The hard part of investing in AI stocks - or steering clear of them - is knowing which companies genuinely belong in that category. At various times, investors have labeled big tech, software makers, chip companies, emerging markets, and energy firms either winners or losers in the AI trade. The result is a market chase built around something that can be difficult to pin down.
?
The first phase of the AI stock frenzy
In 2023 and 2024, ChatGPT set off an early rush into AI stocks, and markets valued companies with the financial judgment of Del Boy. Tech stocks took the hardest hit from that optimism, with many priced as though ordinary investors were about to become millionaires. Doubts began to surface around companies such as Mark Zuckerberg's Meta, along with the big-tech hyperscalers whose enormous balance sheets support the AI infrastructure boom.
Investors keep changing their minds about whether hyperscalers are spending too much on AI. When big tech seems to be burning through cash - what's a couple of hundred billion dollars between friends? - shares sell off; when revenue growth shows up, they rally.
Investors did not seriously entertain the possibility of AI producing losers as well as winners until February 2026. Chatbots looked particularly capable at coding, and short sellers moved into software. During the first half of this year, Adobe and Salesforce both lost 38% of their share value. At the same time, chipmakers - whose products were supposedly making programmers redundant - were caught in a blistering rally. South Korean shares tripled in the 12 months ending 19 June.
The "sell software, buy chips" trade now seems to be over. Korean shares plunged, while the S&P software index regained nearly all of its losses for the year.
Investors remain nervy about AI stocks
Technology shares fell sharply last week, sending the Nasdaq into a correction - a 10% decline from its peak. Trading steadied at the start of this week, but investors were still uneasy, especially after a spy drama involving Nvidia. The immediate trigger was the US midterm blue wave. Washington may now face gridlock, with a Democrat-led Congress clashing with a Republican president, but investors were more focused on the state elections. Democrats who opposed data centres won the governorships in states that host key parts of America's data-centre network. Chipmakers could feel the effect if demand for their products falls below forecasts, Telis Demos writes in The Wall Street Journal.
Talk of tighter AI regulation has pushed up shares in companies such as Amazon and Microsoft. Both already operate large data centres, and fewer new competitors may now enter the market. Investors had mainly worried that big tech would build more data centres than could earn a decent return. The irony is that traders seem to think left-wing regulators could stop Silicon Valley's AI-enthusiastic executives from overreaching.
Power companies could lose the most if US data-centre construction slows. Electricity shortages have held up new projects, and securing a grid connection often takes at least five years. To get power sooner, operators shifted toward off-grid systems that can be installed in roughly 18 months.
That bottleneck now appears less severe. Cheap, plentiful energy and light regulation drew Oracle, OpenAI, CoreWeave and Amazon into Texas, where they announced a wave of projects, Avi Salzman reports in Barron's. Yet Gina Hinojosa, the state's new Democratic governor, seems likely to extend a data-centre moratorium until lawmakers address AI's effect on water use and electricity bills. Goldman Sachs estimates that moratoriums in Texas, Georgia and Ohio could put 45% of planned data-centre capacity at risk.
Shares in Constellation Energy, the largest electricity generator in the United States, have fallen by a quarter this year. Power-plant companies were already losing investor support as politicians pushed back against connecting new data centres to grids that were under strain, Jinjoo Lee reports in The Wall Street Journal. That shift also hurts investors who had bet on equipment makers whose machines can run away from the grid. GE Vernova shares, for example, doubled during the first half of the year, while Caterpillar rose 77%. Both stocks sold off after the election.
"Utilities were once regarded as safe havens," Emily Forgash writes in Bloomberg. That reputation has taken a hit as AI reshapes the sector. Utilities rank as the second-worst-performing of the S&P 500's 11 sub-sectors this year, with rising interest rates adding pressure to an industry that carries heavy debt.
Leave a comment on: AI stocks have new winners and losers after the US midterm elections reshaped Congress. Investors may be chasing something that isn't there. So where should they look now?