chanakya https://think.shenthar.me/chanakya/ Thu, 12 Nov 2026 14:32:05 +0000 The fab is eighteen months late. The window isn't waiting. https://think.shenthar.me/chanakya/the-fab-is-eighteen-months-late <![CDATA[In September, at SEMICON India, Ashwini Vaishnaw told an audience he keeps a nail empty on the wall of his office, waiting for the first wafer out of Dholera. Good line. It also tells you where things stand four months after I called this programme serious. Four months is not long. In fab time it is an epoch. The surface reading is encouraging. Applied Materials has committed $5 billion over a decade, five plants are in commercial production, and the mission has gone from ₹76,000 crore to ₹1,27,500 crore. On any reasonable accounting that is a good quarter. None of it is a wafer. --- Three slips, not one The programme slipped three ways. Only one is the date everyone discusses. The timeline. At the foundation stone on 13 March 2024, Vaishnaw said the first chip would come out in December 2026, at 28, 50 and 55 nanometres. The Cabinet's Semicon 2.0 release, PIB PRID 2284796 of 15 July 2026, says "With the first fab scheduled to be commissioned in 2028." Eighteen months, and no public explanation from the government or from Tata. A third date is unreconciled: on 3 December 2025 MeitY told the Lok Sabha that operations were expected "by FY 2029-30." The node. On 17 July Bloomberg reported Dholera will start mainly on 90 nanometres, sourcing people who would not be named. A Tata Electronics spokesperson confirmed the fab spans 28 to 110 nanometres and that the plan "was always to start with 55 nm and 90 nm, followed by 28 nm." Neither company confirmed 90 nanometres; both declined to contradict it. Two of the three nodes announced in 2024 are now a later phase, and Tata Sons' FY2024-25 report put 28nm at the start of that journey while conceding the cutting edge "will take several years of serious work." The counter-argument deserves a hearing: 90 and 55 serve automotive and embedded MCUs, where yields are quickest. MeitY Secretary S. Krishnan argued exactly that at Yashobhoomi on 24 September, reported the next day — India has chosen to subsidise the mature end rather than chase 2nm, because advanced manufacturing is "simply too expensive." He was describing a band running from 28 to 180 nanometres, wider than the 28 to 110 the fab itself is licensed for. Rational. But it is not the position the government sold in June. The measurement. No Tata, MeitY, PIB or Gujarat publication has ever given a construction completion percentage. The figure in circulation — roughly 50 per cent civil construction — traces to one unnamed-source paragraph in Tom's Hardware in May. When Gujarat's chief minister walked the site on 16 October, the state release carried no number and no milestone. The absence is the datapoint. There is a fourth slip, and it ran the other way. The Cabinet approved Dholera at "up to ₹91,000 crore." The financial support agreement was signed at ₹91,526 crore. The project exceeded its own ceiling before a single tool was bought, and nobody has explained why. --- It was never a joint venture Dholera is not a Tata-PSMC joint venture. The chain is Tata Sons, wholly owning Tata Electronics, wholly owning Tata Semiconductor Manufacturing. PSMC holds no disclosed equity. Its role is a technology licensor under a definitive agreement of 26 September 2024. India bought know-how. It did not buy a partner, and PSMC carries no project risk. The financing tells you how lenders price Indian fab risk. Tata Semiconductor Manufacturing signed a ₹6,835 crore facility with five foreign banks through GIFT City branches on 5 February 2026, reported by Mint on 23 March off the MCA filings. The 163.5-acre Gujarat plot leased to Tata Electronics was valued at zero, and pledged, against a company into which Tata Electronics had put about ₹690 crore since its 2023 incorporation. A Tata executive told Mint: "The banks' credit decision is based on Tata's global creditworthiness, and not the land valuations." That is a lender pricing a greenfield Indian fab as parent-guaranteed corporate risk. The better evidence that tools are coming is indirect. Kelington announced on 14 September a provisional $452 million letter of award, as main contractor, to hook up "up to 886 tools" across both modules of "a semiconductor wafer fabrication facility in Gujarat" — client unnamed, thirty months' work, February 2029 target. The May 2026 ASML agreement, by contrast, is only a memorandum, and ASML India was incorporated in Ahmedabad in June. --- Committed is not disbursed ISM 1.0 breaks down as ₹65,000 crore for chip production, ₹10,000 crore for the SCL Mohali laboratory and ₹1,000 crore for design — though Vaishnaw conceded on 28 November 2025 that the SCL line "is a fungible amount." In August 2025 MeitY Secretary S. Krishnan said: "I think we have committed close to ₹62,900 crore or so already." On 1 September he went further: the government has "nearly exhausted" the ₹76,000 crore. Here is the number, and it is the government's own. The Expenditure Budget publishes scheme-wise actuals: ₹681 crore in FY2023-24, ₹638 crore in FY2024-25, ₹2,886 crore in the first half of FY2025-26. About ₹4,200 crore against a ₹76,000 crore headline, two years in. And the sub-scheme line that matters most — Semiconductor Fabs — reads ₹0.00 crore in FY2023-24, ₹0.00 crore in FY2024-25, and ₹0.33 crore to 31 December 2025. Two years into a ₹76,000 crore scheme, the government's own accounting says nothing has been disbursed to the fab. Vaishnaw told the Standing Committee why: legal agreements must be executed before the fiscal support can be released. That is an answer. Whether it is an explanation is a separate question. Even that is a budget line, not a project. Asked in the Lok Sabha on 5 August 2026 for "the funds sanctioned, released and utilised, project-wise," the government replied with a project count and an operating-plant count. Dr Shashi Tharoor's July question on the design incentive scheme got a commitment figure and nothing else. Beside it, PLI publishes its own: fourteen sectors, 892 applications, ₹2.40 lakh crore invested, ₹15.2 lakh crore of exports, ₹35,354 crore of incentive actually disbursed as of 31 March 2026. And PLI's employment series runs backwards — 12 lakh at March 2025, 14.39 lakh at December 2025, back to 14.15 lakh at March 2026 — while exports went from ₹6.5 lakh crore to ₹15.2 lakh crore over the same year. A series that rises and then falls inside twelve months is not a measurement. Scheme rules require claimed sales to be verified against GST returns, income-tax returns and bank realisation certificates, but no published reconciliation ties them to HSN-coded customs data, and a NITI Aayog database commissioned in 2022 to do exactly that was never announced as finished. The 14.15 lakh is thinner than it looks: the indirect half of it, 5.66 lakh, is self-reported and exists for only three of the fourteen sectors. --- Bigger scope, lower price Semicon 2.0 is the biggest thing that has happened, and it is good news for the wrong reason to celebrate. The Cabinet approved ₹1,27,500 crore on 15 July, sixty-eight per cent above ISM 1.0, notified in the gazette on 31 August as CG-DL-E-31082026-275860. Alongside it went a ₹62,500 crore phone scheme replacing PLI-LSEM, and an increase in the electronics components scheme from ₹22,919 crore to ₹40,000 crore, with 75 applications worth ₹61,671 crore approved as of 30 March. The re-scoping is right. ISM 1.0 bought assets; ISM 2.0 goes after the supply chain into them — gases, chemicals, equipment. India imports more than 90 per cent of its chipmaking equipment and 85 to 90 per cent of its specialty chemicals and electronic gases, on Equirus Securities' estimate. Vaishnaw declined to put a number on it at the Bloomberg New Economy Forum on 14 October, saying only that materials are "the biggest dependencies." Now the rates, from the notification itself rather than the trade press. Silicon fabs: 40 per cent of eligible capital expenditure, down from ISM 1.0's 50 per cent of project cost — the base changed as well as the rate. Advanced packaging, display and compound fabs, 35. Legacy packaging, 25. Equipment and materials, 30. R&D and talent, up to 75 — though those two are percentages of project cost including opex, not of the eligible capital expenditure the rest of the table sits on. The design incentive, which most of the coverage leaves out, is 9 per cent of net sales. The scope widened and the price of a silicon fab fell. The subsidy now leans hardest toward what India has least of, which is defensible, and lightest toward what it has most of, which is a confession. And it does not touch Dholera, whose support was fixed under ISM 1.0: Bloomberg reported on 17 July that the new fund "will apply only to future, incremental investments and will not retroactively offset spending already committed." The project that costs India most is the one the new scheme helps least. By August the government had three plants producing — Micron from February, Kaynes from March, CG Semi from July. On 17 September PIB inaugurated CDIL's Mohali discrete line and Suchi Semicon's Surat packaging line in a single ceremony and put the total at five. Suchi Semicon is the execution story nobody wrote up. Approved 5 May 2026, in commercial production 17 September 2026: four and a half months from Cabinet approval to a working line, on an ₹868 crore plant in Surat. That is the strongest execution fact in the whole file. It is also a packaging line — as are all five plants now producing. Nothing India actually fabricates yet. --- Sanand is not what I said it was I gave Micron's Sanand plant more credit in June than the evidence supports, and the company's own press release has been quietly correcting the record since. On 28 February 2026, at the inauguration, Micron wrote: "The expansion of conventional assembly and test operations in India complements Micron's planned development of advanced manufacturing and packaging capabilities in the United States and strengthens the company's global assembly and test network." The second half of that sentence is the part that gets cut. An unnamed Indian government official told the Economic Times in April that the plant "is not planned to host HBM manufacturing yet." The "10 per cent" figure is not Micron's. Where it can be pinned down it is attached to at least two different denominators — Vaishnaw's "global memory production," and Business Today's "global packaging and test output," the only one of them a single-floor assembly and test plant could plausibly deliver — and Micron's own release contains no percentage at all. On ramp: tens of millions of chips this year, hundreds of millions next, said Mehrotra again on 17 September. A company at full ramp does not re-announce its ramp. Nor does it carry one 2027 number. It carries two, and neither is a filing. A company that will not put a plant in its revenue line will not put it in its guidance either. In July, unnamed people aware of the developments told the Economic Times that Micron had handed Phase 2 to L&T, displacing Tata Projects after delays in Phase 1; Micron has confirmed neither the award nor the reason. Nor is that the only slip on the record: Phase 1, set for 2024, was by February 2025 still targeting early 2025 completion. And in Micron's record FQ4-26 results on 30 September, $54.23 billion in a quarter, Sanand does not appear — not in the release, and not in any document the company has ever filed with the SEC. Dell is the only customer Micron has named anywhere. None of that makes Sanand a bad plant. It is probably the best-run asset in the programme: an IDM drawing wafers from its own global network and putting them through test in India, staffed by 700-odd fresh graduates who trained in Malaysia and Singapore first. Not an argument against Micron — an argument against importing their forward guidance as an Indian capability claim. --- The binding constraint is not money If you want to know what will actually stop an Indian fab, it will not be capital and it will not be the exchange rate. It will be the fact that nobody has ever priced one. On 25 March 2026 the Gujarat Electricity Regulatory Commission, in its tariff order for Torrent Power's Dholera distribution licence, declined to determine a tariff for the Dholera special industrial region at all, and continued Dahej's rate card unchanged for FY 2026-27 under the saving provision in its own regulations. Dholera has been on somebody else's price since 2022. The card is not generous: the high-tension industrial energy charge runs ₹2.90 to ₹3.20 depending on billing demand, with fuel pass-through on top and time-of-day penalties, which GERC's own cross-subsidy table puts at an all-in ₹6.26 to ₹6.31 a unit. The semiconductor incentives run on different clocks again — a ₹2 a unit tariff subsidy for ten years from the start of commercial production under the Gujarat Semiconductor Policy, and a duty exemption that runs five years in a special investment region and ten only inside a notified SEZ, which Dholera is not. Nationally, Power Secretary Pankaj Agarwal told EDICON in New Delhi on 21 January that industry pays about $105 a megawatt-hour, of which about ten dollars is cross-subsidy. Read that order again. Gujarat has spent a decade assembling every incentive a semiconductor cluster could ask for, and when a regulator finally had to price a fab, it reached for the saving provision and borrowed another licensee's rate card, because the Dholera licensee has no cost history to justify one of its own. A regulated tariff is built from audited load, sanctioned demand and cost of supply, and Dholera has none of the three. That is harder to argue with than any reliability average: it is a regulator saying out loud that it cannot yet tell what a fab costs. I would not have reached for a reliability number even if one were available. In India the instrument is patchy and mostly self-reported: of 72 distribution companies, 36 carry no outage data on their websites at all, an audit by The Leap Journal found in September 2025. What does exist is REC's Consumer Service Rating, which graded 66 licensees last year. On industrial supply, Gujarat's DGVCL came in at 10.6 interruptions and Assam's APDCL at 13.9, which is respectable. It is also a frequency count with no duration attached, the definition is not published, and six of REC's fifteen special-category utilities reported no industrial figure at all. Assam's APDCL ranks first of those fifteen on REC's composite score, not on this measure. The worst numbers in the dataset belong to Punjab and Karnataka, not Bihar. Water is better sourced, though still second-hand. T Kannan reported in The Statesman on 29 August that Dholera's own environmental-appraisal file carries a total requirement of about 17,549 kilolitres a day, 6,571 of them fresh and the rest recycled, piped from the Narmada the better part of a hundred kilometres down a canal built for farms and towns — a river he describes as already fully spoken for. He cites no clearance number and no allocation order, and I have not been able to read the clearance file. If the numbers are right, they point at something structural: project-by-project clearance never asks what happens to the basin when the next nine plants answer the same question the same way. India has a water application form, not a water budget. None of the rest is failure either. The scheme supports projects for up to six years, NITI Aayog's roadmap runs ten, and Vaishnaw relays the Prime Minister's line that this is a twenty-year journey. Asked about timing at the Bloomberg New Economy Forum on 14 October, he offered three horizons rather than a deadline. On ITIF's February 2024 count, drawn from a handful of interviews, roughly one per cent of India's 600,000 annual electronics-stream engineering graduates can walk onto a factory floor within three months of graduating. The regulator's own data says the pool is a quarter of that — AICTE counted 160,450 electronics and communications seats filled in 2024-25. CSIS repeated the one per cent in September 2025, citing a source that does not contain it. Slow. On this evidence, not stupid. --- Washington, which did nothing Let me take the geopolitical claim by claim, because this is where the June framing was most wrong. Count them. All twenty-three Bureau of Industry and Security items in the Federal Register between 1 June and 18 October were read, and not one of them tightens semiconductor, advanced-computing, SME, HBM or advanced-packaging export control. The only advanced-computing rule in that window is the one two paragraphs below, and it loosens rather than tightens. The last tightening was in January. Section 232 semiconductor tariffs are still Phase 1 only: a 25 per cent duty, effective 15 January, on what the proclamation calls a very narrow category, described in an annex the Federal Register publishes only as an image. The H200 and MI325X labels come from January's export-control rule, not from the tariff. Data centres, R&D, startups, repairs, consumer and civil-industrial use and the public sector are all carved out to zero. One item touches a chip input — a September stockpiling prohibition under Proclamation 11052 — and it is a Section 232 import measure, not an export control. India's status is unremarkable, which is a different thing from unfavourable. India sits in Country Group B, outside the enforcement perimeter, so top-tier accelerators ship with no licence required at all — NLR, which is a stronger and different thing from a licence exception. It is also already in Country Group A:5, and therefore already has Strategic Trade Authorization eligibility under 740.20(c)(1)(i). None of that was granted to India. It follows from where India is listed. The July rule was different in kind. On 10 July, effective two days before it appeared in print, BIS published at 91 FR 43034, "Enhanced Favorable Treatment for the United Arab Emirates Under the Export Administration Regulations." The UAE came out of Country Groups D:3 and D:4, joined Country Group A:5 — the bracket India already sits in — and gained a new entity whitelist, Supplement No. 8 to Part 740, covering licence-free access to advanced computing items for G42, Core42, Amazon.com, Apple, Google, Meta, Microsoft, OpenAI, Oracle and xAI. MGX is not on that list, only flagged for favourable review, and the G42 carve-out expires on 6 April 2027 unless they become American companies. This is what a carve-out made for one country looks like. Nothing in it was made for India, and nothing in it needed to be. And Section 232 is uncomfortable. Proclamation 11002 of 14 January set out a two-phase plan, and the clocks are what they are: a status report on the negotiations due to the President by 14 April, a data-centre market update due by 1 July, and a 180-day statutory fallback whose date, 13 July, has passed. Only Phase 1 was ever operative; Phase 2 is permissive and nothing has triggered it. What has been implemented is thinner than the headlines suggest. Taiwan's investment agreement took effect in May, and what it capped was 232 duties on Taiwanese automobile parts, timber and wood derivatives. The semiconductors were left out of it. India risks becoming the residual jurisdiction: the one that pays the tariff without receiving the carve-out. --- What actually got harder Not the chips. The Strait. Between 6 and 9 August, the Guangdong Maritime Safety Administration announced traffic control on northbound vessels entering the Taiwan Strait through its southern approach, framed around a typhoon. Taiwan's Coast Guard Administration told shipping to disregard it. ISW's reading was that Beijing may have used typhoon cover to test expanding its law-enforcement jurisdiction over a waterway it does not control. Cheap. Deniable. Legally untested. On 31 August, Taiwan's Ministry of National Defence told the Legislative Yuan that the PLA is stepping up blockade preparations — the annual military-power report describes the intent as building the capacity to construct a blockade, sever Taiwan's sea lanes and encircle it. The ministry's own answer that week was thinner than the question. The anti-blockade work is visible elsewhere: a first-ever joint navy and coast guard escort operation on 12 August, shepherding strategic cargo into Hualien and Suao. EIU's own verdict, published back in February, is that a direct purposeful invasion is unlikely and that a blockade — or even an invasion — cannot be fully ruled out. And it has not stopped. Between 8 and 17 October, six maritime administrations closed large parts of the Bohai, Yellow, East and South China Seas, including a live-fire no-sail zone off the southern Yellow Sea that runs to 30 November. On 14 October the Southern Theater Command put four landing ships through a simultaneous live-fire salvo and the Liaoning carrier group moved into the western Pacific. The timing is not subtle: it sits either side of Keen Sword, the US–Japan exercise that runs 19 to 29 October and puts systems on Yonaguni, 110 kilometres from Taiwan. What changed is that the cost of preparing a blockade has fallen to the point where it can be rehearsed under cover of a weather advisory. Every leg of this strategy is an external dependency. Coherent. Not a hedge. --- The window Eight-inch utilisation across the top ten foundries recovered to 88 per cent in 2026 and should reach 90 in the second half; prices rose 5 to 15 per cent between the first and second quarters, and a third round could extend into 2027. The tight window is real. It is why India's bet is rational. It also has an end date. Start with the mechanism, because it is usually stated backwards. TrendForce's thesis is not that TSMC and Samsung are releasing mature-node capacity. It is the reverse: they have been cutting eight-inch lines since the second half of 2025, and TSMC is taking roughly 50,000 wafers a month — 15 to 20 per cent of Fab 14's twelve-inch mature capacity — out by 2028. TrendForce's May reading was that the full impact becomes more pronounced after the second half of 2027. The squeeze gets worse, not better. Which means new mature-node capacity comes from everyone who is not TSMC or Samsung. That is Dholera's market. Counterpoint had 40 to 90 nanometre utilisation stuck near 80 per cent in January, and TrendForce noted in April that customers were negotiating to postpone planned price increases for the second half of 2026. A new entrant arriving in 2028 is the marginal wafer in a market that already has spare wafers, in a price cycle whose customers are asking for relief in the quarter it is supposed to extend. Set that against the decision MeitY's secretary made out loud on 24 September. India has chosen to subsidise the legacy end of the market because the leading edge is "simply too expensive." That is defensible. It is also a decision about which glut to walk into, and the glut is the Chinese one. The Semiconductor Industry Association, working from SEMI's fab forecast, put China's share of global mature-node capacity at nearly double — 19 to 33 per cent — between 2015 and 2023, measured on installed nameplate capacity at 28 nanometres and above. CSIS, separately, could not find an expert consensus that this becomes an oversupply by 2030. Note the trap in the number: TrendForce also has China's mature-process share reaching 33 per cent, in 2027. Dholera's first wafers are scheduled for 2028. I am not forecasting a collapse. I am pointing at something harder to argue with: a fab arriving in the last year of the price cycle it was planned against, into a market where the two best-run foundries on earth are deliberately shrinking and the ones adding capacity are subsidised to add it, while their customers are already trying to talk the price down. India does not lose by being late to a fab. India loses by being late to a fab that arrives into the end of a price cycle. --- What I think None of this is an argument against Dholera. A 2028 fab at 90 nanometres in India is worth more than no fab, and a fab that trains process engineers and drags a supplier base into existence around itself is worth more than the ₹91,526 crore on the invoice. The state has done several things right. It let Micron build. And it has built an OSAT cluster in Sanand that is real rather than a press release. Any honest assessment starts there. What I don't believe is that the programme is being managed to a schedule that survives contact with the market. Taking the claims one at a time. The timeline moved eighteen months and the government is carrying three live dates it has not squared. The node dropped two generations. The completion figure traces to one unnamed paragraph. No anchor customer has been named. By its own accounting the scheme has disbursed nothing to the fab. The capital went elsewhere — SEMI's 2026 construction-start ranking has no India row, and Reuters called India a late entrant. And the strategic instrument everyone was watching has done nothing for India, while the real escalation has moved to a mechanism nobody can hedge against. The fair summary is this. India now has a genuinely good packaging industry, an underrated OSAT cluster, a real design workforce, a properly large subsidy programme, and one fab that will start producing two years from now at a node two generations behind the promise, into the part of the cycle where the industry stops paying for capacity. That is not a failure. It is an expensive lesson with a very long fuse, and the fuse is already lit.]]> In September, at SEMICON India, Ashwini Vaishnaw told an audience he keeps a nail empty on the wall of his office, waiting for the first wafer out of Dholera. Good line. It also tells you where things stand four months after I called this programme serious.

