Investors use dotcom era playbook to dodge AI bubble risks (reuters.com)

Major investors are quietly reviving a dotcom-era playbook to ride the AI rally while dodging a potential bubble: rather than shorting headline winners like Nvidia — whose market value has topped $4 trillion — they're rotating gains into adjacent, underpriced beneficiaries and hedges. Asset managers from Amundi to Goshawk and Fidelity describe tactics similar to 1998–2000, shedding frothy mega-cap names and redeploying capital into software, robotics, Asian tech, infrastructure suppliers and even power plays (e.g., uranium) that stand to benefit from hyperscaler AI buildouts. The aim is to stay exposed to AI upside without the concentrated risk of the “Magnificent Seven” or the speculative options activity seen on Wall Street. Technically, this strategy mirrors a documented hedge-fund win in the dotcom boom: a study by Brunnermeier and Nagel found nimble funds outperformed by about 4.5% per quarter from 1998–2000 by rotating out before peaks. Investors highlighted practical targets — IT services, Japanese robotics, Taiwan suppliers to TSMC, and Chinese stocks as a regional hedge — and flagged structural risks such as data-center overcapacity and massive, overlapping hyperscaler capex. The trend signals capital is diffusing across the AI value chain, which could prolong sector strength while seeding vulnerabilities typical of past tech bubbles.