Global Stock Sell-Off Deepens as Concerns Over AI Investment Hit Semiconductor Markets

Global financial markets experienced a sharp sell-off on 5 and 6 November 2026 as concerns over the sustainability of artificial intelligence investment triggered heavy losses in technology and semiconductor stocks. The decline, which began on Wall Street, spread across Asian and European markets, wiping trillions of dollars from the value of listed companies and prompting investors to move towards traditionally safer assets such as government bonds.

US equities recorded their steepest daily declines since April 2025. The S&P 500 fell 4.4% on Thursday, the Nasdaq Composite dropped 5.5%, and the Dow Jones Industrial Average declined 3.4%. The Philadelphia Semiconductor Index, which tracks major US chipmakers, plunged 12.2%, marking its largest one-day fall since 2020. The decline left the semiconductor index approximately 16% lower for the week through Thursday's close.

The sell-off reflected growing doubts about the scale of investment being directed towards AI infrastructure and the ability of technology companies to generate sufficient returns from that spending. Major technology groups have committed hundreds of billions of dollars to data centres, advanced processors, networking equipment and electricity infrastructure. Investors are increasingly questioning whether the expected demand for AI computing capacity will justify the costs and whether the construction of new facilities can proceed at the pace anticipated by the market.

Political developments in the United States added to these concerns. The Democratic Party's strong performance in the November midterm elections raised the possibility of increased regulatory scrutiny of data-centre projects. Several successful candidates had campaigned on commitments to slow or restrict new developments, reflecting growing opposition to the facilities among communities concerned about electricity consumption, water use, land requirements and the impact on local infrastructure.

According to market intelligence company Cleanview, approximately 40% of planned US data-centre capacity is located in states where governors or governors-elect have pledged or already imposed pauses on new projects. The prospect of further restrictions has created uncertainty over the timing and scale of future infrastructure investment. Earlier in the year, Senator Bernie Sanders and Representative Alexandria Ocasio-Cortez had also proposed a federal moratorium on new data-centre construction.

Higher borrowing costs have added another layer of pressure. Data centres require substantial upfront capital expenditure, and many projects depend on access to financing before they begin generating revenue. More expensive debt can delay construction, increase operating costs and reduce the expected returns on investment. A slowdown in the development of these facilities could consequently affect demand for semiconductors, electricity, cooling systems and other supporting services.

The market reaction was particularly severe for companies closely associated with the AI investment cycle. Nvidia shares fell 9.3% on Thursday, while Advanced Micro Devices recorded a decline of approximately 17.3%. Qualcomm also came under pressure, falling 9.5% after reporting smartphone chip sales below market expectations. The weakness was not limited to semiconductor manufacturers, with other major technology companies and businesses supplying data-centre infrastructure also experiencing significant losses.

The impact quickly spread to Asian markets, where semiconductor manufacturing and technology companies are important components of national equity indices. Japan's Nikkei 225 fell approximately 6% on Friday, while South Korea's Kospi also recorded a sharp decline after falling as much as 8.8% during trading. The fall in the Korean index triggered a temporary 20-minute trading halt. Taiwan's Taiex closed 4.5% lower after experiencing an even steeper decline earlier in the session.

Major Asian technology companies were among those affected. SK Hynix, a key supplier of advanced memory chips used in AI systems, fell 10.4% in South Korea. Samsung Electronics declined 6.4%, while Taiwan Semiconductor Manufacturing Company (TSMC) closed 4.7% lower after falling considerably more during early trading. In Japan, Kioxia and SoftBank also experienced substantial losses, reflecting investor concerns about both semiconductor demand and the financing of AI-related infrastructure.

European markets followed the decline when trading opened on Friday. The STOXX Europe 600 fell 1.2% in early trading, Germany's DAX declined 1.6%, and France's CAC 40 and the UK's FTSE 100 each lost 1.4%. The losses reflected the transmission of US technology-sector concerns to international markets, although the exposure of European indices to semiconductor companies differs from that of their US and Asian counterparts.

The sell-off also raised concerns about companies providing financing for data-centre development. Dutch bank ING was reported to be preparing the sale of a portfolio of data-centre loans valued at approximately US$500 million. The proposed transaction attracted attention because it came at a time when investors were reassessing the risks associated with lending to businesses whose future performance depends heavily on continued expansion of AI infrastructure.

As equity markets declined, investors sought protection in government bonds. US Treasury yields fell as demand for fixed-income assets increased, while government bonds in several European and Asian markets also strengthened. The yield on the US 10-year Treasury note declined by approximately six basis points during the European morning on Friday. The movement reflected a shift towards safer assets amid concerns about the potential economic consequences of a prolonged technology-sector downturn.

Currency and commodity markets also reacted to the broader uncertainty. The US dollar strengthened against several major currencies, while the euro weakened against the dollar during European trading. Gold prices declined despite the general movement towards safer assets, illustrating that the response across different asset classes was not uniform. Brent crude traded close to US$100 per barrel, with energy markets also responding to developments surrounding the conflict involving the United States and Iran and expectations concerning possible diplomatic progress.

The decline has renewed discussion about whether valuations across the AI sector have become disconnected from the underlying commercial returns generated by the technology. Large-scale investment in data centres and advanced computing has supported strong share-price performance among semiconductor manufacturers and other technology companies. However, the latest market reaction suggests that investors are becoming more sensitive to the possibility that infrastructure spending may not translate into revenue and profitability as quickly as previously anticipated.

The potential economic consequences extend beyond listed technology companies. Data-centre construction has become an important source of investment in the United States and several other economies. A substantial slowdown could affect construction activity, electricity demand, equipment manufacturers, semiconductor exports and businesses providing cloud computing services. Economies with significant exposure to semiconductor production, particularly South Korea and Taiwan, may be especially sensitive to any prolonged reduction in demand.

Nevertheless, the market decline does not establish that investment in artificial intelligence is ending or that the sector is entering a sustained downturn. Some market participants considered the scale of the immediate losses excessive and suggested that the selling reflected a rapid reassessment of risk and investor positioning. The possibility of further volatility remains, however, particularly while uncertainty persists over AI infrastructure spending, financing conditions and the regulatory environment.

For financial institutions and investors, the episode demonstrates how closely connected technology valuations, infrastructure financing and broader market sentiment have become. Concentration of investment in a relatively small number of large technology and semiconductor companies can amplify market movements when expectations change. Banks and other lenders with exposure to data-centre projects may also face greater scrutiny of borrowers' financing structures, future cash flows and dependence on continued demand for AI computing capacity.

The events of 5 and 6 November therefore represent more than an isolated decline in semiconductor share prices. They reflect a wider reassessment of the economic assumptions underpinning the AI investment boom, at a time when political, financing and infrastructure constraints are becoming increasingly relevant to the sector's growth prospects. Whether the correction proves temporary or signals a more fundamental adjustment will depend on the ability of technology companies to demonstrate sustainable returns from their substantial investments.

By fLEXI tEAM