India’s $47 Billion AI Trade Is Hidden in Industrial Stocks

Indian industrial companies supplying the global AI data-center build-out have added roughly $47 billion in combined market value this year on a Bloomberg equal-weighted index tracking 28 firms, rising nearly 50% while India’s benchmark indices shed more than $300 billion over the same stretch. The cohort spans optical fiber makers, transformer manufacturers, switchgear suppliers, and precision cooling companies. In Mumbai’s dealing rooms, traders have settled on a name for it: the AI capex trade.

Sterlite Technologies Ltd., an optical fiber maker owned by the Vedanta Group, is the trade’s most visible beneficiary. The stock surged more than 530% this year after securing a $1.1 billion multi-year product award from a US hyperscaler in May, covering optical connectivity supplies for AI data centers from fiscal 2027 through fiscal 2029. HFCL Ltd. and MTAR Technologies Ltd. have jumped 191% and more than 230% respectively in the same period.

The Companies the Benchmarks Don’t Count

Many of the biggest winners in the cohort are excluded from India’s broadest domestic indexes, so their returns don’t show up in the headline performance data that institutional investors routinely track, according to Bloomberg. The AI capex trade has been running well below the radar of the portfolios that move index numbers.

“We may be on the wrong end of the AI trade, but we could be on the right side of the AI capex trade. One could consider companies benefiting from data centers and the entire value chain associated with this capex,” said R. Sivakumar, chief investment officer at Axis Mutual Fund.

Every AI query runs through a data center, and data centers run on power, fiber connectivity, cooling systems, transformers, and switchgear. India’s capital goods manufacturing base, built up over decades supplying power grids and telecom networks, makes all of these. What changed in 2026 is the scale of who is ordering and at what volume.

Amazon.com Inc. plans to invest $12.7 billion in cloud infrastructure in India through 2030. A Reliance Industries Ltd. joint venture signed an $11 billion pact to build local data centers last year, while AdaniConnex Pvt. has active partnerships with Google and Uber Technologies Inc. on data center construction. Schneider Electric SE, the French energy management company, projects that India’s total data center capacity could reach 6 to 7 gigawatts by 2030, from roughly 0.9 gigawatts in 2023, a shift that puts sustained pressure on every component in the supply chain.

The Billions Committing to Build

Alphabet Inc.’s Google announced its largest India investment of $15 billion in October 2025, committing to a full-stack AI hub in Visakhapatnam, Andhra Pradesh. The project, built in partnership with AdaniConnex and Nxtra by Airtel, includes three data center campuses, a new international subsea cable gateway, and gigawatt-scale compute capacity. Ground broke in April 2026. India’s Union Ministry of Electronics and Information Technology described the project as a game-changer for the country’s digital economy, framing the hyperscaler’s commitment as central to the national AI mission.

India’s Finance Bill 2026 added legislative weight. The bill introduced a data center tax holiday running through 2047 and a related-entity safe harbour provision, giving hyperscalers and their Indian supply chain partners multi-decade cost certainty for capital programs that could otherwise look too long-dated to justify.

The combined committed spend from Amazon, Google, and the Reliance joint venture exceeds $38 billion directed at Indian data center infrastructure. Each of those dollars eventually reaches a supply chain that needs transformers, fiber, switchgear, cooling units, and backup power, the same industrial product lines that the Bloomberg cohort tracks.

Who’s Winning, and by How Much

The trade runs deeper than optical fiber. Hitachi Energy India Ltd., ABB India Ltd., and Cummins India Ltd. have all been re-rated as their transformer, electrification, and backup power products enter data center supply chains that barely existed as a revenue source for them five years ago.

MTAR Technologies Ltd., a precision engineering company based in Hyderabad, manufactures hot boxes and critical assemblies for Bloom Energy Corp.’s fuel cell systems, which data center operators are deploying to address on-site power shortfalls at facilities where grid supply falls short. MTAR’s fiscal 2026 revenue rose 29.6% to Rs 876.2 crore, and the company secured a Rs 2,279 crore international order in May. The stock has more than doubled this year.

Mahesh Viswanathan, chief executive of Finolex Cables Ltd., said on a recent earnings call that this was “the right time to be in this industry.” Finolex shares have gained nearly 36% this year, backed by earnings-driven data center demand rather than thematic speculation.

The outlier is Anant Raj Ltd., the one listed company focused on operating data centers rather than supplying their components. Its stock is up roughly 8% year-to-date. Value Research, an Indian financial research firm, has documented shifting expansion timelines and unclear management communication as factors constraining the stock’s re-rating, even as the company targets 357 megawatts of total capacity by 2032 from 28 megawatts currently operational.

