AI stocks keep selling off. Here’s what is making investors nervous

Investors have become accustomed to abrupt declines in AI-linked stocks, but experts claim these fluctuations stem from fresh data prompting markets to reassess the level of investment companies are making and the time it will take for those investments to yield results. The underlying concerns are well-known, with AI valuations remaining elevated, tech firms pouring substantial resources into infrastructure, and investors eagerly anticipating whether revenue and earnings will eventually validate those expenditures.

Selling often reignites when alterations in earnings, capital expenditure plans, or new technological advancements challenge these assumptions. Exness Financial Markets Strategist Wael Makarem explained that markets continually react to new data points such as news headlines, corporate actions, earnings reports, guidance, or technological advancements, which can impact investor expectations, market direction, and the economics of the sector.

Geopolitical instability, leveraged positions, and crowded trades can exacerbate these moves, causing investors to reduce exposure more rapidly when confidence wanes, stated Makarem. Capital expenditure has emerged as a critical area of concern, as investors seek evidence that the funds allocated to AI infrastructure are generating sufficient returns.

Rising investments without a corresponding increase in revenue can trigger selling, while weaker-than-expected cloud revenue, earnings misses, and higher memory costs can put pressure on companies throughout the AI supply chain. The presence of cheaper, more advanced Chinese AI models has also introduced questions about the extent of infrastructure that needs to be constructed and at what speed.

For many, high valuations are tolerated as long as investors believe the returns will come. However, this tolerance disappears when the expected payoff from that spending changes significantly. For instance, Alphabet's decision to increase its 2026 capital-spending guidance from $180 billion to $205 billion in late July triggered renewed scrutiny of whether future returns would justify the additional investment.

High valuations can be acceptable for months when companies continue to beat earnings expectations and analysts continue to raise profit forecasts. The issue arises when these expectations cease to rise. This sensitivity is particularly pronounced in AI stocks because their prices already incorporate expectations for years of robust growth, leaving less room for disappointment.

Madhur Kakkar, Founder and CEO of Elevate Financial Services, highlighted that expensive valuations are easier to accept while earnings forecasts continue to increase, but this relationship alters when upgrades begin to level off. Behavioral factors also come into play when a strong AI rally occurs; many investors own the leading names, partly because being underweight poses risks.

When momentum shifts, holding these stocks becomes increasingly challenging, leading to accelerated selling. AI-related news can also spread and trigger broader market moves when investors are already concerned about interest rates, economic growth, geopolitical risk, or elevated equity valuations. If only a handful of highly valued AI stocks are declining while earnings expectations for the broader market remain stable, it may indicate profit-taking in a small group of expensive stocks.

However, if the selling extends to profitable technology companies, semiconductors, infrastructure providers, and eventually the wider market, it signifies a more significant warning about risk appetite. Separating companies that are heavily investing in infrastructure from those positioned to earn directly from that investment has become a distinct trend.

Moody's estimates suggest that major hyperscalers could spend around $785 billion this year and nearly $1 trillion in 2027, leaving investors focused on where adequate returns will eventually emerge from this spending. A more prolonged correction would necessitate a deterioration in the fundamentals supporting AI investment, rather than a temporary shift in sentiment.

Material reductions in AI capital spending by hyperscalers combined with slower AI-related revenue growth would provide investors with stronger grounds to question present earnings assumptions.