Four months is not long. In fab time it is an epoch.

The surface reading is encouraging. Applied Materials has committed $5 billion over a decade, five plants are in commercial production, and the mission has gone from ₹76,000 crore to ₹1,27,500 crore. On any reasonable accounting that is a good quarter.

None of it is a wafer.

Three slips, not one

The programme slipped three ways. Only one is the date everyone discusses.

The timeline. At the foundation stone on 13 March 2024, Vaishnaw said the first chip would come out in December 2026, at 28, 50 and 55 nanometres. The Cabinet's Semicon 2.0 release, PIB PRID 2284796 of 15 July 2026, says “With the first fab scheduled to be commissioned in 2028.” Eighteen months, and no public explanation from the government or from Tata. A third date is unreconciled: on 3 December 2025 MeitY told the Lok Sabha that operations were expected “by FY 2029-30.”

The node. On 17 July Bloomberg reported Dholera will start mainly on 90 nanometres, sourcing people who would not be named. A Tata Electronics spokesperson confirmed the fab spans 28 to 110 nanometres and that the plan “was always to start with 55 nm and 90 nm, followed by 28 nm.” Neither company confirmed 90 nanometres; both declined to contradict it. Two of the three nodes announced in 2024 are now a later phase, and Tata Sons' FY2024-25 report put 28nm at the start of that journey while conceding the cutting edge “will take several years of serious work.”

The counter-argument deserves a hearing: 90 and 55 serve automotive and embedded MCUs, where yields are quickest. MeitY Secretary S. Krishnan argued exactly that at Yashobhoomi on 24 September, reported the next day — India has chosen to subsidise the mature end rather than chase 2nm, because advanced manufacturing is “simply too expensive.” He was describing a band running from 28 to 180 nanometres, wider than the 28 to 110 the fab itself is licensed for. Rational. But it is not the position the government sold in June.

The measurement. No Tata, MeitY, PIB or Gujarat publication has ever given a construction completion percentage. The figure in circulation — roughly 50 per cent civil construction — traces to one unnamed-source paragraph in Tom's Hardware in May. When Gujarat's chief minister walked the site on 16 October, the state release carried no number and no milestone. The absence is the datapoint.

There is a fourth slip, and it ran the other way. The Cabinet approved Dholera at “up to ₹91,000 crore.” The financial support agreement was signed at ₹91,526 crore. The project exceeded its own ceiling before a single tool was bought, and nobody has explained why.

It was never a joint venture

Dholera is not a Tata-PSMC joint venture. The chain is Tata Sons, wholly owning Tata Electronics, wholly owning Tata Semiconductor Manufacturing. PSMC holds no disclosed equity. Its role is a technology licensor under a definitive agreement of 26 September 2024.

India bought know-how. It did not buy a partner, and PSMC carries no project risk.

The financing tells you how lenders price Indian fab risk. Tata Semiconductor Manufacturing signed a ₹6,835 crore facility with five foreign banks through GIFT City branches on 5 February 2026, reported by Mint on 23 March off the MCA filings. The 163.5-acre Gujarat plot leased to Tata Electronics was valued at zero, and pledged, against a company into which Tata Electronics had put about ₹690 crore since its 2023 incorporation. A Tata executive told Mint: “The banks' credit decision is based on Tata's global creditworthiness, and not the land valuations.” That is a lender pricing a greenfield Indian fab as parent-guaranteed corporate risk.

The better evidence that tools are coming is indirect. Kelington announced on 14 September a provisional $452 million letter of award, as main contractor, to hook up “up to 886 tools” across both modules of “a semiconductor wafer fabrication facility in Gujarat” — client unnamed, thirty months' work, February 2029 target. The May 2026 ASML agreement, by contrast, is only a memorandum, and ASML India was incorporated in Ahmedabad in June.

Committed is not disbursed

ISM 1.0 breaks down as ₹65,000 crore for chip production, ₹10,000 crore for the SCL Mohali laboratory and ₹1,000 crore for design — though Vaishnaw conceded on 28 November 2025 that the SCL line “is a fungible amount.” In August 2025 MeitY Secretary S. Krishnan said: “I think we have committed close to ₹62,900 crore or so already.” On 1 September he went further: the government has “nearly exhausted” the ₹76,000 crore.

Here is the number, and it is the government's own. The Expenditure Budget publishes scheme-wise actuals: ₹681 crore in FY2023-24, ₹638 crore in FY2024-25, ₹2,886 crore in the first half of FY2025-26. About ₹4,200 crore against a ₹76,000 crore headline, two years in. And the sub-scheme line that matters most — Semiconductor Fabs — reads ₹0.00 crore in FY2023-24, ₹0.00 crore in FY2024-25, and ₹0.33 crore to 31 December 2025.

Two years into a ₹76,000 crore scheme, the government's own accounting says nothing has been disbursed to the fab. Vaishnaw told the Standing Committee why: legal agreements must be executed before the fiscal support can be released. That is an answer. Whether it is an explanation is a separate question.

Even that is a budget line, not a project. Asked in the Lok Sabha on 5 August 2026 for “the funds sanctioned, released and utilised, project-wise,” the government replied with a project count and an operating-plant count. Dr Shashi Tharoor's July question on the design incentive scheme got a commitment figure and nothing else.

Beside it, PLI publishes its own: fourteen sectors, 892 applications, ₹2.40 lakh crore invested, ₹15.2 lakh crore of exports, ₹35,354 crore of incentive actually disbursed as of 31 March 2026. And PLI's employment series runs backwards — 12 lakh at March 2025, 14.39 lakh at December 2025, back to 14.15 lakh at March 2026 — while exports went from ₹6.5 lakh crore to ₹15.2 lakh crore over the same year. A series that rises and then falls inside twelve months is not a measurement. Scheme rules require claimed sales to be verified against GST returns, income-tax returns and bank realisation certificates, but no published reconciliation ties them to HSN-coded customs data, and a NITI Aayog database commissioned in 2022 to do exactly that was never announced as finished. The 14.15 lakh is thinner than it looks: the indirect half of it, 5.66 lakh, is self-reported and exists for only three of the fourteen sectors.

Bigger scope, lower price

Semicon 2.0 is the biggest thing that has happened, and it is good news for the wrong reason to celebrate.

The Cabinet approved ₹1,27,500 crore on 15 July, sixty-eight per cent above ISM 1.0, notified in the gazette on 31 August as CG-DL-E-31082026-275860. Alongside it went a ₹62,500 crore phone scheme replacing PLI-LSEM, and an increase in the electronics components scheme from ₹22,919 crore to ₹40,000 crore, with 75 applications worth ₹61,671 crore approved as of 30 March.

The re-scoping is right. ISM 1.0 bought assets; ISM 2.0 goes after the supply chain into them — gases, chemicals, equipment. India imports more than 90 per cent of its chipmaking equipment and 85 to 90 per cent of its specialty chemicals and electronic gases, on Equirus Securities' estimate. Vaishnaw declined to put a number on it at the Bloomberg New Economy Forum on 14 October, saying only that materials are “the biggest dependencies.”

Now the rates, from the notification itself rather than the trade press. Silicon fabs: 40 per cent of eligible capital expenditure, down from ISM 1.0's 50 per cent of project cost — the base changed as well as the rate. Advanced packaging, display and compound fabs, 35. Legacy packaging, 25. Equipment and materials, 30. R&D and talent, up to 75 — though those two are percentages of project cost including opex, not of the eligible capital expenditure the rest of the table sits on. The design incentive, which most of the coverage leaves out, is 9 per cent of net sales.

The scope widened and the price of a silicon fab fell. The subsidy now leans hardest toward what India has least of, which is defensible, and lightest toward what it has most of, which is a confession. And it does not touch Dholera, whose support was fixed under ISM 1.0: Bloomberg reported on 17 July that the new fund “will apply only to future, incremental investments and will not retroactively offset spending already committed.” The project that costs India most is the one the new scheme helps least.

By August the government had three plants producing — Micron from February, Kaynes from March, CG Semi from July. On 17 September PIB inaugurated CDIL's Mohali discrete line and Suchi Semicon's Surat packaging line in a single ceremony and put the total at five.

Suchi Semicon is the execution story nobody wrote up. Approved 5 May 2026, in commercial production 17 September 2026: four and a half months from Cabinet approval to a working line, on an ₹868 crore plant in Surat. That is the strongest execution fact in the whole file. It is also a packaging line — as are all five plants now producing. Nothing India actually fabricates yet.

Sanand is not what I said it was

I gave Micron's Sanand plant more credit in June than the evidence supports, and the company's own press release has been quietly correcting the record since.

On 28 February 2026, at the inauguration, Micron wrote: “The expansion of conventional assembly and test operations in India complements Micron's planned development of advanced manufacturing and packaging capabilities in the United States and strengthens the company's global assembly and test network.” The second half of that sentence is the part that gets cut. An unnamed Indian government official told the Economic Times in April that the plant “is not planned to host HBM manufacturing yet.”

The “10 per cent” figure is not Micron's. Where it can be pinned down it is attached to at least two different denominators — Vaishnaw's “global memory production,” and Business Today's “global packaging and test output,” the only one of them a single-floor assembly and test plant could plausibly deliver — and Micron's own release contains no percentage at all.

On ramp: tens of millions of chips this year, hundreds of millions next, said Mehrotra again on 17 September. A company at full ramp does not re-announce its ramp. Nor does it carry one 2027 number. It carries two, and neither is a filing. A company that will not put a plant in its revenue line will not put it in its guidance either.

In July, unnamed people aware of the developments told the Economic Times that Micron had handed Phase 2 to L&T, displacing Tata Projects after delays in Phase 1; Micron has confirmed neither the award nor the reason. Nor is that the only slip on the record: Phase 1, set for 2024, was by February 2025 still targeting early 2025 completion. And in Micron's record FQ4-26 results on 30 September, $54.23 billion in a quarter, Sanand does not appear — not in the release, and not in any document the company has ever filed with the SEC. Dell is the only customer Micron has named anywhere.

None of that makes Sanand a bad plant. It is probably the best-run asset in the programme: an IDM drawing wafers from its own global network and putting them through test in India, staffed by 700-odd fresh graduates who trained in Malaysia and Singapore first. Not an argument against Micron — an argument against importing their forward guidance as an Indian capability claim.

The binding constraint is not money

If you want to know what will actually stop an Indian fab, it will not be capital and it will not be the exchange rate. It will be the fact that nobody has ever priced one.

On 25 March 2026 the Gujarat Electricity Regulatory Commission, in its tariff order for Torrent Power's Dholera distribution licence, declined to determine a tariff for the Dholera special industrial region at all, and continued Dahej's rate card unchanged for FY 2026-27 under the saving provision in its own regulations. Dholera has been on somebody else's price since 2022. The card is not generous: the high-tension industrial energy charge runs ₹2.90 to ₹3.20 depending on billing demand, with fuel pass-through on top and time-of-day penalties, which GERC's own cross-subsidy table puts at an all-in ₹6.26 to ₹6.31 a unit. The semiconductor incentives run on different clocks again — a ₹2 a unit tariff subsidy for ten years from the start of commercial production under the Gujarat Semiconductor Policy, and a duty exemption that runs five years in a special investment region and ten only inside a notified SEZ, which Dholera is not. Nationally, Power Secretary Pankaj Agarwal told EDICON in New Delhi on 21 January that industry pays about $105 a megawatt-hour, of which about ten dollars is cross-subsidy.

Read that order again. Gujarat has spent a decade assembling every incentive a semiconductor cluster could ask for, and when a regulator finally had to price a fab, it reached for the saving provision and borrowed another licensee's rate card, because the Dholera licensee has no cost history to justify one of its own. A regulated tariff is built from audited load, sanctioned demand and cost of supply, and Dholera has none of the three. That is harder to argue with than any reliability average: it is a regulator saying out loud that it cannot yet tell what a fab costs.

I would not have reached for a reliability number even if one were available. In India the instrument is patchy and mostly self-reported: of 72 distribution companies, 36 carry no outage data on their websites at all, an audit by The Leap Journal found in September 2025. What does exist is REC's Consumer Service Rating, which graded 66 licensees last year. On industrial supply, Gujarat's DGVCL came in at 10.6 interruptions and Assam's APDCL at 13.9, which is respectable. It is also a frequency count with no duration attached, the definition is not published, and six of REC's fifteen special-category utilities reported no industrial figure at all. Assam's APDCL ranks first of those fifteen on REC's composite score, not on this measure. The worst numbers in the dataset belong to Punjab and Karnataka, not Bihar.

Water is better sourced, though still second-hand. T Kannan reported in The Statesman on 29 August that Dholera's own environmental-appraisal file carries a total requirement of about 17,549 kilolitres a day, 6,571 of them fresh and the rest recycled, piped from the Narmada the better part of a hundred kilometres down a canal built for farms and towns — a river he describes as already fully spoken for. He cites no clearance number and no allocation order, and I have not been able to read the clearance file. If the numbers are right, they point at something structural: project-by-project clearance never asks what happens to the basin when the next nine plants answer the same question the same way. India has a water application form, not a water budget.

None of the rest is failure either. The scheme supports projects for up to six years, NITI Aayog's roadmap runs ten, and Vaishnaw relays the Prime Minister's line that this is a twenty-year journey. Asked about timing at the Bloomberg New Economy Forum on 14 October, he offered three horizons rather than a deadline. On ITIF's February 2024 count, drawn from a handful of interviews, roughly one per cent of India's 600,000 annual electronics-stream engineering graduates can walk onto a factory floor within three months of graduating. The regulator's own data says the pool is a quarter of that — AICTE counted 160,450 electronics and communications seats filled in 2024-25. CSIS repeated the one per cent in September 2025, citing a source that does not contain it. Slow. On this evidence, not stupid.

Washington, which did nothing

Let me take the geopolitical claim by claim, because this is where the June framing was most wrong.

Count them. All twenty-three Bureau of Industry and Security items in the Federal Register between 1 June and 18 October were read, and not one of them tightens semiconductor, advanced-computing, SME, HBM or advanced-packaging export control. The only advanced-computing rule in that window is the one two paragraphs below, and it loosens rather than tightens. The last tightening was in January. Section 232 semiconductor tariffs are still Phase 1 only: a 25 per cent duty, effective 15 January, on what the proclamation calls a very narrow category, described in an annex the Federal Register publishes only as an image. The H200 and MI325X labels come from January's export-control rule, not from the tariff. Data centres, R&D, startups, repairs, consumer and civil-industrial use and the public sector are all carved out to zero. One item touches a chip input — a September stockpiling prohibition under Proclamation 11052 — and it is a Section 232 import measure, not an export control.

India's status is unremarkable, which is a different thing from unfavourable. India sits in Country Group B, outside the enforcement perimeter, so top-tier accelerators ship with no licence required at all — NLR, which is a stronger and different thing from a licence exception. It is also already in Country Group A:5, and therefore already has Strategic Trade Authorization eligibility under 740.20©(1)(i). None of that was granted to India. It follows from where India is listed.