Year-to-date gains across the cohort’s key names, Bloomberg and NSE data as of June 5, 2026:

• +530% Sterlite Technologies (optical fiber and data center connectivity)
• +230%+ MTAR Technologies (fuel cell components for data center power)
• +191% HFCL Ltd. (fiber and networking infrastructure)
• +36% Finolex Cables (industrial cables and wires)
• +8% Anant Raj Ltd. (India’s only listed data center operator)

The Two-to-Four-Year Lead-Time Moat

Nomura Holdings Inc. analysts led by Akash Gupta published a note on June 2 concluding that the most attractive exposure in the AI build-out is India’s industrial supply chain. The report described it as “the ‘picks and shovels’ that build, power, and cool these facilities,” and identified a two-to-four year lead time in supplying certain components as having created “an enviable seller’s market with multi-year backlogs,” with orders secured now generating revenue between 2027 and 2029.

For incumbent suppliers with existing capacity and established customer relationships, those lead times act as a competitive barrier. New entrants can’t quickly replicate a customer relationship that took multiple procurement cycles to build, and the capital cost of new transformer or fiber factory capacity can’t be recouped in the window the current order cycle covers. Pricing power that typically erodes when a sector attracts new supply is, in this cycle, locked in by the physical timeline of the infrastructure itself.

Data center capex has emerged as the single largest contemporary industrial investment cycle, larger than the global wireless 4G rollout, the post-2008 LNG build-out, or the early-2010s shale boom.

Nomura Holdings analysts, June 2, 2026 research note on India AI infrastructure.

Angel One, an Indian brokerage that has tracked the cohort, estimates that global hyperscale data center investments will exceed $1.2 trillion between 2025 and 2027. India’s 28-company cohort sits at the physical base of that spending, supplying the components each new facility requires before it can run a single workload.

The question for the cohort is what the backlog looks like beyond FY29. Grid reliability at scale is what analysts identify as the single most likely constraint on hyperscaler willingness to deepen India infrastructure commitments beyond the current confirmed order window.

Foreign Funds Got Here First

Foreign institutional investors have been net sellers of Indian equities at a record pace this year, unwinding positions as the market lagged global peers on a valuation basis. The AI capex trade is the visible exception. Foreign fund shareholding in Indian industrials rose to 14% as of end-March, the highest level in two years, according to Elara Capital (India) Pvt., a Mumbai-based brokerage. That shift happened while the same investors were simultaneously reducing exposure across the broader Indian market.

The logic follows from how the trade is structured. Multi-year hyperscaler contracts with defined delivery windows give Indian industrial suppliers predictable volumes and contracted pricing before a single unit ships. Technology investors familiar with the Nvidia Corp. supply chain recognize the pattern: infrastructure suppliers often generate more execution-certain cash flows than the platform companies they service.

Domestic fund managers are beginning to apply the same framework. The industrial supply chain thesis is moving from dealing-room shorthand into formal fund positioning, and the next quarterly shareholding disclosures will show how fast domestic institutions have repositioned.

What 70 Times Forward Earnings Requires

The valuation context of the AI capex trade comes into focus in a single comparison. The cohort’s leading gainer currently trades at roughly 70 times its 12-month forward earnings, according to Bloomberg, against the Nifty 500’s 19 times, the benchmark for India’s National Stock Exchange. At that gap, the market is pricing in multi-year contract execution without disruption, with the $1.1 billion product award as the primary visible evidence for that expectation.

Angel One, which has tracked the cohort closely, warned that the rally has left “no room for execution disappointments.” The brokerage noted that the market is rewarding companies with “visible AI-linked earnings rather than just thematic exposure,” a distinction that separates contract-backed winners from capacity-announcement plays.

The risks the current valuations absorb:

• Customer concentration: One supplier in the group derives over 70% of revenue from exports, with a significant portion tied to a single fuel cell manufacturer’s demand growth projections through 2030.
• Lead-time reversal: The multi-year backlogs that protect incumbents today become a vulnerability if hyperscalers reduce or delay their India commitments.
• Foreign-fund dependence: Foreign institutional buying drove a meaningful part of the re-rating, and a reversal in global risk appetite toward emerging markets would unwind the institutional bid quickly.
• Valuation compression: At that multiple, any miss on execution timelines compresses the forward earnings estimate and the valuation simultaneously.

Sterlite’s $1.1 billion product award delivers optical connectivity from FY27 to FY29; those three years are when the 70-times-forward-earnings bet settles.