Moneycontrol Pro Weekender | India is missing the AI trade, and the market knows it

October 2, 2026

Thanks to AI, the world is going through a massive investment boom. Goldman Sachs’ research says the major U.S. hyperscalers are on track to spend roughly $800 billion on capital expenditures in 2026, with total global AI-related investment crossing $1 trillion. Data from Anthropic’s confidential IPO prospectus, leaked this week, show a staggering $518 billion future commitment to cloud computing and AI infrastructure. And according to PwC’s long-range modelling with Oxford Economics, cumulative global investment in AI infrastructure is projected to reach an eye-popping $31.6 trillion through 2050. No wonder AI investment is seen as a big contributor to global resilience, enabling the global economy to combat geopolitics, hot wars, cold wars, high oil prices, inflation et al. with remarkable aplomb. It has also held up the US stock indices, even as the broader market languishes, a sign of how narrowly the rewards are being shared.

AI is now performing a function for global capitalism that railways, housing bubbles, and fibre-optic cables performed in earlier eras. It’s a wonderful outlet for surplus capital—at last we have something really big to bet on. And what enormous bets! Anthropic is targeting an IPO valuation of more than $2 trillion, while OpenAI is reportedly seeking to raise at least $30 billion in a new funding round at a valuation of around $1.4 trillion. The markets are gleefully capitalising the expected future productivity of AI today.

It’s a fascinating loop—expected future growth leads to higher valuations, which creates greater wealth, which fuels more investment, which juices economic growth, and so on and so forth. As the Bank for International Settlements has pointed out, these wealth effects are boosting consumption and financial conditions.

Unfortunately, the rising cost of AI infrastructure is putting pressure on Big Tech free cash flow, and AI firms are increasingly turning to debt to finance their investments.

Where’s the risk? The real bet in the capital markets is not whether AI works, it’s whether the surplus generated by AI will be large enough, and arrive quickly enough, to validate the extraordinary quantity of capital being committed today. All you need is a disappointment relative to expectations to make the entire process go in reverse.

Story continues below Advertisement

For instance, a slowdown in AI investment won’t be merely bad news for Nvidia or Microsoft. It will reduce capital expenditure, construction, electricity demand, employment, and consumption. Debt and circular equity holdings can amplify the subsequent fall. It could well be a spectacular bust, although nobody is losing any sleep thinking about that at present.

AI investment is also geographically concentrated, which has consequences for capital flows. Countries supplying chips, cloud infrastructure, energy and capital can attract enormous investment. Countries that mainly consume AI services, on the other hand, may find capital whizzing by them. While they may capture the productivity gains, the ownership of the intellectual property and infrastructure will determine where the profits ultimately accrue.

That brings us to Indian equities, which have seen foreign portfolio capital deserting them at a prodigious rate, in sharp contrast to AI-heavy Asian markets. As on September 30, 2026, MSCI India, in USD terms, was down 15.65% year to date, while MSCI Emerging Markets was up 21.4%. The crucial distinction: India is an AI user, not yet an AI rentier. Not everyone reads this as a verdict on India. Speaking at the Moneycontrol Startup Conclave 2026, partners from Accel, Elevation Capital and Lightspeed argued that India’s AI opportunity extends well beyond foundation models and compute infrastructure, with applications, enterprise software and AI-led businesses offering significant room for value creation.

Story continues below Advertisement

The first phase of a technological revolution generally rewards whoever owns the bottlenecks. With AI, the bottlenecks are computing power, chips, power, data centres, cloud computing, foundation models. How concentrated that ownership could be was seen in the ASML chief’s remark that “there is only one company making those machines”. The investor is looking for earnings leverage to the new accumulation cycle. And India doesn’t yet have much of it at the listed-company level, although the next generation of AI startups to reach IPO may scale far more quickly.

To be sure, India is increasingly becoming a physical location for AI infrastructure. India’s operational data-centre capacity reached about 1,789 MW by H1 2026, with another 3,860 MW in the pipeline to 2030, according to Cushman & Wakefield. On the hardware side, India is building chip capacity ahead of demand, with 360 ONE preferring EMS plays over pure fabs, while China’s chip strategy offers lessons for India on how a latecomer might climb the value chain.

So, India isn’t outside the AI investment cycle. And if the AI revolution is primarily an infrastructure cycle, then India’s beneficiaries need not be AI companies, they can be the companies supplying the power, transmission, cooling, construction, equipment, for the data centres. Indeed, Goldman Sachs has identified 42 Indian companies as “AI enablers” across power, data centres and semiconductors and many of these stocks have done far better than the headline indices. This week, we took a look at ESDS Software Solution, India’s home-grown integrated AI infrastructure and cloud solutions provider, and Hitachi Energy India, which is seeing strong momentum from India’s expanding power infrastructure cycle.

The bigger question is: Will India capture the rents, or merely provide the land, labour, electricity and market on which somebody else’s AI rents are earned?

There is, arguably, an even more pressing issue. The world economy is being asked to finance the biggest investment boom in decades just as war, energy inflation, fiscal deficits and higher bond yields are making capital more expensive. Bond yields have moved up sharply in the major economies, and oil prices and US Treasury yields are now in their tightest relationship since 1990, which ties the geopolitics of energy directly to the price of money. Data centres, power infrastructure and semiconductor projects are highly capital-intensive. If the global cost of capital rises, the hurdle rate for those investments rises too. India is competing for AI capital at exactly the moment when capital is becoming more discriminating. Franklin Templeton says global liquidity is becoming less accommodative.

India faces the difficult problem of entering the AI race at a time when capital is becoming expensive, the rupee is weak—we spelt out the implications here and here–oil is being weaponised by geopolitics, and other economies are already claiming the most lucrative parts of the AI value chain. Also, the oil risk may not recede soon: a midterm defeat could make Trump more dangerous on Iran.

The AI revolution will undoubtedly be transformative. But if investors have priced a productivity revolution arriving in three years and it takes seven, the technology can be a success while the investment proposition disappoints. The markets are adept at imagining the profits of the future. They are less forgiving when asked to wait for them.

Cheers,

Manas Chakravarty

In case you missed them, here are some of the other stories and insights we published this week, apart from our technical picks in the equity, commodity, and forex markets:

Stocks

SJS Enterprises, Crompton Greaves Consumer, Kotak Mahindra Bank, Sagility, Cholamandalam Investment, Mazagon Dock, Dairy players, Groww, Engineers India, Shringar House of Mangalsutra, Laurus Labs

Markets

Should MF tax norms take a leaf out of IRDAI’s recent playbook?

Gold’s safe haven status in question as prices dip during global conflicts

Sectoral rotation led to underperformance of large-caps: Motilal Oswal

Why the rich invest in bonds

Markets’ top performer: The Taxman

Financial Times

The trouble with China’s bigger, better batteries

UBS should make the positive case for staying Swiss

Midterms and the market