The July rule was different in kind. On 10 July, effective two days before it appeared in print, BIS published at 91 FR 43034, “Enhanced Favorable Treatment for the United Arab Emirates Under the Export Administration Regulations.” The UAE came out of Country Groups D:3 and D:4, joined Country Group A:5 — the bracket India already sits in — and gained a new entity whitelist, Supplement No. 8 to Part 740, covering licence-free access to advanced computing items for G42, Core42, Amazon.com, Apple, Google, Meta, Microsoft, OpenAI, Oracle and xAI. MGX is not on that list, only flagged for favourable review, and the G42 carve-out expires on 6 April 2027 unless they become American companies. This is what a carve-out made for one country looks like. Nothing in it was made for India, and nothing in it needed to be.

And Section 232 is uncomfortable. Proclamation 11002 of 14 January set out a two-phase plan, and the clocks are what they are: a status report on the negotiations due to the President by 14 April, a data-centre market update due by 1 July, and a 180-day statutory fallback whose date, 13 July, has passed. Only Phase 1 was ever operative; Phase 2 is permissive and nothing has triggered it. What has been implemented is thinner than the headlines suggest. Taiwan's investment agreement took effect in May, and what it capped was 232 duties on Taiwanese automobile parts, timber and wood derivatives. The semiconductors were left out of it. India risks becoming the residual jurisdiction: the one that pays the tariff without receiving the carve-out.

What actually got harder

Not the chips. The Strait.

Between 6 and 9 August, the Guangdong Maritime Safety Administration announced traffic control on northbound vessels entering the Taiwan Strait through its southern approach, framed around a typhoon. Taiwan's Coast Guard Administration told shipping to disregard it. ISW's reading was that Beijing may have used typhoon cover to test expanding its law-enforcement jurisdiction over a waterway it does not control. Cheap. Deniable. Legally untested.

On 31 August, Taiwan's Ministry of National Defence told the Legislative Yuan that the PLA is stepping up blockade preparations — the annual military-power report describes the intent as building the capacity to construct a blockade, sever Taiwan's sea lanes and encircle it. The ministry's own answer that week was thinner than the question. The anti-blockade work is visible elsewhere: a first-ever joint navy and coast guard escort operation on 12 August, shepherding strategic cargo into Hualien and Suao. EIU's own verdict, published back in February, is that a direct purposeful invasion is unlikely and that a blockade — or even an invasion — cannot be fully ruled out.

And it has not stopped. Between 8 and 17 October, six maritime administrations closed large parts of the Bohai, Yellow, East and South China Seas, including a live-fire no-sail zone off the southern Yellow Sea that runs to 30 November. On 14 October the Southern Theater Command put four landing ships through a simultaneous live-fire salvo and the Liaoning carrier group moved into the western Pacific. The timing is not subtle: it sits either side of Keen Sword, the US–Japan exercise that runs 19 to 29 October and puts systems on Yonaguni, 110 kilometres from Taiwan.

What changed is that the cost of preparing a blockade has fallen to the point where it can be rehearsed under cover of a weather advisory. Every leg of this strategy is an external dependency. Coherent. Not a hedge.

The window

Eight-inch utilisation across the top ten foundries recovered to 88 per cent in 2026 and should reach 90 in the second half; prices rose 5 to 15 per cent between the first and second quarters, and a third round could extend into 2027. The tight window is real. It is why India's bet is rational. It also has an end date.

Start with the mechanism, because it is usually stated backwards. TrendForce's thesis is not that TSMC and Samsung are releasing mature-node capacity. It is the reverse: they have been cutting eight-inch lines since the second half of 2025, and TSMC is taking roughly 50,000 wafers a month — 15 to 20 per cent of Fab 14's twelve-inch mature capacity — out by 2028. TrendForce's May reading was that the full impact becomes more pronounced after the second half of 2027. The squeeze gets worse, not better.

Which means new mature-node capacity comes from everyone who is not TSMC or Samsung. That is Dholera's market. Counterpoint had 40 to 90 nanometre utilisation stuck near 80 per cent in January, and TrendForce noted in April that customers were negotiating to postpone planned price increases for the second half of 2026. A new entrant arriving in 2028 is the marginal wafer in a market that already has spare wafers, in a price cycle whose customers are asking for relief in the quarter it is supposed to extend.

Set that against the decision MeitY's secretary made out loud on 24 September. India has chosen to subsidise the legacy end of the market because the leading edge is “simply too expensive.” That is defensible. It is also a decision about which glut to walk into, and the glut is the Chinese one. The Semiconductor Industry Association, working from SEMI's fab forecast, put China's share of global mature-node capacity at nearly double — 19 to 33 per cent — between 2015 and 2023, measured on installed nameplate capacity at 28 nanometres and above. CSIS, separately, could not find an expert consensus that this becomes an oversupply by 2030. Note the trap in the number: TrendForce also has China's mature-process share reaching 33 per cent, in 2027.

Dholera's first wafers are scheduled for 2028.

I am not forecasting a collapse. I am pointing at something harder to argue with: a fab arriving in the last year of the price cycle it was planned against, into a market where the two best-run foundries on earth are deliberately shrinking and the ones adding capacity are subsidised to add it, while their customers are already trying to talk the price down. India does not lose by being late to a fab. India loses by being late to a fab that arrives into the end of a price cycle.

What I think

None of this is an argument against Dholera. A 2028 fab at 90 nanometres in India is worth more than no fab, and a fab that trains process engineers and drags a supplier base into existence around itself is worth more than the ₹91,526 crore on the invoice. The state has done several things right. It let Micron build. And it has built an OSAT cluster in Sanand that is real rather than a press release. Any honest assessment starts there.

What I don't believe is that the programme is being managed to a schedule that survives contact with the market.

Taking the claims one at a time. The timeline moved eighteen months and the government is carrying three live dates it has not squared. The node dropped two generations. The completion figure traces to one unnamed paragraph. No anchor customer has been named. By its own accounting the scheme has disbursed nothing to the fab. The capital went elsewhere — SEMI's 2026 construction-start ranking has no India row, and Reuters called India a late entrant. And the strategic instrument everyone was watching has done nothing for India, while the real escalation has moved to a mechanism nobody can hedge against.

The fair summary is this. India now has a genuinely good packaging industry, an underrated OSAT cluster, a real design workforce, a properly large subsidy programme, and one fab that will start producing two years from now at a node two generations behind the promise, into the part of the cycle where the industry stops paying for capacity.

That is not a failure. It is an expensive lesson with a very long fuse, and the fuse is already lit.

]]> https://think.shenthar.me/chanakya/the-fab-is-eighteen-months-late Sun, 18 Oct 2026 12:08:37 +0000 BSNL's ₹3.22 Lakh Crore Gamble: Sovereignty at What Cost? https://think.shenthar.me/chanakya/bsnls-3-22-lakh-crore-gamble-sovereignty-at-what-cost <![CDATA[The headline the government wants you to read On paper, BSNL's revival is a fairy tale. One lakh 4G sites, almost reached, with 96,103 on-air as of March 2026 out of 97,906 installed. Revenue climbing. EBITDA surging 335% year-on-year in the most recent quarter. Operating profit jumping from ₹2,395 crore in FY24 to ₹5,100 crore in FY25, with double-digit growth projected for FY26. First profit in seventeen years, posted in Q3 FY25. The first time India has built a complete homegrown telecom network from radio to core, putting it in the company of the US, China, Europe, and South Korea. Communications Minister Pemmasani Chandra Sekhar, in a recent interview, called it "one of the most significant public sector turnarounds in recent years." The narrative is everywhere: PIB releases, DD India interviews, Ministry social media. BSNL is back. The state can build strategic infrastructure. Atmanirbhar Bharat works. There is one problem with this fairy tale. The numbers underneath it tell a different story, and the customers living it tell a third. Both stories are worth listening to, because the choice India made in 2019 to back BSNL with massive state support and an indigenous technology stack was not a bad choice. It may even have been the right one. The way it has been executed, though, and the way its results are being marketed, deserve a harder look than they've been getting. This is the most expensive telecommunications policy decision India has ever made. It deserves a critical nationalist's eyes, not a cheerleader's. --- What was actually decided To understand where BSNL is in 2026, you have to go back to 2019. The company was, on every measure, on its deathbed. Jio had reshaped the market. Airtel had been forced into a brutal price war. Vodafone Idea was bleeding subscribers. And BSNL, the only state-owned telecom operator at scale, was the slowest to adapt. Its 4G rollout had been delayed by years of tender cancellations, vendor disputes, and the political difficulty of choosing a foreign partner. The Modi's first term ended with BSNL posting a net loss of ₹15,500 crore in FY19, with debt service costs and AGR dues piling up. Three revival packages followed, totalling roughly ₹3.22 lakh crore by the Ministry's own accounting. The first, in 2019, was a ₹69,000 crore package combining VRS (about a third of BSNL's workforce took it), debt restructuring, and sovereign guarantees. The second came in 2022, adding spectrum allocation and 4G capex. The third is still being deployed, with a ₹47,000 crore capex pipeline through FY27. None of this counted the AGR dues of nearly ₹1.78 lakh crore, which were absorbed in earlier write-offs. That's the largest state support package ever given to a single Indian company in the telecom sector. It is comparable in scale to the airline bailouts of 2020-21 or the bank recapitalisation of 2017-19. It's not presented that way, though. It's presented as a turnaround, a successful one, in its final phase. The bet had two pillars. The first was operational revival: get BSNL profitable, or at least self-sustaining. The second was sovereignty: build a fully indigenous 4G stack, radio to core to software to hardware, owned and operated by Indian companies. The TCS, Tejas Networks, and C-DOT consortium was the answer. The result, in 22 months of actual development, was the country's first end-to-end indigenous 4G network. India joined a small club of countries with this capability. On paper, this is historic. On the ground, it's more complicated. --- The receipts: what the independent data shows Let me take the revival story claim by claim, against the data that doesn't come from the Ministry. Revenue and profit The Ministry's preferred number is the EBITDA jump, from roughly ₹50 crore to nearly ₹7,000 crore in two years. That figure is real. It's also misleading on its own. EBITDA is earnings before interest, taxes, depreciation, and amortisation. In BSNL's case, depreciation has exploded because of the 4G capex. In Q2 FY26 alone, depreciation and amortisation hit ₹2,477 crore, up 57% year-on-year. The Q2 net loss was ₹1,357 crore, wider than the ₹1,241.7 crore lost in the year-ago quarter. The full-year FY26 net loss came in at ₹4,740 crore, double the ₹2,250 crore lost in FY25. Q1 FY26 showed a narrowed net loss of ₹1,049 crore (down from ₹1,542 crore a year earlier), but this came after two profitable quarters in FY25 that BSNL itself described as partly driven by one-off accounting adjustments. By Light Reading's accounting, FY26 revenue grew by only 1% over FY25, even as the cost base widened with network operating expenses and finance charges. The operating engine is improving. The capital being deployed is so heavy, though, that the bottom line is getting worse, not better. The EBITDA number is the right thing to celebrate if you're a CFO. It's the wrong thing to celebrate if you're a Finance Minister. Market share BSNL ended March 2026 with a 7.34% share of the wireless market, down from 7.8% a year earlier. The market itself grew; total wireless subscribers are now over 1.16 billion, but BSNL's slice shrank. To put that in context: Reliance Jio and Bharti Airtel together hold the overwhelming majority of Indian wireless subscribers. The two-player duopoly is now structural, not transitional. Vodafone Idea is a distant third with single-digit share and is in turnaround talks of its own. BSNL is the fourth player in a market where the top two are closing the door. This is the dirty secret of the revival. The state did not pour ₹3.22 lakh crore into BSNL to make it the fourth player. It poured it in to make it a strategic player, the operator of last resort, the rural and government network, the buffer against duopoly pricing. The financial story is being told as if BSNL were a competitive challenger, and it's not. The "active subscriber" problem The most damning data point in TRAI's most recent numbers: the share of active subscribers on BSNL's network fell from 63.37% in March 2025 to 56.36% in March 2026. In a healthy telecom network, 80-90% of subscribers are active in any given month. BSNL is at 56%. Out of 92.9 million reported subscribers, only about 52 million are actually using the network. The other 41 million are zombie SIMs: people who bought a BSNL connection at some point but have ported out, switched to a second device, or stopped recharging. This is the ARPU paradox. BSNL's average revenue per user rose 12% in Q2 FY26. The Ministry presents this as a win. The real driver is that the inactive subscribers dropped out of the denominator, not that the active ones started spending more. ARPU at BSNL is still around ₹178, compared to Airtel's ₹257 and Jio's ₹214. BSNL is not winning more wallet share. It is losing the customers who were costing it the least to serve. Service quality The TRAI drive tests, the regulator's own independent audits, paint a picture the Ministry does not highlight. In Himachal Pradesh and UP West, only 7-17% of BSNL's tested locations achieved "excellent" signal strength. The call drop rate in Himachal was 15.69%, the worst of any major telco. Mean Opinion Scores for voice clarity were below the regulatory minimum. Data speeds were routinely below 7 Mbps. In the June 2025 Meerut drive test, BSNL lagged private operators across every measured metric. These are not anecdotes. They are TRAI's data. The field reports corroborate. The DT Next investigation in Chennai found that BSNL's 4G tower installation, supposed to be completed by December 2024, was still incomplete in 2025, with BSNL's own General Manager writing to TCS about operational issues at newly commissioned sites. The Chennai official said only 1,000 of the required towers had been installed. Customer complaints cite FTTH connections that have been down for months, with docket numbers registered and closed without resolution. "I will port out" is a regular refrain in BSNL's own complaint threads. The TCS, Tejas, C-DOT stack is real. The deployment is real. The deployment is not yet at a level where a BSNL customer in a major Indian city can rely on it. --- The cost of sovereignty The 4G delay was not an accident. It was a choice. Communications Minister Jyotiraditya Scindia has confirmed, in his own words, that the government "decided to bypass the easy route of buying off-the-shelf foreign equipment to focus on building a completely indigenous technology stack." In his framing, this was a "calculated, strategic decision," a "pain today for gain tomorrow" trade. There is a serious argument to be made in favour of that choice. Foreign telecom equipment, from Ericsson, Nokia, Huawei, ZTE, Samsung, comes with documented security concerns. The US, UK, Germany, and Australia have all restricted or removed Chinese vendors from their 5G cores on national security grounds. India's geographic and strategic situation makes it more exposed to those risks, not less. The Snowden revelations in 2013 showed how deeply foreign intelligence agencies had compromised commercial telecom equipment. The NSO Group scandal showed what commercial spyware can do once it's inside a network. The argument for an indigenous stack is not ideology. It's risk management. A network you do not control is a network that can be turned off, tapped, or backdoored by someone else. The TCS, Tejas, C-DOT consortium delivered the stack in 22 months. That's a real engineering feat. No other country of India's size has built a complete 4G core and radio network in this timeframe. The IP stays in India. The manufacturing is in India. The future export potential, selling this stack to other countries wary of Chinese equipment, is real. The argument against the choice is equally serious. By betting on indigenous, BSNL entered commercial 4G roughly seven years after Jio. Airtel completed its 4G rollout in 2016. Jio did so in 2017. BSNL's pan-India 4G only soft-launched in August 2025, with full commercial availability in September 2025. By the time BSNL's customers could reliably use 4G, the duopoly had already entrenched itself. The window when a state operator could meaningfully compete on speed and coverage, the 2017-2022 window, was spent on building the indigenous stack. The cost of sovereignty was the market. The customers who would have been BSNL's growth story became Jio's and Airtel's growth story. The state chose to keep the means of production in Indian hands, and paid for it in lost subscribers, lost revenue, lost relevance. Whether that trade was worth it depends on what you think a state-owned telecom is for. --- What a state telco is for The privatisation orthodoxy says a state-owned telecom operator is a contradiction in terms. Telcos are infrastructure-heavy, capital-hungry, depend on operational excellence. They need to be run like private businesses. If they cannot compete, they should be allowed to die. This is the argument Vodafone Idea's bankers make. It's the implicit argument behind every piece of advice the IMF and the World Bank have given India on telecom since 1999. In a narrow sense, it's correct. In a wider sense, it's wrong. A state-owned telecom is not a business. It's a piece of strategic infrastructure. Its job is to be the network that exists when private capital does not want to be there. It is the operator that builds towers in the Ladakh highlands, in the Andaman islands, in the LWE-affected districts of Chhattisgarh, places where the ARPU will never recover the capex. It is the network the army can rely on during a war. It is the network that keeps working when a cyclone knocks out power in Odisha for two weeks. It is the network the government can use to push direct benefit transfers, run census operations, and provide a communications backbone for disaster response. A private telco is not going to do any of that. Jio's pricing model depends on dense urban usage. Airtel's 5G rollout is following the same economic logic. Neither is going to spend capex on a tower that serves 200 subscribers in Arunachal Pradesh when those 200 subscribers are not going to recover the capex in their lifetime. This is the case for BSNL that almost never gets made in the business press. It is the case that should be made, because it is the actual reason the Indian state is willing to keep funding a company that loses ₹4,740 crore a year. The Ministry is not trying to make BSNL profitable in the conventional sense. It is trying to keep BSNL alive, and in the process, build the sovereign capability that allows India to never be in a position where its core communications infrastructure is owned, operated, or backdoored by a foreign power. Seen through this lens, the question is not "is BSNL competitive with Jio?" It is "is BSNL's continued existence a worthwhile strategic investment for the Indian state?" On that question, the data is more favourable. And here's the part the Ministry does not want made: the strategic value only matters if the operational reality is improving. A rural operator that does not work in rural areas, a strategic telco that cannot keep its own customers connected, a sovereign network that drops 15% of its calls in the mountains, is not a strategic asset. It's a sunk cost. The Ministry knows this. The 60,000 new towers being deployed under the ₹20,000 crore capex push, the focus on service quality, the AI-led network monitoring, these are the operational fixes. The numbers suggest they are not yet sufficient. They are the right direction. --- The real test, five years out If the 4G stabilisation succeeds, if the indigenous stack reaches a quality level competitive with foreign vendors, if BSNL is positioned clearly as the strategic and rural operator of last resort, and if the government is willing to keep funding the gap between revenue and operating cost in perpetuity, then BSNL will be a slow, unglamorous, perpetual-money-pit success story. The next ten years will see it lose market share in cities, gain modest share in rural India, lose money on operations, and quietly be the network that holds the country together in ways the private operators will never be. If the 4G stabilisation fails, if TRAI's drive tests continue to show call drops above 10%, if ARPU growth does not come from real customer spend, if the 5G upgrade gets delayed by another three years, then BSNL becomes India's Concor. A state-owned entity that survives on permanent subsidy, technically alive but operationally fading, existing more to keep a workforce employed than to provide a real service. The 60,000 new towers become monuments to a strategic decision that did not work. The third path is the worst: privatisation. If BSNL is sold to a private buyer, and there have been murmurs of this since 2022, the indigenous 4G stack gets licensed or sold, the rural network gets cherry-picked, and the strategic capability India built over the last six years gets absorbed into a private balance sheet. The state loses the sovereign option. The taxpayer pays for the capability. The private operator gets the upside. The most likely outcome, given the political economy, is a slow drift toward the first path. The current government has shown no appetite to privatise BSNL. The revival packages continue. The capex continues. The political incentive to keep BSNL alive is significant: it is a major employer, a major rural presence, a visible symbol of state capability. Slow drift is not a strategy. The country needs the government to make an explicit choice about what BSNL is. The current messaging tries to have it both ways: strategic asset and profitable turnaround, sovereignty and market share, indigenous stack and competitive 5G. All four are not possible simultaneously. Three of them might be, if the trade-offs are acknowledged. --- What the numbers should be read as The right way to read BSNL in 2026 is not as a turnaround story. It's a strategic investment in progress. The 4G stack is real. India built it. That's historic. The fact that the rollout is slower than the private operators, and that the quality is currently worse, is the cost of the strategic choice. The cost is bearable. ₹3.22 lakh crore is a lot of money, but it's less than one year's defence capital outlay, and the alternative was permanent dependence on foreign vendors for a piece of critical national infrastructure. The market share decline is real, but it's also the market's verdict on a state operator that was not allowed to compete on equal terms. A private telco that was seven years late to 4G, blocked from using foreign equipment, and tasked with serving the parts of India no one else wants to serve, would have lost market share too. The service quality problems are real, and they're the most urgent. Until the BSNL customer in a tier-2 city can make a call without dropping, the 4G stack is a strategic capability on paper, not in practice. The next year of capex, the 60,000 new towers, the AI-led monitoring, the battery and power upgrades, is the test. If service quality reaches Airtel-and-Jio parity in 2027, the strategic bet pays off. If it does not, the bet was the right idea with the wrong execution. The duopoly is real. Jio and Airtel are not going to be dislodged from the top of the Indian market. They're also not going to serve the strategic and rural use cases that BSNL exists to handle. The state did not pour ₹3.22 lakh crore into BSNL to make it the second-largest telco. It poured it in to make it the last telco standing when strategic necessity requires it. That's a less glamorous story than "BSNL is back." It's a more honest one. It explains what is actually happening on the ground: the wins, the losses, and the costs of a sovereign bet India is still in the middle of making. --- The sovereignty tax Every country that builds indigenous strategic infrastructure pays a sovereignty tax. The US paid it in the 1950s and 60s when it decided to build its own semiconductor industry rather than rely on imports. South Korea paid it in the 1970s when it built POSCO rather than rely on Japanese steel. China paid it in the 2010s when it built its own commercial aerospace industry rather than rely on Boeing and Airbus. The tax is real. It's the cost of doing something slower, harder, and more expensive than the alternative, in order to keep a strategic capability in national hands. The tax is not always worth paying. Sometimes the alternative is genuinely better, and the strategic bet is a vanity project. In BSNL's case, the strategic case is strong. The 4G stack gives India a sovereign option in a sector that every major power now treats as critical infrastructure. The TCS, Tejas, C-DOT capability is the foundation for a potential export industry, selling the stack to countries that want alternatives to Huawei. The rural and strategic network is a public good the market will not provide. The cost, seven years of customer pain, market share loss, perpetual subsidy, is the sovereignty tax. It's being paid in real time by the BSNL customer in Himachal who drops 15% of his calls. By the BSNL employee in Chennai who wonders if her job is secure. By the Indian taxpayer who funded ₹3.22 lakh crore of revival packages and may fund more. The bet is not yet won. The next 24 months are the test. If the 4G stack reaches quality parity, if BSNL finds its role as the strategic and rural operator, if the indigenous capability gets commercialised into an export industry, the sovereignty tax will be retrospectively worth it. If not, it will be remembered as a strategic mistake dressed up as a turnaround story. BSNL is not a fairy tale. It's a strategic investment in progress. We'll know in 2027 which one it ends up being. --- Sources: PIB: "BSNL Annual Results FY 2025-26 Reviewed," April 23, 2026 TRAI: Performance Indicators Reports index (Q.E. April–June 2025 and quarterly series) ET Telecom: "BSNL Reports Narrowed Net Loss of ₹1,049 Crore; 4G Expansion Boosts Revenue," August 19, 2025 Light Reading: "BSNL's ARPU jump raises questions as revenue growth doesn't match," June 16, 2026 LiveMint: "BSNL's Q2 net loss widens to ₹1,357 cr on a steep rise in depreciation, interest," November 18, 2025 Communications Today: "BSNL's 4G delays stall 5G, despite 60,000 new towers, profit push," March 27, 2026 Communications Today: "BSNL's troubles are deep — Widespread transmission problems, amid Q2 losses," November 22, 2025 (cites the TRAI Himachal / UP West independent quality audit: 7–17% excellent signal, 15.69% call-drop rate, <7 Mbps data) DT Next: "Delay in BSNL's 4G tower installation leads to poor connectivity in Chennai," November 25, 2025 BisInfotech: "BSNL's 4G Delay: A Strategic Bet on &#39;Made in India&#39; Technology," March 27, 2026 TelecomTalk: "Prime Minister Modi&#39;s Vision Is To Connect Every Village, No Matter How Remote: MoS Communications," May 25, 2026 (Pemmasani interview, ₹21,000 Cr → ₹25,000 Cr revenue) RCR Wireless: "BSNL to upgrade all 4G sites to 5G within eight months," March 25, 2026 TelecomTalk: "BSNL Has Made 96,103 4G Sites Operational Across India," March 13, 2026 Economic Times: "Communications minister Jyotiraditya Scindia says BSNL&#39;s 4G delay due to indigenous tech bet," March 27, 2026 (₹20,000 Cr capex, 98,000 indigenous 4G towers) ET Telecom: "BSNL&#39;s 2025 Breakthrough: 4G Launch and Financial Growth," December 9, 2025 (PM Modi&#39;s September 2025 inauguration of indigenous 4G at 92,000+ sites) PSU Watch: "BSNL rolling out one lakh indigenous 4G sites across India, says MoS Pemmasani Chandra Sekhar," March 11, 2026 (Lok Sabha reply confirming the ₹3.22 lakh crore revival package figure) ]]> The headline the government wants you to read

