The global artificial intelligence frenzy has reached a critical crossroads. In India, a basket of low-profile engineering, telecom infrastructure, and tech stocks has quietly morphed into the market’s hottest play. Dubbed the “AI Capex Trade,” companies like Sterlite Technologies (up 350-500% year-to-date), HFCL, Netweb Technologies, MTAR Technologies, and Aeroflex Industries have delivered staggering multibagger returns.
However, as capital floods into the infrastructure that powers AI, prominent global voices and market experts are sounding the alarm: The AI thesis may be running too far ahead of financial reality.
The Global Catalyst: Michael Burry’s “Beginning of the End”
Michael Burry, the legendary contrarian investor who famously shorted the 2008 housing bubble, has officially taken aim at the AI boom. Burry recently increased his short positions on global semiconductor giants and AI proxies like Nvidia and Palantir.
The primary trigger for his latest warning? South Korea’s massive corporate layout. When tech titans Samsung Electronics and SK Hynix announced an eye-popping $500+ billion (896 trillion KRW) mega-investment to build a comprehensive “AI semiconductor hub” in South Korea, the Nasdaq celebrated. Burry, however, called it the “Beginning of the End.”
The Structural Risk: Massive infrastructure capital expenditure (Capex) assumes a permanent shortage. However, memory (like the High-Bandwidth Memory chips controlled by SK Hynix and Samsung) is a deeply cyclical commodity. If end-user demand for consumer AI tools doesn’t monetize fast enough to justify these half-trillion-dollar bets, supply will quickly flip from a “shortage” to a “surplus”—bursting the profit expectations currently priced into the tech sector.
The Domestic Front: India’s Hidden AI Capex Winners
While the world focused on US software and Taiwanese foundries, Indian mid-cap companies captured a massive portion of the physical hardware build-out. Every AI query routed through a power-hungry graphic processing unit (GPU) cluster requires a vast physical footprint of high-speed fiber optics, specialized cooling, and stable power infrastructure.
| Company | 2026 Context & Driver | Valuation Status |
|---|---|---|
| Sterlite Technologies (STL) | Bagged a landmark $1+ Billion multi-year contract with a US-based hyperscaler for its Neuralis AI Data Center optical portfolio. | Trading at an elevated 66x forward earnings post-rally. |
| HFCL Ltd. | Massive multi-year order book (~₹21,200 crore) driven by data center upgrades and state-backed BharatNet connectivity projects. | Significant run-up based on long-term visibility. |
| MTAR Technologies | Delivering precision liquid-cooling and clean energy components for high-density data centers. | Stock has more than trebled on niche engineering demand. |
| Netweb & Aeroflex | Benefiting directly from domestic high-performance computing (HPC) server architecture and flexible piping systems for heavy cooling infrastructure. | High institutional crowding and premium valuations. |
While these companies have genuine, tangible order books backing their growth, their stock prices have baked in multiple years of flawless execution. Any global slowdown in hyperscaler spending leaves them highly vulnerable to aggressive profit-taking.
The CLSA View: How a Global Unwind Triggers an “FII U-Turn” to India
If the global AI bubble does deflate, it may paradoxically offer a silver lining for the broader Indian equity market. According to insights from global brokerage CLSA, an AI cooling period could trigger a massive realignment of Foreign Institutional Investor (FII) capital.
Earlier this year, a relentless wave of global liquidity exited the Indian market to chase pure-play AI semiconductor exposure in Taiwan (via TSMC) and South Korea (via Samsung/SK Hynix). This flight of capital briefly pushed the South Korean and Taiwanese stock markets past India in terms of global market capitalization, dropping India to the 7th largest spot.
However, CLSA and market analysts note that a correction in crowded AI bets would trigger a reversal:
- The Volatility Exits: Massive localized single-day drops in the South Korean and Taiwanese indices have already shown that global money is sitting on fragile, crowded profits.
- The Flight to Macro Stability: As FIIs look to derisk out of pure-play tech hardware, India emerges as a primary destination. Powered by a stabilizing Rupee (rebounding to around 94.40 against the USD from mid-May lows) and a sharp cooling in global crude oil prices below $73, India’s broader macroeconomic balance of payments looks highly resilient.
- The Valuation Dilemma: While Indian equities are admittedly not cheap relative to historical earnings, global funds hunting for secular growth outside of the volatile tech ecosystem simply do not have many large-scale choices.
The Verdict: Better Safe Than Sorry?
For investors holding onto 3x or 5x gains in domestic data center, optical fiber, and industrial engineering stocks, caution is highly warranted.
The underlying data center theme in India is structurally real—the country’s IT load capacity has expanded at a ~29% CAGR over the last few years. However, when a stock trades at 60x+ forward earnings on the back of global AI hype, it ceases to be an infrastructure play and becomes a momentum play.
With global tech visionaries flashing warning signs and macro capital already shifting back into defensive domestic sectors, locking in partial profits on over-extended multibaggers might just be the safest play in the current market lifecycle.
To better understand the structural risks and look closely at the data driving these warnings, you can watch this analysis on Michael Burry’s AI bubble warnings. It details exactly why major global economists are concerned about corporations over-extending their capital on artificial intelligence infrastructure before seeing proof of sustainable profits.