Kyro Ventures Research (@kyroventures) header
Kyro Ventures Research (@kyroventures) profile picture

Building a deep dive research library & intelligence on top Indian businesses, themes and sectors No recommendations. Just pure research & learning

India
Joined May 2026
Photo by Kyro Ventures Research (@kyroventures)Photo by Kyro Ventures Research (@kyroventures)Photo by Kyro Ventures Research (@kyroventures)Photo by Kyro Ventures Research (@kyroventures)Photo by Kyro Ventures Research (@kyroventures)Photo by Kyro Ventures Research (@kyroventures)Photo by Kyro Ventures Research (@kyroventures)Photo by Kyro Ventures Research (@kyroventures)Photo by Kyro Ventures Research (@kyroventures)Photo by Kyro Ventures Research (@kyroventures)
The market pays the most for gaps that close soonest. We plotted each company's P/E against how many months its supply gap has left. The months are our base-case estimate of when new capacity closes each gap. The highest multiples (PTC ~291x, MTAR ~140x, Hitachi Energy ~117x, Azad ~116x) sit on gaps we expect to close around 2028–29. The Leela has the longest runway on the map, with luxury-room demand growing 12–14% a year against 5–7% supply into FY31, and it trades at ~36x. The median P/E across 19 profitable names is ~57x (StockScans, 16 Nov), and four still trade above 100x. Q2 also showed how a high multiple can come down without the price falling. STL's P/E dropped from about 183x to about 47x in six weeks because its Q2 margin hit 32.1% and earnings caught up with the price. HFCL's went from 59x to 36x. A high multiple on a short clock is paying today for margins after the gap closes. Earnings can justify it, as STL showed, but only while the gap lasts. Educational only. Not investment advice. Explore more about us at kyroventures.com
7
521
Three textile guidances needed a big H2. Q2 split them. After Q1, our textiles book calculated what each company needed to average in Q2–Q4 to hit its own FY27 guidance. Q2 results are now out (StockScans): ✅ Indo Count: revenue ₹1,673 Cr against ~₹1,430 Cr needed (117%) ✅ Sanathan: EBITDA ₹155 Cr against ₹137–144 Cr needed (110%) ❌ SP Apparels: revenue ₹452 Cr against ~₹533 Cr needed (85%) ⏳ Welspun Living: margin 11.9% against ~13% needed in H2 Pearl Global, which our book rates as having the best guidance track record, grew revenue 37.2% to ₹1,802 Cr. This matters because the sector's leaders trade at 23–28x EV/EBITDA, against a 10-year average of about 14x. At nearly double the usual multiple, missing guidance costs a stock more than beating it earns. Q3 is now the test for SP Apparels, and H2 margins are the test for Welspun. Educational only. Not investment advice. We keep decoding new sectors, themes and businesses at kyroventures.com
10
595
One factory. Four completely different ways to get paid. 1. Assembly: high-mix industrial work earns 10–20%. Phone and appliance volume earns single digits. 2. Box build: 60% of a box's cost is the board you already made. Revenue jumps, but profit barely moves. 3. Design wins the stickiest customer. For component plants, the state pays 65–70% of the plant. 4. And margin isn't return: a 3.8%-margin model earns 42.9% on capital, while a 15.8%-margin one earns 11.0%. Full sector deep dive on kyroventures.com Educational, not investment advice. #Kyro #EMS #ElectronicsManufacturing #MakeInIndia
1
1
11
920
AI is hitting Indian IT through price before volume. Q2 FY27 constant-currency growth (year on year): Wipro +0.6% · HCLTech +2.6% · TCS +2.8% · Tech Mahindra +7.2% · Persistent +20.8% HCLTech's CEO puts AI-led price deflation at 2–3% a year, and TCS's CEO says 10–15% of productivity gains are passed back to clients. The key question is whether this is a one-time reset or a 2–3% drag every year. A one-off cut is a bad year. A recurring drag changes what the business is worth, and that is why TCS trades at about 16.9x earnings, down from 30.6x. There is an upside. AI work earns 7–10 points more margin, according to HCLTech, and it added 32 bp to HCLTech's Q2 margin. TCS's AI revenue is at a $3.1 bn annual run rate. But much of it replaces older work rather than adding to it. Our test for turning positive on large-cap IT: growth back above 5% for several quarters. That hasn't happened yet. The specialists (Persistent, Coforge, eClerx) are still compounding. Educational only. Not investment advice. We keep decoding new sectors, themes and businesses at kyroventures.com
1
1
5
545