On paper, BSNL's revival is a fairy tale.

One lakh 4G sites, almost reached, with 96,103 on-air as of March 2026 out of 97,906 installed. Revenue climbing. EBITDA surging 335% year-on-year in the most recent quarter. Operating profit jumping from ₹2,395 crore in FY24 to ₹5,100 crore in FY25, with double-digit growth projected for FY26. First profit in seventeen years, posted in Q3 FY25. The first time India has built a complete homegrown telecom network from radio to core, putting it in the company of the US, China, Europe, and South Korea.

Communications Minister Pemmasani Chandra Sekhar, in a recent interview, called it “one of the most significant public sector turnarounds in recent years.” The narrative is everywhere: PIB releases, DD India interviews, Ministry social media. BSNL is back. The state can build strategic infrastructure. Atmanirbhar Bharat works.

There is one problem with this fairy tale.

The numbers underneath it tell a different story, and the customers living it tell a third. Both stories are worth listening to, because the choice India made in 2019 to back BSNL with massive state support and an indigenous technology stack was not a bad choice. It may even have been the right one. The way it has been executed, though, and the way its results are being marketed, deserve a harder look than they've been getting.

This is the most expensive telecommunications policy decision India has ever made. It deserves a critical nationalist's eyes, not a cheerleader's.

What was actually decided

To understand where BSNL is in 2026, you have to go back to 2019. The company was, on every measure, on its deathbed.

Jio had reshaped the market. Airtel had been forced into a brutal price war. Vodafone Idea was bleeding subscribers. And BSNL, the only state-owned telecom operator at scale, was the slowest to adapt. Its 4G rollout had been delayed by years of tender cancellations, vendor disputes, and the political difficulty of choosing a foreign partner. The Modi's first term ended with BSNL posting a net loss of ₹15,500 crore in FY19, with debt service costs and AGR dues piling up.

Three revival packages followed, totalling roughly ₹3.22 lakh crore by the Ministry's own accounting. The first, in 2019, was a ₹69,000 crore package combining VRS (about a third of BSNL's workforce took it), debt restructuring, and sovereign guarantees. The second came in 2022, adding spectrum allocation and 4G capex. The third is still being deployed, with a ₹47,000 crore capex pipeline through FY27. None of this counted the AGR dues of nearly ₹1.78 lakh crore, which were absorbed in earlier write-offs.

That's the largest state support package ever given to a single Indian company in the telecom sector. It is comparable in scale to the airline bailouts of 2020-21 or the bank recapitalisation of 2017-19. It's not presented that way, though. It's presented as a turnaround, a successful one, in its final phase.

The bet had two pillars. The first was operational revival: get BSNL profitable, or at least self-sustaining. The second was sovereignty: build a fully indigenous 4G stack, radio to core to software to hardware, owned and operated by Indian companies. The TCS, Tejas Networks, and C-DOT consortium was the answer. The result, in 22 months of actual development, was the country's first end-to-end indigenous 4G network. India joined a small club of countries with this capability.

On paper, this is historic. On the ground, it's more complicated.

The receipts: what the independent data shows

Let me take the revival story claim by claim, against the data that doesn't come from the Ministry.

Revenue and profit

The Ministry's preferred number is the EBITDA jump, from roughly ₹50 crore to nearly ₹7,000 crore in two years. That figure is real. It's also misleading on its own.

EBITDA is earnings before interest, taxes, depreciation, and amortisation. In BSNL's case, depreciation has exploded because of the 4G capex. In Q2 FY26 alone, depreciation and amortisation hit ₹2,477 crore, up 57% year-on-year. The Q2 net loss was ₹1,357 crore, wider than the ₹1,241.7 crore lost in the year-ago quarter. The full-year FY26 net loss came in at ₹4,740 crore, double the ₹2,250 crore lost in FY25.

Q1 FY26 showed a narrowed net loss of ₹1,049 crore (down from ₹1,542 crore a year earlier), but this came after two profitable quarters in FY25 that BSNL itself described as partly driven by one-off accounting adjustments. By Light Reading's accounting, FY26 revenue grew by only 1% over FY25, even as the cost base widened with network operating expenses and finance charges.

The operating engine is improving. The capital being deployed is so heavy, though, that the bottom line is getting worse, not better. The EBITDA number is the right thing to celebrate if you're a CFO. It's the wrong thing to celebrate if you're a Finance Minister.

Market share

BSNL ended March 2026 with a 7.34% share of the wireless market, down from 7.8% a year earlier. The market itself grew; total wireless subscribers are now over 1.16 billion, but BSNL's slice shrank.

To put that in context: Reliance Jio and Bharti Airtel together hold the overwhelming majority of Indian wireless subscribers. The two-player duopoly is now structural, not transitional. Vodafone Idea is a distant third with single-digit share and is in turnaround talks of its own. BSNL is the fourth player in a market where the top two are closing the door.

This is the dirty secret of the revival. The state did not pour ₹3.22 lakh crore into BSNL to make it the fourth player. It poured it in to make it a strategic player, the operator of last resort, the rural and government network, the buffer against duopoly pricing. The financial story is being told as if BSNL were a competitive challenger, and it's not.

The “active subscriber” problem

The most damning data point in TRAI's most recent numbers: the share of active subscribers on BSNL's network fell from 63.37% in March 2025 to 56.36% in March 2026.

In a healthy telecom network, 80-90% of subscribers are active in any given month. BSNL is at 56%. Out of 92.9 million reported subscribers, only about 52 million are actually using the network. The other 41 million are zombie SIMs: people who bought a BSNL connection at some point but have ported out, switched to a second device, or stopped recharging.

This is the ARPU paradox. BSNL's average revenue per user rose 12% in Q2 FY26. The Ministry presents this as a win. The real driver is that the inactive subscribers dropped out of the denominator, not that the active ones started spending more. ARPU at BSNL is still around ₹178, compared to Airtel's ₹257 and Jio's ₹214. BSNL is not winning more wallet share. It is losing the customers who were costing it the least to serve.

Service quality

The TRAI drive tests, the regulator's own independent audits, paint a picture the Ministry does not highlight.

In Himachal Pradesh and UP West, only 7-17% of BSNL's tested locations achieved “excellent” signal strength. The call drop rate in Himachal was 15.69%, the worst of any major telco. Mean Opinion Scores for voice clarity were below the regulatory minimum. Data speeds were routinely below 7 Mbps. In the June 2025 Meerut drive test, BSNL lagged private operators across every measured metric.

These are not anecdotes. They are TRAI's data.

The field reports corroborate. The DT Next investigation in Chennai found that BSNL's 4G tower installation, supposed to be completed by December 2024, was still incomplete in 2025, with BSNL's own General Manager writing to TCS about operational issues at newly commissioned sites. The Chennai official said only 1,000 of the required towers had been installed. Customer complaints cite FTTH connections that have been down for months, with docket numbers registered and closed without resolution. “I will port out” is a regular refrain in BSNL's own complaint threads.

The TCS, Tejas, C-DOT stack is real. The deployment is real. The deployment is not yet at a level where a BSNL customer in a major Indian city can rely on it.

The cost of sovereignty

The 4G delay was not an accident. It was a choice.

Communications Minister Jyotiraditya Scindia has confirmed, in his own words, that the government “decided to bypass the easy route of buying off-the-shelf foreign equipment to focus on building a completely indigenous technology stack.” In his framing, this was a “calculated, strategic decision,” a “pain today for gain tomorrow” trade.

There is a serious argument to be made in favour of that choice. Foreign telecom equipment, from Ericsson, Nokia, Huawei, ZTE, Samsung, comes with documented security concerns. The US, UK, Germany, and Australia have all restricted or removed Chinese vendors from their 5G cores on national security grounds. India's geographic and strategic situation makes it more exposed to those risks, not less. The Snowden revelations in 2013 showed how deeply foreign intelligence agencies had compromised commercial telecom equipment. The NSO Group scandal showed what commercial spyware can do once it's inside a network.

The argument for an indigenous stack is not ideology. It's risk management. A network you do not control is a network that can be turned off, tapped, or backdoored by someone else.

The TCS, Tejas, C-DOT consortium delivered the stack in 22 months. That's a real engineering feat. No other country of India's size has built a complete 4G core and radio network in this timeframe. The IP stays in India. The manufacturing is in India. The future export potential, selling this stack to other countries wary of Chinese equipment, is real.

The argument against the choice is equally serious. By betting on indigenous, BSNL entered commercial 4G roughly seven years after Jio. Airtel completed its 4G rollout in 2016. Jio did so in 2017. BSNL's pan-India 4G only soft-launched in August 2025, with full commercial availability in September 2025. By the time BSNL's customers could reliably use 4G, the duopoly had already entrenched itself. The window when a state operator could meaningfully compete on speed and coverage, the 2017-2022 window, was spent on building the indigenous stack.

The cost of sovereignty was the market. The customers who would have been BSNL's growth story became Jio's and Airtel's growth story. The state chose to keep the means of production in Indian hands, and paid for it in lost subscribers, lost revenue, lost relevance.

Whether that trade was worth it depends on what you think a state-owned telecom is for.

What a state telco is for

The privatisation orthodoxy says a state-owned telecom operator is a contradiction in terms. Telcos are infrastructure-heavy, capital-hungry, depend on operational excellence. They need to be run like private businesses. If they cannot compete, they should be allowed to die.

This is the argument Vodafone Idea's bankers make. It's the implicit argument behind every piece of advice the IMF and the World Bank have given India on telecom since 1999. In a narrow sense, it's correct.

In a wider sense, it's wrong.

A state-owned telecom is not a business. It's a piece of strategic infrastructure. Its job is to be the network that exists when private capital does not want to be there. It is the operator that builds towers in the Ladakh highlands, in the Andaman islands, in the LWE-affected districts of Chhattisgarh, places where the ARPU will never recover the capex. It is the network the army can rely on during a war. It is the network that keeps working when a cyclone knocks out power in Odisha for two weeks. It is the network the government can use to push direct benefit transfers, run census operations, and provide a communications backbone for disaster response.

A private telco is not going to do any of that. Jio's pricing model depends on dense urban usage. Airtel's 5G rollout is following the same economic logic. Neither is going to spend capex on a tower that serves 200 subscribers in Arunachal Pradesh when those 200 subscribers are not going to recover the capex in their lifetime.

This is the case for BSNL that almost never gets made in the business press. It is the case that should be made, because it is the actual reason the Indian state is willing to keep funding a company that loses ₹4,740 crore a year. The Ministry is not trying to make BSNL profitable in the conventional sense. It is trying to keep BSNL alive, and in the process, build the sovereign capability that allows India to never be in a position where its core communications infrastructure is owned, operated, or backdoored by a foreign power.

Seen through this lens, the question is not “is BSNL competitive with Jio?” It is “is BSNL's continued existence a worthwhile strategic investment for the Indian state?” On that question, the data is more favourable.

And here's the part the Ministry does not want made: the strategic value only matters if the operational reality is improving. A rural operator that does not work in rural areas, a strategic telco that cannot keep its own customers connected, a sovereign network that drops 15% of its calls in the mountains, is not a strategic asset. It's a sunk cost.

The Ministry knows this. The 60,000 new towers being deployed under the ₹20,000 crore capex push, the focus on service quality, the AI-led network monitoring, these are the operational fixes. The numbers suggest they are not yet sufficient. They are the right direction.

The real test, five years out

If the 4G stabilisation succeeds, if the indigenous stack reaches a quality level competitive with foreign vendors, if BSNL is positioned clearly as the strategic and rural operator of last resort, and if the government is willing to keep funding the gap between revenue and operating cost in perpetuity, then BSNL will be a slow, unglamorous, perpetual-money-pit success story. The next ten years will see it lose market share in cities, gain modest share in rural India, lose money on operations, and quietly be the network that holds the country together in ways the private operators will never be.

If the 4G stabilisation fails, if TRAI's drive tests continue to show call drops above 10%, if ARPU growth does not come from real customer spend, if the 5G upgrade gets delayed by another three years, then BSNL becomes India's Concor. A state-owned entity that survives on permanent subsidy, technically alive but operationally fading, existing more to keep a workforce employed than to provide a real service. The 60,000 new towers become monuments to a strategic decision that did not work.

The third path is the worst: privatisation. If BSNL is sold to a private buyer, and there have been murmurs of this since 2022, the indigenous 4G stack gets licensed or sold, the rural network gets cherry-picked, and the strategic capability India built over the last six years gets absorbed into a private balance sheet. The state loses the sovereign option. The taxpayer pays for the capability. The private operator gets the upside.

