’s AUM but contributed 21% of its arrear bucket. The more lenders a borrower had, the more likely they were to default — and the industry had spent two years adding lenders to each borrower’s profile.

The most important observation of Q1 came from Vasudevan, who named what nobody else was willing to say clearly: for the first time in living memory, this crisis wasn’t caused by something that happened to microfinance. It was caused by microfinance itself.

“For a change, the crisis or it’s not the word crisis, but at least a stress levels in microfinance for a change is not induced by an external one. This is the first time, I think, in the last 15 years that I know or 17 years that I have been in this business that I’m saying that it’s not an external event trigger, but it’s something inherently coming up from the sector itself.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q1 FY25

Not everyone agreed with this framing. Spandana’s MD Shalabh Saxena pushed back on the idea that overleveraging was systemic.

“I personally do not believe there is a systemic over-leverage issue across the industry. I firmly believe that. There could be a few pockets, a few institutions here and there.”

— Shalabh Saxena, MD & CEO, Spandana Sphoorty | Q1 FY25

Within two quarters, the data would prove him wrong.

• • •

The Guardrails

In July 2024, MFIN — the Microfinance Institutions Network, which functions as the sector’s self-regulatory organization — rolled out what became known as the Guardrails. These were a set of industry-wide lending norms, agreed upon by members representing roughly 87% of the microfinance market, aimed at curbing the overleveraging that was now clearly driving the stress.

The key rules: no borrower should have loans from more than four microfinance lenders. Total unsecured indebtedness, including microfinance, should not exceed Rs 2 lakh. Lenders were required to check bureau data before every disbursement, not just at onboarding.

The guardrails were a significant step, but they were forward-looking — they could prevent new overleveraged loans from being created, but they couldn’t fix the existing book. Borrowers who already had five or six lenders would continue to struggle. The stress already embedded in portfolios would need to be absorbed through provisions, write-offs, and time.

The impact on new business was immediate. At Muthoot Microfin, about 11% of loan applications that would previously have been approved were now being rejected under the new rules. Satin Creditcare reported that at the time the guardrails kicked in, just 1% of their existing clients exceeded the four-lender cap — a sign that some companies had been more disciplined than others.

Fusion stopped disbursements entirely in 104 of its branches. CreditAccess tightened credit filters. Spandana paused new member acquisition in 230 branches and stopped sourcing new-to-credit customers altogether. The industry was simultaneously trying to fix the new pipeline while absorbing damage from the old one.

The CreditAccess team tried to quantify the timeline.

“We however, expect the delinquency trend to stabilise in the coming quarter and credit cost within the guided range of 2.2% to 2.4% for the year.”

— Udaya Kumar Hebbar, Managing Director, CreditAccess Grameen | Q1 FY25

That guidance would not survive the next ninety days.

• • •

Into the Storm

By September 2024, the scale of the problem was undeniable. What had started as “stress in some pockets” was now a full-blown industry credit cycle. The heat waves and elections were months behind, but collection efficiencies were still deteriorating. Equitas’s Vasudevan captured the moment of realization.

“The industry and us included were quite caught by surprise when the collection efficiency dipped further in Q2 for no apparent external reasons. In our case, it dropped from 98.9% of Q1 down to 98.2% in Q2. And when this happened across geographies, it was clear that an important reason for the stress was as much internal as external.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q3 FY25

CreditAccess Grameen, which six months earlier had guided for 23-24% portfolio growth and 5.4% ROA, rewrote its entire outlook.

“In light of the current industry landscape and short-term challenges encountered, we have revised our estimate for FY25 annual performance guidance. We anticipate loan portfolio growth of 8-12%, NIM of 12.8-13.0%, credit cost of 4.5-5.0%, ROA of 3.0-3.5%, and ROE of 12.0-14.0%.”

— Udaya Kumar Hebbar, Managing Director, CreditAccess Grameen | Q2 FY25

Growth guidance halved. ROA guidance cut by nearly 40%. Credit cost guidance doubled. And CreditAccess was among the better-positioned players — it had been conservative relative to its peers, with a diversified geographic footprint and a long track record. For the more concentrated and leveraged players, the numbers were far worse.

Fusion Finance reported a quarterly credit cost of Rs 693 crore — a number large enough to trigger covenant breaches with lenders. This is worth pausing on: Fusion’s borrowing costs and credit lines depended on maintaining certain financial ratios. When its losses blew through those thresholds, its own access to capital came under threat. The stress wasn’t just flowing from MFI borrowers to MFI companies — it was threatening to flow from MFI companies to their lenders, creating a potential second-order shock.

“I would like to mention that post the review of our auditors, our actual credit cost for Q2 FY25 has been determined at Rs. 693 crore. The elevated provisioning has been due to us proactively taking accelerated provisions... Looking at our portfolio performance and challenges faced by the sector, our auditor wants to bring to the attention that there are covenant breaches that will require waivers.”

— Devesh Sachdev, MD & CEO, Fusion Finance | Q2 FY25

IDFC First Bank’s V. Vaidyanathan, who had spent the previous quarter framing the Tamil Nadu floods as a one-off episode, acknowledged the broader reality and took an aggressive provisioning stance.

“Now, what we have done this quarter is because microfinance is an issue and we cannot wish it away because the microfinance portfolio is disturbing in many parts of India. So, what we have done is that in SMA-1 and SMA-2, we have taken provisions and therefore... almost 99% of SMA-1 plus SMA-2 has been provided for by the Bank, fully.”

— V. Vaidyanathan, MD & CEO, IDFC First Bank | Q2 FY25

Ujjivan Small Finance Bank, one of the larger microfinance-focused SFBs with a significant presence in southern states, saw its PAR — the portfolio at risk, measuring any loan that’s even one day overdue — climb sharply.

“On asset quality, as mentioned earlier, we are observing stress in the microfinance segment, due to which our PAR has increased to 5.1% in September 24 vs 4.2% in June 24. PAR 0 for our group loan portfolio has increased to 5.5% in September 24 versus 4.1% in June 24.”

— Sanjeev Nautiyal, MD & CEO, Ujjivan SFB | Q2 FY25

IndusInd Bank’s CEO Sumant Kathpalia — who had claimed to spot the trouble eleven months early — was now projecting confidence that the worst would be over within weeks.

“In my opinion, if everything goes well, within two months we should start seeing the flow rates coming down in this business... I am very bullish on the microfinance segment, and I think you will see the stability coming in very soon.”

— Sumant Kathpalia, MD & CEO, IndusInd Bank | Q2 FY25

That timeline would prove wildly optimistic. Kathpalia himself would not see it through — within months, he would exit the bank amid a governance crisis triggered in the bank’s derivative portfolio.

Vasudevan, characteristically more measured, offered the hardest truth of the quarter.

“This is the first time I’m seeing where an extended period of nearly 9, 10 months, we have been seeing high slippages and it’s by all of us and in many markets. So it’s very difficult to tell you within what time frame we expect it to come back to normalcy.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q1 FY25

• • •

The Darkest Quarter — and the Turn

Q3 FY25 — October through December 2024 — was when the stress peaked for most companies. Bandhan saw microfinance slippages of Rs 1,196 crore in a single quarter. Its collection efficiency outside West Bengal dropped to 96.3%. Satin Creditcare’s PAR-1 rose to 6.8%. RBL Bank described conditions on the ground as being “in a flux.”

But this quarter also contained the single most important inflection point of the entire cycle: December 2024.

For months, the data had been uniformly grim. Then, almost simultaneously, company after company reported the same thing — the first material improvement in the current bucket, the pool of borrowers who are not yet overdue. When this pool stops deteriorating, it means fewer borrowers are falling into trouble for the first time. It doesn’t fix the existing NPAs, but it signals that the pipeline of future stress is narrowing.

CreditAccess’s MD, who had initially expected the peak in September, acknowledged the delay but confirmed the reversal.

“Our initial assessment of the current delinquency cycle being transitory in nature has come true as we see the new delinquency addition rate slowing down across various geographies in mid-November 2024... The new delinquency trend reversal was materially visible across various markets beginning mid-November, getting further stronger in December and January.”

— Udaya Kumar Hebbar, Managing Director, CreditAccess Grameen | Q3 FY25

RBL Bank saw the same inflection.

“In the JLG segment, the situation on the ground has been in a flux for most of the past months, but December has seen the first material uptick in collection efficiency and the recoveries of old NPAs.”

— R. Subramaniakumar, MD & CEO, RBL Bank | Q3 FY25

Fusion’s Devesh Sachdev, whose company had been among the hardest hit, was watching a different metric — the new loans disbursed under tighter guardrails since August.

“Even I can tell you one more data, that the new sourcing which we are doing, which we changed in the middle of August. And now when I look it’s a five-month MOB, though it is too early, it’s just a 5 months MOB, but I believe the numbers are infant or early delinquency is clearly showing trends which we have seen in year 2022, ’23.”

— Devesh Sachdev, MD & CEO, Fusion Finance | Q3 FY25

This was quietly one of the most significant data points of the cycle. Loans written under the new guardrails weren’t just marginally better — they were performing like pre-crisis vintage. The guardrails were working.

But Bandhan’s new MD Partha Pratim Sengupta — who had replaced Ratan Kesh in yet another leadership change during the crisis — offered a note of caution that tempered the optimism. The SMA-0 book was improving, but the backlog of loans already deeper in arrears still needed to work through the system.

“So, maybe the level of slippage could not be 1,196 in EEB book, what we have witnessed in Q3, but it will be substantial... But slippages, we are seeing the trend, but at the same time as you are witnessing, you see that our SMA-0 book is improving... But the trend is reversing.”

— Partha Pratim Sengupta, MD & CEO, Bandhan Bank | Q3 FY25

The green shoots were real. But the cleanup was far from over.

• • •

The Reckoning

Q4 FY25 — January through March 2025 — became the quarter of reckoning. The turn in collection efficiencies was confirmed, but the damage from the preceding three quarters had to be absorbed. Companies took massive write-offs to clear the decks. The numbers were staggering.

Equitas SFB revealed the full toll. Its microfinance credit cost went from 2.3% in FY24 to 11.37% in FY25 — a nearly fivefold increase that wiped away Rs 630 crore of profit and dragged the bank’s overall ROA down to 0.32%.

“Last year turned out to be a tough year. We had credit cost in Microfinance portfolio moving up from 2.3% in FY ’24 to 11.37% in FY ’25, wiping away about INR 630 crores of profit of the bank.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q4 FY25

“In response to these headwinds, the bank slowed down its fresh disbursements in Microfinance, leading to a drop in the MFI advances from INR 6,265 crores in March ’24 to just around INR 4,500 crores in March ’25.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q4 FY25

Spandana Sphoorty posted a net loss of Rs 1,035 crore for the full year. CreditAccess wrote off Rs 1,124 crore, including Rs 479 crore in accelerated write-offs of non-paying 180+ DPD accounts. Muthoot Microfin saw its GNPA climb to 4.84% — but its CEO called this the peak.

“So our GNP levels are at 4.84%, which is at an elevated number. But this is the peak of these GNPAs. We feel that in coming quarters, these numbers will only come down because we are seeing better collection efficiency.”

— Sadaf Sayeed, CEO, Muthoot Microfin | Q4 FY25

RBL Bank took perhaps the most aggressive provisioning step of the entire cycle — writing its JLG net NPA to zero in a single quarter.

“In the JLG business, we normally take 25% provisioning each quarter on NPAs, but we have now taken 100% provisioning on the NPA as at March 31, 2025. This means we have a nil net NPA in the JLG business. We have also taken 75% provisioning amounting to INR 283 crores on SMA-0, 1 and 2.”

— R. Subramaniakumar, MD & CEO, RBL Bank | Q4 FY25

And then there was IndusInd Bank. In an industry already reeling from credit stress, IndusInd disclosed something that went beyond a credit cycle — it was a governance failure. An internal review found that microfinance loans at its subsidiary BFIL had been misclassified, concealing Rs 1,885 crore in under-provisioning and unrecognized NPAs.

“The review has also identified the misclassification of certain microfinance loans has resulted in under-provisioning and non-recognition of NPAs aggregating to Rs 1,885 Crores. The Bank has addressed the underlying cause and is in the process of taking actions for staff accountability.”

— Sunil Mehta, Chairman, IndusInd Bank | Q4 FY25

The disclosure triggered a leadership overhaul. Rajiv Anand replaced Kathpalia as MD & CEO. The microfinance subsidiary’s governance framework was restructured for “greater transparency.” It was a stark reminder that credit cycles don’t just stress balance sheets — they expose operational weaknesses that good times had papered over.

Meanwhile, the industry’s aggregate footprint was shrinking. This is a number that didn’t get much attention in individual earnings calls but told a dramatic story at the sector level.

“The overall loan outstanding at the sector, which was about INR 4.4 lakh crores in March ’24, came down to about INR 3.75 lakhs in March ’25 and further down to INR 3.5 lakh crores in June ’25.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q2 FY26

From Rs 4.4 lakh crore to Rs 3.5 lakh crore in fifteen months — the entire Indian microfinance industry shrank by 20%. That’s roughly Rs 90,000 crore of credit that was withdrawn from some of the poorest households in the country. The human cost of an industry correcting its own excesses.

IDFC First Bank’s Vaidyanathan, whose bank had taken provision hits early and heavily, offered a framing that captured what this quarter was — not the start of a new problem, but the crest of a wave.

“You think of this 2025 as a year that this has happened because of microfinance. You think of it that 2026, we will stage a smart recovery. And FY 27, FY 28, FY 29, we should be back winning ways.”

