Preparing for an AI bubble crash
With a potential AI bubble crash in 2026, what should investors do to ensure they are not over exposed to tech stocks, and what other types of investments could they consider?
Copy article
4 responses from the Newspage communityCopy all
Copy all
Copy
If you are worried about a potential AI bubble bursting in 2026, the first step is simply to look under the bonnet of your portfolio and ask a very basic question: how much of my future depends on one story going right? Over the past three years, returns have been dominated by a small group of AI linked giants and the “picks and shovels” providers of chips and infrastructure. Prices have been driven more by hope than by proven profits, at a time when overall valuations (CAPE, Buffett indicator) are already at levels last seen just before the dot-com crash. That combination should at least prompt a sober sense-check rather than blind faith that “this time is different”. Not being over-exposed does not mean abandoning AI altogether; it means not letting one theme quietly swell to an outsized share of your wealth. Trimming back concentrated positions in the obvious AI winners and rebalancing into areas where valuations still reflect real-world cash flows, rather than blue-sky projections
Copy
The first thing to do is accept that markets go down as well as up. You cannot always avoid market falls. The second thing to do is to have a look at your investments and see what exposure you have. If you have a passive US exposure, such as an S&P 500 tracker, that's going to get you about 34% in tech. If you feel that the AI boom is very real that might be fine but if you're worried, it may be worth considering something more diversified. You could look for an active manager who has limited their exposure to AI stocks, but you may find their recent performance hasn't been so good. That's the thing with bubbles, they do so well that they drag lots of people along with them and those that don't follow can look silly, until the bubble bursts.
Copy
If the AI bubble pops in 2026, investors should do what generations before them have done in times of panic: pretend they ‘always planned’ to diversify and casually move some money into boring things like bonds, gold, or anything that doesn’t say disruptive in the brochure.
Copy
Over the past three years, a peculiar financial mania has taken root in markets, where a handful of technology giants and their supporting cast of chip manufacturers have delivered returns that bear precious little relationship to actual profits and everything to do with collective hope that this time really is different. Valuations have inflated to levels not witnessed since the dot-com crash of 2000, where the Buffett indicator and cyclically adjusted price-to-earnings ratios scream that something has become dangerously disconnected from economic reality.
Investors have poured capital into AI not because the business models are proven but because missing out feels worse than the risk of losing everything when sentiment eventually reverses. The maths of bubble correction is utterly indifferent to your conviction or your narrative.
When a small group of stocks dominates returns, and those returns depend on hope rather than cash flows, eventual reversion does not ask for permission.
Investors have poured capital into AI not because the business models are proven but because missing out feels worse than the risk of losing everything when sentiment eventually reverses. The maths of bubble correction is utterly indifferent to your conviction or your narrative.
When a small group of stocks dominates returns, and those returns depend on hope rather than cash flows, eventual reversion does not ask for permission.