India's grid shortage has split at 220 kV. "Grid shortage" now describes two different markets. **Below 220 kV,** transformer makers run at 60–65% utilisation and margins have halved. Shilchar went from 33% to 16%, and polymer insulators (Olectra) from 39% to 24%. **Above 220 kV** (400–765 kV and HVDC), it's still tight. The tightest part is the porcelain insulator on the transmission tower, and only 3–4 Indian makers can supply qualified product. Hindusthan Insulators' operating margin went 5.7% (Q1 FY26) → 37.8% (Q1 FY27) → 43.6% (Q2 FY27), on Q2 revenue of ₹172 Cr, up 147%. The reasons it stays tight: insulator delivery runs 18–24 months, a new 1,200 kV porcelain plant takes 3–4 years to build and type-test, and India's FY27 grid transformation target is ~40% above FY26. Before treating that margin as permanent, check whether Q2 was lumpy dispatch or a new run-rate. Polymer insulators should ease within 12–18 months. And the insulator makers carry governance flags, including related-party loans and an auditor qualification, which partly explains why they trade at low multiples. Educational only. Not investment advice. Get exclusive access to the Kyro Library at kyroventures.com
7
742
Kyro Ventures Research retweeted
India has too much power at noon… and too little at 8 PM. Solar floods the grid at midday, so much that some of it is wasted. At sunset it drops to zero, just as cities switch on. A grid battery charges on cheap midday power (~₹1/unit) and sells into the evening peak (~₹8/unit). What a state utility pays for a 2-hour battery fell from ₹10.83 lakh to ₹2.1–2.4 lakh per MW per month (2022 → 2026). India has ~9–10 GWh built, with ~15× that in the pipeline. Full thematic deep dive on kyroventures.com Educational, not investment advice. #EnergyStorage #BESS #RenewableEnergy #SolarIndia #PowerSector #IndianStockMarket
3
7
1,084
DEE Development Engineers | The ₹1,500 crore target needs a very different second half DEE’s Q2 FY27 revenue rose 8.5% YoY to ₹293 crore, but was nearly flat versus Q1. Operating EBITDA grew 33.7% YoY to ₹58.9 crore, and the margin expanded to 20.1% from 16.3%. Consolidated PAT rose 31.8% to ₹23.5 crore. Here’s what stood out in the presentation and concall: 1. Contract mix can lift revenue faster than tonnage In job work, the customer supplies the material and DEE bills for fabrication. When DEE supplies the material, more of the contract value flows through reported sales. Management expects more material-inclusive work in H2. That can lift revenue faster than physical output, but management described job work as carrying a higher percentage margin. The test is whether better factory utilisation keeps operating margins healthy as the mix changes. 2. The order book is ₹2,586 crore. The H2 hurdle is ₹913 crore. Management retained its ₹1,500 crore FY27 revenue guidance. After ₹587.4 crore in H1, that implies ₹912.6 crore in H2, about 55% more than H1. The October-end order book gives DEE work to execute, but the November call highlighted an important exception: management said the backlog includes a letter of intent for GE’s heat-recovery steam generator project. The formal purchase order is still awaited, and management expects that work to start next financial year. That is separate from GE gas-turbine piping already underway. 3. Lower debt and better cash conversion are separate tests DEE raised ₹300 crore through a preferential issue. Finance costs fell to ₹15.3 crore from ₹17.2 crore in Q1, though they were still above the ₹13.7 crore a year earlier. Meanwhile, H1 used ₹6.8 crore of operating cash despite ₹39.5 crore of PAT. Receivables and inventory absorbed cash. With DEE supplying materials on more contracts, watch whether higher billing turns into collections or ties up more working capital. 4. Two longer-term items deserve different treatment The domestic nuclear EPC framework is already announced. The proposed global nuclear JV is separate, and the presentation targets closure by FY27-end. Neither warrants counting projected nuclear revenue as booked work. Separately, the auditor qualified its limited review because management had not assessed ₹50.6 crore of Malwa Power assets for impairment. The auditor could not determine whether an adjustment was needed. No loss of that amount has been established. Kyro Read: The margin improvement is already visible. The next proof is faster dispatches, healthy margins on material-inclusive work, and collections keeping pace with billing. That will tell us more than the order-book headline alone. Disc: Not a recommendation Explore more about us at kyroventures.com
8
633