— V. Vaidyanathan, MD & CEO, IDFC First Bank | Q4 FY25

• • •

The Proof

The first sign that the recovery was real — not just hoped for — came from CreditAccess Grameen’s Q1 FY26 results. The company that had slashed guidance a year earlier now reported its highest-ever first-quarter disbursement.

“Our Q1 FY’26 performance has created a new benchmark, achieving the highest ever 1st Quarter disbursement in our history. This is a testament to our resilience and agility that define us given that we are coming off the back of a challenging credit cycle.”

— Ganesh Narayanan, CEO and MD Designate, CreditAccess Grameen | Q1 FY26

Note the leadership change here too — Ganesh Narayanan had taken over from founding MD Udaya Kumar Hebbar, whose guidance had shaped the company through both the confidence of Q4 FY24 and the crisis that followed. CreditAccess wasn’t alone. Bandhan had moved from Ghosh to Kesh to Sengupta in the span of two years. Spandana’s MD Shalabh Saxena — who had pushed back on the overleveraging thesis — had exited, replaced by Ashish Damani as interim CEO. Fusion brought in Sanjay Garyali. The cycle had churned leadership across the sector.

But the data was now more important than the names on the door. The deleveraging that the MFIN guardrails had set in motion was showing up in every company’s investor presentation with striking consistency. The overleveraged borrower pool that had caused the crisis was being systematically drained.

“GLP of borrowers with greater than 3 pre-lenders stood at 11.1% in June 2025 versus 25.3% in August 2024. GLP of borrowers with greater than 2 lakh unsecured indebtedness stood at 9.5% as of June 2025 compared to 19.1% in August 2024.”

— Ganesh Narayanan, CEO and MD Designate, CreditAccess Grameen | Q1 FY26

Muthoot Microfin saw the same pattern from the industry-level data.

“If you look at the overall industry, the leverage customers were around 20%. They have come down to around 8%, which is more than 4 loans.”

— Sadaf Sayeed, CEO, Muthoot Microfin | Q1 FY26

And the proof that the guardrails were producing a fundamentally better book — not just a smaller one — came from Equitas.

“We had implemented the MFIN Guardrail 2.0 from Jan ’25. Out of the portfolio created between Jan to June of ’25, the X-bucket efficiency is about 99.6%, which is more or less what we used to have before this whole crisis started sometime in the first quarter of last year.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q1 FY26

New loans, written under new rules, were performing at pre-crisis levels. Spandana’s data was even more definitive — its FY26 disbursements were tracking at 99.9% collection efficiency. The problem had been in the origination standards, the guardrails had fixed the origination standards, and the new book was proving it.

Fusion Finance, which less than a year earlier had reported covenant breaches and a Rs 693 crore quarterly credit cost, showed a steady trajectory of decline.

“Credit costs have steadily declined QoQ from Rs. 571 Cr in Q3 FY25 to Rs. 253 Cr in Q4 FY25 and further to Rs. 178 Cr in Q1 of FY26. GNPA improved from 7.92% in Q4 FY25 to 5.43% in this quarter.”

— Devesh Sachdev, Managing Director, Fusion Finance | Q1 FY26

Muthoot Microfin, which had posted a loss-making FY25, reported a small but symbolically significant quarterly profit — Rs 6.2 crore. Modest. But the direction mattered more than the magnitude.

“Though it’s a very modest profit, more importantly, it indicates a firm ushering of a turnaround within the operation and the financial performance of the company.”

— Sadaf Sayeed, CEO, Muthoot Microfin | Q1 FY26

• • •

The Industry Exhales

In April 2025, Guardrail 2.0 kicked in — tightening the lender cap from four to three, and further restricting indebtedness limits. Some companies, like Spandana, had already implemented stricter rules from January. The immediate effect was another spike in rejection rates and a temporary pause in the recovery of collection sentiment. Utkarsh Small Finance Bank’s CEO noted the disruption.

“From 1st of April 2025, they changed it to 3 lender caps. Because of that, we saw some more slippages or little elevated credit cost during this quarter. And you must have seen that is more or less industry phenomena.”

— Govind Singh, MD & CEO, Utkarsh SFB | Q1 FY26

But by Q2 FY26, the adjustment was absorbed. And the tone in earnings calls shifted from cautious optimism to something closer to relief. IDFC First’s Vaidyanathan, who had spent five quarters absorbing microfinance losses, said what many were feeling.

“Looking ahead, our own sense is that this microfinance issue is behind us. It’s really taken a lot out of us in the last 5 or 6 quarters. Many of you may have almost begun to lose confidence in us.”

— V. Vaidyanathan, MD & CEO, IDFC First Bank | Q2 FY26

By Q3 FY26 — December 2025 — the normalization was broad-based. Collection efficiencies across the sector had returned to 99.4-99.7%, close to or at pre-crisis levels. Monthly PAR accretion — the rate at which new loans fall overdue — had collapsed. CreditAccess reported the number at 18 basis points in December, down from 47 bps in September.

Fusion Finance returned to profitability — a PAT of Rs 14 crore. Its new CEO described it as an inflection point.

“Q3 represents an important inflection point for Fusion. The business has now entered a phase of controlled stabilization and disciplined execution. I’m pleased to share that we have returned to profitability this quarter.”

— Sanjay Garyali, MD & CEO, Fusion Finance | Q3 FY26

Lenders who had pulled back from the sector during the crisis were returning. Fusion noted that “several banks have freshly opened up” and institutions that were in “wait-and-watch mode” had started actively re-engaging. The wholesale funding squeeze — the second-order effect of the credit cycle — was easing.

RBL Bank’s JLG disbursals had crossed Rs 700 crore monthly, with early bucket collection efficiency at 99.5%.

“The disbursal in the JLG segment is at a run rate of INR 700 crores per month versus INR 550 crores in the previous quarter. The good news is that the early bucket collection efficiency is 99.5%. This is as good as one has got in this segment for a long time.”

— Jaideep Iyer, Head of Strategy, RBL Bank | Q3 FY26

Bandhan declared the de-growth phase in its microfinance book over. Muthoot was back to normalized disbursements of Rs 850 crore a month. Ujjivan SFB reported bucket-X collection efficiency of 99.7% in December — its highest in the current fiscal — with 10 out of 10 states at 99.6% or above.

Satin Creditcare’s HP Singh, who a year earlier had been deploying 1,100 dedicated collection staff to manage overdue buckets, offered a simple verdict.

“The headwinds are practically over. You can see green shoots now coming in. I think overall it bodes well for the industry as well as for us.”

— Dr. HP Singh, CMD, Satin Creditcare | Q3 FY26

And Equitas’s Vasudevan, who had been the first to diagnose this crisis as self-inflicted, delivered the closing line for the cycle.

“As we sign off, I can only reiterate that I think the issues of microfinance is clearly behind not just us but the industry. And going forward, let’s hope for better times for all of us.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q3 FY26

While the MFI players had been fighting for survival, the large diversified banks had watched from a very different vantage point. HDFC Bank’s microfinance exposure was less than 1% of its total book. SBI’s chairman noted their MFI portfolio of Rs 10,000-11,000 crore “does not really add up to anything.” By Q3 FY26, HDFC Bank’s deputy MD was using language that felt like it belonged to a different industry entirely.

“The banking industry right now to borrow a term is going through a Cinderella phase where you’ve got very strong balance sheets... We have the lowest accretion of gross NPAs and net NPAs are at decadal lows.”

— Kaizad Bharucha, Deputy MD, HDFC Bank | Q3 FY26

A Cinderella phase for large banks. A near-death experience for several MFIs. The same economic cycle, experienced at entirely different altitudes depending on how much microfinance a company carried on its books.

• • •

What Changed

The industry that emerged from this cycle was not the one that entered it. The changes were structural, deliberate, and in some cases, permanent.

Every major player shrank its microfinance portfolio as a share of total assets. Equitas took its MFI mix from 20% to 8.5% and intended to stabilize at 10%. IDFC First expected its microfinance book to bottom at Rs 7,500 crore — half its peak. Bandhan’s non-microfinance book now accounted for 63% of advances, up from 45% two years earlier. RBL shrank its JLG book by 23% and targeted keeping it at 6-7% of advances, down from 9%. The consensus: microfinance at 7-10% of the balance sheet, not 20%+.

Companies were also insuring against the next cycle. CGFMU — the Credit Guarantee Fund for Micro Units — had existed before the crisis but was lightly used. Now it was becoming standard.

“The intent is to take that to 100% and thereby eliminating the tail risk on the microfinance business. So, if you are able to do that and manage the proportionality of the business somewhere between 7% - 8% of the asset side, I do believe that we can build a more predictable and profitable microfinance business going forward.”

— Rajiv Anand, MD & CEO, IndusInd Bank | Q3 FY26

RBL reached 80%+ CGFMU coverage on its standard JLG book. IDFC First had been insuring since January 2024 — one of the earliest movers.

Kotak Mahindra went further than portfolio caps and insurance. It did something more fundamental — it replaced the joint liability group model itself.

“What we have done is that we have replaced the joint liability group model with individual underwriting risk-based models. And therefore, the way we are dispersing and where we are not dispersing has a lot to do with what the model says are the propensity for repayment.”

— Ashok Vaswani, MD & CEO, Kotak Mahindra Bank | Q1 FY26

It was a quiet admission that the JLG model — the foundational innovation of microfinance — had weakened. Social pressure among group members was no longer enough to ensure repayment in an environment where borrowers had loans from five different institutions. Individual credit assessment, powered by bureau data and machine learning, was being layered on top of or replacing the group guarantee.

And the baseline expectations for the business had permanently shifted. The “old normal” of 1.5-2% credit costs was gone. Vasudevan was explicit about what the “new normal” looked like.

“Post corona, we thought around 3% is a normal credit cost for microfinance. Now post this 2024 overleveraging crisis, probably anywhere between 3% to 4% could be a normal credit cost for microfinance.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q1 FY26

Three to four percent, not two. The industry had recovered, but to a worse steady state than before. The premium for lending to this segment had been repriced — not by a regulator, but by experience.

Vasudevan also offered the most honest structural assessment of the sector — one that explained why every company was simultaneously declaring the crisis over while also reducing its exposure to the business that had built them.

“And in spite of that, we have been seeing a series of crisis in the Microfinance sector time and again for various reasons. The reasons change, but the repetitiveness of the crisis doesn’t change.”

— P.N. Vasudevan, MD & CEO, Equitas SFB | Q4 FY25

The industry had survived. The cycle had turned. But the companies that came out the other side were building for a different future — one where microfinance remained in the portfolio but never again dominated it. Smaller books, government guarantees, individual underwriting, tighter guardrails, higher steady-state credit costs. The cycle didn’t just create losses. It created a new architecture.

• • •

What to Watch

• The “new normal” credit cost. Most companies guide for BAU credit costs of 2.5-3.5% going forward, up from the pre-crisis expectation of 1.5-2%. Whether this holds through FY27 will confirm the repricing is real and not just post-crisis caution.

• Karnataka. The last major geography to stabilize. Several companies still flag it as lagging behind the rest of the country. CreditAccess expects Karnataka’s PAR accretion to normalize by end of Q1 FY27.

• CGFMU as industry standard. IndusInd targeting 100%, RBL at 80%+, IDFC First already covered. If credit guarantee becomes universal, the tail risk profile of microfinance changes permanently — but it also adds ~1% to the cost of lending, which has to be passed on somewhere.

• Guardrail 2.0’s growth ceiling. The 3-lender cap is structurally limiting the addressable borrower pool. Several managements now expect industry growth to settle at 10-15% — a far cry from the 25%+ that was standard before the crisis. Whether this is a temporary adjustment or a permanent reset will shape the sector’s economics.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

That’s when we reached out to Phillip Capital, who walked us through everything. This article is based on that conversation. It was originally published on .

Subtext by Zerodha

GIFT City, Explained: What We’ve Learned Running a Fund There

Hi, I’m Kashish. GIFT City keeps coming up in conversations among colleagues, in the news, on investor forums but most of the coverage stays at the policy level. I wanted to understand how it actually works from the perspective of someone running money through it. Specifically, the wealth management side with details about onboarding, fund structures, t…

Read more

2 months ago · Zerodha and Ankush Datar

You can also listen to it on Spotify or Apple Podcast.

──────────────────────────────────────────────────────

We, through our presence in GIFT City, namely, Phillip Ventures IFSC Pvt. Ltd , run the Phillip India Billion Opportunities Fund, an open ended CAT-III AIF for onboarding NRI and foreign national clients for investments into India and our firm also runs a Global PMS managing outbound investments as well.. Here’s what we’ve learned, stripped of the jargon.

Why does GIFT City even exist?

The short version: India didn’t have its own offshore financial jurisdiction. For decades, if large foreign capital wanted to enter India — or if Indians wanted to invest globally — the investments had to be made through Mauritius, Singapore, Luxembourg, or Dubai. Those jurisdictions offered something India’s domestic system simply couldn’t — a single-window, low-friction way to move capital.

If a foreign investor wanted to invest directly into India, they’d have to apply for a PAN card, register as an FPI, open local bank accounts, and start filing Indian tax returns. For an NRI, it meant getting an OCI card, opening NRE/NRO accounts and setting up a PIS account. This is a maze of paperwork across multiple regulators. Most foreign capital looked at that and routed through Mauritius or Singapore instead, where the fund structures were simpler, onboarding was faster, and the regulatory interface was a single entity rather than a patchwork of coordinating with multiple regulators and authorities.

GIFT City wanted to solve that. It was originally the brainchild of Mr. Modi during his time as Chief Minister of Gujarat. When he became Prime Minister, it took real shape. The idea was straightforward — why should India’s financial gateway sit in someone else’s country?

