The iPhone Moved to India. The Value Stayed in China.

The next AI trade may not be another chatbot.

It may be surgical robots, automated warehouses, smart factories, and machine vision systems already reshaping how companies operate.

MarketBeat’s new 7 Stocks to Buy Before the Robotics Revolution report reveals seven companies positioned across the automation boom, from robot builders and AI chip leaders to machine vision providers and factory automation giants.

This is where AI gets a body.

And as labor shortages, wage pressure, and supply chain stress push more companies toward automation, these stocks could move before the robotics story becomes impossible to ignore.

The report normally sells for $29.97, but it is free for a limited time.

Send My Free Report

India has won the final-assembly layer of global smartphone manufacturing in about five years, while the components, tooling, and value that sit underneath it remain overwhelmingly Chinese and move on a decade clock that no policy has yet bent.

Two policy breaks made this thesis live, and one of them reversed itself in February.

The first was the tariff whipsaw. Through 2025, the emergency-powers tariff regime taxed China-assembled phones far more heavily than India-assembled ones, and US-bound assembly rerouted accordingly. On February 20, 2026, the Supreme Court struck those emergency-powers tariffs in a 6 to 3 decision. The replacement took effect on February 24: a uniform 10% global tariff under Section 122 of the Trade Act of 1974, a balance-of-payments authority that lets a president impose a temporary import surcharge. It does not reach phones at all. Smartphones sit on the proclamation's exemption list, carry no China-specific duty, and enter at a zero base rate under the international agreement that covers information-technology products. A US-bound phone from either origin now pays roughly nothing at the border. The wedge that pulled assembly to India has closed behind it. Whether the shift survives without it is the question the next two quarters of customs data will answer.

The second break was Apple's supply-chain pivot. According to Bloomberg reporting in April 2025, citing people familiar with the plans, Apple aims to source most of the iPhones it sells in the United States from India by the end of 2026, up from roughly 3 to 4% of global iPhone output made in India as recently as 2021 (Counterpoint). That reported plan predates the February 2026 tariff reset by ten months. It began as a corporate strategy under the old regime and outlived the policy that first made the math work.

The old world was simple: China assembles for everyone. The 2025 world added a footnote: China assembles for everyone, except where Washington taxes it, and India was the shelter. The 2026 question is narrower and more reversible than "the factory of the world is moving." It is whether capacity built for a tax advantage stays built once the advantage is gone. Early evidence leans toward yes, because a committed assembly line does not un-build the day a tariff expires, but one quarter of post-wedge data is thin.

Exhibit 1: The import flip is real, fast, and measured in customs data. The headline flip shows up in realized customs data. In the second quarter of 2025, India accounted for 44% of smartphones imported into the United States while China's share fell to 25%, reversing a 13%-against-61% split from a year earlier (Canalys). The volume of made-in-India smartphones shipped to the US grew 240% year over year in that quarter. On monthly customs data the swing is starker still: between September 2024 and May 2025, India's share of US phone imports ran from 4.9% to a 65.2% peak while China's fell from 91.3% to 16.1%.

BCR's own modeling of 99 months of US customs data (from January 2018 through March 2026) finds that the entire flip sits inside the tariff window. Every one of the ten largest shifts in India's trend lands between October 2024 and January 2026, and none appears in the prior 81 months. The first month after the wedge closed, India still held 56.3% of US phone imports, above the 40% line that would signal the migration was purely tax-driven.

The upshot: the assembly migration is one of the best-documented supply-chain moves on record, and so far it is holding even after the tax reason for it disappeared.

Figure 1: US smartphone imports flipped source countries in a single year. Source: Canalys (Omdia), Q2 2025.

Exhibit 2: The value stayed behind. Assembly share and value captured are two different measurements, and they read in opposite directions. India keeps roughly 20% of the export value it books, on industry estimates of domestic value addition, the share of a product's value actually created inside the country. GTRI's stricter count, which nets out the imported parts first, puts the figure near 6%: about $30 of a $500 iPhone leaving the factory gate. Roughly 71% of the component shipments for those phones arrive from China (first quarter of 2025). The assembly fee itself is the thinnest slice. On Farok Contractor's teardown of an iPhone 16, Apple pays Foxconn about $14 for final assembly on a $563.73 total manufacturing cost, and the Chinese portion of that cost is $38.89.

For the portfolio, the practical read is that a headline export number overstates the domestic economics by about five times on the industry measure, and by roughly sixteen times on GTRI's stricter count. The country that assembles the phone captures the smallest slice of its value.

