The brutal selloff sweeping through global tech and AI stocks may not be the end of the story, but it is a necessary and overdue correction, according to Jonathan Schiessl, Deputy CIO at Westminster Asset Management.
Speaking on ET Now as Korean tech stocks and the Mag-7 came under sharp pressure, Schiessl offered a measured but candid read on where the AI trade stands, what a full unwind could mean globally, and why India is surprisingly well-positioned to ride it out.
The selloff was coming. It may still have teeth
For Schiessl, the scale of the correction was not a surprise. Stocks like Samsung and SK Hynix had "gone vertical," he said, and a growing crowd of investors had been waiting for something — anything — to trigger a pullback.
His diagnosis: certain parts of the market had entered what he described as a "blowoff top" phase, that final, frenzied push higher before a steep reversal. Heavy leverage and crowded positioning into the same trade meant that when the unwind began, it was always going to hurt.
But he stopped short of calling it a collapse. "I am not sure the game is over in this area," he said. Dramatic rises in sectors are typically followed by steep selloffs - that is normal market behaviour. The bigger question is whether the underlying thesis has broken down. On that, Schiessl is not yet convinced it has.
US earnings are holding the floor
One reason Schiessl is reluctant to call this a full bubble burst is the earnings picture. US corporate earnings have been "quite extraordinary," he said, and there is a plausible case that current American market valuations are reasonably justified by actual earnings growth, not just hype.
The Mag-7 names, he noted, actually have valuation support and had been quietly underperforming the broader market even as memory chipmakers were being re-rated. That divergence matters when assessing how much structural risk actually sits in the AI trade.
The real threat: The global cost of capital
Where Schiessl sees a genuine structural risk is not in the technology itself but in the capital environment surrounding it. The sheer scale of money being deployed globally, AI data centres, defence spending, government borrowing across western economies, energy transition, and potential Middle East reconstruction, means one thing: the cost of capital is heading in only one direction.
"Globally, capital is only going one way, which is up," he said, "and that could put a blocker on the funding for the whole data centre story."
There is a counter-argument, he acknowledged. AI models coming out of China are getting cheaper, and cheaper technology typically drives higher adoption. The long-term trend is intact. But the sheer concentration of investor money in the sector needs to cool before the next leg up can be sustained.
Why India is better placed than most
India's lack of direct exposure to the AI trade, often seen as a missed opportunity, is now functioning as a form of insulation. With no major domestic AI hardware or chip story to unwind, Indian markets are largely sheltered from the forced selling hitting Korea and the US.
Schiessl's broader view on India is constructive, particularly if Middle East tensions ease and oil prices remain well-behaved, both significant inputs for the Indian economy.
The one sector he is firmly avoiding, however, is Indian IT. Despite meaningful derating, he sees too much uncertainty around the business model to justify stepping back in. The threat of AI disruption to traditional outsourcing is real, the earnings visibility for the next four to five years is poor, and at roughly 18 times earnings, Indian IT is still more expensive than comparable Chinese internet stocks trading at 12 times, which have also been underperforming.
Until the business model question is resolved, Westminster is staying away.