Think of GIFT City as India’s answer to the DIFC in Dubai or the financial centres of Singapore and Mauritius. It sits physically between Ahmedabad and Gandhinagar, but legally, it operates as a Special Economic Zone with its own regulator — the IFSCA (International Financial Services Centres Authority). Though technically on Indian soil, it functions as an offshore jurisdiction.

The real problem it solves

Forget the grand narrative for a moment. The single biggest thing GIFT City does is eliminate friction.

Let us give you a concrete example from our own business. At Phillip Capital, over 30% of our PMS assets under management comes from NRI clients. To onboard an NRI through the traditional domestic route, the process takes 30 to 45 days . As we shared earlier, the client needs a PAN card, an OCI card, a savings bank account, a PIS account, a trading account, a demat account, and finally a PMS account. That’s 50 to 60 signatures across forms. If someone signs in blue ink one time and black the next, the bank rejects it. It’s a grind.

Through our GIFT City fund? Onboarding takes two to three days. No PAN card needed. No OCI card needed. No PIS account. Minimal documentation. For second and third-generation NRIs who may not even hold an OCI card, this is a game-changer.

Now scale that logic to a foreign national, say, someone sitting in London or New York who’s bullish on India. Before the inception of GIFT City, they would need to register as an FPI, get a PAN card, open local bank accounts, and start filing Indian tax returns when choosing a route to invest. Most of them look at that and say: why bother? In contrast, through a GIFT City fund structure, all of that goes away. The fund handles the operational complexity. The investor just invests.

Adding to that, in India, you need to deal with multiple regulators & authorities. In GIFT City, there’s one unified regulator: IFSCA. For operators like us, that alone simplifies life furthermore.

What’s encouraging is that IFSCA has been proactive. They’ve been allowing innovation for new products, new regulations, and studying global best practices. They’re stringent — they don’t want GIFT City to earn a bad reputation — but they’re also pragmatic and forward-looking. The traction from global banks, family offices, and domestic fund houses setting up shop there is the proof of things going well. After recent geopolitical events in the Middle East, we even saw a rush of NRIs wanting to open a bank account in GIFT City as a safer alternative.

One more operational advantage worth mentioning is cost. Setting up a fund in, say, Singapore or Mauritius means paying legal counsel in dollars. Skilled manpower is expensive and harder to access. But, GIFT City, being in India, brings all of those costs down significantly while still operating under the highest international standards.

Money going out: The outbound story

For Indian residents wanting global exposure, GIFT City has opened doors for investors to invest globally through regulated fund management entities in India, which was not possible earlier due to there being no platform for the fund managers to set up shop.

Before GIFT City, your options were limited. You could invest in a domestic mutual fund that ran an international fund-of-fund — say, a Nasdaq ETF or a Hang Seng fund. That worked until the RBI’s aggregate industry limits were breached, after which many fund houses stopped accepting fresh inflows into international schemes. Or, if you had serious money — $100,000 or $200,000 — you could open an account in Singapore or the US directly. For smaller amounts there were limited avenues.

GIFT City changes that.

Through our platform, for example, we now offer access to US-listed equities. Even with $5,000 or $10,000, you can start building a global portfolio. Previously you would have to open an account with a broker who would be having a third party arrangement with an external broker to give you access to the global markets. LRS limits still apply though — you’re capped at $250,000 per person per year for remittances — but those limits are generous enough for most individual investors.

And since your investments in GIFT City are denominated in dollars, you automatically get a currency hedge. The Indian rupee has historically depreciated against the dollar by roughly 3-4% annually on average — and over the last 15 months, the depreciation has been steeper. That tailwind adds up over time.

Beyond individual stock access, the fund management ecosystem is growing as well

Let us explain. Say, an Indian AMC sets up a fund in GIFT City, and that fund — the “feeder” — pools investor money and channels it into a larger global fund or directly into overseas securities. As an investor, your experience feels familiar, since you’re investing in a fund run by a brand you recognize, in a regulated setup. But behind the scenes, your money is being deployed globally through GIFT City’s infrastructure. It’s essentially the same idea as the Nasdaq or S&P fund-of-funds that domestic AMCs used to run, except it’s domiciled in GIFT City instead of onshore India.

And that distinction matters, because unlike domestic fund-of-funds, these aren’t subject to the same RBI industry-wide caps that forced so many international schemes to stop accepting fresh money. And as the ecosystem matures, investors should expect thematic exposure too, not just broad S&P 500 or Nasdaq funds, with narrower slices like AI, cloud computing, or medical devices that don’t have listed equivalents in India.

Money coming in: The inbound story

Local AMCs have been especially active because they already have NRI client bases and are now offering these clients a far smoother onboarding and investment experience through GIFT City structures. As we explained in the start of the article, there are multiple operational benefits for a foreign national or NRI to choose the GIFT city route to invest.

The vehicles being used are primarily AIFs — Category I, II, and III. We specifically run a Category 3 III open ended AIF. Global family offices are also increasingly setting up in GIFT City, drawn by the presence of global banks, the regulatory clarity, and — frankly — by how well India’s growth story is selling internationally right now, especially with recent valuation corrections making the market more attractive.

A practical advantage for inbound investors currently is reporting. When an NRI invests through our domestic PMS, we report returns in rupee terms. But the client remitted dollars. If the rupee depreciates between the time they invested and the time they exit, their dollar-denominated return is lower than what our INR reporting shows. This creates confusion.

Through our GIFT City fund, the NAV is published in dollar terms, post-tax, post-expenses. What you see is what you actually earned. No mental currency gymnastics.

The Tax Reality

This is the single biggest misconception we encounter, and we partly blame WhatsApp University for it. GIFT City is not tax-free.

If you’re an Indian resident investing through a GIFT City fund into global securities, you still owe capital gains tax in India. You must report your income and pay up, same as any other investment. What is different though, is that there’s no GST on management fees and other charges levied within GIFT City. That’s some savings for investors, but it’s not “zero tax.”

For NRIs, the advantage is more tangible. When an NRI invests directly in Indian stocks, TDS gets deducted by the bank, and they must file Indian tax returns at year-end. Through a GIFT City fund, the fund itself handles Indian taxation at the fund level. The NRI investor doesn’t need to file returns in India, but file in their country of residence.

If that country is the UAE, where there’s no capital gains tax, they effectively pay only the Indian tax at the fund level and nothing more. If they’re based in any other country, where for example the short-term capital gain tax is 25%, and the Indian tax already paid at the fund level is 20%, they only owe the 5% differential in that country as India has DTAA with several countries. This is just a broad example of how it works and clients should consult with their tax advisors for a more granular understanding.

So, Should You, a Retail Investor, Care?

Yes, but with a caveat.

We always tell our clients: it’s not India or global. It’s India and global. Over the last 15-18 months, global markets have significantly outperformed Indian markets. That’s a reminder of why diversification matters.

GIFT City gives you a regulated, accessible vehicle to get that global exposure. You can open a bank account there, hold assets in dollar terms, and invest in a growing range of funds and direct equity options. For retail investors, the mutual fund feeder route through GIFT City is probably the most practical starting point, especially since many domestic international fund schemes remain closed to new inflows.

But don’t go overboard. A small, consistent allocation to global assets is the right approach. The temptation to shift everything offshore when global markets are running is a mistake. Think of it as diversification, not a switch.

The proof of concept is established, the volumes are growing across products, and the regulator has earned credibility. There have been talks of some companies even setting up their IPOs through GIFT City. For anyone — NRI, foreign national, or Indian resident — looking to move capital in or out of India through a regulated, low-friction, cost-efficient structure, it’s the most compelling option available today.

Nishit Shah is Principal Officer & Vice President PMS at Phillip Capital India Pvt. Ltd. and has been with the firm for over 10 years. Ankush Datar is part of Nishit’s team, working in the PMS team and assisting in fund management, for over 7 years at Phillip Capital.

Over the past few weeks, the rupee has been moving around in ways that most people haven’t been able to fully explain. Some of it is straightforward — India imports a lot of oil, oil is now significantly more expensive, and that means we need a lot more dollars than we did before. But there has been another layer to the story, a more technical one, and that is what we wanted to dig into. So we got Deepak Shenoy on a call.

You can listen to the podcast on Spotify and Apple Podcasts.

The heart of the conversation was about something called the NDF market — non-deliverable forwards. It is essentially a market for betting on the rupee that operates entirely outside India, beyond the RBI’s direct reach. No rupees change hands, no dollars change hands — it is a cash-settled bet on where the rupee will land. Deepak called it the suited-booted equivalent of a dabba trade.

When the rupee came under pressure in recent weeks, this offshore market started showing higher dollar-rupee rates than the onshore market in India. Indian banks, seeing a gap between the two prices, started buying dollars cheaply in India and selling them at a higher price in the NDF market, pocketing the difference. The problem was that all this dollar-buying onshore wasn’t genuine import demand — it was pure arbitrage, and it was pushing the rupee down in the very market the RBI was trying to stabilise.

The RBI figured it out and capped each bank’s position at $100 million. Banks then tried to quietly transfer their positions to corporate importers to avoid taking losses on the unwind. The RBI caught on and shut that door too. What followed was a messy unwind that explains much of the rupee’s erratic movement over the past week.

The deeper point Deepak made is that none of this actually solves anything. Every time the RBI plugs one hole, the market finds another, because the incentives don’t disappear. His argument is that the real fix is to open the rupee up — more participants, fewer restrictions, genuine internationalisation. But that requires changes to law that are outside the RBI’s hands entirely.

This was a really good conversation. I hope you enjoy it.

Share

And, if you are one of the few who prefers to read this instead of watching or listening, here’s the transcript.

Subtext by Zerodha

Deepak Shenoy on what's happening with the rupee

Hi folks, Krishna here…

Listen now

2 months ago · 2 likes

Hi folks, Krishna here.

Since February 28th, a 34-kilometer stretch of water has upended the global oil market in a way that most people still haven’t fully absorbed. The Strait of Hormuz, which carries roughly 20% of the world’s oil supply, has been closed for over a month. Asian governments are in full panic mode. Flights are being cancelled and routes cut because jet fuel is running short. India, which was one of the first countries to feel the acute stoppage of supply, is already dealing with a serious LPG crunch — and LPG is not a premium product, it is a basic cooking fuel for hundreds of millions of households in this country.

Yes, there is a ceasefire now, and we are glad for it. But this conversation was never really about the war. It is about the system underneath — how the world built itself into a position where one 34km stretch of water could do all of this, and what that means going forward. That does not have an expiry date.

We wanted to understand what is actually happening here, so we got Rory Johnston on a call. It was very kind of him to reply to our tweet :)

You can listen to the podcast on Spotify and Apple Podcasts.

Rory runs Commodity Context and has spent years closely tracking the oil market. He describes himself as a natural bear — someone who usually believes the market will find a way to adapt, and has the reputation to back that up. The oil market has been through a lot over the past five years: the attacks on Saudi Aramco facilities in 2019, COVID, Russia’s invasion of Ukraine, the Houthis in the Red Sea. Each time, the market found a way through. Which is why it is significant that Rory is genuinely alarmed right now. He believes this is not a version of those previous shocks. It is something categorically different.

The numbers explain why. About 20 million barrels of oil a day flows through the Strait of Hormuz. Even after accounting for every available pipeline reroute — the Saudi East-West pipeline, Emirati and Iraqi alternatives — you are still left with a net shortfall of about 13 million barrels a day. And critically, this is not just lost exports with oil quietly accumulating in storage somewhere. Between Iraq, Kuwait, Saudi Arabia, the UAE and Qatar, production has been physically shut in. Those barrels were never produced. They do not exist. Over the first month of the crisis alone, the global system lost more than 200 million barrels of oil. April’s losses, Rory estimates, will be equivalent to the entirety of the IEA’s record coordinated strategic petroleum reserve release — the largest such release ever conducted — and that will happen in a single month.

So where are prices?

Brent crude was sitting at around $109 when we spoke. Rory thinks it should already be in the $125 to $135 range at minimum, and that if the strait stays closed, it could go to $200 or beyond. The question we wanted to ask — and which turns out to be the most interesting question in the whole conversation — is why the financial market hasn’t got there yet.

Part of the answer is structural. Before this crisis, the oil market was in oversupply. Inventories were building, prices were under downward pressure, and the base case assumption for most market participants was that this soft, bearish environment would persist. That prior is hard to shake. But the bigger factor, Rory says, is Trump. Every week of this war, someone from the White House has said the conflict is nearly over — just one or two more weeks. Markets believe it every time, and oil sells off. There is also a simpler, more human explanation: a rapidly rising oil price is bad for equities, and everyone in the market would rather it be over. As Rory put it, any opportunity to sell down crude and bid equities higher just feels better, so that’s where the money goes.

Meanwhile, the physical market is telling a completely different story. Refined products — jet fuel, gasoline, gas oil — are trading well above $200 to $250 a barrel at the Singapore market. Dated Brent, which is the price for physical crude for near-immediate delivery, is already above $120, even as the futures market — which prices oil for June delivery — sits at $109. That gap is what’s called backwardation, and it is worth understanding because it is one of the clearest signals of how bad things actually are. When the futures curve slopes downward like this, it doesn’t mean the market thinks prices are going to fall. It means the market is paying an enormous premium to get oil right now, today, because supply is so critically short in the present moment. The futures price is reflecting the hope of a resolution. The physical price is reflecting reality.