Exhibit 3: The buildout is real and paid for on both ends. The capacity is already built and running. India's smartphone incentive scheme is the standout of its whole industrial-policy program: electronics accounts for about ₹15,554 crore of the ₹28,748 crore paid out across all fourteen incentive sectors as of December 31, 2025 (Press Information Bureau; a crore is ten million rupees). Apple's iPhone exports from India crossed roughly $23 billion in calendar 2025. Foxconn has invested more than $2.5 billion in Indian plants. Micron's Sanand facility, India's first semiconductor packaging site operating at scale, began commercial production in February 2026 (inaugurated February 28).

What this means on the ground: the state is subsidizing both the assembly work and the first pieces of the supply chain around it, and phones are shipping in the tens of millions. Apple assembled about 55 million iPhones in India in 2025, up roughly 53% from about 36 million the year before (Bloomberg).

Exhibit 4: The money that would confirm industrialization did not come. The flow that would validate the deeper story is foreign direct investment: long-term capital that builds factories and supply chains, as opposed to fast-moving portfolio money. It collapsed. Net FDI fell from $44 billion in fiscal 2020-21 to $0.96 billion in fiscal 2024-25, under a billion dollars, before recovering to just $7.6 billion in fiscal 2025-26 (Reserve Bank of India), with individual months still negative into December 2025. The assembly story kept accelerating while the capital that would turn assembly into a domestic supply chain went the other way.

The bottom line for investors: the export headlines and the investment data tell different stories, and the investment data is the one that governs the decade clock.

Figure 2: Net FDI — the investment that would deepen the supply chain never arrived. Source: Reserve Bank of India.

Figure 3: Two clocks — assembly share migrated, value capture did not. Source: Canalys; industry reporting and GTRI, 2025.

Stage 1: Washington taxed Chinese assembly, then stopped. Through 2025 the emergency-powers regime taxed China-assembled phones at the US border far more heavily than India-assembled ones, and the all-in cost of landing a US-bound iPhone flipped in India's favor while the gap stayed open. Since February 24, 2026, that gap is roughly zero: smartphones are exempt from the new baseline tariff, carry no China-specific duty, and enter at a zero base rate from either origin. Committed capacity does not un-build when the tax reason for it closes, so the lines built during the wedge keep running.

Stage 2: Apple reroutes final assembly, and only final assembly. India's iPhone build rose to about a quarter of global output, and its share of US smartphone imports jumped to 44% in a single year. Final assembly is the mobile layer of the process: a screwdriver-and-test line reproduces inside an existing supplier network in two to three years, which is why this clock runs fast.

Stage 3: The components keep following the old map. Roughly 71% of the component shipments for India-assembled phones arrive from China, and India's imports from China grew about 10% in 2024 and 14% in 2025 even as its exports to the US surged (Federal Reserve FEDS Notes, April 2026). The Fed's economists put it plainly: “As has been the case in countries that have seen an increased share of U.S. imports, India's imports from China have also risen.” Peer-reviewed research on post-COVID supply-chain moves finds the same shape everywhere. Freund, Mattoo, Mulabdic and Ruta (Journal of International Economics, 2024) find China's share of US imports fell from 22% to 16% while the countries replacing China "tend to be deeply integrated into China's supply chains and are experiencing faster import growth from China, especially in strategic industries." Baldwin, Freeman and Theodorakopoulos (2023) measure US look-through exposure to China, the reliance that shows up once you trace components back through the supply chain, at almost four times the face-value number. Every assembled-in-India phone still buys the Chinese stack; the trade flow reroutes through Chennai without moving away from Shenzhen.

Stage 4: Value capture lags share capture, and the lag is built in. India's domestic value addition sits near 20% of export value on industry reporting, and closer to 6% on GTRI's stricter count. The government's own 2020 target, in its incentive-scheme launch, was raising mobile-phone value addition "from the current 15-20% to 35-40%" (Press Information Bureau, October 6, 2020). Six years on, the industry measure still reads about 20%, and the follow-on components program now carries the 35% ambition. Assembly fees are structurally the smallest part of the bill, so the headline export figures overstate the domestic economics several times over.

Stage 5: The stack moves only where the state pays a decade of tuition. India's first real chip fabrication plant, the Tata-PSMC project at Dholera, is a ₹91,000 crore build targeting first silicon on a 28-nanometer process late in 2026 and volume around 2028. That process generation is roughly fifteen years behind the leading edge: the industry's most advanced foundry reached 28 nanometers in 2011 and 2 nanometers in 2025. Broad manufacturing is growing, just not faster than the rest of the economy. Its real output nearly doubled over the decade to 2025 (up about 82%, or roughly 6% a year), yet its share of GDP still slipped to about 13% in 2024, well short of a 25% national target now pushed out to 2035. A flat share means factory output merely kept pace with GDP. Real industrializers outrun it: China and Vietnam ran manufacturing one to two-and-a-half points a year faster than their economies through their boom decades. The flagship $23 billion incentive scheme was allowed to lapse in March 2025 having hit 37% of its production target with under 8% of its incentives actually paid out. The assembly win is narrow; the industrialization claim is off-trajectory on the government's own scoreboard.

