AI's reality check: a temporary selloff or a shift in market leadership?
Key takeaways
AI stocks are falling despite strong earnings, as high valuations, massive spending and political uncertainty collide.
Strong demand suggests the AI boom remains intact, but financing pressures and uncertain returns raise fundamental questions.
Investors should distinguish a temporary correction from a lasting rotation towards other sectors and more profitable businesses.
On Wednesday, Wall Street celebrated the Democratic sweep in the US midterm elections. By Thursday, investors were running for the exits, particularly in AI stocks.
On 5 November, the Nasdaq 100 plunged 6.36%, closing at 26,629.81. Semiconductor companies suffered particularly heavy losses, alongside data-centre developers and other AI-related businesses.
The selloff is striking because AI demand remains strong. Companies continue reporting impressive earnings and announcing enormous investments.
So, what changed? And more importantly, are we witnessing a temporary reassessment of expensive stocks, or the beginning of a more fundamental rotation in market leadership?
What triggered the AI selloff?
The weakness did not start with the elections. AI stocks had already been under pressure throughout October, as investors questioned whether share prices had moved too far ahead of profits.
Three forces are now colliding.
First, expectations. Companies such as Nvidia, AMD and Broadcom have benefited enormously from AI investment. But when valuations already assume years of exceptional growth, even strong earnings can disappoint.
Second, spending. Microsoft, Amazon, Alphabet and Meta are committing extraordinary amounts to AI infrastructure. Investors increasingly want evidence that this spending will generate attractive returns, rather than simply higher costs.
Third, politics. The Democratic sweep raises concerns about tighter oversight of data-centre construction, electricity consumption and environmental pressures.
The election may have accelerated the selloff, but the underlying doubts were already there.
Is something fundamentally changing?
Not necessarily, at least not in AI demand.
Nvidia's latest quarterly data-centre revenue reached 89 billion USD, more than doubling from a year earlier. That hardly suggests an industry running out of customers.
Yet strong demand does not automatically translate into attractive investment returns.
Interestingly, Microsoft, Amazon, Alphabet and Meta have shown relative resilience compared with chipmakers and AI infrastructure suppliers. This could signal an early rotation within the AI trade, rather than a broader retreat from the technology.
These technology giants still face challenges. However, their profitable existing businesses, diversified revenues and stronger balance sheets provide a cushion that smaller AI companies often lack.
For infrastructure developers such as CoreWeave and Nebius, financing costs and construction delays present greater risks. Slower data-centre expansion could hurt hardware suppliers while potentially benefiting established operators with existing capacity.
This leaves two possibilities.
If demand remains strong and AI investments deliver attractive returns, the selloff may prove a temporary valuation reset. If construction slows and profits disappoint, investors may increasingly favour financially stronger businesses, potentially signalling a lasting shift in market leadership.
The question is shifting from how fast AI can grow to who can turn that growth into sustainable profits.
Beyond AI: is the market changing its favourites?
Perhaps the most interesting development is happening beyond AI.
October already offered signs of a rotation. While the S&P 500 technology sector fell 6.8%, nine of the index's eleven sectors advanced, and more than 70% of its companies gained ground. Investors were not necessarily abandoning equities, but increasingly looking beyond AI for opportunities.
That suggests a potential broadening of market leadership, rather than simply a collapse in confidence.
Even within technology, differences are emerging. Software companies such as ServiceNow and Salesforce face questions about AI monetisation, while AppLovin's earnings-driven rally on Thursday showed that the market can still reward strong execution.
But Thursday's broad market decline suggests investors were also reducing risk generally, not merely switching between sectors.
The distinction matters. A healthy rotation means other businesses take the lead. A broader selloff means investors are becoming less willing to take risks altogether.
A lasting change in leadership would require more than a few difficult sessions for AI stocks. It would mean other sectors consistently outperforming, even as market conditions stabilise.
An investor playbook
- If AI earnings remain strong: compare valuations with realistic profit expectations. Falling shares do not necessarily mean deteriorating businesses.
- If infrastructure spending slows: monitor customer orders, financing costs and free cash flow, particularly among smaller developers.
- If market leadership broadens: review portfolio concentration and consider diversification beyond AI, without abandoning the theme solely because of volatility.
The AI revolution and the AI trade are different stories
Markets sometimes sell first and ask questions later. Thursday's selloff may prove excessive, particularly if demand remains strong and companies continue delivering impressive results.
But investors should not dismiss every concern as temporary nervousness.
AI could transform the economy for decades while individual AI stocks struggle because expectations, valuations or spending commitments become unsustainable.
The coming quarters will help distinguish a temporary correction from a lasting shift in market leadership. Earnings, cash generation and evidence of sustained strength outside technology will matter more than any single trading session.
For investors, the lesson is simple: the AI revolution may continue for years, but that does not mean the same stocks will lead the market throughout.