There is also what Rory described as an air pocket moving through the global supply chain, and this is perhaps the most useful image in the entire conversation for understanding the timing of what is happening. When the Strait closed, there were still ships at sea — tankers that had left the Gulf weeks earlier, laden with crude and petroleum products, still making their way to their destinations. Those ships have been arriving one by one. India felt the shortage first because of its proximity. Then East Africa. Then East Asia. Europe is next, and North America after that. But once that last tanker completes its journey and docks, as Rory put it, nothing will be behind it but air. That is the moment, he believes, when the financial market will have no more cover.

We also spoke about who is winning and losing in all of this. The Gulf states — whose oil revenues have essentially evaporated — are among the biggest losers. So are Asian importing nations, who are bearing the sharpest end of the supply shock. The single largest winner, by Rory’s reckoning, is Russia. The Trump administration had spent months successfully tightening sanctions on Russian oil, and had actually managed to reduce Indian imports of Russian crude from over two million barrels a day down to around one million. Almost all of that progress has now been undone. India has pivoted back to Russian crude, and the US Treasury has explicitly eased restrictions on Russian floating storage to help facilitate the flow. There is a grim irony to it — Ukraine has noticed, and has responded by dramatically intensifying attacks on Russian export infrastructure.

And then we spoke about what comes after, whenever this eventually ends. Rory doesn’t think the market snaps back cleanly. The production that has been shut in will take months to restart. The inventories that have been drawn down will take longer still to replenish. And beyond the immediate recovery, he thinks this crisis will accelerate two things in Asia simultaneously: a faster push toward energy diversification and renewables, and a structural overbuild of strategic reserves, as governments try to ensure they are never this exposed again.

This was one of the best conversations we’ve had. I hope you enjoy it.

Share

And, if you are one of the few who prefers to read this instead of watching or listening, here’s the transcript.

Subtext by Zerodha

Rory Johnston (oil expert) on what's happening in the oil markets

Hi folks, Krishna here…

Listen now

3 months ago · 2 likes

The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

Things don’t look good in the diamond market

I came across this tweet yesterday.

Source

It shows that diamond prices have fallen to their lowest level this century. And it reminded me that it’s been a while since we gave a proper update on what’s been happening with the diamond industry on this show, because quite a lot has happened since we last covered it. So today I want to catch everyone up and talk through a quote from Anglo American’s CEO.

But before that, let me give you some quick context.

A quick recap for those who are new

De Beers is the company that essentially invented the modern diamond industry as we know it.

In 1888, a British financier named Cecil Rhodes set it up with one specific goal — to control the global supply of diamonds so completely that prices could never collapse. The logic was elegant in a somewhat unsettling way: diamonds don’t have real intrinsic value, and when massive deposits were discovered in South Africa in the 1870s, they risked becoming worthless. So De Beers built a cartel that at its peak controlled nearly 90% of the world’s rough diamond supply, and paired that with one of the most effective marketing campaigns in history — “A Diamond Is Forever” — which convinced entire generations that a diamond ring was the only acceptable way to express lifelong love.

That cultural norm, which feels like it’s always existed, was manufactured by De Beers. We covered this in much more detail in one of our Daily Brief pieces if you want the full story.

Then came lab-grown diamonds, and everything started unravelling. Scientists figured out how to grow diamonds in a laboratory — stones that are physically, chemically, and optically identical to mined ones, which a professional jeweller cannot tell apart without specialist equipment. Because they can be mass-produced, their prices collapsed. According to data tracked by diamond analyst Paul Zimnisky, a one-carat lab-grown diamond that cost around $5,000 in 2016 now costs under $1,000.

Source

The American market shifted dramatically toward them, with more than half of all US engagement rings now carrying a lab-grown stone. China — the other great engine of global diamond demand — essentially exited the market, with Chinese diamond consumption falling from around 100 billion yuan in 2021 to just 43 billion yuan by 2024 as consumers moved toward gold instead. For a company whose entire model rested on manufactured scarcity, this was an existential problem, and the Barchart chart captures exactly what happened to prices as a result.

Anglo American and three straight years of writedowns

De Beers doesn’t operate independently — it’s 85% owned by a British mining giant called Anglo American, which most people haven’t heard of even though its products show up in virtually everything we use, from the copper in your phone to the iron ore in steel. The remaining 15% is held by the government of Botswana, which supplies roughly 70% of De Beers’ rough diamonds and has historically funded its schools and hospitals from diamond revenues.

Source

As the market deteriorated, Anglo American started formally acknowledging on its books that De Beers was worth less than it had been carrying it for — what’s called a writedown. They wrote it down by $1.6 billion in 2024, by another $2.9 billion in early 2025, and by a further $2.3 billion this past February, bringing the total across three consecutive years to $6.8 billion. Anglo has also put De Beers up for sale since 2024, wanting to refocus on copper and iron ore, but the sale has been complicated because potential buyers are valuing the company at far less than even its already-slashed book value.

It was after announcing this third writedown that Anglo’s CEO got on a call with journalists and said something I found quite remarkable. When asked whether things would get better from here, he said: “I certainly hope this is a low point.“ Not I believe, not the fundamentals support a recovery. Just hope. From the CEO of one of the world’s largest mining companies, after three straight years of writing down the same asset by a combined $6.8 billion, the best he could offer was that he hoped it had stopped getting worse.

What the sightholder cuts are really saying

Alongside the writedown, something else happened last month that tells the story even more clearly. De Beers quietly informed more than 20 of its 69 so-called sightholders that their contracts would not be renewed when current agreements expire in July, reducing the list to roughly 45 names. A sightholder is essentially a member of the exclusive club De Beers built to control how rough diamonds flowed through the market. De Beers has never sold diamonds on an open exchange — instead, ten times a year, it invited approved buyers to receive a parcel of rough diamonds. You didn’t choose what was in the parcel, you didn’t negotiate the price, you took what De Beers gave you and paid what De Beers asked, and you felt fortunate to be included. Banks extended credit to sightholders. Manufacturers sought them as partners. The list itself was a signal of standing in the industry. In the 1970s, it had over 350 names on it. Then it fell to 69. And now it stands at roughly 45. From 350 to 69 to 45 — that is the contraction of the natural diamond industry in three numbers.

The surface explanation for the cuts is that De Beers is producing fewer diamonds and needs fewer buyers, having reduced its 2026 production guidance to 21–26 million carats from a previous estimate of 26–29 million. But the deeper reason it’s producing less is that it’s sitting on approximately $2 billion worth of unsold rough diamonds it cannot move — levels not seen since the 2008 financial crisis — because demand for natural stones has simply fallen away. The sightholder cuts are not a tactical adjustment while the company waits for things to bounce back. They are De Beers getting honest about the actual size of the business it now has.

Even Botswana is backing down

There’s one more thing that completes this picture.

When Anglo put De Beers up for sale, Botswana’s president Duma Boko was unambiguous about what he wanted — not a larger stake, but control. His exact words were: “It’s our people who are running this country, and we said we want De Beers, and we are going to take it.“ A few weeks ago, The Economist reported that Botswana has now quietly dropped that position, and is pursuing “at least 25%” — a minority stake. An executive at Debswana, the joint venture between the Botswana government and De Beers, told the publication: “It probably doesn’t make sense to go all out.” The fact that even the country which mines most of De Beers’ diamonds has pulled back from the idea of taking control tells you something about how people closest to this business actually feel about its trajectory.

What India means for all of this

The bull case for natural diamonds runs mostly through India. De Beers CEO Al Cook said in January that India grew 11% in natural diamond demand in 2025 — the strongest of any country in the world — and has overtaken China to become the world’s second-largest market after the US. Cook argues that Indian consumers understand the difference between natural and lab-grown, and tend to prefer the real thing for occasions that carry emotional weight. And beyond consumption, Surat is where more than 90% of the world’s diamonds are cut and polished, making India the backbone of the entire supply chain. We covered Titan’s recent entry into the lab-grown space after years of holding out, but that was targeted at the everyday fashion segment — the premium natural diamond category remains relatively insulated in India for now.

But what the sightholder cuts tell you is that De Beers is not waiting for the India story to fully play out before restructuring. It is building its business today for the smaller market that exists today. Forty-five elite buyers, $2 billion of unsold inventory, a third consecutive writedown, and a CEO who, when asked whether things will improve, says he hopes so. The ice, as the Barchart chart makes clear, is still very much melting.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

Today’s Who Said What? episode is going to be a little different. I am not going to pick a comment from a business leader or a fund manager and contextualize things for you.

We recently started something called Subtext. The idea behind it is simple — there’s a lot of interesting thinking happening in markets, regulation, energy, technology, and everything adjacent to all of that, and most of it never reaches people who aren’t deep inside those worlds. So we’ve been trying to create a place where you can read extremely high-quality opinions from people who are actually in the thick of it. We have some exciting stuff lined up there, and we will start releasing it next week.

The other thing we’ve been doing on Subtext is something a little more personal. It’s called What We’re Reading, where every weekend, we send a long piece as to what all of us at Markets by Zerodha is reading. I wanted to experiment and see if this is something you like. If it is, let me know in the comments.

In the last edition, all of us went through a bunch of interesting reads, all of which you’ll find links to in the description below. I’m going to walk you through a few of them here today. The ones that I thought were worth talking through out loud. But if you want the full piece, it’s in the description and it’s worth reading in full.

Let me start with mine.

I’ve been reading a piece by Ed Zitron, who runs a newsletter called Where’s Your Ed At? It’s fun, it’s a little provocative, and he regularly goes places that mainstream tech journalism mostly won’t. This piece is called The Beginning of History, and it starts where you wouldn’t expect a piece about the AI bubble to start: the Strait of Hormuz.

The Strait of Hormuz is a 24-mile-wide channel between the UAE and Iran. That’s it — 24 miles at its narrowest point. And roughly 20% of the world’s oil and LNG flows through it every year.

By now of course you know all of this. At the end of February, Iran’s Revolutionary Guard closed it off, telling merchant ships that passage was no longer allowed. Iran can’t legally do this — it’s international waters, governed by UN convention — but merchant captains don’t particularly want to find out how seriously Iran means it. Neither do their insurers. And so the tankers are sitting put.

What that means for oil prices is fairly direct. We saw a 30% overnight spike, with the price of a barrel crossing $100 for the first time since 2022. And oil prices feed into everything: shipping costs, construction materials, medicines, basically anything that has petrochemicals in it or needs to be moved anywhere. Inflation goes up. When inflation goes up, central banks raise interest rates to cool things down.

And here’s where the AI bubble enters the picture. The entire infrastructure of AI — data centres, chip shipments, the VC funds backing startups, the private credit sitting on top of all of it — runs almost entirely on debt. When interest rates rise, even by a point or two, financing the hundreds of billions this bubble requires gets significantly more expensive. Investors in VC funds can suddenly earn better returns with far less risk by parking money in a government bond. The debt spigot, which has been running wide open, starts to tighten.

Then there’s the physical layer. The data centres powering AI are largely gas-powered, and natural gas is up as much as 50%. Flying racks of NVIDIA chips from Taiwan to Texas becomes more expensive. Every contractor building these facilities will pad their fees to account for their own rising costs. The squeeze comes from every direction at once.

All of that is about external pressure — macro forces making the bubble more expensive to sustain. But Zitron’s second argument is different, and in some ways worse. It’s about whether the foundation that’s supposed to justify all this spending was ever real in the first place.

Press reports have been citing Anthropic’s annualised revenue at anywhere from $14 billion to $19 billion — but those figures take a single month of revenue and multiply it by 12. Then a court filing from Anthropic’s own CFO, submitted as part of a suit against the Department of Defense, stated that Anthropic’s lifetime revenue as of March 9th, 2026 is $5 billion. When Zitron adds up all the monthly figures that have been publicly reported, they already exceed $6 billion — which makes the $5 billion lifetime figure, and the $4.5 billion annual figure that’s been widely cited for 2025, very hard to believe.

His broader point is that journalism keeps reaching for the same historical comfort: dot com, crypto, now AI — and in each case building a narrative around the assumption that it’ll work out because things have worked out before. He thinks that’s lazy, and that this era is different enough that the comparison fails. I’m not sure I have a fully settled view on that yet. But the Anthropic numbers alone make this worth reading.

Now, Pranav has been reading two things, and I’ll tell you about both.

The first is a Paul Graham essay called The Brand Age. He starts with the Swiss watch industry. After World War Two, the Swiss were the best in the world at making watches. Then the Japanese started closing the gap, and within a generation they’d caught up, and a lot of Swiss manufacturers went bankrupt. The ones that survived did something interesting — they stopped competing as tool-makers and started competing as jewellers. They sold you not a watch, but a brand. And that brand saved them.

Graham’s conclusion isn’t a comfortable one. He argues that brand is often more important than actually making a good product. That good design — sensible, functional, honest design — can actually be bad for branding. That good branding might need to be loud and gaudy, which is the opposite of what anyone with taste would want to make. He’s not celebrating this. He’s just describing it plainly, which is what makes it sting a little.

However, one of our own stories might not entirely agree with Paul Graham. Last year, we narrated a tale of how the company that followed this playbook for natural diamonds to the highest possible extreme is now fighting for bare survival. And that’s because its fundamental business is becoming more commoditized every single day.

The second thing Pranav has been reading is Ruchir Sharma in the FT on China’s growth targets. Most countries discover their GDP growth rate after the fact — the year ends, the data comes in, and you find out how you did. China does it the other way around. They announce a target, and then they hit it. Ruchir’s point is about what it costs to do that, and who ends up paying.

Hitting a target, year after year, requires constant investment whether or not that investment makes economic sense. You build factories, you build capacity, you produce. Over time, this creates a structural glut — too much of everything, produced at a cost lower than anyone else can match. The first version of that reaches you as the cheap Chinese goods that have made life genuinely more convenient. The second version is the hollowing out of industries in countries that simply cannot keep up. That’s bred a deep resentment across much of the world. The targets make China’s problems the world’s problems — and there’s only so long the world will absorb that quietly.