Stage 6: China's counter-levers set the ceiling on the pace. Engineer recalls, equipment holds, and rare-earth licensing raise the cost of every next rung India tries to climb. The weak link in this chain is the slope of the second clock: whether India's component localization accelerates past the roughly 20% value-added mark is the open question. The evidence establishes the two clocks cleanly. It does not yet fix the exact speed of the slower one.

Figure 4: India's factory output nearly doubled while its share of GDP stayed flat. Source: World Bank WDI; Benjamin Capital Research model.

China itself is the closest parallel, and it is an uncomfortable one for the hype case. On the Xing and Detert teardown, Chinese firms captured about 3.6% of an iPhone's manufacturing cost in 2009. By the iPhone X in 2018, that share had climbed to 25.4% (Xing series). Moving from assembly to meaningful value capture took China a full decade, and it did so with the largest coordinated industrial policy in modern history, a near-universal-literacy workforce (China's adult literacy was about 97% in 2020), and infrastructure India has not built.

Figure 5: China climbed from assembly to value over a decade. Source: Xing & Detert (2010); Xing & Huang (2021).

India in 2026 sits almost exactly where China sat in 2009: booking large gross exports on thin domestic value. The similarity is the starting point. The differences cut in both directions.

India had a tariff wedge that China never had, though that wedge has now closed. India also faces an incumbent actively defending the stack, through engineer recalls and equipment holds, which China in 2009 never had to contend with. And there is a structural headwind China never faced. Dani Rodrik's work on premature deindustrialization, the tendency of late developers to run out of factory jobs early (Journal of Economic Growth, 2016), finds that countries industrializing after 1990 reach their peak manufacturing employment at about a third of the income levels earlier industrializers enjoyed. India, on his numbers, appears to have already hit its peak manufacturing employment share "at income levels of $700." The ladder India is climbing may have fewer rungs at the top than China's did.

BCR's own modeling puts numbers on the pace. India's iPhone-assembly migration is running at about 4.6 percentage points a year, a pace statistically indistinguishable from Japan's rare-earth diversification after 2010, the best-documented case of a state-backed supply-chain move on record. Japan went from 90% Chinese dependence to a floor near 58% over about seven years, and never to zero. Even at that record pace, China keeps a majority of iPhone assembly until around 2032, and the probability that China falls below half of iPhone assembly by 2030 is under 1%.

Translated into calendar risk: the fastest corporate-led migration anyone has measured is running at exactly history's fastest state-backed pace, and it still leaves China holding half the assembly base for another six years.

The commentary in this section describes how asset classes have historically behaved in similar macro environments, for educational purposes only. Nothing here is a recommendation to buy, sell, or hold any security.

Equities. In past supply-chain migrations, the narrow set of sectors that touch the physical buildout (contract manufacturers, industrials, and logistics) have historically seen orders, employment, and export receipts move years before national value-added statistics catch up. The headline-index story is different. Indian equities traded at a steep premium to emerging-market peers until that premium peaked in July 2024; over the twelve months to June 2026 India lagged the emerging-market complex by 40.5%, leaving its relative valuation near the bottom of its trailing five-year range. Much of the froth has already come out, which changes the entry math on the broad index from where it stood a year ago without pointing to a timed re-entry. The differentiated exposure the thesis describes sits in the manufacturing and industrial verticals; the broad index is a separate question. On the other side of the trade, the data does not support pricing Chinese component makers for a displacement, because the component layer is precisely where the migration is not happening.

Rates and Fixed Income. Local-currency emerging-market bonds have historically drawn support from index-inclusion flows and a manufacturing-led growth narrative. Indian government bonds began phased inclusion in JPMorgan's emerging-market bond index in June 2024, which obliges funds tracking that index to buy them on a schedule, a structural tailwind that runs independent of this thesis. The counterweight is the fiscal cost of the industrial subsidies themselves. This thesis carries no direct read-through to US interest rates.

Credit. State-anchored projects such as the ₹91,000 crore Dholera fab behave as quasi-sovereign risk: close to government credit even when the borrower is nominally a corporation. The private-credit leg of Indian capital spending remains thin, and debt tied to the fab buildout is a story for 2028 and beyond as those projects mature. The slow pace of incentive disbursement, under 8% of the flagship program actually paid out, is a reminder that subsidy cash arrives later than the announcements imply.

