Global Market Today: Asian shares bend under AI stress; Apple earnings loom

Asian shares fell as investors worried that the AI investment boom may be becoming too expensive relative to the profits it generates. The article links the anxiety to negative cash flow, the huge scale of planned AI spending and earnings growth that is strong but no longer accelerating. These concerns made investors more cautious about technology-heavy markets.

The pressure was visible across the region. Japan’s Nikkei fell 0.5%, South Korea lost 0.8% after shedding 16% in October, and MSCI’s broadest Asia-Pacific index outside Japan slipped 0.4%. Intel’s nearly 29% share-price tumble intensified concerns ahead of major technology earnings.

The market is now watching Apple, AMD and Palantir for evidence that AI investment is producing adequate returns. Goldman Sachs expects hyperscaler capital spending to reach $800 billion this year and $1.2 trillion in 2027. Such figures raise the stakes if earnings growth slows while investment keeps climbing.

AI capital expenditure, or capex, is spending on long-lasting assets used to build AI systems and services. It includes data centres, advanced chips, servers, networking equipment and related computing infrastructure. This spending matters because AI requires large amounts of computing capacity, so companies must expand physical infrastructure before generating expected returns.

Hyperscalers are committing especially large sums. Goldman Sachs expects their capex to almost double to $800 billion this year and rise to $1.2 trillion in 2027. That money supports the facilities and equipment needed for expanding AI services. It is recorded as investment rather than an ordinary day-to-day expense, but it still requires major cash outflows.

The risk is that spending can grow faster than profits. Goldman expects AI investment to continue rising, but says its growth rate is peaking. Investors therefore want evidence that this infrastructure will produce earnings, rather than only larger bills and weaker near-term cash flow.

The expected spending is enormous: Goldman Sachs forecasts hyperscaler capex of almost $800 billion this year, rising to $1.2 trillion in 2027. Hyperscalers are the large technology companies operating extensive online computing infrastructure. Their spending is aimed at expanding AI capacity through data centres, chips and other equipment.

The key mechanism is the gap between investment and returns. Companies pay for infrastructure now, while the revenue and profits expected from AI may arrive later. If demand and earnings grow quickly enough, the spending can be justified. If not, negative cash flow and concerns about profitability can spread through equity markets.

The scale also raises the market’s sensitivity to earnings news. Goldman expects S&P 500 earnings-per-share growth near 30% in 2026, slowing to 18% the following year. Investors are therefore watching whether AI investment supports that growth or becomes a sign that spending is accelerating beyond profits.

Technology stocks are valued partly on the profits investors expect in the future. If AI spending keeps rising but earnings growth stops accelerating, investors may decide that the investment is producing weaker returns than expected. That can reduce confidence in companies whose share prices already reflect strong future growth.

The mechanism is straightforward. More data centres, chips and computing equipment require cash today, while profits may come later. If revenue does not keep pace, negative cash flow becomes more concerning. Goldman Sachs described the core challenge as strong earnings growth that is no longer accelerating, alongside AI capex whose growth rate is peaking.

The immediate result could be greater volatility or lower technology share prices, although the article does not predict a specific outcome. The Nasdaq already lost almost 7% in October, its worst month since April 2025. Apple, AMD and Palantir earnings may show whether AI spending is supporting profits.

These earnings reports matter because they may provide fresh evidence about technology demand, profitability and the financial effects of AI investment. Investors are already worried about negative cash flow and whether earnings growth is keeping pace with rapidly increasing capital expenditure. Results from prominent companies can therefore influence confidence across technology markets.

Intel’s nearly 29% share-price fall on Friday created a difficult backdrop. Apple, AMD and Palantir are scheduled to report afterward, so investors will compare their results and outlooks with the concerns exposed by Intel’s decline. The reports may show whether companies are converting AI-related demand and investment into stronger earnings.

The broader context is demanding. Goldman expects hyperscaler capex to reach $800 billion this year and $1.2 trillion in 2027, while S&P 500 EPS growth is forecast to slow from nearly 30% in 2026 to 18% next year. That makes each report more important to market expectations.

A credit-default swap, or CDS, is a financial contract that transfers the risk of a borrower defaulting from one party to another. The buyer pays a premium for protection if the borrower fails to meet debt obligations. The article refers to the cost of insuring against corporate defaults, which is the price of that protection.

The mechanism is based on perceived risk. When investors become more worried about a company’s ability to repay, they demand more protection. The premium, or CDS price, rises. A broad increase can indicate that markets see greater financial stress, especially when concerns include negative cash flow and very large investment commitments.

The article says this insurance cost “blew out” last week as investors questioned the scale and profitability of AI capex. It also says stressed investors may have sold gold to raise money for margin calls. Together, these developments contributed to an uneasy backdrop for technology earnings and wider markets.

Interest rates affect how investors value future profits. When rates rise, future cash flows are generally worth less in today’s money because investors can earn more from safer assets. Companies whose expected profits lie far ahead are therefore more sensitive to rate increases than companies generating substantial profits now.

Bond yields provide a visible market measure of those borrowing and discounting conditions. In the article, the 10-year Treasury yield was 4.952%, while the two-year yield was 4.612%. Futures still implied a 72% chance of a quarter-point Federal Reserve hike in December, although markets priced 55 basis points of total tightening, down from more than 70 a month earlier.

This matters for AI and technology shares because their prices reflect ambitious future growth. Higher yields can pressure those valuations even when business prospects remain strong. The article also notes that a sustained stock selloff could affect expectations about further Federal Reserve hikes.