And then there’s what Manie has been reading, which I think is one of the more interesting things in this edition — a long piece by Kevin Xu on China’s open-source software history.

Kevin Xu is someone worth knowing. He led GitHub’s international business strategy, worked at the White House during the Obama years, and runs a newsletter called Interconnected that covers Chinese technology and entrepreneurship. His piece starts from a question that I suspect most people haven’t thought to ask: DeepSeek, MiniMax, and Kimi — the Chinese large language models that have been making headlines — are all open-source. Did that come from nowhere? Or is there a much longer story behind it?

It turns out it’s the latter. China’s open-source community wasn’t created by a government mandate. It grew the same way it did everywhere else — spontaneously from the bottom-up, from people who cared about it. The government only truly committed to it once they realised how strong the movement already was, and how much it could add to China’s economic and technological strength.

The first big inflection point Xu traces is Alibaba. They were entirely dependent on Western providers — Oracle, IBM — for their server infrastructure. As their e-commerce business scaled, those bills kept growing. So Jack Ma brought in someone to fix it: a man named Wang Jian, who had a doctorate in psychology and, notably, did not know how to code. Wang Jian ended up playing the biggest role in helping Alibaba wean itself off that dependency.

What makes the piece stick is that it doesn’t treat open-source as just a software story. It puts it inside a frame of national security and sovereignty — who controls the tools, who can be cut off, and who can’t. Manie says he doesn’t like most of the long essays that get published on Twitter these days, but this one is a gem. I’d agree.

Thats it for this episode, do let me know if you liked this and if we should do more of this :)

The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

What’s going on with private credit?

In the past few days, apart from the ongoing war coverage, there’s one — actually, two — words that keep quietly popping up in financial news: private credit. Wild things are happening in the US because of it. Some people are already calling it the next financial crisis. Funds are freezing withdrawals, assets are being marked down, and companies are going bust.

What really got us curious was a specific warning from Steffen Meister, the chair of Partners Group — a Swiss private markets firm managing more than $180 billion in assets. He told the Financial Times this week that default rates in private credit could double in the next few years. When someone managing that much money says something like that publicly, it’s worth paying attention.

Source

But since private credit is still a relatively new concept in India, let’s take a step back first — understand what it actually is and how it works — before getting into what’s going wrong.

Source

Where private credit came from

To understand private credit, you first need a rough picture of how traditional banking works. A bank takes deposits from ordinary people like you and me — money that can be withdrawn at any time — and lends it out to businesses for years at a stretch. This is an inherently unstable arrangement, because the bank’s liabilities are short-term but its assets are long-term. If too many depositors ask for their money back at once, the bank cannot pay. That’s what a bank run looks like.

Because of this fragility, banks are heavily regulated. After the 2008 global financial crisis, that regulation got significantly tighter. In the US, new rules forced banks to hold much more capital against risky loans, which made lending to mid-sized businesses far less attractive. They pulled back, and a large portion of the market was left without a reliable source of funding.

Now, here’s where private credit stepped in.

Instead of going to a bank, a mid-sized company could now borrow directly from a private credit fund — a large pool of capital raised from institutional investors like pension funds and insurance companies, who agreed to lock their money away for several years. Because these investors weren’t demanding their money back at short notice, the fund could lend longer, lend more flexibly, and lend to borrowers that banks were no longer willing to touch. Over the decade and a half since 2008, this market has grown into a $2 trillion global industry.

How it actually works

When a private credit fund lends to a company, it isn’t going through any public market. There are no bond ratings, no standardised terms, no thousands of anonymous creditors. It’s a direct deal negotiated privately between the borrower and usually just one lender, which means every element — the repayment schedule, the interest rate, the conditions attached — can be customised. If a company needs to fund an acquisition quickly, or has a complex business model that a bank’s standard process wouldn’t know what to do with, a private credit fund can structure something that works. The trade-off is price: private credit is significantly more expensive than a traditional bank loan. In India, where bank loans might cost 8–9% a year, private credit deals routinely target lender returns of roughly 18%.

Source

These loans also sit at the top of what’s called the capital structure.

If a company goes bankrupt and its assets are liquidated, the private credit lender gets paid back before anyone else — before other bondholders, before shareholders, before almost everyone. That combination of high returns and first-in-line repayment priority is why large institutional investors found private credit so attractive. There is one more feature worth understanding, though, because it becomes central to what’s happening right now. Unlike a fixed-rate bond, most private credit loans carry a floating interest rate, meaning the interest the borrower pays rises and falls with prevailing market rates. When rates are low, this is manageable. When rates rise sharply, the burden on the borrower can increase dramatically.

When the tide turned

For most of the decade after 2008, private credit was a success story. But two things have happened recently that are beginning to unwind it simultaneously.

The first is interest rates. After central banks raised rates aggressively from 2022 onwards to fight inflation, the floating-rate structure that defines most private credit suddenly became far more burdensome for borrowers.

Source

Companies that had borrowed at manageable rates found their interest payments climbing faster than their revenues. The consequences are clear in the data: according to Fitch Ratings, the US private credit default rate hit 9.2% in 2025, a record high, up from 8.1% the year before. Among the smallest borrowers, those with annual earnings under $25 million, the default rate was closer to 13%.

The second issue is AI. Over the last decade, a significant share of private credit lending went to mid-sized software and technology companies, often because private equity firms were using borrowed money to buy and consolidate these businesses. The model worked well when software companies had stable, recurring revenues. But AI tools have introduced a risk nobody priced in — the very services many of these companies sold can now be replaced or undercut. What makes this particularly painful for private credit, as opposed to private equity, is the asymmetry of the investment.

A private equity firm that owns a software company can still benefit if AI ends up helping that specific company thrive — the equity goes up, and the firm wins. A private credit fund doesn’t have that option. It lent money and needs it back with interest. It doesn’t share in the upside if a company flourishes, but it bears the full loss if the company fails. In a world where AI is bifurcating outcomes — some companies surging, many others getting wiped out — private credit is structurally exposed to the losing side. This is exactly what the fund manager I mentioned in the beginning described to the FT, and it’s why JPMorgan has already begun marking down its private credit exposure in software — a notable signal from a bank that lends heavily to these funds.

The structural cracks

Beyond the macro picture, two features of how private credit has been operating are amplifying the stress. The first is what the industry calls “covenant-lite” loans. A covenant is a rule a borrower must follow as a condition of the loan — say, a minimum cash balance they must maintain. These rules give lenders early warning if a borrower starts deteriorating, so they can intervene before things spiral. But as private credit became increasingly competitive, funds started dropping these protections to win deals. More than 30% of private credit deals today are covenant-lite, meaning a lender may not find out a borrower is in serious trouble until the company has already run out of money.

The second is opacity, and this one is directly causing the withdrawal freezes you’ve been reading about. Because private credit loans aren’t traded on any exchange, there’s no daily market price for them. Funds value their loan portfolios internally, typically at or near original value even as conditions change. This was a theoretical concern for years — until private credit firms began launching a specific type of fund called a Business Development Company, or BDC, to let wealthy retail investors into the market. Unlike traditional funds that lock capital away for a decade, BDCs offered semi-liquidity: investors could typically withdraw up to 5% of the fund’s value each quarter.

The trouble is that partial liquidity creates a classic bank-run dynamic the moment investors lose confidence. If people suspect the actual value of the loans inside a fund has fallen, but the fund is still allowing withdrawals at the original stated value, rational investors will rush to take their money out before everyone else does. Once requests hit the 5% cap, the fund freezes further redemptions — and that announcement tends to confirm exactly the fears that triggered the rush. Last week, BlackRock limited redemptions from a flagship debt fund after a surge in requests. Blackstone disclosed that its private credit fund BCRED saw a significant jump in withdrawal requests in the first quarter of this year. These are not obscure players — they are among the largest asset managers in the world.

So what’s actually happening on the ground?

The most vivid example of all this coming together is Blue Owl Capital, one of the biggest names in private credit with over $300 billion under management. Blue Owl had built a large portion of its lending business around software companies — exactly the sector most exposed to AI disruption. When retail investors in one of its BDC funds started asking for their money back in large numbers, Blue Owl did two things. First, it sold $1.4 billion worth of loans from its portfolio to institutional investors to raise cash. Then it permanently shut the gates on quarterly redemptions for that fund — meaning investors who expected to be able to take their money out every three months no longer can. Blue Owl’s stock has since fallen sharply, entering an 11-day losing streak after the announcement, its worst run since going public. Dan Rasmussen of Verdad Capital called it “a canary in the coal mine.” Others have drawn comparisons to the early warning signs of 2007.

Source

To be clear, none of this means a collapse is imminent. But there are warning signs.There are many more layers to this story — how these funds are structured, how fraud has crept in through some of the bankruptcies, what continuation vehicles and CLOs are, and what all of this means for India’s own private credit market, which is growing fast. We’ll get into all of it, either through future episodes of Who Said What or on The Daily Brief. For now, though, you have enough to understand why two words are suddenly everywhere.

What we’re reading

Our team at Markets is always reading, often much more than what might be considered healthy. So, we thought it would be nice to have an outlet to put out what we’re reading that isn’t part of our normal cycle of content.

So we’re kickstarting “What We’re Reading”, where every weekend, our team outlines the interesting things we’ve read in the past week. This will include articles and even books that really gave us food for thought.

Subtext by Zerodha

What we're reading #3

Hi folks, hope you’ve had a great week…

Read more

3 months ago · Zerodha

The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

Infosys held its annual Investor Day recently. And a quote from the company’s management looked particularly interesting to us:

“I am going to talk about an interesting segment called energy, utilities, resources and services, and I call this interesting because AI has created a circular economy in energy, utilities and resources sector. While these sectors are heavy users and consumers of AI, they are also critical enablers of AI. If you look at utilities today, particularly electric utilities, they power and decide where the next data center should be and how fast the AI data centers can grow. In fact, there are views that electricity is the only limiting factor in growth of AI.”

There’s a lot to unpack here.

We’ve heard a lot about the immense power requirements of AI data centers. And, in our recent Daily Brief story on data center financing, we briefly highlighted the importance of getting power grid permits for an AI data center. But we didn’t really get into what entails a power grid connection for an AI data center.

A data centre operator usually scouts sites not just based on where there’s a lot of land, but also where power substations are close by. Once they decide on a site, they submit an interconnection request to the appropriate utility company, asking for a certain amount of power capacity over a period of time, as well as ramp-ups. And when a data center is built, the operator signs a long-term power purchase agreement, either with a utility, or even directly with a power-generating company.

The bottlenecks lie in power management. For one, there’s only so much non-renewable energy in the world. And without adequate substations and transmission lines in place, even that will be heavily mismanaged. In the US, data centers draw electricity from a grid that is still heavily powered by fossil fuels, especially natural gas. And policymakers are worried that this might already be pushing up household power prices in some regions.

Naturally, renewables are expected to play a huge role here. In fact, they already power nearly half of India’s new data centers. However, renewables aren’t spot-free either. Without ample energy storage in place, it’ll be very hard to run a data center on solar panels at night.

However, beyond energy being a bottleneck for AI, what we actually want to address is the fusion of two business models that, at first glance, have little resemblance.

Tech has historically been faster-moving, more speculative, and more tolerant of risk, with profits driven by scale. Utilities, by contrast, are slow-moving, politically regulated, and heavily dependent on capex in long-lived physical infrastructure. Their revenues are closer to annuity-like cash flows, and they typically recover investments gradually through tariffs over decades, earning only a regulated return on that asset base.

Based on how fast they want to move, tech presents its power forecasts to utilities, who accordingly have to decide their investments. But if those forecasts turn out to be wrong, who bears the risk?

However, this fusion has created an interesting result. Now that AI has forced tech to spend far more on data center capex than they ever have before, AI data centers share many features with power utilities.

Think about it. Running an AI data center forces a tech company to start thinking like an infrastructure operator. It has to forecast demand for computing power, plan capacity additions, secure reliable electricity supply, and build redundancy through backup power systems.

More importantly, the end product of an AI data center is computing power. While in electricity you pay per kilowatt-hour, with AI, the conversation shifts to paying per token, or per API call, or even per GPU-hour. Much like power grids, data centers have to manage supply of compute (through GPUs and servers) with demand for it (which comes from AI labs).

And when you take many data centers together, you essentially have built a grid for computing power. Perhaps, it’s because of this realization that China is attempting to build a national, state-led network of cloud computing centers. After all, there will always be fluctuations in the demand and supply of computing power across regions. A national network will essentially ensure that there is a functioning market for compute, which ensures transfer from excess regions to deficit ones. All of this sounds quite a bit like electricity, if not entirely.

So far, we’ve spoken about how utilities fit into the AI data center story. However, we want you to bring back to the original quote that got us to explore these concepts. Take the circular economy side of things that Infosys talks about. In essence, it is a thesis they’re betting on.

In Infosys’ view, AI growth is driving massive investment into energy, utilities, and resources companies. These companies themselves need AI to optimize grids, predict loads, manage complex supply chains, do predictive maintenance on assets, etc. They also have enormous ERP-heavy, legacy-heavy IT estates that need modernization.

Infosys is positioning themselves as the AI services partner for these companies, and they seem to have made considerable progress on this, saying:

“And the proof of the pudding is we are the AI partners for 15 of our top 25 clients in this segment. We do work across the AI framework. We have created digital twins for a very large oil and gas major to take asset telemetry and make the assets more intelligent, more automated, reduce their downtime. We have worked in AI-grade data engineering for a very large electricity provider to predict the load on the grid and ensure they invest on the grid where there is congestion to provide electricity to the data centers.”

