The sell-off on Wall Street was exceptionally severe Thursday. The S&P 500 Index lost 4.4% (in USD), the Dow Jones fell 3.4% and the Nasdaq Composite tumbled 5.5%, pushing the Nasdaq firmly into correction territory (-12% from its recent peak).
At its lowest point, the Nasdaq was down around 6.4%, before recovering some of its losses towards the close. Stocks closely linked to AI, semiconductors and data centres were hit hardest. The Philadelphia Semiconductor Index plunged 12%, highlighting the intensity of the risk-off move across the technology sector.
"}},{"componentType":"paragraph","richBody":{"value":"The immediate catalyst was the Democratic victory in the midterm elections. The main concern is that the physical build-out of AI infrastructure could become more expensive, slower and more complex.
Attention is particularly focused on Texas, one of the key regions for new data centres, where tighter conditions could emerge around electricity and water consumption, grid capacity, local energy costs and tax incentives.
However, the market reaction deserves some perspective. Greater scrutiny of data-centre development was already emerging before the elections. More importantly, restrictions in Texas would primarily affect the cost, location and timing of investment rather than the underlying demand for computing capacity.
Hyperscalers still need that capacity and can redirect part of their investment towards other regions. Developers in Texas are also exploring solutions based on dedicated or nearby power generation.
Regulation can therefore create delays and increase costs, but it does not mean that planned AI investment will suddenly disappear.
"},"title":"Midterm elections add uncertainty, but do not change the underlying demand for AI"},{"componentType":"sectionTitle"},{"componentType":"paragraph","richBody":{"value":"The scale of the market correction also needs to be put into perspective against corporate fundamentals. Around 90% of reporting IT companies beat earnings expectations in the third quarter, and earnings growth remains strong.
The debate is increasingly shifting away from pure revenue growth towards free cash flow and the return on the enormous amounts of capital being invested in AI. Hyperscalers continue to generate substantial operating cash flows. However, rising capital expenditure and depreciation mean that investors will increasingly want evidence that AI can generate sufficient additional revenue and cash flow to justify these investments. This is a genuine risk and is likely to become an increasingly important focus in the coming quarters.
At the same time, valuations have fallen significantly following the recent declines. Share prices therefore appear to be deteriorating considerably faster than the underlying earnings outlook.
In the short term, further position unwinding could generate additional volatility and push risk premia higher. But as long as earnings growth, margins and cash flows remain robust, it seems premature to interpret the current correction as the end of the AI investment cycle.
The risks have clearly increased, but the speed and magnitude of the sell-off also suggest that technical factors and positioning are playing an important role, potentially amplifying the decline.
"},"title":"Share prices are falling faster than fundamentals"},{"componentType":"sectionTitle"},{"componentType":"paragraph","richBody":{"value":"After months in which almost every decline in AI-related stocks automatically attracted new buyers, markets are clearly becoming more demanding. This is an evolution we had anticipated by maintaining a neutral stance on equities.
Volatility is nothing unusual in the technology sector and should be put into perspective. The Nasdaq has now fallen by more than 12% from its early-October peak. But since 2000, the index has experienced an average intra-year decline of around 21%, while still delivering a positive annual return in 20 of the past 26 years.
The current volatility therefore does not mean that enthusiasm for AI has disappeared. Instead, it shows that investors are becoming significantly more demanding when it comes to valuations, capital expenditure and future earnings growth.
In that respect, it is worth remembering that the earnings of the major AI companies surged by 96% in euro terms during the first half of the year. The outlook also remains encouraging, with average earnings growth expected to reach around 95% for the full year 2026, compared with 89% for the energy sector and 35% for major global equities.
The sharp return of market volatility is therefore not, in our view, a reason to abandon AI exposure. AI remains a structural investment theme that is likely to continue attracting substantial capital over the short, medium and long term.
Rather, the current correction provides active investors with an opportunity to reassess their positioning and focus more selectively on technology companies with the strongest fundamentals, balance sheets, cash-flow generation and ability to translate AI investment into sustainable earnings growth.
We also remain convinced that data-centre construction will continue. The pace may vary, tax incentives may change and some US data-centre projects may have to move from one state to another. Regulation may become tighter and additional safeguards may be introduced. But the underlying demand for computing power remains substantial.
"},"title":"We remain positive, but selective, on the AI theme"},{"componentType":"sectionTitle"},{"componentType":"paragraph","richBody":{"value":"Against this backdrop, ING’s investment strategy is to use periods of market weakness to selectively and gradually rebuild portfolio exposure to IT and AI, particularly in the United States and emerging markets.
The recent correction does not mean that the AI story is over. But it does mark an important transition: from enthusiasm towards greater selectivity, and from investing in the AI theme broadly towards identifying the companies best placed to turn massive investment into sustainable cash flows and earnings.
"},"title":"Our investment approach"}]},"hasMacro":false,"id":"be7538bd-832d-47a8-9de3-9721b0dcca16","localeString":"en-GB","mainHeaderZone":{"backLink":{"textLink":{"text":"News","url":"/en/individuals/news/economy-and-financial-markets"}},"componentType":"editorialHeader","coreHeader":{"body":"Sometimes, markets need just one day to change their mind completely. On Wednesday, investors celebrated the clarity provided by the US midterm elections with a strong rally. On Thursday, that sense of relief proved short-lived. Just one day after the celebrations, investors began asking what the election result could mean for AI, data centres and the enormous investment required to support them. Wall Street’s response was anything but subtle...","headerImage":{"extension":"jpg","original":"https://assets.ing.com/asset/78225b33-ba2f-4386-9b0f-6b2c60f985dd/side-view-of-woman-using-laptop-while-sitting-at-office.jpg","publishedAt":"2026-11-06T11:59:00Z","transformBaseUrl":"https://assets.ing.com/transform/78225b33-ba2f-4386-9b0f-6b2c60f985dd/side-view-of-woman-using-laptop-while-sitting-at-office","type":"image","updatedAt":"2026-11-06T11:59:00Z","width":5700},"title":"Important Market Update"},"date":"2026-11-06","readingTime":5},"publishDate":"2026-11-06T14:20:47.428+01:00"}}