Macquarie: AI Theme Shifting from Chips to Robotics and Biotech
Taylor Wilson
Macquarie global strategist Viktor Shvets argues AI investment hype is rotating in a "rolling bubble" — the chip bubble is fading, and the next beneficiaries are application-layer plays like robotics, biotech, and quantum computing. The window for betting solely on infrastructure is narrowing.
What is a "rolling bubble" — and where is the money going?
Shvets describes the AI trade as a bubble that keeps changing venue: software ran hot and pulled back; chips ran hot and are now fading; next up is the application layer — robotics, biotech, 3D printing, the metaverse, and quantum computing.
In plain terms = AI itself isn't deflating — hot money is moving within AI, from companies that build infrastructure to companies that build products on top of it.
He advises investors to identify themes likely to perform over the next two to four years, and to prepare for one theme exiting fast as another takes over.
Is the chip bubble really fading — what's the evidence?
Semiconductor stocks in the Nasdaq 100 had rallied roughly 80%, adding about $500 billion in market cap to the index.
But last week the VanEck Semiconductor ETF dropped nearly 10% intraday; it has since partially recovered, yet a pattern of high-level volatility is now visible.
This means → the "easy money" phase for chips may be over, and the risk-reward of staying all-in on semis is deteriorating.
How is Macquarie positioned — still overweight Korea and Taiwan?
Macquarie's June 1 report notes the current bubble is concentrated in AI infrastructure stocks that drive the Korean and Taiwanese markets.
Application-layer sectors — robotics and automation, 3D printing, metaverse, biotech, quantum computing — have not yet formed their own bubbles.
Macquarie is still willing to overweight Korea and Taiwan for now, but warns investors to watch closely for the signal that AI-bubble winners are switching — and to rebalance quickly once it appears.
What does this mean for the ordinary investor?
This reflects a critical timing window: the application-layer bubble has not yet formed, but the likely directions have been named — the window to position early is opening.
In plain terms = if you are fully loaded on chip stocks, the strategy risk isn't that chips get worse — it's that when market attention moves on, capital follows it out the door.
Whether investors can position in the application layer before its own bubble inflates is the key test of this rotation thesis.
Content is for reference only, not financial advice.