Now, this is certainly an interesting thesis. However, is this the only angle that Indian IT is thinking about when it comes to AI data centers? After all, Infosys’ most important competitor seems to have a slightly different view.

A few months ago, TCS announced that they were building their own AI data center with 1 GW capacity. And just recently, TCS said that they aren’t stopping at one data center, either. Here’s what their CEO, K Krithivasan, said:

“We are having discussions with multiple other hyperscalers. We are very, very bullish. There is going to be lot of latent demand or unmet demand by 2030 so there is going to be a lot of investment required.”

As we covered before, the understanding is that this is a defensive move from TCS. With LLM usage increasing each day, AI labs will require more computing power. However, it would be far-fetched to expect TCS to build their own LLM, considering that they aren’t a product company. Instead, they decided to capture a share in computing power.

TCS does not publicly detail a separate power procurement model for its data centers. But its disclosures show that the company draws on a broader operational energy mix which is renewable-heavy, while also including 10.2 MW of rooftop solar across campuses.

The question, then, is whether Infosys will go down the same route. They do already have non-AI data centers, 45% of which also run on solar energy. It can also be argued that by owning their own AI data centers, they may be essentially doubling down on their thesis. However, they’ve yet to make any announcement on AI data centers like TCS.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

India tiptoes (forward) around Chinese investments

Over the last five years, India’s posture toward Chinese investments has only become more complicated. Since the Galwan Valley clash in 2020, India enforced the “Press Note 3” norms, which limited Chinese investments into India. And since then, India has adopted a precautionary stance with China.

However, at Business Standard’s event, Piyush Goyal seemed to hint that that strategy won’t forever be closed off. Here’s what he said:

“Also, in cases where our relationships with neighbouring countries have improved, we are keeping an open mind about how we can attract better technology and more investments, from China too. We are in dialogue, and it will be an evolving situation. We are open to newer ideas.”

In the same conversation, he also said:

“The government’s approach at the moment is to accelerate the approval process.”

Now, for all you could say, this could have been a statement meant just for display. But over the past six months, the thaw between Delhi and Beijing has gone from symbolic to substantive.

Prime Minister Modi met President Xi Jinping on the sidelines of the BRICS summit in October 2024, the first formal bilateral meeting between the two in five years. The two sides agreed to resume direct flights, ease visa restrictions, and — crucially — work toward normalizing economic ties.

And that diplomatic thaw has already started producing deals on the ground. Last year seemed to show some level of opening up. In batteries, for instance, we covered how 2025 saw a lot of deals. Sectors like EMS and pharma didn’t face a lack of activity either. But these seemed like more isolated incidents, rather than a broad national strategy.

But here’s where it gets really interesting. Goyal’s statement didn’t arrive in a vacuum. It came just weeks after India signed an interim trade deal with the United States — the first phase of which involves India lowering tariffs on American goods in exchange for the US easing some tariff pressure on Indian exports. Of course, this trade deal was effectively reset by a US Supreme Court order, but that doesn’t negate the fact that the ball has begun rolling.

And prior to that, India and the EU signed a landmark trade deal — the largest for both sides. In the past 3 years, India has signed trade agreements with Australia, the UK, UAE, Oman, and New Zealand.

So India is now simultaneously deepening trade ties with the US, accelerating a trade pact with Europe, and signaling an openness to Chinese investment, besides making connections with the rest of the world. One could even say that India might be positioning itself as the country that plays with everyone.

The US wants India as a counterweight to China in Asia and as an alternative manufacturing destination. Europe wants a trade partner that isn’t fully in the American orbit. And Chinese firms, with their declining domestic profits, need markets and manufacturing bases to reroute their supply chains, especially as Western tariffs bite.

India is trying to figure out how to be all of those at once. Each of these deals effectively acts as a hedge against the other.

Now, none of this means India is throwing the doors wide open. Goyal was careful to say “calibrated steps“. The government is only likely to ease restrictions on a sector-by-sector basis — starting with areas where India desperately needs Chinese expertise (EVs, batteries, solar, electronics manufacturing) and keeping walls up in areas considered sensitive (telecom, defence, critical infrastructure).

The model seems to be what some policy watchers call “controlled engagement” — let Chinese capital and technology in where it serves India’s industrial goals, but under structures that keep Indian partners in the driver’s seat and limit Chinese control. We’ve spoken about this in a previous edition of The Daily Brief.

The risk, of course, is that this balancing act gets harder as the trade war intensifies. The US has reportedly expressed discomfort with Indian overtures toward Chinese manufacturers. In fact, in the interim trade deal, India was somewhat forced to choose who we could trade with. One of the terms of the deal was that we had to significantly reduce our imports of Russian oil. One can assume that if that could happen, walking the tightrope between Washington and Beijing is going to require increasingly nimble diplomacy.

However, from a national security standpoint, it’s also difficult to assess some of these risks in many sectors. For instance, EVs have supposedly not yet opened up to Chinese investments because they can be exploited remotely by digital kill switches.

For now, India seems to have decided that the geopolitical benefit of keeping it out is higher than the discomfort of letting some of it in. Indeed, that will mean slower EV adoption, lagging capacity in key upstream parts of the solar value chain, and so on. But from a national security point of view, this looks rational.

Naturally, China’s own role in this equilibrium doesn’t escape scrutiny, either. Their willingness to use strong export curbs on rare earths makes them a fairly untrustworthy trade partner. Furthermore, the negative effects of their overcapacity on the rest of the world makes trade difficult with them.

India’s open-mindedness with respect to China might sound like gentle diplomatic language. But in the context of five years of heightened economic distrust with China, it’s a significant shift. India is recalibrating, all the while signing trade deals with everyone else in the room.

In a world that’s fracturing along trade lines, hedging might just be the best play India can make. But the window to do so is increasingly getting smaller.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

Will India get a self-driving car?

A bunch of things happened at the AI Impact Summit last week—the memes, the drama, and the billion-dollar commitments. But one comment from Sundar Pichai stuck with me. Here’s what he said:

“Taking my parents for a fully autonomous car ride in San Francisco. Seeing a Waymo ride through my 83-year-old dad’s eyes, I saw the progress in a whole new light. Of course, he said he’d be more impressed if it worked on India’s busy roads. Still working on that one day.”

It was said lightly, but this comment is actually very interesting…at least to me.

When most people hear “self-driving car,” they picture a car that can drive itself anywhere—any road, any city, any country. That version of the technology doesn’t exist yet. What actually exists today is considerably more specific. A self-driving car can only operate in cities where it has been extensively prepared in advance. Before one of these cars ever carries a passenger somewhere new, teams spend months mapping that city in exhaustive detail—every lane, every traffic light, every road marking. The car uses this map constantly while driving, comparing what its cameras and sensors see in real time against what it already knows should be there. If you take it somewhere that hasn’t been set up this way, it cannot operate.

The pursuit of this goes back about 20 years. In 2004, a US defense agency called DARPA offered a million-dollar prize to any team that could build a vehicle to complete a 142-mile course through a desert with no human driver. In that challenge, every vehicle had failed; the best one made it 7.32 miles.

DARPA ran it again in 2005, and this time, a Stanford team’s car called Stanley completed the course.

By 2007, they raised the difficulty, asking vehicles to navigate a simulated city with traffic laws, intersections, and other moving cars. Six teams finished. Google recruited the engineers who did well and launched its self-driving car project in 2009. That project became Waymo, which spun out as a separate company in 2016 and is today the most advanced commercial robotaxi operation in the world.

Waymo currently runs 2,500 cars across six American cities — San Francisco, Phoenix, Los Angeles, Austin, Atlanta, and Miami — completing around 450,000 rides per week. In 2025 alone, it did 15 million rides, three times its volume from the year before, and it just raised $16 billion in a new funding round. By the end of 2026, it plans to expand to 20 more cities, including London and Tokyo, its first markets outside America.

Tesla is also running a robotaxi pilot in Austin, but its cars still have human supervisors inside — it isn’t driverless in the same sense. China is the other major story, with Baidu’s Apollo Go operating across 22 cities and a company called Pony.ai claiming fully driverless commercial service across four of China’s largest cities simultaneously. The technology is real, it is scaling, and the money flowing into it is serious.

The difficult part of this story is that the gap between “working in specific cities” and “working everywhere” turns out to be enormous.

In 2023, a company called General Motors’s robo taxi division, Cruise, had a serious incident where one of its cars struck a pedestrian who had already been hit by another vehicle and then dragged her. California’s regulators suspended Cruise’s permits almost immediately, and within a year, GM had shut the entire operation down, writing off billions.

Beyond safety, there is also a straightforward political problem. For example, New York State recently dropped a plan to legalize robotaxis, not because anyone demonstrated the technology couldn’t handle New York roads, but because taxi unions and transit workers opposed it strongly enough that the government walked away before it was even tested.

Which is what makes Sundar Pichai’s comment so pointed. The autonomous vehicles operating today were built on certain assumptions about what roads look like—lane markings exist, traffic lights are in predictable positions, and vehicles generally stay in their lanes. The system handles deviations from this as exceptions, unusual events it has been trained to deal with occasionally.

Indian roads don’t work this way. Lanes are frequently nonexistent or treated as suggestions. Cars, two-wheelers, auto-rickshaws, pedestrians, and animals all share the same space without any fixed hierarchy between them. Vehicles make U-turns from the middle of traffic. And yet it all functions, because Indian drivers are doing something that cameras and sensors trained on American roads cannot yet do — they are constantly reading the intentions of everyone around them through eye contact, the direction of a front wheel, the pattern of a honk, tiny physical signals that form an informal shared language of the road. This is not a problem that better maps or more sensors can straightforwardly solve. It requires rethinking, at a fairly deep level, how the system models and interprets the world.

There’s another thing: India’s transport minister, Nitin Gadkari, has said clearly and repeatedly that he will not allow driverless cars in India because they would displace between 7 and 8 million driversThe technology arriving in India isn’t just an engineering problem — it is an economic and political one, and no improvement in software addresses that.

And that’s just the surface of it.

The story around self-driving cars goes much deeper—the economics of running a robotaxi fleet, the technical battle between companies like Waymo and Tesla over how these systems should actually be built, and the question of whether any of this ever becomes a real business. I haven’t even gotten into most of that because, well, that’s the nature of this show. But if we find the right trigger, I’ll do a deeper dive on this in The Daily Brief at some point. For now, that’s a wrap on this one.

See you in the next episode.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

For those of you who are new here, let me quickly set the context for what this show is about. The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

Moving up the value chain ft. EMS companies

Some of the largest EMS companies in India seem to be making the same bet. They’re spending hundreds to thousands of crores to move on from being mere assemblers and start making components or trying to move ahead in the value chain.

The logic is straightforward. If you’re just assembling phones or medical devices or industrial equipment, you’re taking components made elsewhere, putting them together, and earning a thin service margin. Dixon’s CFO laid out the economics plainly:

“Mobile business margins are closer to 3.5%. On top of that, the PLI benefit comes to about 5-6%, which is consolidated in this margin.” Take away the government incentive, and you’re left with brutally thin margins on a business that depends entirely on component prices you don’t control and customer demand you can’t predict.”

[Before I go ahead, just a quick tidbit. I went through all the concalls on Concall Monitor by Tijori. Our team uses it to track concalls as soon as they’re released, speeding up research times and even generating a quick sector report. Do check it out if you are an active investor.]

So the only way out is backward integration. Make the components yourself - displays, camera modules, printed circuit boards. Capture more of the value chain, improve margins, and reduce import dependence. Dixon’s management put it this way:

“Finally, we see almost 70-80% of our business being integrated into the component landscape by FY28, which will lead to margin expansion, even without PLI.”

That “even without PLI” bit matters. PLI is propping up margins right now, and there’s uncertainty about whether it will extend beyond FY27. But the good news for the industry is that in the recent budget, the allocation for component manufacturing was increased to ₹40,000 crores in the latest budget.

Kaynes is making an even more explicit pitch. Their CFO explained:

“This would mean a business potential of about INR15,000 crores for the group from the customers from our current investment of INR1,500 crores in the HDI PCB manufacturing operations, assuming the general thumb rule of PCB to PCBA of 10% based on our own data. From the same clients, INR1,500 crores of PCB revenues and INR13,500 crores of EMS revenues are likely.”

The logic is that if you make the PCB, you win the contract to assemble the full product. One pulls the other. It’s not just margins on the PCB; it’s the EMS business that follows.

Syrma’s MD echoed the same thing:

“In the first phase, ending December 2026, we should spend approximately 360 to 400 crores. This provides capacity for 720,000 square meters of multi-layer and 480,000 square meters of single-layer lines. We believe the additional multi-layer lines will be required sooner than initially planned due to strong customer interest.”

Customer interest is the key phrase there. It suggests customers actually want them to make these components, which means more business is on the table if they can deliver.

The capex isn’t small. Dixon is putting ₹1,100–1,200 crore into display manufacturing and scaling camera modules from 40 million to 190–200 million units annually. Kaynes is committing ₹1,400 crore for PCB manufacturing and ₹1,700–1,800 crore for OSAT, plus ₹250–300 crore on camera modules. Syrma is investing ₹360–400 crore in the first phase of its PCB plant.

But here’s the thing. All of this is supposed to pay off in FY28. Dixon’s management said it explicitly:

“Additional margins will come from our component play, which will largely play out in FY28. Both in terms of capacity expansion and deepening manufacturing, it will take six to eight months.”