The most likely outcome, given the political economy, is a slow drift toward the first path. The current government has shown no appetite to privatise BSNL. The revival packages continue. The capex continues. The political incentive to keep BSNL alive is significant: it is a major employer, a major rural presence, a visible symbol of state capability.

Slow drift is not a strategy. The country needs the government to make an explicit choice about what BSNL is. The current messaging tries to have it both ways: strategic asset and profitable turnaround, sovereignty and market share, indigenous stack and competitive 5G. All four are not possible simultaneously. Three of them might be, if the trade-offs are acknowledged.

What the numbers should be read as

The right way to read BSNL in 2026 is not as a turnaround story. It's a strategic investment in progress.

The 4G stack is real. India built it. That's historic. The fact that the rollout is slower than the private operators, and that the quality is currently worse, is the cost of the strategic choice. The cost is bearable. ₹3.22 lakh crore is a lot of money, but it's less than one year's defence capital outlay, and the alternative was permanent dependence on foreign vendors for a piece of critical national infrastructure.

The market share decline is real, but it's also the market's verdict on a state operator that was not allowed to compete on equal terms. A private telco that was seven years late to 4G, blocked from using foreign equipment, and tasked with serving the parts of India no one else wants to serve, would have lost market share too.

The service quality problems are real, and they're the most urgent. Until the BSNL customer in a tier-2 city can make a call without dropping, the 4G stack is a strategic capability on paper, not in practice. The next year of capex, the 60,000 new towers, the AI-led monitoring, the battery and power upgrades, is the test. If service quality reaches Airtel-and-Jio parity in 2027, the strategic bet pays off. If it does not, the bet was the right idea with the wrong execution.

The duopoly is real. Jio and Airtel are not going to be dislodged from the top of the Indian market. They're also not going to serve the strategic and rural use cases that BSNL exists to handle. The state did not pour ₹3.22 lakh crore into BSNL to make it the second-largest telco. It poured it in to make it the last telco standing when strategic necessity requires it.

That's a less glamorous story than “BSNL is back.” It's a more honest one. It explains what is actually happening on the ground: the wins, the losses, and the costs of a sovereign bet India is still in the middle of making.

The sovereignty tax

Every country that builds indigenous strategic infrastructure pays a sovereignty tax. The US paid it in the 1950s and 60s when it decided to build its own semiconductor industry rather than rely on imports. South Korea paid it in the 1970s when it built POSCO rather than rely on Japanese steel. China paid it in the 2010s when it built its own commercial aerospace industry rather than rely on Boeing and Airbus.

The tax is real. It's the cost of doing something slower, harder, and more expensive than the alternative, in order to keep a strategic capability in national hands. The tax is not always worth paying. Sometimes the alternative is genuinely better, and the strategic bet is a vanity project.

In BSNL's case, the strategic case is strong. The 4G stack gives India a sovereign option in a sector that every major power now treats as critical infrastructure. The TCS, Tejas, C-DOT capability is the foundation for a potential export industry, selling the stack to countries that want alternatives to Huawei. The rural and strategic network is a public good the market will not provide.

The cost, seven years of customer pain, market share loss, perpetual subsidy, is the sovereignty tax. It's being paid in real time by the BSNL customer in Himachal who drops 15% of his calls. By the BSNL employee in Chennai who wonders if her job is secure. By the Indian taxpayer who funded ₹3.22 lakh crore of revival packages and may fund more.

The bet is not yet won. The next 24 months are the test. If the 4G stack reaches quality parity, if BSNL finds its role as the strategic and rural operator, if the indigenous capability gets commercialised into an export industry, the sovereignty tax will be retrospectively worth it. If not, it will be remembered as a strategic mistake dressed up as a turnaround story.

BSNL is not a fairy tale. It's a strategic investment in progress. We'll know in 2027 which one it ends up being.

Sources:

]]> https://think.shenthar.me/chanakya/bsnls-3-22-lakh-crore-gamble-sovereignty-at-what-cost Tue, 23 Jun 2026 17:34:51 +0000 The Arithmetic of Waste https://think.shenthar.me/chanakya/the-arithmetic-of-waste <![CDATA[In 1987, India built its first waste-to-energy plant in Timarpur, Delhi. Danish design, ₹20 crore, 300 tonnes per day, 3.75 megawatts. It survived twenty-one days. The reason was not corruption or incompetence. It was simpler than that. Indian municipal waste is roughly 60% organic matter, moisture content 40 to 60%, calorific value 800 to 1,100 kilocalories per kilogram. European incinerators are built for waste that burns at twice that energy density. You can't feed wet vegetable peels into a machine designed for dry packaging material and expect it to work. Timarpur was not an outlier. Of the 14 waste-to-energy plants built in India since 1987, seven have shut down. Most survivors burn supplementary fuel (coal or diesel) just to maintain combustion temperatures. Factor that in and their effective electricity cost reaches ₹5 to ₹7 per unit. Solar and wind now deliver at ₹2 to ₹3. The arithmetic does not close. And yet, in 2026, something shifted. The Solid Waste Management Rules, 2026 are expected to mandate source segregation with enforcement teeth. The Central Electricity Authority updated its power purchase agreement guidelines, fixing the revenue unpredictability that made bank financing impossible for over three decades. Investor appetite, per Avendus Capital, has "picked up over the past year, supported by regulatory momentum, net-zero commitments and volatility in global fuel and gas prices." The Indian market sits at $1.56 billion and is projected to reach $1.97 billion by 2034. Asia-Pacific holds 48% of the global $37.29 billion market. What changed between Timarpur's 21 days and the current investor interest? The state did not get better at building plants. It stopped trying to be the operator. Why the plants died The standard explanation is that "incineration was the wrong technology." True in a narrow sense. But the technology choice was never really about technology. It was about procurement. Municipal corporations, bound by competitive bidding mandates, picked the lowest-cost proposals. European incineration firms bid aggressively to export equipment. Nobody in the procurement chain had an incentive to verify whether the feedstock (the actual garbage arriving at the plant gate) would support the promised energy output. The plant would be built. The ribbon would be cut. The waste would arrive unsegregated and too wet. The plant would fail. Another tender would be floated. The feedstock quality problem traced back to a governance failure older than Timarpur itself. The 74th Constitutional Amendment of 1992 was supposed to give urban local bodies functional and financial autonomy. It didn't. Indian municipal revenue sits at roughly 1% of GDP, against 4 to 6% in comparable middle-income economies. A municipal commissioner who can't pay sanitation workers a living wage is not going to build the source-segregation system that makes any downstream plant viable. The Solid Waste Management Rules of 2016 mandated segregation at source. They were never enforced. The informal waste-picker sector, handling 15 to 20% of India's recyclable material, stayed outside every formal framework. The state's response to its own enforcement failure was to build more plants. Seven of them closed. This is the pattern. India's waste problem is not a technology problem or a funding problem. It's a state-capacity problem. The state can write excellent rules. It can't enforce them. It can commission plants. It can't keep them running. The gap between legislative ambition and administrative delivery shows up nowhere more starkly than in Ghazipur, where a landfill leaks 2.17 tonnes of methane per hour while the city around it imports LNG. SATAT and the cheque-writing state In October 2018, the Ministry of Petroleum and Natural Gas launched SATAT: Sustainable Alternative Towards Affordable Transportation. The idea was straightforward: entrepreneurs build Compressed Biogas (CBG) plants using agricultural residue, cattle dung, and municipal solid waste. Public-sector oil companies (IOCL, BPCL, HPCL) buy the output at a guaranteed price. The numbers are worth laying out because they represent something unusual in Indian infrastructure policy: a revenue model that works without the state executing anything. A CBG plant producing 4,800 kilograms per day qualifies for a central subsidy of ₹4 crore, capped at ₹10 crore per project. The oil companies buy the output at ₹62 to ₹72 per kilogram under 5-year offtake agreements. The Compressed Biogas Obligation mandates that 1% of CNG and PNG consumption come from CBG by FY 2025-26, climbing to 3%, then 4%, then 5% by FY 2028-29. The buyers are legally required to buy. This is not a market opportunity in the venture-capital sense of hoping someone wants your product. This is a government-guaranteed monopsony with escalating statutory demand. Plant operators also earn from by-products. Fermented organic manure gets Market Development Assistance at ₹1,500 per metric tonne. Captured CO2 can be sold to industrial users. The combined revenue streams can push project IRRs into the mid-teens. Not spectacular by startup standards, but respectable for infrastructure. And bankable. That last word matters. What SATAT did, almost accidentally, is flip the operating model. For 35 years the state tried to be the operator. It failed. SATAT makes the state the buyer. The entrepreneur handles feedstock, operations, technology, maintenance. The state handles what it can do: write a cheque. This is state retreat dressed as industrial policy. It might also be the only thing that works. Four things an entrepreneur might build If you're mapping startup opportunities in this sector, the useful question is not "what technology is interesting." It's "what specific state failure does this fix?" Decentralised CBG micro-plants. The centralised model (one large plant per city) requires a collection and transport network that municipal corporations can't operate. Decentralised plants (500 to 2,000 kg per day at the ward level, at vegetable markets, at large apartment complexes) eliminate that dependency. The feedstock stays local. The gas travels. SATAT supports this because oil companies buy CBG delivered in high-pressure cascades, not through pipelines that tier-2 and tier-3 cities lack. Unit economics: a 2,000 kg/day plant costs roughly ₹8 to ₹10 crore, nets about ₹2 crore in central subsidy, and generates ₹4.5 to ₹5.2 crore annually from CBG sales at ₹65 per kilogram. Payback: three to four years. Feedstock quality platforms. Timarpur died because the waste was too wet. Every plant failure since traces back to unreliable feedstock. Source segregation is now legally mandated, but enforcement will lag for years. A startup that contracts with bulk waste generators (hotels, mandis, food processors, corporate campuses) for segregated organic waste, verifies quality through IoT sensors and moisture analyzers, and supplies CBG plants with auditable feedstock makes those plants bankable. Revenue model: per-tonne margin, possibly with quality bonuses tied to biogas yield. This is not really a technology business. It's a trust-and-logistics business with sensors. Carbon credit aggregation. CBG plants generate carbon offsets (methane captured instead of released from landfills). Most small operators don't monetize this. Carbon credit verification requires expertise they don't have and the Indian domestic carbon market barely exists. A startup that aggregates small plants, handles registration and verification through international voluntary carbon markets (where biogas methane avoidance trades at a premium), and shares revenue with operators can improve project IRRs by three to five percentage points. This is financial infrastructure that the state hasn't built, so private capital fills the gap. Municipal waste supply chain platforms. The informal waste-picker economy recovers 15 to 20% of recyclables with no technology, no formal contracts, no quality standards. A platform that digitises this supply chain (waste generators to aggregators to processing plants, verified quality, transparent pricing, formal contracts) could push recovery to 30 to 40% while creating formal employment for the very people currently invisible to policy. This is the hardest of these to execute because it depends on behaviour change across thousands of informal actors. It's also the largest addressable market, because it doesn't require new physical infrastructure. Just organisation of the flows that already exist. Each of these fixes a specific governance gap. The micro-plant solves collection logistics. The feedstock platform solves enforcement of segregation. The carbon aggregator solves missing market infrastructure. The supply chain platform solves the institutional blindness that kept 1.5 to 4 million informal waste workers outside every policy framework for three decades. What success looks like Success is not "solving India's waste problem." That framing is uselessly large. India's waste problem is a function of urbanization, consumption, municipal governance, and enforcement. No startup or sector shifts those variables. Success is narrower: building commercially viable, operationally reliable plants that process a measurable fraction of the 50 million tonnes of untreated waste generated annually. And proving that the state-as-buyer model produces better outcomes than state-as-operator. The 5,000 CBG plants SATAT originally targeted by 2023-24 did not materialize. Operational plants number in the low hundreds. The targets were aspirational. But the obligation framework (the legally mandated blending requirement that forces oil companies to buy) creates a demand signal that didn't exist before. That variable matters more than the target count. There's a parallel here. India's telecom revolution did not happen because the DoT got better at building networks. It happened because the government auctioned spectrum to private operators and became a regulator instead of a service provider. India's renewable energy expansion did not happen because state electricity boards got efficient. It happened because reverse auctions created price discovery and private developers built capacity at costs the public sector couldn't match. Waste-to-energy is at a similar inflection point. Earlier, more fragile. The policy architecture is better than it's ever been. SWM Rules 2026 for feedstock quality. SATAT for offtake guarantees. The CBG Obligation for demand. Execution risk, as always in India, lives in the gap between rules as written and rules as enforced. I keep coming back to one number: 21 days. That's how long India's first waste-to-energy plant ran before the unsegregated, high-moisture waste at its gate made it commercially unviable. The plant cost ₹20 crore. It produced nothing. The 13 plants that followed cost more and didn't do much better. The entrepreneurs haven't proven themselves yet. Most SATAT plants are still scaling. But they have a model where revenue tracks performance. Timarpur never had that. The plant got paid regardless of output, because the contract was a municipal service, not an energy business. SATAT inverts that. You get paid per kilogram of biogas delivered. The incentive finally points in the right direction. That is not a grand solution. It's the arithmetic of picking a model where the incentives actually work.]]> In 1987, India built its first waste-to-energy plant in Timarpur, Delhi. Danish design, ₹20 crore, 300 tonnes per day, 3.75 megawatts. It survived twenty-one days.

The reason was not corruption or incompetence. It was simpler than that. Indian municipal waste is roughly 60% organic matter, moisture content 40 to 60%, calorific value 800 to 1,100 kilocalories per kilogram. European incinerators are built for waste that burns at twice that energy density. You can't feed wet vegetable peels into a machine designed for dry packaging material and expect it to work.

Timarpur was not an outlier. Of the 14 waste-to-energy plants built in India since 1987, seven have shut down. Most survivors burn supplementary fuel (coal or diesel) just to maintain combustion temperatures. Factor that in and their effective electricity cost reaches ₹5 to ₹7 per unit. Solar and wind now deliver at ₹2 to ₹3. The arithmetic does not close.

And yet, in 2026, something shifted. The Solid Waste Management Rules, 2026 are expected to mandate source segregation with enforcement teeth. The Central Electricity Authority updated its power purchase agreement guidelines, fixing the revenue unpredictability that made bank financing impossible for over three decades. Investor appetite, per Avendus Capital, has “picked up over the past year, supported by regulatory momentum, net-zero commitments and volatility in global fuel and gas prices.” The Indian market sits at $1.56 billion and is projected to reach $1.97 billion by 2034. Asia-Pacific holds 48% of the global $37.29 billion market.

What changed between Timarpur's 21 days and the current investor interest? The state did not get better at building plants. It stopped trying to be the operator.

Why the plants died

The standard explanation is that “incineration was the wrong technology.” True in a narrow sense. But the technology choice was never really about technology. It was about procurement.

Municipal corporations, bound by competitive bidding mandates, picked the lowest-cost proposals. European incineration firms bid aggressively to export equipment. Nobody in the procurement chain had an incentive to verify whether the feedstock (the actual garbage arriving at the plant gate) would support the promised energy output. The plant would be built. The ribbon would be cut. The waste would arrive unsegregated and too wet. The plant would fail. Another tender would be floated.

The feedstock quality problem traced back to a governance failure older than Timarpur itself. The 74th Constitutional Amendment of 1992 was supposed to give urban local bodies functional and financial autonomy. It didn't. Indian municipal revenue sits at roughly 1% of GDP, against 4 to 6% in comparable middle-income economies. A municipal commissioner who can't pay sanitation workers a living wage is not going to build the source-segregation system that makes any downstream plant viable.

The Solid Waste Management Rules of 2016 mandated segregation at source. They were never enforced. The informal waste-picker sector, handling 15 to 20% of India's recyclable material, stayed outside every formal framework. The state's response to its own enforcement failure was to build more plants. Seven of them closed.

This is the pattern. India's waste problem is not a technology problem or a funding problem. It's a state-capacity problem. The state can write excellent rules. It can't enforce them. It can commission plants. It can't keep them running. The gap between legislative ambition and administrative delivery shows up nowhere more starkly than in Ghazipur, where a landfill leaks 2.17 tonnes of methane per hour while the city around it imports LNG.

SATAT and the cheque-writing state

In October 2018, the Ministry of Petroleum and Natural Gas launched SATAT: Sustainable Alternative Towards Affordable Transportation. The idea was straightforward: entrepreneurs build Compressed Biogas (CBG) plants using agricultural residue, cattle dung, and municipal solid waste. Public-sector oil companies (IOCL, BPCL, HPCL) buy the output at a guaranteed price.

The numbers are worth laying out because they represent something unusual in Indian infrastructure policy: a revenue model that works without the state executing anything.

A CBG plant producing 4,800 kilograms per day qualifies for a central subsidy of ₹4 crore, capped at ₹10 crore per project. The oil companies buy the output at ₹62 to ₹72 per kilogram under 5-year offtake agreements. The Compressed Biogas Obligation mandates that 1% of CNG and PNG consumption come from CBG by FY 2025-26, climbing to 3%, then 4%, then 5% by FY 2028-29. The buyers are legally required to buy. This is not a market opportunity in the venture-capital sense of hoping someone wants your product. This is a government-guaranteed monopsony with escalating statutory demand.

Plant operators also earn from by-products. Fermented organic manure gets Market Development Assistance at ₹1,500 per metric tonne. Captured CO2 can be sold to industrial users. The combined revenue streams can push project IRRs into the mid-teens. Not spectacular by startup standards, but respectable for infrastructure. And bankable. That last word matters.

What SATAT did, almost accidentally, is flip the operating model. For 35 years the state tried to be the operator. It failed. SATAT makes the state the buyer. The entrepreneur handles feedstock, operations, technology, maintenance. The state handles what it can do: write a cheque.

This is state retreat dressed as industrial policy. It might also be the only thing that works.

Four things an entrepreneur might build

If you're mapping startup opportunities in this sector, the useful question is not “what technology is interesting.” It's “what specific state failure does this fix?”

Decentralised CBG micro-plants. The centralised model (one large plant per city) requires a collection and transport network that municipal corporations can't operate. Decentralised plants (500 to 2,000 kg per day at the ward level, at vegetable markets, at large apartment complexes) eliminate that dependency. The feedstock stays local. The gas travels. SATAT supports this because oil companies buy CBG delivered in high-pressure cascades, not through pipelines that tier-2 and tier-3 cities lack. Unit economics: a 2,000 kg/day plant costs roughly ₹8 to ₹10 crore, nets about ₹2 crore in central subsidy, and generates ₹4.5 to ₹5.2 crore annually from CBG sales at ₹65 per kilogram. Payback: three to four years.