Beijing's calendar sets the price of every Indian battery. Every Indian battery part is priced against one benchmark: what the Chinese equivalent costs once it lands here. Beijing keeps moving that benchmark. 15 Jul 2025: cathode technology added to China's export-restriction list 9 Oct 2025: export licences required for cells, LFP cathode, graphite anode and the machines that make them 7 Nov 2025: those controls suspended to 10 Nov 2026 1 Apr 2026: export rebate cut from 9% to 6% 1 Sep 2026: 2% domestic battery tax (exports exempt) 9 Nov 2026: suspension extended, but only to 10 Jan 2027 1 Jan 2027: export rebate goes to zero 10 Jan 2027: suspension ends unless extended again Sep 2027: domestic tax rises to 4% This cuts both ways for India: • Assemblers using imported cells get squeezed as their input cost rises. • Indian cell and chemical makers get a wider price window. Chinese storage-cell prices have already gone from about CNY 0.28/Wh to 0.36–0.42/Wh. The catch is that India isn't ready to fill the window yet. Only about 1–1.4 GWh of the 40 GWh of cell PLI awards has been installed. Exide's CEO still sees a 5–10% cost gap against Chinese imports. And the same restrictions are slowing technology transfer: one of Reliance's Chinese partners has already pulled out. A more expensive China helps Indian cells and chemicals, and hurts anyone reselling imported cells. Educational only. Not investment advice. We keep decoding new sectors, themes and businesses at kyroventures.com
7
467
Spinners are full. Home textiles are not. Capacity utilisation across Indian textiles: Spinners: Vardhman ~100%, Sanathan Punjab ~99%, GHCL 98–99%, Nitin 98% Garments: KPR 88.9%, Gokaldas ~85% Home textiles: Welspun 77.8%, Himatsingka 70.3%, Indo Count 61.5%, Trident 47% The difference matters. Spinners can only grow by building new plants. Home-textile companies can grow by filling looms they already own. On our illustrative model, going from 65% to 80% utilisation adds 4–5 points of EBITDA margin with no new capex. Q2 FY27 shows both effects (StockScans): • Indo Count revenue +57.5% to ₹1,673 Cr, with margin up to 13.9%. Idle looms are filling. • Vardhman's margin rose to 19.6% from 17.5%. Management had said it "will moderate", and it hasn't yet. The risk is that utilisation only helps if prices hold. Last year US buyers pushed tariff costs back onto suppliers, home-textile exporters are 59–61% exposed to the US, and US payment terms have stretched to 90–120 days. Educational only. Not investment advice. Get exclusive access to the Kyro Library at kyroventures.com
11
883
India has too much power at noon… and too little at 8 PM. Solar floods the grid at midday, so much that some of it is wasted. At sunset it drops to zero, just as cities switch on. A grid battery charges on cheap midday power (~₹1/unit) and sells into the evening peak (~₹8/unit). What a state utility pays for a 2-hour battery fell from ₹10.83 lakh to ₹2.1–2.4 lakh per MW per month (2022 → 2026). India has ~9–10 GWh built, with ~15× that in the pipeline. Full thematic deep dive on kyroventures.com Educational, not investment advice. #EnergyStorage #BESS #RenewableEnergy #SolarIndia #PowerSector #IndianStockMarket
3
7
1,084
Ola opened the EV scooter market. The incumbents took it. Electric scooters now make up more than 11% of two-wheeler sales. Here's who sells them (Sep 2026, Vahan): TVS 26.5% · Bajaj 23.8% · Ather 15.0% · Hero Vida 11.9% · Ola 6.6% Ola had 51% in April 2024. The old companies didn't just take share. Their margins went up while they did it: TVS 9% to 13% · Bajaj 16% to 20% · Hero 11.5% to 14.7% (FY22 to FY26) EVs are now about 30% of Bajaj's domestic revenue. Ola went the other way. Revenue fell from ₹5,010 Cr in FY24 to ₹2,253 Cr in FY26, the store network is being cut from about 4,000 to a planned ~550, and the company has 10,644 complaints on the national consumer helpline. China played out the opposite way, with new brands winning. In India the shift was slower (EVs were 6.5% of two-wheeler sales in FY26), which gave incumbents time to use their dealers and service networks. Being first was not a moat. Distribution and after-sales were. Ather is the disruptor that is still winning, and it's because it built service. Educational only. Not investment advice. Get exclusive access to the Kyro Library at kyroventures.com
4
578
India's longest supply gaps are the world's queues. How long global buyers are waiting (Oct 2026): Airbus jets: ~10.6 years of backlog GE Vernova gas turbines: ~5.8 years Merchant ships: ~4 years US power transformers: up to 4 years Optical fibre (small buyers): up to 1 year Indian companies win this overflow as qualified second sources: TD Power (generators), Azad (turbine airfoils), PTC (aero castings), MTAR (fuel-cell hot boxes). That gives them a long clock, since TD Power had 93% of its Q1 orders from exports and raised FY27 guidance from ₹2,400 Cr to ₹2,600 Cr. But it is borrowed pricing power. GE, Siemens and Bloom are expanding their own capacity through 2028–30, and outsourced volume is the first thing a customer cuts when its own queue shrinks. Defence shows the same pattern: 155mm shell prices are 4x+ their 2022 level, and Goodluck's shells earn ~38% EBITDA against 3–5% on its pipes. Europe may make ~2 mn shells a year by 2027. Second sources get the overflow first and lose it first. Watch the customers' own capex, not just the size of the queue. Educational only. Not investment advice. We keep decoding new sectors, themes and businesses at kyroventures.com