Kaynes said almost the exact same thing. But the timelines keep moving. Dixon’s display JV with HKC was supposed to start trials in Q4 FY26. Now that’s pushed even further. These things take longer than expected, and that’s normal in manufacturing, but it does mean the margin improvement story keeps getting pushed out.

There’s also a genuine external shock hitting some of these companies, particularly Dixon. Memory prices have spiked because of what’s happening in AI. We wrote about this at length in one of our earlier Daily Brief piece. Dixon’s MD gave the clearest explanation:

“The potential cutback in smartphone shipments comes as brand struggle with an unfolding super cycle in the global memory sector. Top suppliers are shifting capacity for large intelligent applications, resulting in a supply squeeze for the smartphone segment.”

AI and data centers need massive amounts of memory chips. Memory manufacturers are redirecting supply away from consumer electronics to serve that demand. The result is memory prices for smartphones have shot up significantly over the last two quarters.

And here’s why that’s a problem for someone like Dixon:

“Memory has moved from being a relatively small line item to one of the most sensitive parts of the bill of materials, especially for lower-priced devices.”

If memory prices spike, companies face a choice—absorb the cost or pass it on. Dixon says the cost is a pass-through for them, but higher phone prices can hurt demand in the lower and mid-tier segments where they operate.

Strip it down, and the story is this: India’s EMS companies know assembly alone cannot carry them very far. The margins are thin, the bargaining power is limited, and the comfort of PLI will not last forever. Moving into components is not an optional upgrade; it is an attempt to change the structure of the business itself — to own more of the product, earn more per unit, and rely less on incentives.

But this shift comes with heavier capital, longer timelines, and far less room for error. Component manufacturing demands scale and steady demand to justify the investment, and both depend on forces these companies still do not fully control.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

For those of you who are new here, let me quickly set the context for what this show is about. The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

India's largest IT firm deals with ghosts old and new

With every new development in AI, the pressure on traditional Indian IT services seems to be clamping down.

In an interview with the Financial Times, K Krithivasan, the CEO of TCS, said the following:

“There will always be some segment within some small pockets within the company who will not fit in the scheme of things. AI is not going to create lay-offs by itself.”

In the context of AI itself, this statement is bold about the future. But we’re more interested in how this statement fits in the story that TCS is trying to write for itself.

See, over the past year, TCS has fired about 3% of its workforce. But none of its other peers — not Infosys, Tech Mahindra, or HCLTech — has undertaken any firing of such scale. All of them have already hinted at AI potentially changing not just their own workflows, but also how they charge clients. But any firing that has taken place so far has not been attributed to AI. In fact, a few months ago, Krithivasan explicitly said so:

“No, this is not because of AI giving some 20 percent productivity gains. We are not doing that. This is driven by where there is a skill mismatch, or, where we think that we have not been able to deploy someone.”

In other words, why TCS fired so many people is because they couldn’t deploy a lot of their existing employee base on the right projects.

What does this mean? Let’s start with who got fired. Many of the people who seem to have been fired are project managers, who generally have 10-20 years of experience. Basically, they handle client escalations, assign tickets to developers and testers, manage the bench rotation, and coordinate many different parts of a single large-scale project. This is generally not perceived to be a very technical role where domain specialization is necessary to have.

The idea that AI did not have a huge role to play in these layoffs then makes sense, too. Ideally, entry-level employees would likely have been the first to be asked to leave. Instead, those who were let go are far more senior.

In the pandemic, where large-cap IT stocks like TCS had enjoyed outperformance, over-hiring was pretty common. However, most of the hires were more experienced people from other firms, and not freshers. Rediff spoke to Kamal Karanth, a founder of a specialist staffing company called Xpheno, who said:

“The big-fat middle layer is a net outcome of the hyper-hiring phase, combined with three cycles of low to no fresher intakes. A large chunk of lateral hiring made during the demand surge now sits in the 5 to 13 year experience range -- the very layers seeing restructuring now. Ironically, these were also the layers with the highest churn during the hiring boom.”

In fact, as per Xpheno’s data, the middle layer bloat is quite real. The top 7 Indian IT firms host nearly 4.9 lakh employees who had 9-17 years’ of experience. Many of them — though not all — are the project managers we mentioned above. In comparison, there were 4.6 lakh employees with less than 5 years of experience. This less-experienced group also got paid much less.

When you think about it, what it would imply is that the number of freshers was not that far off from the number of more experienced people. That’s a very lopsided ratio, primarily a consequence of that over-hiring. This inefficiency did not necessarily need AI to be exposed, although AI has now made that clearer than ever

In fact, there were some signs that TCS had recognized this lopsidedness quite early on. In April 2024, more than a year before the mass layoffs, TCS reported a decline in its full-year employee count. This was the first time their headcount declined in nearly 2 decades. In simple terms, they had basically stopped hiring.

Moreover, TCS didn’t seem to be able to utilize its employees well enough. At various points of time, TCS would have a bench of employees who wouldn’t be staffed anywhere. This benching makes up nearly 10% of TCS’ payroll costs. This is why, at the same time as the layoff announcement last year, TCS made its bench policy more restrictive — no employee could be on bench for more than 35 days a year.

It didn’t help that, at the same time, the geography that mostly demanded TCS’ services was going through a rough economic patch. The US was suffering from weak consumer demand, and fears of a recession were high. Now, a recession was eventually avoided, but a slowdown still exists. In fact, right now, besides AI-related capex, the US doesn’t have any strong source of economic growth.

Interestingly, while TCS was lagging, mid-cap IT firms were having some of the best time in their entire history. They had increasingly begun competing for the same deals as TCS and its peers. And more importantly, by virtue of how they specialize in domains, midcap IT firms never really suffered from a bloat of project managers who didn’t have enough technical depth.

So, there is plenty of reason to believe that TCS’ layoffs weren’t due to AI as we know it today. For a long time, this layoff didn’t take place because, after all, TCS is one of the largest providers of stable jobs in India. Any layoff decision would be incredibly political and damaging not just to TCS, but to the Tata brand as a whole.

All of this was way before AI got good at writing code. And now, it can.

TCS wants to be the world’s largest AI-enabled services company. They’re undertaking an organization-wide training exercise on all things AI. They’re also going through a huge restructuring to prepare for the AI age.

What does this restructuring mean? For one, it would mean smaller teams that would further make pure project managers redundant. Instead, domain specialists would play a much bigger role. And, of course, as we covered in The Daily Brief recently, AI is already changing how Indian IT bills its clients.

So, to a very large degree, why the CEO of TCS said that AI won’t cause layoffs at scale by itself is because, so far, TCS has not made a decision to layoff people because of technological advancements. But TCS has often been late to technological advancements that aren’t AI-related. If anything, TCS bet heavily on scaling its existing model, and couldn’t course-correct quickly unlike its peers.

Now that AI is here, and the problems that TCS was plagued with 2-3 years ago still exist, how does TCS figure out how to change? That is, quite literally, a billion-dollar question for the company that kick-started India’s IT journey.

We’ve spent a lot of time at The Daily Brief writing about the less-visible parts of electronics—semiconductors, PCBs, and everything around them.

We recently released a podcast with someone who’s spent years actually making PCBs, talking through how they work, why they matter, and where India still struggles. If you’re curious, you can listen to the conversation here on YouTube.

If you prefer to listen to the audio, you can check it out on Spotify and Apple Podcasts. Now, back to The Daily Brief.

For those of you who are new here, let me quickly set the context for what this show is about. The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

Why is cement demand so strong?

Today, I’ll start by talking about cement. UltraTech’s results came out last week, and the management sounded extremely bullish about the economy. I wanted to understand why. What I found was a larger story about public infrastructure and private capex — and the gap between them.

Let me start with what UltraTech’s CFO, Atul Daga, said on the earnings call:

“Let me get to the core topic for discussion: demand. That is the most important aspect of our business. Everything else becomes secondary and falls in line.”

For the next several minutes, he proceeded to lay out an exhaustive, region-by-region catalogue of infrastructure projects across India. Punjab is spending Rs 16,000 crores on road development. Delhi Metro is announcing new corridors worth Rs 12,000 crores. And, the list went on.

One analyst joked that it sounded like a Union Budget speech.

But he wasn’t just listing projects for effect. He was making a specific argument about what these projects mean for cement demand. Elevated metros for example require 11,000 metric tons per kilometer. So, when you read that a city is adding 80 kilometers of elevated metro track, that’s potentially 880,000 tons of cement.

And he was explicit about where this demand is coming from:

“As we see the progress, government’s focus on infrastructure is translating into a robust pipeline of new projects nationwide, with several marquee investments announced across every region, translating into solid demand.”

Government’s focus on infrastructure. Remember this point. I’ll come back to it.

See, if UltraTech is seeing the demand picture from the cement side, JSW Steel is seeing it from the side of steel — a material just as important to infrastructure. And what JSW Steel described in their latest results was similar. When asked which sectors would lead this growth, management’s response echoed UltraTech’s thesis:

“We are seeing growth across sectors in the India story. This includes construction, infrastructure, and commercial real estate. We are seeing strong growth in industrial sectors and, post-GST, in consumption sectors like automotive and appliances. Another major area is renewable energy.”

JSW Steel also provided a useful piece of context that explains why October and November were weak for both cement and steel:

“Central government capex was low in October and November but is up 28% from April to November due to a strong H1 performance. The annual capex target appears to be on track.”

Government spending was front-loaded in the first half of the fiscal year, then slowed in October-November. When government projects paused, contractors stopped ordering materials, demand softened, and prices fell. But with the full-year capex target still on track, both companies expect a strong Q4 as spending catches up.

So far, this is a straightforward story. India’s two largest building materials companies are seeing strong demand, driven by government infrastructure spending. Both are confident about the future. Both are investing heavily to capture it.

However, when it comes to private capex, the confidence isn’t as strong as for infrastructure or real estate. Now, JSW Steel seems to indicate otherwise when they said:

“Conditions for private capex are increasingly conducive, supported by healthy balance sheets, the RBI’s recent rate cuts, and lower inflation.”

Now, conditions may be conducive. But is private capex actually picking up?

Before I answer that, let me explain why private capex matters a lot.

See, there’s a meaningful difference between an economy powered by government spending and one powered by private investment. When the government builds a highway, it creates demand for cement and steel and labor. But once the highway is built, that demand stops. The government has to keep announcing new projects, and keep borrowing to fund them, just to maintain the same level of activity.

Private investment works differently. When a company builds a new factory, it does so because it expects to sell products for years. That factory employs workers who earn steady wages. Those workers spend their wages on goods and services. That spending creates demand for other businesses. Those businesses then invest to meet that demand. It’s a self-reinforcing cycle that doesn’t require the government to keep borrowing.

Projects Today is a firm that tracks new investment announcements across India. Their CEO, Shashikant Hegde, highlighted his concerns on private capex to Business Standard. Private sector investment intentions have fallen for two consecutive quarters. Government investment announcements, meanwhile, rose sharply — the public sector’s share of new investment proposals climbed from under 30% to nearly 36% in just one quarter. The government is stepping in to compensate for private sector hesitancy.

He explained what’s happening:

“While domestic macro conditions remain supportive, often described as a Goldilocks phase with stable demand and manageable inflation, the willingness of private promoters to commit to large new capex may remain cautious in the near term. Global headwinds, including renewed tariff-related risks, geopolitical uncertainty and a softening export environment, are likely to keep risk appetite restrained and delay final investment decisions in private-led sectors.”

This is worth unpacking. By the textbook metrics, conditions look good — it’s a “Goldilocks” environment, not too hot, not too cold. But private companies aren’t investing anyway. They’re in wait-and-watch mode, uncertain about global trade, worried about geopolitical risks, and unconvinced that demand will be there when their new capacity comes online.

Remember JSW Steel’s comment that conditions for private capex are “increasingly conducive”? The Projects Today data suggests that conducive conditions aren’t translating into actual investment. The ability is there. The willingness isn’t.

This gap between conducive conditions and actual investment is not new. In fact, it has persisted for over a decade.

We wrote about this a few months ago.

Private investment as a share of GDP peaked at around 27% in 2007-08, during India’s pre-financial crisis boom. Even after the global crisis hit, it recovered to about 27% by 2011-12. But since then, it has been on a steady decline — sliding to around 21% by 2015-16 and hovering there ever since.

The government has tried everything to revive it. After the NPA crisis that peaked in 2018, it cleaned up bank balance sheets and recapitalized public sector banks, hoping that freely flowing credit would unlock investment. It didn’t. In 2019, it slashed corporate tax rates — one of the steepest cuts in India’s history — hoping that higher profits would translate into more investment. They didn’t. It launched Production-Linked Incentive schemes to directly subsidize investment in targeted sectors. And when none of that worked, it ramped up public capex to historic levels, hoping that government spending would eventually “crowd in” private investment.

The result?

The private sector’s share of total fixed investment has actually shrunk — from over 40% in 2015-16 to just 33% in 2023-24. Not only has overall investment as a share of GDP declined, but the private sector’s slice of that shrinking pie has gotten smaller too.

The problem, as we noted in that piece, was never liquidity or incentives. It was demand and confidence. Companies invest when they believe customers will be there to buy what they produce. If they’re not convinced that demand will materialize, no amount of tax cuts or cheap credit will get them to build new factories.

Writing in the Financial Times earlier this month, Ruchir Sharma put it bluntly:

“It is no coincidence that domestic private investment in India has also been anaemic over the past decade, held back by the same regulatory maze and overzealous bureaucracy that foreigners complain about. And boosting investment, both domestic and foreign, is the key to creating jobs and stemming the exodus.”