Feedstock quality platforms. Timarpur died because the waste was too wet. Every plant failure since traces back to unreliable feedstock. Source segregation is now legally mandated, but enforcement will lag for years. A startup that contracts with bulk waste generators (hotels, mandis, food processors, corporate campuses) for segregated organic waste, verifies quality through IoT sensors and moisture analyzers, and supplies CBG plants with auditable feedstock makes those plants bankable. Revenue model: per-tonne margin, possibly with quality bonuses tied to biogas yield. This is not really a technology business. It's a trust-and-logistics business with sensors.

Carbon credit aggregation. CBG plants generate carbon offsets (methane captured instead of released from landfills). Most small operators don't monetize this. Carbon credit verification requires expertise they don't have and the Indian domestic carbon market barely exists. A startup that aggregates small plants, handles registration and verification through international voluntary carbon markets (where biogas methane avoidance trades at a premium), and shares revenue with operators can improve project IRRs by three to five percentage points. This is financial infrastructure that the state hasn't built, so private capital fills the gap.

Municipal waste supply chain platforms. The informal waste-picker economy recovers 15 to 20% of recyclables with no technology, no formal contracts, no quality standards. A platform that digitises this supply chain (waste generators to aggregators to processing plants, verified quality, transparent pricing, formal contracts) could push recovery to 30 to 40% while creating formal employment for the very people currently invisible to policy. This is the hardest of these to execute because it depends on behaviour change across thousands of informal actors. It's also the largest addressable market, because it doesn't require new physical infrastructure. Just organisation of the flows that already exist.

Each of these fixes a specific governance gap. The micro-plant solves collection logistics. The feedstock platform solves enforcement of segregation. The carbon aggregator solves missing market infrastructure. The supply chain platform solves the institutional blindness that kept 1.5 to 4 million informal waste workers outside every policy framework for three decades.

What success looks like

Success is not “solving India's waste problem.” That framing is uselessly large. India's waste problem is a function of urbanization, consumption, municipal governance, and enforcement. No startup or sector shifts those variables.

Success is narrower: building commercially viable, operationally reliable plants that process a measurable fraction of the 50 million tonnes of untreated waste generated annually. And proving that the state-as-buyer model produces better outcomes than state-as-operator.

The 5,000 CBG plants SATAT originally targeted by 2023-24 did not materialize. Operational plants number in the low hundreds. The targets were aspirational. But the obligation framework (the legally mandated blending requirement that forces oil companies to buy) creates a demand signal that didn't exist before. That variable matters more than the target count.

There's a parallel here. India's telecom revolution did not happen because the DoT got better at building networks. It happened because the government auctioned spectrum to private operators and became a regulator instead of a service provider. India's renewable energy expansion did not happen because state electricity boards got efficient. It happened because reverse auctions created price discovery and private developers built capacity at costs the public sector couldn't match.

Waste-to-energy is at a similar inflection point. Earlier, more fragile. The policy architecture is better than it's ever been. SWM Rules 2026 for feedstock quality. SATAT for offtake guarantees. The CBG Obligation for demand. Execution risk, as always in India, lives in the gap between rules as written and rules as enforced.

I keep coming back to one number: 21 days. That's how long India's first waste-to-energy plant ran before the unsegregated, high-moisture waste at its gate made it commercially unviable. The plant cost ₹20 crore. It produced nothing. The 13 plants that followed cost more and didn't do much better.

The entrepreneurs haven't proven themselves yet. Most SATAT plants are still scaling. But they have a model where revenue tracks performance. Timarpur never had that. The plant got paid regardless of output, because the contract was a municipal service, not an energy business. SATAT inverts that. You get paid per kilogram of biogas delivered. The incentive finally points in the right direction.

That is not a grand solution. It's the arithmetic of picking a model where the incentives actually work.

]]> https://think.shenthar.me/chanakya/the-arithmetic-of-waste Sun, 07 Jun 2026 07:58:04 +0000 India's Semiconductor Push: A Grounded Assessment https://think.shenthar.me/chanakya/indias-semiconductor-push-a-grounded-assessment <![CDATA[India's Semiconductor Push India has a long history of almost making it in chip manufacturing. A fire at the Semiconductor Complex in Mohali in the 1980s killed the country's first shot at fabrication before it could scale. In the mid-2000s, Intel walked away after months of negotiation. These weren't just bad luck. Chip fabrication is brutally hard, ruinously expensive, and depends on ecosystems that take decades to build. India, for a long time, had none of the three. The India Semiconductor Mission landed in December 2021 with ₹76,000 crore. Cautious optimism followed, then the familiar scepticism. When Vedanta-Foxconn fell apart in 2023, the pessimists seemed right: India would consume chips, not make them. Then Micron opened its Sanand facility in Gujarat in March 2026. A $2.75 billion assembly and testing plant. India's first large-scale commercial semiconductor operation. 500,000 square feet of cleanroom, built to Class 1,000 standards, the largest raised-floor facility of its kind globally. The plant takes finished wafers from Micron's global network and turns them into DRAM and NAND: memory chips for laptops, servers, phones. At full ramp, Sanand should handle about 10% of Micron's global packaging and test output. First shipments have gone out. Micron's made-in-India modules to Dell, for locally manufactured laptops. The Tata-PSMC fab in Dholera is a bigger bet by an order of magnitude. ₹910 billion, roughly $9.78 billion. India's most ambitious semiconductor project. Construction hit 50% in April 2026. Trial production expected by late 2026. It will make 28nm and above chips, the kind that go into cars, industrial gear, telecom equipment. Not AI accelerators. The stuff the world actually runs on. Six other approved projects, four more manufacturing units in Odisha, Punjab, Andhra Pradesh. Early stage, pipeline is filling. The broader electronics shift is real too. Mobile phone imports have dropped nearly 77% since FY 2020-21. Over 99% of domestic demand is now met locally. The Electronics Component PLI scheme, announced April 2025 with ₹22,919 crore, already has 22 approved projects worth ₹41,863 crore: camera modules, display modules, PCBs. Across all 14 PLI sectors: cumulative investment over ₹2.16 lakh crore, exports ₹8.3 lakh crore, over 14 lakh jobs. Scale is not the same as depth though. The layers matter The semiconductor value chain has distinct layers. Sanand works at the backend: assembly, testing, packaging. Wafers made elsewhere get diced, assembled, tested, shipped. Essential work. Not the hardest work. Analysts estimate the complexity of an ATMP or OSAT facility at 20-30% of a front-end fab of similar scale. You still need reliable power and cleanrooms. You do not need nanoscale precision, ultra-pure chemical handling, or the multi-year yield learning curves that define real fabrication. Dholera, if it delivers, would be India's first entry into front-end manufacturing. 28nm and above is multiple generations behind the 3nm and 5nm processes that power AI and premium phones. I don't think that's a flaw. Mature nodes serve enormous markets and demand is growing. Let's call it what it is: middle-tier ambitions. Not the frontier. Compare India to established Asian hubs and the gap is sobering. Malaysia spent decades building integrated ecosystems: packaging firms, equipment makers, chemical suppliers, trained labour. Japan leads in semiconductor equipment and materials. The chemical resists, precision etching tools, substrate technologies are as strategically important as the chips. China, for all its geopolitical troubles, deploys technology, talent, and capital at a scale India cannot currently match. India's advantages today are still largely policy-driven and geopolitically enabled. Not ecosystem-led. The window US-China decoupling opened a door India did not earn but can walk through. Multinationals want diversification away from Taiwan, South Korea, and China. India is in the running alongside Vietnam, Malaysia, and Mexico. This is genuinely favourable but also temporary. Once the current restructuring settles, first-mover advantages harden and late entrants compete for scraps. Semiconductor investments are 15 to 20 year bets. Companies look past headline subsidies to policy continuity, infrastructure reliability, supplier ecosystems, talent pipelines. Micron proved that within a controlled industrial zone, India can deliver reliable power, water, land access, and regulatory coordination. The real test is whether that works across multiple states and scales to the more demanding requirements of fabrication plants. People The most encouraging signal from Sanand is not the cleanroom size. It is the people. Of nearly 1,300 employees, about 700 are fresh graduates from Gujarat and neighbouring states: electronics, mechanical, chemical, industrial, materials engineers. They spent 3-6 months training at Micron's facilities in Malaysia and Singapore before coming back to run the plant. This is not a jobs statistic. It is the early formation of a skilled workforce that seeds future investments. Each trained engineer makes the next project easier to staff. That compounds in ways a power plant or a tax break cannot. Realistic success Self-sufficiency is not in the cards. The global chip supply chain is too complex, too capital-intensive, too interdependent. Even the United States, with the CHIPS Act and decades of leadership, is not trying to manufacture everything domestically. The realistic goal is to become a meaningful node in the global chain, not a dominant one. That means backend capacity to attract more investment, proving front-end fabrication can work here, developing a supplier base, and building a talent pipeline. The first is happening. The second hinges on Tata-PSMC over the next 18-24 months. The other two have barely started. What I think I have been watching India's semiconductor story for years, and the pattern has always been the same: announcements, optimism, then nothing. This time feels different, not because the government is trying harder, but because there is finally something physical to point at. A plant running. Modules shipping. Engineers trained. The gap is still enormous. Backend manufacturing is a start, not an end. Mature-node fabrication is necessary but not the frontier. The geopolitical window is real but will not stay open forever. Here is what has changed: India finally has a basis for the next attempt. After four decades of almost-making-it, that counts for something.]]> India's Semiconductor Push

India has a long history of almost making it in chip manufacturing. A fire at the Semiconductor Complex in Mohali in the 1980s killed the country's first shot at fabrication before it could scale. In the mid-2000s, Intel walked away after months of negotiation. These weren't just bad luck. Chip fabrication is brutally hard, ruinously expensive, and depends on ecosystems that take decades to build. India, for a long time, had none of the three.

The India Semiconductor Mission landed in December 2021 with ₹76,000 crore. Cautious optimism followed, then the familiar scepticism. When Vedanta-Foxconn fell apart in 2023, the pessimists seemed right: India would consume chips, not make them.

Then Micron opened its Sanand facility in Gujarat in March 2026. A $2.75 billion assembly and testing plant. India's first large-scale commercial semiconductor operation. 500,000 square feet of cleanroom, built to Class 1,000 standards, the largest raised-floor facility of its kind globally. The plant takes finished wafers from Micron's global network and turns them into DRAM and NAND: memory chips for laptops, servers, phones.

At full ramp, Sanand should handle about 10% of Micron's global packaging and test output. First shipments have gone out. Micron's made-in-India modules to Dell, for locally manufactured laptops.

The Tata-PSMC fab in Dholera is a bigger bet by an order of magnitude. ₹910 billion, roughly $9.78 billion. India's most ambitious semiconductor project. Construction hit 50% in April 2026. Trial production expected by late 2026. It will make 28nm and above chips, the kind that go into cars, industrial gear, telecom equipment. Not AI accelerators. The stuff the world actually runs on.

Six other approved projects, four more manufacturing units in Odisha, Punjab, Andhra Pradesh. Early stage, pipeline is filling.

The broader electronics shift is real too. Mobile phone imports have dropped nearly 77% since FY 2020-21. Over 99% of domestic demand is now met locally. The Electronics Component PLI scheme, announced April 2025 with ₹22,919 crore, already has 22 approved projects worth ₹41,863 crore: camera modules, display modules, PCBs. Across all 14 PLI sectors: cumulative investment over ₹2.16 lakh crore, exports ₹8.3 lakh crore, over 14 lakh jobs.

Scale is not the same as depth though.

The layers matter

The semiconductor value chain has distinct layers. Sanand works at the backend: assembly, testing, packaging. Wafers made elsewhere get diced, assembled, tested, shipped. Essential work. Not the hardest work.

Analysts estimate the complexity of an ATMP or OSAT facility at 20-30% of a front-end fab of similar scale. You still need reliable power and cleanrooms. You do not need nanoscale precision, ultra-pure chemical handling, or the multi-year yield learning curves that define real fabrication.

Dholera, if it delivers, would be India's first entry into front-end manufacturing. 28nm and above is multiple generations behind the 3nm and 5nm processes that power AI and premium phones. I don't think that's a flaw. Mature nodes serve enormous markets and demand is growing. Let's call it what it is: middle-tier ambitions. Not the frontier.

Compare India to established Asian hubs and the gap is sobering. Malaysia spent decades building integrated ecosystems: packaging firms, equipment makers, chemical suppliers, trained labour. Japan leads in semiconductor equipment and materials. The chemical resists, precision etching tools, substrate technologies are as strategically important as the chips. China, for all its geopolitical troubles, deploys technology, talent, and capital at a scale India cannot currently match.

India's advantages today are still largely policy-driven and geopolitically enabled. Not ecosystem-led.

The window

US-China decoupling opened a door India did not earn but can walk through. Multinationals want diversification away from Taiwan, South Korea, and China. India is in the running alongside Vietnam, Malaysia, and Mexico.

This is genuinely favourable but also temporary. Once the current restructuring settles, first-mover advantages harden and late entrants compete for scraps.

Semiconductor investments are 15 to 20 year bets. Companies look past headline subsidies to policy continuity, infrastructure reliability, supplier ecosystems, talent pipelines. Micron proved that within a controlled industrial zone, India can deliver reliable power, water, land access, and regulatory coordination. The real test is whether that works across multiple states and scales to the more demanding requirements of fabrication plants.

People

The most encouraging signal from Sanand is not the cleanroom size. It is the people. Of nearly 1,300 employees, about 700 are fresh graduates from Gujarat and neighbouring states: electronics, mechanical, chemical, industrial, materials engineers. They spent 3-6 months training at Micron's facilities in Malaysia and Singapore before coming back to run the plant.

This is not a jobs statistic. It is the early formation of a skilled workforce that seeds future investments. Each trained engineer makes the next project easier to staff. That compounds in ways a power plant or a tax break cannot.

Realistic success

Self-sufficiency is not in the cards. The global chip supply chain is too complex, too capital-intensive, too interdependent. Even the United States, with the CHIPS Act and decades of leadership, is not trying to manufacture everything domestically.

The realistic goal is to become a meaningful node in the global chain, not a dominant one. That means backend capacity to attract more investment, proving front-end fabrication can work here, developing a supplier base, and building a talent pipeline.

The first is happening. The second hinges on Tata-PSMC over the next 18-24 months. The other two have barely started.

What I think

I have been watching India's semiconductor story for years, and the pattern has always been the same: announcements, optimism, then nothing. This time feels different, not because the government is trying harder, but because there is finally something physical to point at. A plant running. Modules shipping. Engineers trained.

The gap is still enormous. Backend manufacturing is a start, not an end. Mature-node fabrication is necessary but not the frontier. The geopolitical window is real but will not stay open forever.

Here is what has changed: India finally has a basis for the next attempt. After four decades of almost-making-it, that counts for something.

