How you can prove to clients that any AI bubble will be manageable within their diverse portfolio.

The AI bubble is now the number one perceived risk to markets, according to a widely watched Bank of America survey. That sounds alarming. Less alarming, perhaps, is the fact that investors are still buying equities enthusiastically, largely because nobody wants to be the only person not invited to the AI party.

When asked in October what posed the biggest tail risk to markets, 33 per cent of global investors pointed to an artificial intelligence equity bubble. That is up sharply from 11 per cent the previous month. Concern is rising quickly. Behaviour, however, has not changed very much at all.

Some prominent voices are starting to sound the alarm. JP Morgan chief Jamie Dimon has most recently raised concerns, followed by the Bank of England. Market attention is increasingly concentrated on a relatively small number of companies, and the risk that bad news in one name sends ripples through the wider market is clearly growing.

Ben Barringer, global head of technology research at Quilter Cheviot, says the noise around a potential AI bubble and a broader market correction is getting louder by the day.

“There are arguably good reasons to be cautious,” he says. “The ten biggest firms in the S&P 500 account for 41 per cent of the index, but only 35 per cent of the profits. There is a mismatch there, although it is not as extreme as we have seen in previous tech bubbles.”

Retail investor momentum is another warning sign. Demand from investors desperate not to miss out has been accelerating. History suggests that when fear of missing out replaces analysis, markets tend to drift towards more dangerous territory.

Recent deal activity has also raised eyebrows. OpenAI is spending aggressively in a bid to emerge as the ultimate winner of the AI race. How this expansion is funded remains unclear. Nvidia and AMD are among those helping to finance this growth, which introduces its own set of risks.

“Being financed by your customers and your wider ecosystem is not a positive sign, nor is it sustainable,” Barringer notes.

Not every company will win in this market. Barringer expects a small number of dominant players to emerge, with the rest seeing their market share shrink. That is not unusual in technology revolutions, but it does make broad assumptions about winners risky.

Valuations, however, remain well below previous bubble territory. Comparisons with the 1999–2000 tech bubble are tempting but imperfect. Back then, valuations reached around 60 times earnings. Today, they are roughly half that.

“Certainly expensive, but not necessarily a screaming bubble,” Barringer says.

For now, earnings and profit growth remain strong, driven by sustained demand and increased government spending. In an environment where interest rates are expected to fall, technology stocks also tend to perform relatively well.

The overall picture remains uncertain. That uncertainty, rather than outright optimism or fear, should shape investor behaviour.

“Diversification is crucial,” Barringer says. “Not just across sectors, but within technology itself. Investors need to avoid overexposure to a single chip manufacturer or AI platform.”

Key signals to watch include corporate IT spending trends over the next year. A slowdown in digital advertising, capital expenditure, or enterprise technology budgets could quickly unsettle markets.

“For now, it is time to stay level-headed,” Barringer concludes. “We are not yet in bubble territory, but with warning flags beginning to appear, a degree of caution would be sensible.”

For advisers, that caution needs to be practical rather than theatrical.

• Pressure-test client expectations. Ask what returns clients are assuming from AI-linked investments, and whether those assumptions are realistic or simply optimistic headlines in disguise.

• Check concentration risk. Many portfolios are more exposed to a handful of tech names than clients realise, often via passive funds.

• Diversify within tech. Avoid portfolios that effectively hinge on one chip maker, one platform, or one narrative.

• Reframe “missing out”. Help clients understand that long-term outcomes are built on disciplined allocation, not perfect timing.

• Agree a plan before volatility hits. Clients make better decisions when the rules are set in advance, not when markets are wobbling.

AI may well reshape the global economy. That does not mean every AI-related stock will reward investors equally. Advisers who help clients separate long-term opportunity from short-term excitement will add more value than any chatbot ever could.

‍