Foreign Exchange and Emerging Markets. The assembly boom has not translated into currency strength, and there is no directional view on the rupee or the dollar to be drawn from this thesis. Gross exports rise, but import content rises alongside them as components come in from China, so the net external accounts improve far less than the export headlines suggest. Weak foreign direct investment removes the steady inflow of capital a genuine industrialization story would bring. Trade-flow headlines are not a currency signal, and this report draws no de-dollarization conclusion of any kind.

Commodities. Electronics parks, fabs, and logistics infrastructure are metals-intensive, so India's buildout is an incremental demand story for industrial metals and energy at the margin. It is also a live exposure to China's rare-earth lever on the input side. China's April 2025 export-licensing regime covers seven rare-earth elements upstream of electronics manufacturing, and an October 2025 expansion to five more heavy elements is suspended until November 10, 2026, a standing re-imposition trigger that sits directly on the industry's input costs.

Counter-Argument 1: With the wedge already closed, the migration stalls or reverses

Since February 24, 2026, a US-bound iPhone enters at roughly zero duty from either origin, so the pure cost case tilts back toward China's deeper and cheaper ecosystem. The path to a restored advantage is explicit: alongside the Section 122 proclamation, the administration announced new trade investigations that trade lawyers expect to produce more durable tariffs to replace the time-limited Section 122 surcharge, and the president has signaled an intent to raise that surcharge toward its 15% ceiling. A successor regime that re-creates a China-specific rate on phones would re-open the advantage in India's favor. One that also targets India, and Washington has already pressured Apple over the India buildout, would turn the calculus negative. The assembly claim is hostage to a policy coin-flip in both directions, though committed capacity keeps most of what is already built either way. The first post-wedge month, with India holding 56.3% of US phone imports, leans mildly against this risk, but a single month is too thin to lean on.

Estimated probability counter-argument is correct: 35%

Counter-Argument 2: China's chokepoint squeeze caps the ramp

Engineer recalls, equipment-export holds, and rare-earth licensing are already reported to have disrupted India's roughly $32 billion smartphone-export target for the fiscal year, a projection the industry body ICEA put to the government. Foxconn recalled hundreds of its own Chinese engineers from its India plants in July 2025 (Bloomberg), a move widely linked to Beijing's informal pressure. If China escalates to formal export controls on assembly equipment or key components bound for India, the five-year assembly claim itself degrades, because India's lines currently run on Chinese tooling and mid-stream inputs. The offsetting logic is that a hard squeeze against Apple's supply chain risks accelerating exactly the diversification China is trying to prevent, which caps how far Beijing can push before the lever erodes its own leverage. This is a real drag, and an unlikely full stop.

Estimated probability counter-argument is correct: 25%

Counter-Argument 3: India's own policy and capital environment stalls the climb

The stack claim requires a decade of sustained investment, and the flow evidence is weak: net FDI near zero in fiscal 2024-25, record repatriation of foreign profits, the broad incentive scheme allowed to lapse, manufacturing growing only in step with GDP, and agriculture still 43% of employment. The government's own scoreboard is the strongest evidence here: the 25% manufacturing target has slipped from 2022 to 2035, and the flagship incentive scheme reached 37% of its production target before lapsing. The peer-reviewed base rate leans the same way, with Rodrik's finding that post-1990 latecomers peak far earlier in their development. BCR's modeling puts India's value-added slope at about half a percentage point a year, with a confidence range that includes zero; one plausible reading of the data is no measurable value progress at all. Against all of this, the smartphone incentive vertical delivered, which is why this scores as the highest-probability drag on the decade claim, with little power to unwind the assembly base already in place.

Estimated probability counter-argument is correct: 45%

If India's domestic value addition breaks sustainably above 30% or its Chinese component share falls below 55%, the decade clock is running ahead of schedule and the thesis accelerates. If a successor tariff regime targets India directly, or component dependence stays pinned above 70% through 2027, it is time to reassess.

Methodology note: value-addition comparisons name both the industry domestic-value-addition measure (~20%) and GTRI's stricter net-accrual measure (~6%), because they measure different things and the memo carries both. Migration-pace and probability estimates draw on BCR internal models built on US customs data (99 months, January 2018 to March 2026), Monte Carlo simulation of assembly-share paths, and a relative-valuation study of Indian versus emerging-market equities. Pre-2020 academic work (Rodrik 2016, Xing and Detert 2010) serves only as a long-run structural base rate and a historical anchor; the current-regime claims rest on post-2020 data and BCR's own models.

This report is for informational and educational purposes only. It does not constitute investment advice, a recommendation, or a solicitation to buy or sell any security. All asset class commentary reflects historical patterns and educational analysis, not personal investment advice. Past performance does not guarantee future results. Readers should consult a qualified financial advisor before making investment decisions.

Benjamin Capital Research | July 8, 2026