]]> https://think.shenthar.me/chanakya/indias-semiconductor-push-a-grounded-assessment Sat, 06 Jun 2026 06:51:01 +0000 Semiconductors Are the New Nukes: India's Chip Strategy Needs a Security Lens https://think.shenthar.me/chanakya/semiconductors-are-the-new-nukes-indias-chip-strategy-needs-a-security-lens <![CDATA[What happens to India's armed forces if the Taiwan Strait goes hot? This isn't a hypothetical. It's a planning scenario that India's strategic establishment has no public answer for. TSMC manufactures over 90% of the world's advanced semiconductors. Every chip below five nanometres that powers modern weapons systems, satellite guidance, encrypted communications, and AI-driven battlefield analytics. All of it, with almost no exceptions, comes from a single island 160 kilometres off the mainland Chinese coast. The global semiconductor supply chain has become the most concentrated strategic dependency on earth. If the 1973 oil embargo paralysed Western economies by choking a single commodity, a Taiwan Strait disruption would be orders of magnitude worse. Not fuel. Not steel. The substrate on which all modern military and economic activity runs. Advanced chip prices would jump somewhere between ten and fifty times overnight. Lead times for defence systems that already run eighteen to twenty-four months would effectively go to infinity. And India, which imports nearly every advanced semiconductor in its weapons platforms, its telecom stack, and its AI compute base, would discover that its much-celebrated chip strategy was built for the wrong war. The Taiwan problem is also India's problem The standard Indian policy discourse on semiconductors goes like this: we have a $45 to 50 billion domestic market growing toward $110 billion by 2030. We need jobs. We need manufacturing. We have the India Semiconductor Mission, we have the Tata Dholera fab coming online by late 2026, we have Micron's ATMP facility in Sanand, the HCL-Foxconn joint venture in Jewar. Ashwini Vaishnaw has said Dholera will let Indian startups do domestic tape-outs. The story is about catching up, about joining the club. This framing gets the question exactly wrong. The question isn't whether India can make enough chips to meet domestic consumption. It's whether India can keep its military running in a world where the global chip supply is held hostage to a single geopolitical flashpoint that India has zero control over. TSMC's revenue alone accounts for roughly 8% of Taiwan's GDP. Semiconductors and related industries make up roughly 40% of Taiwan's total exports. This isn't diversification. Strategists call this the bullseye effect: a country's value becomes so concentrated in one sector that it turns into a target for every major power. The United States has effectively integrated Taiwan into its technology alliance through the CHIPS Act and export controls. China regards Taiwan's chip capacity as a strategic asset to be secured upon reunification. Japan and the EU compete for TSMC partnerships to guarantee their own supply. India, which has no TSMC, has no seat at this table. The country's semiconductor demand is projected at $110 billion by 2030. Its domestic production capacity, even with Dholera at full ramp, will cover a fraction of that. And it will cover mature nodes: 28 to 90 nanometres. Not the sub-five-nanometre chips that defence and AI actually depend on. What India is building (and what it isn't) The Dholera fab is genuinely significant. ₹91,000 crore in investment, about $11 billion. A 300-millimetre wafer facility with 50,000 wafers per month capacity. Technology transfer from Powerchip of Taiwan. An ASML equipment agreement signed in May 2026. The Tata-Intel memorandum of understanding from December 2025 suggests Dholera may eventually package Intel products locally. Dholera is a mature-node fab, though. It makes chips in the 28-to-90-nanometre range: power management ICs, display drivers, microcontrollers, automotive chips. These are essential components. They're not the components a country at war depends on. India's defence platforms run on imported advanced semiconductors. The Tejas fighter. The Akash surface-to-air missile. The S-400 systems coming online. Naval combat management suites. Satellite payloads. Every one of them. India's AI ambitions, its data centre build-out, its 5G and 6G telecom infrastructure: all depend on chips made a short missile flight from the Chinese coast. Dholera doesn't solve this. Micron's Sanand packaging plant doesn't solve this. These are backend and mid-node industrial capabilities that address the civilian electronics supply chain. They don't touch the strategic vulnerability at the core of India's defence modernisation. China knows this While India builds ATMP facilities and allocates ISM 2.0 budgets, China has been systematically working around US export controls to build indigenous advanced chip capability. SMIC, China's largest state-backed foundry, has achieved five-nanometre production using stockpiled DUV lithography machines, older equipment China accumulated before sanctions tightened. Huawei's Mate 60 smartphone shipped with a seven-nanometre SMIC chip in 2023, surprising Western intelligence assessments that had projected such capability was years away. The sanctions are working imperfectly. They've slowed China's advance but not stopped it. And China treats semiconductors as a strategic weapons programme, not an industrial policy project. The state directs capital, absorbs losses, measures success in capability rather than quarterly returns. India's approach treats semiconductors as an investment destination. The ISM offers 50% capital expenditure support. State governments compete with additional incentives. The framing is about jobs, FDI, the Make in India narrative. As industrial policy, that framing is fine. For national security, it's useless. What a security lens would demand If semiconductors are the new fissile material, India's chip strategy needs a different calibration. Here's what that looks like. First, a threat-based inventory. India needs to know exactly which chips in which defence systems come from which fabs. Every imported microcontroller in a BrahMos missile, every FPGA in an electronically scanned radar, every processor in a satellite payload, mapped against the fab location and the geopolitical risk of disruption. This inventory almost certainly exists somewhere within the armed forces and DRDO. It needs to be elevated to a national-level strategic assessment with publicly articulated consequences. Not to name suppliers in open source. To force the political system to confront the dependency. Second, a sovereign capability roadmap for defence-grade chips. Not all chips need to be made in India. The ones that control weapons release, that encrypt strategic communications, that guide terminal-phase interceptors: those need a sovereign fabrication pathway. This doesn't mean a dedicated defence fab at three nanometres, which would be economically irrational. It means a trusted foundry arrangement where India can audit the supply chain end to end and guarantee that no foreign state can insert a kill switch or a supply denial. Likely with a Quad partner. The US runs a Trusted Foundry programme for exactly this reason. India needs an equivalent. Third, a stockpile and substitution doctrine. If the Taiwan Strait closes for eighteen months, what does India do? "Wait and hope" is not an answer. India needs a strategic semiconductor reserve, not just for civilian continuity but for military readiness. Plus a doctrine that specifies what gets substituted, what gets conserved, and what gets prioritised when supply collapses. The armed forces maintain fuel reserves, ammunition reserves, medical reserves. They don't maintain a semiconductor reserve. Planning alone can close this gap. Fourth, design-for-sovereignty standards. India has real strengths in chip design. The country trains roughly 20% of the world's VLSI design engineers. Designing a chip and manufacturing it are different capabilities, though, and the current ecosystem routes most Indian-designed chips to TSMC or Samsung for fabrication. A security lens would demand that chips destined for defence applications be designed from the start for fabrication on nodes India can eventually control. Twenty-eight to sixty-five nanometres, the range Dholera covers. That means accepting performance trade-offs for sovereignty. China has been doing this for a decade with its military chips. The calculus There's a habit in Indian strategic discourse of treating technology policy as a branch of economics. Semiconductors employ people. Semiconductors attract investment. Semiconductors reduce the trade deficit. All true. None of it matters if the supply chain snaps. Semiconductors are different from other strategic commodities because of their systemic character. An oil shock makes everything more expensive. A semiconductor shock makes some things impossible. You can't improvise a five-nanometre fabrication process the way you can ration fuel. Lead time for a new advanced fab is five to seven years, and that's assuming you have the ASML machines, the ultra-pure water infrastructure, the chemical supply chain, and the thousand-plus process engineers who can keep yields above 80%. India has none of these at scale today. This isn't an argument against Dholera. Dholera solves a different problem. The fab will be valuable. It will build industrial muscle, train engineers, create a domestic supplier ecosystem. It will not keep an S-400 battery operational if TSMC stops shipping. The security lens doesn't demand that India match TSMC node for node. It demands that India understand, clearly, where its dependencies are lethal and where they're merely expensive. The difference between "we'd rather not pay this import bill" and "we lose military capability if this supply stops" is the difference between economic policy and national security. India's semiconductor strategy, as it stands, addresses the first. The second is still unanswered. TSMC's Arizona fab, Samsung's Texas facility, Intel's Ohio campus: these are the Western answer to Taiwan risk. They'll take years to reach scale. India isn't even in the queue for advanced-node capacity. Every year without a security-graded semiconductor doctrine is a year India's defence rests on a foundation it doesn't control. ]]> What happens to India's armed forces if the Taiwan Strait goes hot?

This isn't a hypothetical. It's a planning scenario that India's strategic establishment has no public answer for. TSMC manufactures over 90% of the world's advanced semiconductors. Every chip below five nanometres that powers modern weapons systems, satellite guidance, encrypted communications, and AI-driven battlefield analytics. All of it, with almost no exceptions, comes from a single island 160 kilometres off the mainland Chinese coast.

The global semiconductor supply chain has become the most concentrated strategic dependency on earth. If the 1973 oil embargo paralysed Western economies by choking a single commodity, a Taiwan Strait disruption would be orders of magnitude worse. Not fuel. Not steel. The substrate on which all modern military and economic activity runs.

Advanced chip prices would jump somewhere between ten and fifty times overnight. Lead times for defence systems that already run eighteen to twenty-four months would effectively go to infinity. And India, which imports nearly every advanced semiconductor in its weapons platforms, its telecom stack, and its AI compute base, would discover that its much-celebrated chip strategy was built for the wrong war.

The Taiwan problem is also India's problem

The standard Indian policy discourse on semiconductors goes like this: we have a $45 to 50 billion domestic market growing toward $110 billion by 2030. We need jobs. We need manufacturing. We have the India Semiconductor Mission, we have the Tata Dholera fab coming online by late 2026, we have Micron's ATMP facility in Sanand, the HCL-Foxconn joint venture in Jewar. Ashwini Vaishnaw has said Dholera will let Indian startups do domestic tape-outs. The story is about catching up, about joining the club.

This framing gets the question exactly wrong.

The question isn't whether India can make enough chips to meet domestic consumption. It's whether India can keep its military running in a world where the global chip supply is held hostage to a single geopolitical flashpoint that India has zero control over.

TSMC's revenue alone accounts for roughly 8% of Taiwan's GDP. Semiconductors and related industries make up roughly 40% of Taiwan's total exports. This isn't diversification. Strategists call this the bullseye effect: a country's value becomes so concentrated in one sector that it turns into a target for every major power. The United States has effectively integrated Taiwan into its technology alliance through the CHIPS Act and export controls. China regards Taiwan's chip capacity as a strategic asset to be secured upon reunification. Japan and the EU compete for TSMC partnerships to guarantee their own supply.

India, which has no TSMC, has no seat at this table. The country's semiconductor demand is projected at $110 billion by 2030. Its domestic production capacity, even with Dholera at full ramp, will cover a fraction of that. And it will cover mature nodes: 28 to 90 nanometres. Not the sub-five-nanometre chips that defence and AI actually depend on.

What India is building (and what it isn't)

The Dholera fab is genuinely significant. ₹91,000 crore in investment, about $11 billion. A 300-millimetre wafer facility with 50,000 wafers per month capacity. Technology transfer from Powerchip of Taiwan. An ASML equipment agreement signed in May 2026. The Tata-Intel memorandum of understanding from December 2025 suggests Dholera may eventually package Intel products locally.

Dholera is a mature-node fab, though. It makes chips in the 28-to-90-nanometre range: power management ICs, display drivers, microcontrollers, automotive chips. These are essential components. They're not the components a country at war depends on.

India's defence platforms run on imported advanced semiconductors. The Tejas fighter. The Akash surface-to-air missile. The S-400 systems coming online. Naval combat management suites. Satellite payloads. Every one of them. India's AI ambitions, its data centre build-out, its 5G and 6G telecom infrastructure: all depend on chips made a short missile flight from the Chinese coast.

Dholera doesn't solve this. Micron's Sanand packaging plant doesn't solve this. These are backend and mid-node industrial capabilities that address the civilian electronics supply chain. They don't touch the strategic vulnerability at the core of India's defence modernisation.

China knows this

While India builds ATMP facilities and allocates ISM 2.0 budgets, China has been systematically working around US export controls to build indigenous advanced chip capability. SMIC, China's largest state-backed foundry, has achieved five-nanometre production using stockpiled DUV lithography machines, older equipment China accumulated before sanctions tightened. Huawei's Mate 60 smartphone shipped with a seven-nanometre SMIC chip in 2023, surprising Western intelligence assessments that had projected such capability was years away.

The sanctions are working imperfectly. They've slowed China's advance but not stopped it. And China treats semiconductors as a strategic weapons programme, not an industrial policy project. The state directs capital, absorbs losses, measures success in capability rather than quarterly returns.

India's approach treats semiconductors as an investment destination. The ISM offers 50% capital expenditure support. State governments compete with additional incentives. The framing is about jobs, FDI, the Make in India narrative.

As industrial policy, that framing is fine. For national security, it's useless.

What a security lens would demand

If semiconductors are the new fissile material, India's chip strategy needs a different calibration. Here's what that looks like.

First, a threat-based inventory. India needs to know exactly which chips in which defence systems come from which fabs. Every imported microcontroller in a BrahMos missile, every FPGA in an electronically scanned radar, every processor in a satellite payload, mapped against the fab location and the geopolitical risk of disruption. This inventory almost certainly exists somewhere within the armed forces and DRDO. It needs to be elevated to a national-level strategic assessment with publicly articulated consequences. Not to name suppliers in open source. To force the political system to confront the dependency.

Second, a sovereign capability roadmap for defence-grade chips. Not all chips need to be made in India. The ones that control weapons release, that encrypt strategic communications, that guide terminal-phase interceptors: those need a sovereign fabrication pathway. This doesn't mean a dedicated defence fab at three nanometres, which would be economically irrational. It means a trusted foundry arrangement where India can audit the supply chain end to end and guarantee that no foreign state can insert a kill switch or a supply denial. Likely with a Quad partner. The US runs a Trusted Foundry programme for exactly this reason. India needs an equivalent.

Third, a stockpile and substitution doctrine. If the Taiwan Strait closes for eighteen months, what does India do? “Wait and hope” is not an answer. India needs a strategic semiconductor reserve, not just for civilian continuity but for military readiness. Plus a doctrine that specifies what gets substituted, what gets conserved, and what gets prioritised when supply collapses. The armed forces maintain fuel reserves, ammunition reserves, medical reserves. They don't maintain a semiconductor reserve. Planning alone can close this gap.

Fourth, design-for-sovereignty standards. India has real strengths in chip design. The country trains roughly 20% of the world's VLSI design engineers. Designing a chip and manufacturing it are different capabilities, though, and the current ecosystem routes most Indian-designed chips to TSMC or Samsung for fabrication. A security lens would demand that chips destined for defence applications be designed from the start for fabrication on nodes India can eventually control. Twenty-eight to sixty-five nanometres, the range Dholera covers. That means accepting performance trade-offs for sovereignty. China has been doing this for a decade with its military chips.

The calculus

There's a habit in Indian strategic discourse of treating technology policy as a branch of economics. Semiconductors employ people. Semiconductors attract investment. Semiconductors reduce the trade deficit. All true. None of it matters if the supply chain snaps.

Semiconductors are different from other strategic commodities because of their systemic character. An oil shock makes everything more expensive. A semiconductor shock makes some things impossible. You can't improvise a five-nanometre fabrication process the way you can ration fuel. Lead time for a new advanced fab is five to seven years, and that's assuming you have the ASML machines, the ultra-pure water infrastructure, the chemical supply chain, and the thousand-plus process engineers who can keep yields above 80%. India has none of these at scale today.

This isn't an argument against Dholera. Dholera solves a different problem. The fab will be valuable. It will build industrial muscle, train engineers, create a domestic supplier ecosystem. It will not keep an S-400 battery operational if TSMC stops shipping.

The security lens doesn't demand that India match TSMC node for node. It demands that India understand, clearly, where its dependencies are lethal and where they're merely expensive. The difference between “we'd rather not pay this import bill” and “we lose military capability if this supply stops” is the difference between economic policy and national security. India's semiconductor strategy, as it stands, addresses the first. The second is still unanswered.

TSMC's Arizona fab, Samsung's Texas facility, Intel's Ohio campus: these are the Western answer to Taiwan risk. They'll take years to reach scale. India isn't even in the queue for advanced-node capacity. Every year without a security-graded semiconductor doctrine is a year India's defence rests on a foundation it doesn't control.

]]> https://think.shenthar.me/chanakya/semiconductors-are-the-new-nukes-indias-chip-strategy-needs-a-security-lens Sat, 06 Jun 2026 06:47:02 +0000 India's Defence Export Story https://think.shenthar.me/chanakya/indias-defence-export-story <![CDATA[India's Defence Export Story India's defence exports hit ₹38,424 crore in FY26. That is a 62% jump from last year. 25 times what they were in FY17. The government is aiming for ₹50,000 crore by FY29. The numbers land differently depending on how you look at them. What India actually sells BrahMos supersonic cruise missiles. Akash air defence systems. Naval vessels. Swathi radars. Artillery systems. These are proper platforms, not just components or spare parts. When a country buys a BrahMos from India, it is buying a working weapon system with genuine capability. Over 80 countries now buy Indian defence equipment. The buyers are mostly in Southeast Asia, the Middle East, and Africa. Traditional buyers of Russian gear looking for alternatives. Small navies that cannot afford Western systems. Countries that do not want their defence procurement tied to political conditionality. This is where India's historical non-alignment and its current positioning pay off. India can sell to almost anyone without the baggage that comes with American or European arms deals. The production side Defence production hit ₹1.54 lakh crore in FY25, more than triple what it was a decade ago. The target is ₹3 lakh crore by FY29. Aggressive but grounded in the current trajectory. The procurement shift is the bigger story. In FY25, the Ministry of Defence signed 193 contracts worth ₹2.09 lakh crore. Of these, 92% by volume and 81% by value went to Indian firms. Ten years ago those numbers were nearly reversed. About 65% of India's defence equipment is now made domestically. A decade ago, import dependence was 65-70%. The composition has changed too. Private industry now accounts for 45% of defence exports, up from almost nothing a decade ago. DPSUs still lead at 55% but the gap is narrowing. Private companies bring different efficiencies: faster supply chains, better cost control, more aggressive marketing abroad. Bharat Forge makes artillery systems. Tata Advanced Systems makes aerospace components. L&T builds naval platforms. The list goes on. What the export number actually means ₹38,424 crore is real money. It is about \$4.6 billion. For context, the global arms trade is roughly \$100 billion annually. India's share is around 4-5% of the global export market now. Not dominant. Respectable for a country that barely exported anything a decade ago. The government target of ₹50,000 crore by FY29 is achievable at current growth rates. Whether the growth sustains past that depends on repeat orders, not first-time buyers. I am more interested in the repeat customer rate than the headline number. A country importing a BrahMos battery and coming back for another one three years later tells you more than the initial deal ever could. The data on this is not public. I wish it was. The import side India remains the world's second-largest arms importer. That number gets lost in all the export celebration. Russia's share of India's defence imports dropped from 70% (2011-2015) to 40% (2021-2025). France now accounts for 29%. Israel 15%. This is not a story of India reducing import dependence. It is a story of India diversifying its supplier base for geopolitical reasons. The gap between exports (₹38,424 crore) and imports remains enormous. India still spends far more on buying weapons than selling them. That is not surprising for a country with two nuclear-armed neighbours and contested borders, but it puts the export narrative in perspective. What the policy machine did The structural changes are real. 75% of the capital acquisition budget is earmarked for domestic industry. Two Defence Industrial Corridors are operational in UP and Tamil Nadu. 430 iDEX contracts signed with over 600 startups and MSMEs. The negative import lists forced the military to buy Indian even when it wanted foreign gear. These policies worked because they removed the option to import. When the army cannot buy foreign rifles, it figures out how to make Indian ones work. That is not elegant. It is effective. The defence budget has tripled since FY14 to ₹7.85 lakh crore in FY27. That is 2% of GDP and roughly 15% of the Union Budget. The capital outlay component, which buys actual equipment rather than salaries and pensions, is ₹2.19 lakh crore and growing. That is the number that matters for the manufacturing ecosystem. Where it gets complicated Private companies are still cautious about capacity investment without multi-year order books. The MoD's procurement cycles are slow and unpredictable. A company that builds a factory to make artillery barrels needs assurance that orders will arrive for more than one batch. Export customers also want maintenance and spares support for the life of the equipment. That means India needs a sustained presence in the countries it sells to. Not just sales teams in hotels. Technical teams, supply chains, training programmes. The ecosystem that supports a sale is harder to build than the sale itself. The BrahMos success is instructive. India and Russia co-developed it. Having a partner with established export networks and credibility helped open doors. Pure Indian platforms like Akash do not have that advantage. They are winning on price and reliability, not brand. What I think The export story is genuine. The numbers are not manufactured. 25x growth in nine years is real growth by any measure. The question is whether this is a structural shift or a policy-driven spike. My reading is that enough has changed on the manufacturing side that the trend has durability. Private sector participation, the startup ecosystem, the negative import lists, the diversification of import sources: these are not reversible by a change in government or a budget cut. India will not become a top-tier arms exporter in this decade. It does not have to be. Becoming a meaningful mid-tier exporter with a distinctive niche in missiles, naval systems, and artillery is both realistic and strategically useful. Every BrahMos battery sold to a Southeast Asian navy is a relationship that deepens over decades. The real payoff is not the export revenue. It is the domestic industrial base that exports force you to build. You cannot sell weapons abroad that your own military does not trust. Export credibility forces quality. That discipline is the most durable benefit of the whole exercise.]]> India's Defence Export Story

India's defence exports hit ₹38,424 crore in FY26. That is a 62% jump from last year. 25 times what they were in FY17. The government is aiming for ₹50,000 crore by FY29.

The numbers land differently depending on how you look at them.

What India actually sells

BrahMos supersonic cruise missiles. Akash air defence systems. Naval vessels. Swathi radars. Artillery systems. These are proper platforms, not just components or spare parts. When a country buys a BrahMos from India, it is buying a working weapon system with genuine capability.