2
21
1,577
A trading battery earns ~2x a contracted one. Until everyone builds one. There are two ways to make money from a grid battery. **Contracted:** a utility pays a fixed monthly rent. At the 2026 tariff of about ₹2.3 lakh per MW a month, a two-hour battery earns about ₹13.8 lakh per MWh a year. **Merchant:** the battery buys power at midday for about ₹1 a unit and sells it in the evening for about ₹8. After 15% losses and one cycle a day, that is about ₹25 lakh per MWh a year. Adani Green's own rule of thumb is ₹25–30 lakh. That gap explains why ACME Solar reported an 80% EBITDA margin on ₹550 Cr of H1 FY27 battery revenue. But every battery added narrows the spread it trades on. In Australia, battery price spreads fell 85% in one year as the fleet passed 9 GW. Our bear case has India's spreads halving, which would take merchant earnings below what a contract pays. Contracted tariffs have already been through their own cycle: ₹10.83 lakh per MW a month in 2022, a low of ₹1.48 lakh in 2025, and ₹2.1–2.4 lakh today. IEEFA–JMK judged about three-quarters of the two-hour capacity awarded in 2025 to be "risky". How we read it: value contracted megawatts at face value, and apply a haircut to merchant megawatts. Educational only. Not investment advice. Get exclusive access to the Kyro Library at kyroventures.com
1
4
1,185
Gufic: filling a factory is not the same as building a better business. Q2 standalone revenue rose 25.2% to ₹288.4 crore and PAT grew 57.3% to ₹23.5 crore. But EBITDA margin slipped from 18.10% in Q1 to 17.23%. The more interesting question is what goes into Indore, and when it starts paying for itself. 1. Breakeven has layers. Management says Indore has crossed EBITDA breakeven; covering interest and depreciation is targeted for Q4 FY27. Higher imported-API and freight costs squeezed this quarter’s margin. 2. Approval is not immediate revenue. In August, management hoped to hear from Europe within one or two months. Certification remains pending. The November call explains the delay: smoke-study modifications while production continued. Management targets European supplies around September 2027, after certification and site addition; US timing also depends on partners. 3. The product mix matters more than a utilisation headline. Pranav Choksi: “we just don’t want to clutter the facility with me-too products”. In August, he explained why multiplying utilisation into revenue is misleading: different products carry vastly different revenue per vial. The planned mix includes long-acting and liposomal injectables, not simply more low-value volume. 4. Better factory economics will also fund the next pipeline. Management plans to reinvest surplus into R&D. Do not assume every improvement in plant profitability flows straight into reported margins. Inventory is another checkpoint: management targets roughly ₹325–330 crore by March, versus about ₹370 crore now. The opportunity is better products earning more from existing capacity. The execution questions are approvals, commercial conversion and cash release. Disc: Not a recommendation Explore more about us at kyroventures.com
1
4
578
China gave up 18 points. India picked up two. Share of US knitted-apparel imports, CY16 to CY25: China −18 pts (33% to 15%) · Vietnam +7 · Cambodia +4 · Bangladesh +3 · India +2 (4% to 6%) Of the share China lost, India took about one-ninth. World textile and apparel trade grows only about 1% a year (US$839 bn in CY19 to US$904 bn in CY25), so every point of share has to be taken from someone. 2026 made it worse. US apparel imports from India fell 25.7% in Jan–Jul while imports from Cambodia rose 10.9%. India's US tariff is now 10%, the same as Bangladesh and 2.5 points below Vietnam and China. Gokaldas's MD doesn't see that as "a remarkable advantage", and a new US sanctions law allows tariffs of up to 100%. China+1 was never automatic. Tariff parity alone hasn't moved share; India needs an edge on cost, scale or trade deals to take it. Educational only. Not investment advice. Explore more about us at kyroventures.com
3
441
