By Ramachandran Rajeev Kumar — 2026-05-27
India Doesn't Have an FDI Problem. It Has a Conditions Problem.
The net FDI collapse is real. So is the record gross inflow. The question hiding between those two facts is the one that matters.
Start with what actually happened, because the two dominant takes in Indian commentary are both wrong.
Take one: net FDI collapsed 96.5 percent — from $10.1 billion in FY24 to $0.4 billion in FY25 — and this is a five-alarm crisis. This is badly incomplete. Gross FDI inflows rose fourteen percent to $81 billion in FY25 and hit a record ~$94.5 billion in FY26. Capital is arriving at the highest volumes in India's history.
Take two: repatriation is "a sign of a mature market" — investors entering and exiting smoothly. This is the line RBI Governor Sanjay Malhotra offered in June 2025. It is the comfortable official gloss, and it is insufficient. Repatriation hit ~$53.6 billion in FY26 — the highest on record, over half the gross inflow. Outward FDI by Indian companies ran ~$29.2 billion in FY25, up 75 percent year-on-year. Portfolio investors pulled out a net ~$15.5 billion in FY26. Combined, direct and portfolio capital produced a net negative of roughly $9 billion. The RBI sold approximately $30 billion defending the rupee in March 2026 alone. The rupee is down around five percent.
A current account deficit funded by debt and reserve drawdowns is not maturity. It is a warning.
The honest diagnosis: gross FDI is genuinely strong, reflecting India's scale and the China+1 search for alternatives. But rising repatriation alongside portfolio flight signals something specific is prompting capital to leave faster than it should. That something is fixable. Fixing it is urgent.
Why FDI Is Not a Scoreboard Number
The instinct to wave off FDI as a vanity metric is worse than wrong.
FDI is not primarily about the capital. It is about what travels with it: technology transfer, management practice, supply chain integration, and wage-paying jobs at the end of that chain. China's transformation from agrarian economy to global factory floor took three decades and required FDI-led capital formation that domestic savings alone could not have produced.
India's Make in India programme targets manufacturing at 25 percent of GDP and ~143 million new jobs. The sector sits at roughly 17 percent today. That gap is not closeable through domestic consumption and public capex alone — as the private-capex signal in steel demand illustrates, a domestic investment upturn still needs global capital and technology to reach manufacturing scale.
India adds roughly one million workers to its labour force every month. Services cannot absorb them all; agriculture cannot absorb more. The young person in Jharkhand or Rajasthan who cannot find formal work is not failed by a trade balance figure. She is failed by a factory not built because an investor examined the conditions and chose Vietnam instead.
Vietnam drew approximately $36 billion in FDI in 2025. India captured the premium end of the China+1 bet — Apple's supply chain accounts for roughly 25 percent of global iPhone output, ~$23 billion in exports, ~350,000 PLI-linked jobs. That is genuine. It is also narrow. The broad manufacturing wave India was positioned to win largely went elsewhere. The reason lies in four fixable failures.
The Four Fixable Failures
Infrastructure and Land
A factory needs land, power, water, roads, and a port within reach. In India, across too many states and sectors, that combination remains genuinely difficult to assemble. Land acquisition is slow, contested, and politically volatile. Industrial corridor development has improved connectivity in select zones, but coverage is patchy and logistics cost per unit stays structurally higher than in competing destinations. The 18% tariff outcome gives India a real pricing edge over Vietnam and Bangladesh. That edge erodes if logistics costs eat the margin before the goods reach the port.
Taxation: The Deepest Scar
This is where India has done the most self-inflicted damage, and where the reputational overhang is longest-lived.
In 2012, India enacted retrospective tax legislation to reverse a Supreme Court ruling in favour of Vodafone, taxing a transaction already judged legal. Cairn Energy was assessed under the same framework and ultimately won a ~$1.2 billion international arbitration award — which India resisted enforcing. The cases settled, but the signal to global capital did not: India will change the rules after the game and charge you for playing by the old ones.
India then terminated most of its bilateral investment treaties in 2016-17, replacing them with a 2015 Model BIT widely characterised as among the most investor-unfavourable frameworks in the world — limiting international arbitration access, requiring exhaustion of domestic remedies, and narrowing the definition of covered investments.