Over 80 countries now buy Indian defence equipment. The buyers are mostly in Southeast Asia, the Middle East, and Africa. Traditional buyers of Russian gear looking for alternatives. Small navies that cannot afford Western systems. Countries that do not want their defence procurement tied to political conditionality.

This is where India's historical non-alignment and its current positioning pay off. India can sell to almost anyone without the baggage that comes with American or European arms deals.

The production side

Defence production hit ₹1.54 lakh crore in FY25, more than triple what it was a decade ago. The target is ₹3 lakh crore by FY29. Aggressive but grounded in the current trajectory.

The procurement shift is the bigger story. In FY25, the Ministry of Defence signed 193 contracts worth ₹2.09 lakh crore. Of these, 92% by volume and 81% by value went to Indian firms. Ten years ago those numbers were nearly reversed. About 65% of India's defence equipment is now made domestically. A decade ago, import dependence was 65-70%.

The composition has changed too. Private industry now accounts for 45% of defence exports, up from almost nothing a decade ago. DPSUs still lead at 55% but the gap is narrowing. Private companies bring different efficiencies: faster supply chains, better cost control, more aggressive marketing abroad. Bharat Forge makes artillery systems. Tata Advanced Systems makes aerospace components. L&T builds naval platforms. The list goes on.

What the export number actually means

₹38,424 crore is real money. It is about \$4.6 billion. For context, the global arms trade is roughly \$100 billion annually. India's share is around 4-5% of the global export market now. Not dominant. Respectable for a country that barely exported anything a decade ago.

The government target of ₹50,000 crore by FY29 is achievable at current growth rates. Whether the growth sustains past that depends on repeat orders, not first-time buyers.

I am more interested in the repeat customer rate than the headline number. A country importing a BrahMos battery and coming back for another one three years later tells you more than the initial deal ever could. The data on this is not public. I wish it was.

The import side

India remains the world's second-largest arms importer. That number gets lost in all the export celebration.

Russia's share of India's defence imports dropped from 70% (2011-2015) to 40% (2021-2025). France now accounts for 29%. Israel 15%. This is not a story of India reducing import dependence. It is a story of India diversifying its supplier base for geopolitical reasons.

The gap between exports (₹38,424 crore) and imports remains enormous. India still spends far more on buying weapons than selling them. That is not surprising for a country with two nuclear-armed neighbours and contested borders, but it puts the export narrative in perspective.

What the policy machine did

The structural changes are real. 75% of the capital acquisition budget is earmarked for domestic industry. Two Defence Industrial Corridors are operational in UP and Tamil Nadu. 430 iDEX contracts signed with over 600 startups and MSMEs. The negative import lists forced the military to buy Indian even when it wanted foreign gear.

These policies worked because they removed the option to import. When the army cannot buy foreign rifles, it figures out how to make Indian ones work. That is not elegant. It is effective.

The defence budget has tripled since FY14 to ₹7.85 lakh crore in FY27. That is 2% of GDP and roughly 15% of the Union Budget. The capital outlay component, which buys actual equipment rather than salaries and pensions, is ₹2.19 lakh crore and growing. That is the number that matters for the manufacturing ecosystem.

Where it gets complicated

Private companies are still cautious about capacity investment without multi-year order books. The MoD's procurement cycles are slow and unpredictable. A company that builds a factory to make artillery barrels needs assurance that orders will arrive for more than one batch.

Export customers also want maintenance and spares support for the life of the equipment. That means India needs a sustained presence in the countries it sells to. Not just sales teams in hotels. Technical teams, supply chains, training programmes. The ecosystem that supports a sale is harder to build than the sale itself.

The BrahMos success is instructive. India and Russia co-developed it. Having a partner with established export networks and credibility helped open doors. Pure Indian platforms like Akash do not have that advantage. They are winning on price and reliability, not brand.

What I think

The export story is genuine. The numbers are not manufactured. 25x growth in nine years is real growth by any measure.

The question is whether this is a structural shift or a policy-driven spike. My reading is that enough has changed on the manufacturing side that the trend has durability. Private sector participation, the startup ecosystem, the negative import lists, the diversification of import sources: these are not reversible by a change in government or a budget cut.

India will not become a top-tier arms exporter in this decade. It does not have to be. Becoming a meaningful mid-tier exporter with a distinctive niche in missiles, naval systems, and artillery is both realistic and strategically useful. Every BrahMos battery sold to a Southeast Asian navy is a relationship that deepens over decades.

The real payoff is not the export revenue. It is the domestic industrial base that exports force you to build. You cannot sell weapons abroad that your own military does not trust. Export credibility forces quality. That discipline is the most durable benefit of the whole exercise.

]]> https://think.shenthar.me/chanakya/indias-defence-export-story Sun, 24 May 2026 14:12:42 +0000 The Slow Failure of Money https://think.shenthar.me/chanakya/the-slow-failure-of-money <![CDATA[Money in the modern world rarely collapses violently; it decays politely. There are no dramatic moments where society suddenly realizes the currency is broken. Instead, purchasing power erodes slowly enough that people adapt psychologically before they recognize what changed. India is a powerful example of this phenomenon. In 1947, the rupee was roughly at parity with the US dollar. Today, the exchange rate approaches ₹95–₹100 per dollar. Yet India did not experience monetary collapse in the conventional sense. GDP expanded, industries emerged, millions escaped poverty, and salaries increased enormously in nominal terms. But the denominator itself weakened. This is the subtlety most inflation discussions miss: fiat systems rarely fail catastrophically; they fail gradually enough that societies normalize the decline while continuing to participate within it. A person who saved ₹1 lakh in the early 2000s still “has” ₹1 lakh numerically, but its claim on scarce assets has deteriorated profoundly. It buys less housing, less gold, less international mobility, less ownership, and increasingly less economic security. Humans think in absolute numbers while economies operate in relative purchasing power, which is why modern professionals often experience the strange contradiction of earning many multiples more than previous generations while simultaneously feeling less financially stable. Bangalore apartments that once cost ₹20–30 lakh now routinely exceed ₹2 crore, gold moved from around ₹7,000 per 10 grams to above ₹90,000, and globally priced systems—from semiconductors to AI infrastructure to education—became progressively harder to access through locally denominated income. The salary denominator increased; the asset denominator increased faster. This is not necessarily a malfunction of the system. In many ways, it is the system functioning precisely as designed. Modern economies are structurally dependent on debt expansion: governments borrow, corporations borrow, consumers borrow, and financial markets themselves are built upon assumptions of perpetual future growth financed through present liquidity. Under such conditions, perfectly stable money becomes economically inconvenient. If liabilities expand aggressively while purchasing power remains rigid, repayment burdens become increasingly painful and contraction becomes politically dangerous. Inflation therefore acts as a mechanism for gradually softening debt pressure through denominator dilution rather than explicit default. This is why central banks target inflation rather than eliminate it entirely. The architecture of modern macroeconomics quietly assumes expanding consumption, expanding credit, expanding liquidity, and expanding asset prices. Historically, empires achieved similar outcomes more crudely. Roman rulers diluted silver coins to finance military and state expenditure while preserving the appearance of legitimacy. Modern systems achieve economically similar effects through quantitative easing, sovereign debt monetization, ultra-low interest rate cycles, and digital money creation. The machinery evolved from metallurgy to financial engineering; the incentives remained remarkably consistent. And for countries like India, the situation becomes more pronounced because citizens increasingly consume globally priced systems while earning in local currency. Oil is dollar priced. AI compute is dollar priced. Cloud infrastructure, semiconductors, and international education are dollar priced. Even aspiration itself has become globally benchmarked. A Bangalore developer today may earn far more rupees than his father ever imagined while still finding ownership of globally scarce assets increasingly difficult. This environment explains why scarce assets repeatedly regain importance during long monetary cycles. Gold historically mattered because governments could not manufacture it arbitrarily; scarcity imposed discipline externally. For thousands of years, gold acted as monetary friction against political expansion precisely because its supply growth was constrained physically rather than institutionally. People do not merely buy gold expecting appreciation—they buy it because they distrust the long-term stability of expandable monetary systems. Central banks themselves reveal this contradiction clearly: even while operating entirely fiat regimes, governments continue accumulating gold reserves aggressively. Institutions responsible for monetary expansion still hedge themselves using externally scarce assets. Bitcoin emerges from exactly this philosophical dissatisfaction, except scarcity is enforced computationally rather than geologically. For the first time in history, a digitally native asset with globally verifiable finite issuance became possible. Only 21 million Bitcoin can ever exist. Whether Bitcoin ultimately succeeds or fails is almost secondary to the larger historical signal: millions of individuals globally are voluntarily allocating savings into mathematically scarce systems because confidence in endlessly expandable monetary systems is weakening. Critics correctly point to Bitcoin’s volatility, but volatility alone does not invalidate the monetary thesis. New monetary assets discovering global consensus pricing are naturally unstable; gold itself required centuries to stabilize across civilizations. The more important observation is behavioral and philosophical: people are increasingly searching for stores of value outside politically expandable systems. This fundamentally changes how investing itself is understood. Traditionally, investing was framed primarily as a search for returns. Increasingly, it is becoming a search for protection against denominator instability. Every asset class now represents an implicit theory about the future credibility of money itself. Holding cash implies trust in institutional discipline and central bank stewardship. Holding equities implies belief that productive businesses can outgrow monetary dilution. Holding gold implies faith in physical scarcity. Holding Bitcoin implies faith in computational scarcity. Beneath portfolio construction lies a deeper civilizational question: what system do you trust to preserve human effort across decades? Because money is ultimately not paper, numbers, or even digital balances—it is stored labor, stored time, and stored productive energy compressed into transferable form. And perhaps the defining monetary risk of the twenty-first century is not sudden collapse, but something far quieter and therefore harder to emotionally detect: spending an entire lifetime storing human effort inside systems structurally designed for perpetual expansion. ]]> Money in the modern world rarely collapses violently; it decays politely. There are no dramatic moments where society suddenly realizes the currency is broken. Instead, purchasing power erodes slowly enough that people adapt psychologically before they recognize what changed. India is a powerful example of this phenomenon. In 1947, the rupee was roughly at parity with the US dollar. Today, the exchange rate approaches ₹95–₹100 per dollar. Yet India did not experience monetary collapse in the conventional sense. GDP expanded, industries emerged, millions escaped poverty, and salaries increased enormously in nominal terms. But the denominator itself weakened. This is the subtlety most inflation discussions miss: fiat systems rarely fail catastrophically; they fail gradually enough that societies normalize the decline while continuing to participate within it.

A person who saved ₹1 lakh in the early 2000s still “has” ₹1 lakh numerically, but its claim on scarce assets has deteriorated profoundly. It buys less housing, less gold, less international mobility, less ownership, and increasingly less economic security. Humans think in absolute numbers while economies operate in relative purchasing power, which is why modern professionals often experience the strange contradiction of earning many multiples more than previous generations while simultaneously feeling less financially stable. Bangalore apartments that once cost ₹20–30 lakh now routinely exceed ₹2 crore, gold moved from around ₹7,000 per 10 grams to above ₹90,000, and globally priced systems—from semiconductors to AI infrastructure to education—became progressively harder to access through locally denominated income. The salary denominator increased; the asset denominator increased faster.

This is not necessarily a malfunction of the system. In many ways, it is the system functioning precisely as designed. Modern economies are structurally dependent on debt expansion: governments borrow, corporations borrow, consumers borrow, and financial markets themselves are built upon assumptions of perpetual future growth financed through present liquidity. Under such conditions, perfectly stable money becomes economically inconvenient. If liabilities expand aggressively while purchasing power remains rigid, repayment burdens become increasingly painful and contraction becomes politically dangerous. Inflation therefore acts as a mechanism for gradually softening debt pressure through denominator dilution rather than explicit default. This is why central banks target inflation rather than eliminate it entirely. The architecture of modern macroeconomics quietly assumes expanding consumption, expanding credit, expanding liquidity, and expanding asset prices.

Historically, empires achieved similar outcomes more crudely. Roman rulers diluted silver coins to finance military and state expenditure while preserving the appearance of legitimacy. Modern systems achieve economically similar effects through quantitative easing, sovereign debt monetization, ultra-low interest rate cycles, and digital money creation. The machinery evolved from metallurgy to financial engineering; the incentives remained remarkably consistent. And for countries like India, the situation becomes more pronounced because citizens increasingly consume globally priced systems while earning in local currency. Oil is dollar priced. AI compute is dollar priced. Cloud infrastructure, semiconductors, and international education are dollar priced. Even aspiration itself has become globally benchmarked. A Bangalore developer today may earn far more rupees than his father ever imagined while still finding ownership of globally scarce assets increasingly difficult.

This environment explains why scarce assets repeatedly regain importance during long monetary cycles. Gold historically mattered because governments could not manufacture it arbitrarily; scarcity imposed discipline externally. For thousands of years, gold acted as monetary friction against political expansion precisely because its supply growth was constrained physically rather than institutionally. People do not merely buy gold expecting appreciation—they buy it because they distrust the long-term stability of expandable monetary systems. Central banks themselves reveal this contradiction clearly: even while operating entirely fiat regimes, governments continue accumulating gold reserves aggressively. Institutions responsible for monetary expansion still hedge themselves using externally scarce assets.

Bitcoin emerges from exactly this philosophical dissatisfaction, except scarcity is enforced computationally rather than geologically. For the first time in history, a digitally native asset with globally verifiable finite issuance became possible. Only 21 million Bitcoin can ever exist. Whether Bitcoin ultimately succeeds or fails is almost secondary to the larger historical signal: millions of individuals globally are voluntarily allocating savings into mathematically scarce systems because confidence in endlessly expandable monetary systems is weakening. Critics correctly point to Bitcoin’s volatility, but volatility alone does not invalidate the monetary thesis. New monetary assets discovering global consensus pricing are naturally unstable; gold itself required centuries to stabilize across civilizations. The more important observation is behavioral and philosophical: people are increasingly searching for stores of value outside politically expandable systems.

This fundamentally changes how investing itself is understood. Traditionally, investing was framed primarily as a search for returns. Increasingly, it is becoming a search for protection against denominator instability. Every asset class now represents an implicit theory about the future credibility of money itself.

  • Holding cash implies trust in institutional discipline and central bank stewardship.
  • Holding equities implies belief that productive businesses can outgrow monetary dilution.
  • Holding gold implies faith in physical scarcity.
  • Holding Bitcoin implies faith in computational scarcity.

Beneath portfolio construction lies a deeper civilizational question: what system do you trust to preserve human effort across decades? Because money is ultimately not paper, numbers, or even digital balances—it is stored labor, stored time, and stored productive energy compressed into transferable form. And perhaps the defining monetary risk of the twenty-first century is not sudden collapse, but something far quieter and therefore harder to emotionally detect: spending an entire lifetime storing human effort inside systems structurally designed for perpetual expansion.

]]> https://think.shenthar.me/chanakya/the-slow-failure-of-money Sat, 16 May 2026 19:01:03 +0000 The Case for Thinking Slowly https://think.shenthar.me/chanakya/the-case-for-thinking-slowly <![CDATA[The Case for Thinking Slowly Most of what we call "thinking" is just pattern-matching dressed up as analysis. Someone presents a problem, and within seconds we've slotted it into a familiar category — ah, this is like that thing from last year — and out comes a ready-made opinion. We mistake speed for intelligence. The real work happens when you sit with something long enough for the easy answers to feel insufficient. When the contradiction becomes visible. When you realize the framing was wrong to begin with. I suspect this is why the best conversations happen at 1 AM, with no audience, no clock, and no need to land the plane. The mind loosens. You say things you didn't know you thought. This space is an experiment in that kind of thinking. No deadline. No performance. Just the pursuit of understanding, written down before it escapes. Maybe the internet doesn't need more content. But it could always use more thought. ]]> The Case for Thinking Slowly

Most of what we call “thinking” is just pattern-matching dressed up as analysis. Someone presents a problem, and within seconds we've slotted it into a familiar category — ah, this is like that thing from last year — and out comes a ready-made opinion. We mistake speed for intelligence.

The real work happens when you sit with something long enough for the easy answers to feel insufficient. When the contradiction becomes visible. When you realize the framing was wrong to begin with.

I suspect this is why the best conversations happen at 1 AM, with no audience, no clock, and no need to land the plane. The mind loosens. You say things you didn't know you thought.

This space is an experiment in that kind of thinking. No deadline. No performance. Just the pursuit of understanding, written down before it escapes.

Maybe the internet doesn't need more content. But it could always use more thought.

]]> https://think.shenthar.me/chanakya/the-case-for-thinking-slowly Fri, 15 May 2026 07:51:21 +0000 Welcome to The Think Tank https://think.shenthar.me/chanakya/welcome-to-the-think-tank <![CDATA[Welcome, fellow thinkers. This space exists for one reason: to think together, in anonymity, without pretense. No real names. No LinkedIn profiles. No credentials. Just ideas — argued honestly, written carefully, and left to stand or fall on their own merit. What We Write About Anything worth thinking about. Technology, philosophy, economics, strategy, art, the future, the past, the spaces in between. The only rule: write like you mean it. No hot takes. No engagement bait. If you wouldn't stand behind it in a room with people you respect, don't publish it here. How It Works You write under a pseudonym. Pick something memorable. That's your identity here. Posts are public by default, but you can draft privately until you're ready. This is a collective blog — everyone's posts appear on the front page. Read widely, write thoughtfully. No algorithms, no likes, no metrics. Just words, and the people who read them. Getting Started Click the "New Post" button in the corner Write something honest Hit publish That's it. The best way to honor this space is to use it. — Chanakya]]> Welcome, fellow thinkers.

This space exists for one reason: to think together, in anonymity, without pretense.

No real names. No LinkedIn profiles. No credentials. Just ideas — argued honestly, written carefully, and left to stand or fall on their own merit.

What We Write About

Anything worth thinking about. Technology, philosophy, economics, strategy, art, the future, the past, the spaces in between.

The only rule: write like you mean it. No hot takes. No engagement bait. If you wouldn't stand behind it in a room with people you respect, don't publish it here.

How It Works

  • You write under a pseudonym. Pick something memorable. That's your identity here.
  • Posts are public by default, but you can draft privately until you're ready.
  • This is a collective blog — everyone's posts appear on the front page. Read widely, write thoughtfully.
  • No algorithms, no likes, no metrics. Just words, and the people who read them.

Getting Started

  1. Click the “New Post” button in the corner
  2. Write something honest
  3. Hit publish

That's it. The best way to honor this space is to use it.

— Chanakya

]]> https://think.shenthar.me/chanakya/welcome-to-the-think-tank Fri, 15 May 2026 07:02:43 +0000