INDO-MIM: the value is in the complexity, not just the sector label. H1 FY27 revenue rose 11.8% to ₹2,491 crore. Underneath that headline, aerospace revenue grew 31.5% to ₹224 crore, while the machining business grew 55.4% to ₹188 crore. These are different views of the business, end market versus manufacturing process, not numbers to add together. The important point from the call: a larger medical or aerospace mix does not automatically mean higher margins. The complexity, weight and finishing requirements of individual components matter. As CFO Bala put it: “It depends on the complexity in that particular component.” That also explains the investment priorities. Management says primary moulding and casting capacity is available; spending is directed towards secondary operations and finishing for developing programs. The opportunity is to do more value-added work on the part, not simply install more primary capacity. There are two different growth clocks here. The existing business is already delivering. Aerospace, machining and investment casting are growing. Investment-casting revenue reached ₹477 crore, up 36%. But electronic component manufacturing fell 40.5% to ₹268 crore, so the growth is not uniform. The newer programs are still being developed. Humanoid and satellite opportunities are active, but management has not quantified their revenue contribution. Customer forecasts are not INDO-MIM orders. The earnings comparison needs care too. Q2 PAT almost doubled to ₹215 crore, but the prior-year quarter included a ₹93 crore exceptional charge. PBT before exceptional items grew about 28% year on year and declined about 5% sequentially. That is the cleaner comparison. And execution has a working-capital cost: the presentation reports inventory days of 441 versus 340 a year earlier. Management expects roughly 75 days of reduction by year-end under its current calculation, not an inventory level of 75 days. Our read: judge the opportunity by programs entering production, the economics of the components shipped, and inventory converting into cash. Sector excitement alone cannot tell us those things. Disc: Not a recommendation Explore more about us at kyroventures.com
1
3
14
844
DMart's profit rose 21%. Its P/E fell from 116x to ~72x. Disruption hits an incumbent in a set order: 1. Growth slows. DMart's same-store growth fell from 7.1% to 5.5% (Q1 FY26 to Q1 FY27). 2. Pricing power weakens. Not yet: gross margin has held around 12.1% for three years. 3. The multiple falls. Trailing P/E went from 116x (Mar 2024) to ~72x today (StockScans). 4. Earnings fall. Not yet: profit rose from ₹2,536 Cr (FY24) to ₹3,057 Cr (TTM). The market priced in quick commerce long before the P&L showed anything. If you look for disruption in earnings, you'll see it last. TCS followed the same pattern: its P/E went from 30.6x to about 16.9x while its FY26 operating margin of 25.0% was a four-year high. There's an honest counterpoint. DMart's Q2 FY27 same-store growth bounced to 9.5%, from 6.8% a year earlier, and it opened a record 85 stores in FY26. One quarter doesn't reverse a trend, but two more would. Meanwhile Blinkit went from 639 dark stores to 2,443 in eight quarters. In a disruption, watch the multiple, not the P&L. Educational only. Not investment advice. Explore more about us at kyroventures.com
4
441
Kyro Ventures Research retweeted
Tonbo Imaging owns 65%+ of India's thermal-sight installed base, earns 60–65% gross margins on exports, and grew platforms from 3% to 43% of revenue in two years (82% of today's order book). It also has 517 working-capital days and negative operating cash flow for 2.5 years. Defence deep-tech compounder, or a lumpy contractor with a call option? Read about it here ↓ kyroventures.substack.com/p/… Disc: Not a recommendation
5
17
1,487
103 GWh of battery assembly announced. 45–50 GWh to install. Indian listed companies have announced about 103 GWh a year of battery container and pack capacity: Reliance 40 · GPIL 20 · Pace Digitek 10 · Vikram Solar 7.5 · Premier (with RCT) 6 · Waaree 5.15 · JSW Energy 5 · Amara Raja 5 · Solarworld 3.4 · Prostarm 1.2 Crisil expects India to commission 45–50 GWh over FY27 and FY28 combined. The reason so many companies are building this is that it's cheap. A container plant costs ₹20–60 Cr per GWh. A cell plant costs about ₹460–500 Cr per GWh, roughly 8–10x more. Amara Raja shows what that means in practice. Its ₹280 Cr assembly plant is guided to ₹2,700–2,800 Cr of revenue, so it pays for itself in under two years. But the guided margin is only 6–7%. Across the sector, guided assembly margins range from 6% to 15%. Exide's CEO put it bluntly: people are busy drilling and welding freight containers, and if that is localisation, Exide is not in that game. There are offsets. Exports, smaller commercial-and-industrial cabinets and building containers in-house can all help. But a business that is cheap to enter is also easy to crowd. The value in batteries sits in cells and chemistry. Assembly will end up priced like assembly. Educational only. Not investment advice. Explore more about us at kyroventures.com
903