Then came March 2026. Customs authorities issued a $601 million demand against Samsung: $520 million in alleged unpaid duties on Remote Radio Head telecom components imported 2018-21, plus $81 million in personal fines on seven executives. Samsung contests the classification and notes that Reliance Jio sourced identical components without similar demands. Whatever the technical merits, the optics are unambiguous — a classification dispute on components imported years earlier, pursued against a global manufacturer with personal criminal exposure attached. This is precisely the pattern that investment officers in Seoul, Taipei, and Houston cite when advising against committing capital to India.
Regulatory Complexity
The Economic Survey 2025 called explicitly for "Ease of Doing Business 2.0" — a useful admission that the first version did not finish the job. Approvals, licences, and registrations required to establish and operate a manufacturing facility create friction that compounds at every stage: setup, operations, exports, exit.
Regulatory unpredictability is as damaging as complexity. A factory is a ten-year bet. An investor making that bet needs to know the rules will not be materially rewritten in year three.
Bureaucratic Discretion
Related to regulatory complexity but distinct from it: the discretionary, opaque, slow exercise of official authority across multiple levels. Not an argument against a strong state — an argument against a state whose strength is expressed through unpredictable power rather than clear, consistently applied rules. The real cost of doing business in India lives not in the headline tax rate but in the gap between what the rules say and what officials can require. Every discretionary step is a transaction cost. At sufficient density, those costs make a location non-competitive regardless of its other advantages.
The Prescription: Remove the Friction
India does not need to bribe capital with subsidies. Production-Linked Incentives have worked in targeted sectors — the Apple supply chain is the evidence — but subsidies distort allocation and attract rent-seekers as readily as genuine manufacturers. What India needs is to stop creating reasons for capital to leave.
The reform signals already in motion deserve acknowledgment. The equalisation levy was scrapped in August 2024. Insurance was opened to 100 percent FDI in February 2025. A Task Force on Compliance Reduction was constituted in January 2025. The PMO has asked commerce to re-examine the Model BIT text. Real moves — and too slow and too partial for the scale of the problem.
Four things need to happen with the clarity of a commitment, not the gradualism of a review:
Stabilise taxation and end retrospective exposure. The retrospective tax era formally ended in 2021 — but Samsung in 2026 looks, to a foreign observer, remarkably similar to Cairn in 2014. The principle that tax demands will not be applied to historical transactions under reclassified frameworks needs to be operative, not merely declared. Disputes of the Samsung type should go through transparent adjudication with defined timelines, not demands calibrated to the size of the target.
Restore investor protection through BIT reform. The PMO's instruction to revisit the Model BIT is the right instinct. The revision needs genuine access to international arbitration and protections as foreign investors define them — not a framework requiring a decade of domestic litigation before international recourse is available. The trade-deal execution phase is moving forward commercially; investor protection must move at the same pace.
Cut compliance burden and discretion — specifically. Not in principle. Defined lists of approvals eliminated, defined timelines for what remains, defined limits on official discretion at each stage. The Task Force has the mandate. It needs to produce numbers, not a report.
Build the infrastructure broadly. The industrial corridor model works where executed. It needs to reach the states where land is available and labour is ready but logistics is the binding constraint — a project that started a decade ago and needs to accelerate.
None of this is sufficient without education reform — the workforce quality question belongs in a companion piece. But the tax, regulatory, protection, and infrastructure failures repel capital from an economy whose workforce is available and whose market is already large. Fix those first.
The Cost of Getting This Wrong
The macroeconomic pressure is real — a current account deficit funded by debt and reserve drawdowns, a rupee that needed $30 billion in RBI defence in a single month, a net capital position that went negative in FY26. Not a crisis today; it becomes one if the repatriation trend continues and the structural deterrents remain.
The deeper cost is human. Every percentage point manufacturing fails to reach as a share of GDP is a cohort of young Indians with narrowing formal employment options. The jobs FDI-led manufacturing creates are not abstract — they are the first formal wage in a family, the first provident fund contribution, the first rung on an industrial career. Domestic consumption sustains growth. Government capex builds roads. Neither builds a manufacturing economy at the scale India's demography demands.
The conditions problem is fixable. The question is whether the pace of fixing matches the pace at which India's young enter the labour force. It does not — yet.
Make the conditions. The capital will come.
BarathVector covers Indian economic policy with the conviction that what the government chooses to do — and chooses not to fix — has names and addresses.