India’s AI Paradox: Booming Economy, Tumbling Stocks

Here is a number that should not make sense: India’s artificial intelligence market is expanding at nearly 39 percent a year. And here is another number, sitting right next to it, that flatly contradicts it — the Nifty IT index, home to the companies actually building that AI economy, is down roughly 24 percent in 2026 alone, making it the worst-performing sector on Dalal Street this year. Tata Consultancy Services has shed close to 30 percent of its value. Infosys is down about a quarter. Wipro has lost nearly a third. Even Tata Elxsi, a smaller, AI-forward engineering-services player, has fallen close to 30 percent over the past twelve months.

Somewhere in that gap between “the sector is growing” and “the sector is being sold off” lies one of the more interesting economic stories of the year — and almost nobody has told it properly.

Borrowed and Inverted

Earlier this year, the US Federal Reserve made an unusually candid observation about its own economy: financial markets have responded to artificial intelligence with near-euphoric enthusiasm, driving valuations to historic extremes, even as the real economy — output, productivity, employment — shows only patchy, concentrated evidence of transformation. In plain terms: Wall Street is pricing in an AI revolution that Main Street has not yet fully lived through.

It is a genuinely useful lens. And it travels to India — just not in the direction most people would expect.

In America, the imbalance runs one way: markets ahead of reality. In India, the same framework, applied honestly, produces something closer to the opposite. Here, it is the real economy — hiring patterns, government investment, enterprise adoption — that is quietly reorganising around AI, while the stock market has decided to punish, rather than reward, the one sector best positioned to capture that shift.

Start with what is actually happening inside India’s technology companies, as opposed to what is happening to their share prices.

Demand for agentic AI engineers — the specialists who build AI systems capable of executing multi-step tasks autonomously — has risen roughly 260 percent compared with last year, according to industry hiring data. The National Association of Software and Service Companies estimates the country will need close to a million AI-skilled professionals by 2027, against a trained talent pool that currently sits somewhere between five and six-and-a-half lakh. That is not the profile of an industry in decline. That is the profile of an industry racing to keep up with its own demand curve.

At the same time, headline hiring volumes at the country’s largest IT services firms have fallen — the top five companies cut a net of roughly 7,000 positions in the last fiscal year, having added around 12,000 the year before. Read carelessly, that looks like confirmation of the “AI is destroying IT jobs” narrative dominating financial television. Read carefully, it looks like something more precise: a recompositing of the workforce, not a collapse of it. The real economy is reorganising exactly the way the Fed described — concentrated, structural, visible in the composition of employment rather than in the aggregate jobs number.

None of this is showing up favourably in how the market prices these companies.

Visible Patterns

The IT sector isn’t the only place this pattern shows up — it’s just the most visible one. Look at the broader growth data released alongside it, and the “real economy moving faster than sentiment allows” story gets harder to dismiss as a one-sector quirk.

India’s real GDP grew 7.8 percent in the April-June quarter of this fiscal year, comfortably ahead of the Reserve Bank of India’s own 7 percent forecast, according to Ministry of Statistics data. What stands out inside that print is the composition: real Gross Fixed Capital Formation rose 11.9 percent, nearly double the 5.8 percent pace recorded a year earlier, lifting investment’s share of GDP to roughly 34.3 percent from 31.4 percent. Manufacturing gross value added grew 9.2 percent in real terms, its fastest pace in three quarters, with capital goods and electrical equipment production leading the gain.

Credit data tells a related story. Bank lending to industry has been outpacing lending to households through much of 2026, even as retail credit growth cools — a pattern that typically shows up when companies start committing to capacity expansion after a long period of balance-sheet caution, rather than consumers simply borrowing more. Layered on top of this is the government’s push through the India Semiconductor Mission, which has committed substantial fresh funding toward domestic chip fabrication and design capacity — infrastructure that only makes sense if planners expect sustained, AI-linked compute demand over the coming decade.

None of this proves the economy is being “transformed” by artificial intelligence in any single, clean sense — senior economists, have pointed out that manufacturing’s rising real growth has coincided with its nominal share of GVA slipping to a multi-decade low, a reminder that strong volume growth and strong value creation are not the same thing, and that India’s investment cycle has disappointed before. But taken together with the IT-sector hiring data, the picture is consistent: investment, manufacturing activity and industrial credit are all moving in the same direction, at a pace the broader market’s caution doesn’t fully reflect.

The Valuation Puzzle

This is where the story becomes a genuine financial anomaly rather than just a sentiment story.

As of mid-2026, the Nifty 50 as a whole was trading at a price-to-earnings ratio of around 20.9 — roughly 14 percent below its own ten-year average. That is not a market pricing in a bubble. If anything, it is a market pricing in caution, even after a year in which India remained among the fastest-growing large economies on earth, with GDP expanding 7.8 percent in the first quarter of this fiscal year against a Reserve Bank forecast of 7 percent.

Within that fairly valued broader market, IT services stocks have been de-rated even more sharply than the sector-wide numbers suggest. Analysts covering Tata Elxsi, for instance, have noted the stock now trades at roughly 25 times projected FY28 earnings, down from a peak multiple north of 55 times earlier this year — a compression driven not by a collapse in earnings, but by a collapse in the multiple investors are willing to pay for those earnings. The company is producing more revenue. Investors are paying less for each rupee of it.

The stated reasons for the sell-off are real enough — anxiety that large language models will hollow out the outsourcing model that built Indian IT, a freeze in US corporate technology budgets, and disappointing quarterly guidance from Infosys and Wipro through the middle of the year. But notice what those reasons have in common: they are all reasons to worry about the future, priced against companies that are, right now, expanding their AI capabilities faster than at any point in their history. The market is trading the threat. It has not yet had to price the adaptation.

Put the two halves together and the picture that emerges is not the Silicon Valley story of a market too excited about AI. It is closer to the reverse: a real economy that is quietly, measurably reorganising around artificial intelligence — in hiring composition, in government investment through initiatives like the Semiconductor Mission, in enterprise adoption running at nearly 39 percent annual growth — while the equity market remains unconvinced, treating India’s most AI-exposed companies as casualties of disruption rather than beneficiaries of it.

That gap will not persist indefinitely. Either the earnings eventually confirm the market’s scepticism — AI-driven margin pressure finally shows up in the numbers, and the sell-off turns out to have been prescient — or the earnings confirm what the hiring data already suggests, and investors currently sitting out Nifty IT at a 24 percent discount end up having missed the re-rating of a decade.

For readers watching their portfolios rather than the headlines, that is the actual question worth asking this year. Not whether AI is a bubble. But whose price — the market’s, or the economy’s — is currently wrong.

(The writer is a bio-fintech consultant and columnist. Views personal.)