These aren’t problems that government infrastructure spending can solve. You can build all the highways you want, but if companies don’t trust that the rules of the game will remain stable, they won’t invest.

None of this means UltraTech and JSW Steel are wrong about what they’re seeing.

But the composition of that demand matters. When Ultratech lists dozens of government infrastructure projects and says demand is everything, he’s describing an economy where the government is doing the heavy lifting. When JSW Steel says conditions for private capex are conducive, but the data shows private investment falling, that’s telling you something about the sustainability of this growth.

The infrastructure wave is real. But for it to translate into durable, self-sustaining growth, private investment has to follow which hasn’t happened yet.

And governments, unlike private companies, eventually run out of room to borrow.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

For those of you who are new here, let me quickly set the context for what this show is about. The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

What happened at Davos?

This past week, the World Economic Forum held its annual conference at Davos. It is a gathering of some of the most powerful people in the world — heads of state, economists, billionaire entrepreneurs, and so on.

It’s the most exclusive of exclusive gatherings, full of meetings, private dinners, and, well, partying. Here, people have to be very nice to each other, no matter what. But, in such a situation, what they say to each other might have more meaning than what’s on the surface. In fact, FT came up with a pretty funny list of how the code of Davos speech works:

Let me pause for a bit so that you can read it.

However, this year, Davos didn’t really have many of these niceties. In fact, it was extremely blunt. World leaders were being brutally honest about their needs and wants — even going as far as berating other leaders.

In particular, many implicitly voiced how uncertain they were about the US being a reliable partner or ally. Of late, under the Trump administration, the US has been trying to shore up manufacturing within its own borders. For them, globalization is now a mistake.

As the US retreats from worldwide economic integration, so are other nations. They are all prioritizing indigenizing supply chains and reducing dependence on other countries, even if it’s not economically efficient to do so. It’s clear that on a global level, trust is broken.

To start with, let’s take Howard Lutnick, the US Secretary of Commerce. On a Davos panel, Lutnick went on a rant against Europe’s energy ambitions to go net-zero. Here’s what he said:

They are not in the middle — they are the flag, whichever way the wind blows. So if the wind blows one way, you’re told you should have solar; if it blows another way, you should have wind.

Why are you going to do solar and wind? Why would Europe agree to be net-zero by 2030 when they don’t make batteries? They don’t make batteries. By committing to 2030, they are effectively deciding to be subservient to China, which does. Why would you do that?

Why would the United States of America — which has oil and natural gas — try to convert entirely to electricity? China does not have oil and natural gas. Electricity and electric cars make perfect sense for them. That is practical and logical.

The point I want to make — and I want people to think about this — is that America First means the job of the government is to take care of its workers, to make sure their lives are better for it. That doesn’t mean America alone — but it does mean America first.

That is a lot to take in — even for Lutnick’s fellow panelists, who were clearly in shock with what he said. In essence, Lutnick is making a statement about what US foreign policy is today. For them, any dependence on China is not desirable, even if it could improve our chances in the battle against climate change.

But besides China, there are other details worth dissecting in his quote. For one, while reliance on China might not be nice to have, does that make an overreliance on the US for oil & gas inherently better? In 2024, oil production in the US reached an all-time high. Last year, the US swung into becoming a leader in natural gas production. The US basically controls the global markets for these commodities.

Countries that depend on procuring their oil needs from global markets know how volatile prices are and how that hurts their trade balances. So, they are trying to diversify away from that issue by scaling up renewables. With solar panels, the thinking goes, all the power is generated in-house, even if you import the panels from elsewhere. That inherently reduces the price volatility of renewables.

And the world is answering accordingly, just as solar prices keep dropping. India has jumped to becoming one of the biggest solar players in the world. In the 12 months since August 2024, the entire African continent imported 60% more solar panels than the preceding year. These imports are helping the world’s poorest nations reduce dependence on fossil fuels they do not have.

In other words, what other countries seem to think is: just because you only import the panels doesn’t make you subservient to the host country. In any case, this is better than being subservient to oil markets. Renewables give them much more control over their energy infrastructure. But oil & gas, on the other hand, is subject to massive price volatility, a finite supply, and often cartelisation by the oil-producing countries.

The other question that Lutnick’s quote raises is: Does Europe really not make any batteries of its own? Well, while it trails much behind China, Europe does have a battery manufacturing base. It is actively trying to scale the same, even attracting foreign investments to build within its borders. While not dominant, Europe has a meaningful output of batteries today.

Lutnick (and the Trump administration as a whole) believes that globalization has been bad, primarily for the US. In his view, it is the cause for why dependence on China has increased — even though it was American companies who initiated the current form of globalization, by shifting their production to low-cost labor countries like China. And now, the US is trying to reverse it, even if it means sacrificing important climate change goals.

Anyhow, Lutnick was not met without a response from the target of his quote — if not directly, then indirectly. Emmanuel Macron, the President of France, made a speech while wearing an expensive pair of shades though there was no sunlight glaring at him. But, if he didn’t score too many points on style, he probably did on substance. Here’s what he said:

China is welcome, but what we need is more Chinese foreign direct investment in Europe — particularly in key sectors — to contribute to our growth and enable technology transfer. Not merely exporting devices or products into Europe, especially when some of these do not meet the same standards or are far more heavily subsidized than those produced in Europe.

Compared to what Lutnick said, Macron (and the EU overall) is clearly far more open to taking the help of China. Battery manufacturing is one of the sectors where Chinese firms like CATL have set up shop in Europe. Macron also mentioned why he doesn’t trust the US as strongly anymore, highlighting how the US kneecapped European exports:

Competition from the United States of America through trade agreements that undermine our export interests, demand maximum concessions, and openly aim to weaken and subordinate Europe — combined with an endless accumulation of new tariffs — is fundamentally unacceptable, even more so when tariffs are used as leverage against partners.

The drama doesn’t end in the speeches we did get to see, though. Apparently, at a private post-conference dinner in Davos, Christine Lagarde — the President of the European Central Bank — abruptly walked out when Lutnick continued to make a mockery out of Europe’s economy. The geopolitical conflict of our time, it seems, presented itself in petty fights at fancy private dinners.

The allyship that the US and Europe were known for has clearly badly fractured, to the extent that Europe is willing to create new, even if uneasy, alliances with a supposed rival like China.

Nowhere is this more evident than the speech given by Mark Carney, the Prime Minister of Canada, hailed by many as the star of the Davos show. In his speech, he declares an end to the old world order, away from globalization, and into a world far more uncertain and siloed. Even if you may not fully agree with them, it is how the world seems to be moving now.

This is a rupture, not a transition. Over the past two decades, a series of crises — in finance, health, energy, and geopolitics — have laid bare the risks of extreme global integration. More recently, great powers have begun to weaponize that integration: tariffs used as leverage, financial infrastructure as a tool of coercion, and supply chains turned into vulnerabilities to be exploited.

You cannot continue living under the illusion of mutual benefit through integration when integration itself becomes the source of your subordination.

India, of course, has not stayed silent in Davos. At a panel, IMF Chief Kristina Georgieva called India a “second-tier AI power”:

So, Ashwini, with an incredible talent base, the fastest-growing major economy in the world, and an emerging AI player — but clearly in the second grouping that Cristina is talking about — we have to think carefully about the path ahead.

To which Ashwini Vaishnaw did not agree, stating that India fell in the first-tier of AI powers, citing a 5-pillar approach:

Actually, clearly in the first group. And the reason for that is simple: there are five layers in the AI architecture — the application layer, the model layer, the chip layer, the infrastructure layer, and the energy layer. We are working across all five, and making strong progress in each of them.

At the application layer, we will probably be the largest supplier of services to the world.

Now, how true is this? Well, India is certainly trying to vie for a strong position in the AI value chain. But it’s still very far from success in many of the 5 pillars mentioned. For one, our chip capabilities are very nascent, though we’re attracting foreign investments now. Secondly, we still don’t have a state-of-the-art LLM, nor do we have the infrastructure yet. All of this combined wouldn’t put us in the first-tier by any means.

However, on two pillars, the minister may have a point. India is indeed proving to be a giant in renewable energy adoption. And on the application layer, while Indian IT is currently struggling to integrate AI, it has always been a valuable provider of services to the world and may continue to be so.

Some individual Indian states also converted many Davos meetings into worthwhile investment commitments. Delegations from Telangana and Uttar Pradesh have been trying to attract global capital, even signing MoUs at Davos.

In sum, Davos threw up a lot of questions. Its very existence was threatened by multiple speeches that went against the ethos of a globalized, peaceful world, where multilateral institutions like itself and the UN have failed. We haven’t even covered what Donald Trump himself was up to at Davos — but perhaps, the speeches of Carney and Macron told us enough. Yet, countries did get what they wanted out of the gathering, signing deals and agreements.

It seems that at the world’s most famous gathering of elites, everyone got what they wanted, even at the cost of harmony.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.

For those of you who are new here, let me quickly set the context for what this show is about. The idea is that we will pick the most interesting and juiciest comments from business leaders, fund managers, and the like, and contextualize things around them. Now, some of these names might not be familiar, but trust me, they’re influential people, and what they say matters a lot because of their experience and background.

So I’ll make sure to bring a mix—some names you’ll know, some you’ll discover—and hopefully, it’ll give you a wide and useful perspective.

For all the sources mentioned in this video, don’t forget to check out our newsletter; the link is in the description.

With that out of the way, let me get started.

How different are US and Chinese AI, really?

Ren Zhengfei, the founder of Huawei, doesn’t like to be in the public eye very often. However, last year, with Huawei being front and center of China’s AI ambitions, Zhengfei has made many media appearances.

Here’s one such interesting quote that had us very intrigued:

“The US is seeking AGI and ASI to get answers to questions like ‘What is human?’ and ‘What is the future of society?’ They are trying to solve the whole problem [at once], but it takes time for [humanity] to know what the problem is. China is focusing on how to get things done [with AI] to create value and fix development issues.”

Let’s get into the nitty-gritties of Zhengfei’s quote.

For one, the US and China do seem to run technically different playbooks. Many American AI labs run on closed-source weights — there is no way for you to find out or change what your model should prioritize. American AI labs are, of course, at the absolute frontier of the AI curve, thanks to NVIDIA’s cutting-edge GPUs. If our last AI round-up on The Daily Brief didn’t make it clear, for us, Claude Code has already gone into magic territory. We don’t really care too much about sentient AGI beings anymore.

Chinese labs, meanwhile, are open-source: you can, to a large extent, open up the hood of a Chinese LLM and see how it works. They run on GPUs that are inferior to NVIDIA’s best offerings, and therefore fall a little behind the frontier. However, their ability to maximize more juice out of existing resources has been second to none.

Secondly, both countries seem to run different business playbooks, too. DeepSeek, for instance, is already integrated into Chinese cars, phones, hospitals — things regular people use. Consumer revenue is first-priority for China’s AI labs, and they’re leveraging China’s mobile-first internet economy to get there. Additionally, their progress in robotics shows how they’ve used their manufacturing prowess to integrate AI and automation in industries.

For the US, though, revenues are nowhere to be seen relative to how much is being spent in the name of AI. In The Daily Brief, we’ve covered how AI capex made up 40% of US economic growth last year, while its other industries stagnated. Their data center boom far surpasses that of China, which has been more conservative. There have indeed been attempts to make money — image generation was one of them. But so far, not too fruitful.

To a degree, Zhengfei may not be wrong. But something seems to be changing now. Recently, tech professional Grace Shao wrote a great essay on how China and the US might be converging with their AI approaches. Or, at least, if it’s not convergence, then it’s mutual respect for each other’s approaches.

In her essay, Shao talks about how many in the US are finally talking about moving towards integrating with consumer (and industrial) applications. For instance, in October last year, Sam Altman said at a Bloomberg panel that:

“We’re seeing AI move from labs to living rooms, with real-world adoption accelerating as tools become intuitive for everyday users.”

Or Jensen Huang, who seemed to hint that a lot of effort was being shifted from developing LLMs to building AI agents that would integrate within companies. Either way, the US AI ecosystem is now facing pressure to prove ROI on its huge capex, which, in Shao’s view, is shifting the narrative.

On the other hand, Chinese AI labs have recognized how important being at the frontier is. For instance, Zhilin Yang, the founder of Moonshot AI (which trains the Kimi LLMs), seems as enthusiastic about AGI as Sam Altman is. In fact, that’s why he named his company Moonshot:

“For us, it’s about exploring the unknown. Just like AGI, you usually only see the illuminated side of the moon, but the dark side remains mysterious. It’s challenging, yet full of potential. That aligns with our mission.”

The need for Chinese models to get to the frontier is also shown by their clamoring for NVIDIA’s advanced chips. Recently, China banned the import of all NVIDIA chips — perhaps, in an effort to promote its own (like Huawei’s). However, Chinese AI labs still don’t think there’s a Chinese GPU that matches NVIDIA’s best. Maybe, as a response to this demand, China is now redrafting rules to allow NVIDIA H200 chips into the country again — but with conditions.

We haven’t even touched on what is perhaps the biggest similarity, one we’ve already written about on The Daily Brief — national sentiment. Both the governments of the US and China are formulating industrial policy to set up full AI supply chains within their borders. They have made AI part of a large nation-wide push to be coordinated by them. Both are using similar policy tools to incentivize production and block each other’s efforts.

By hook or by crook, there is indeed plenty of similarity between both AI ecosystems. They don’t seem to lie in a binary anymore. We can’t say who will win the AI race and how, but increasingly, it seems to be a clash of titans with similar armors.

That’s it for this edition. Thank you for reading. Do let us know your feedback in the comments.