DeepSeek’s AI Technology Causes Sell-Off in U.S. Power Stocks
By Laila Kearney and Liz Hampton
NEW YORK/HOUSTON (Reuters) – On Monday, U.S. power, utility, and natural gas companies experienced significant sell-offs, marking some of the largest one-day drops on record. This was due to new AI technology developed by Chinese start-up DeepSeek, which cast doubt on the projected surge in U.S. electricity demand and tech spending.
Power producers had been among the top performers in the S&P 500 in the previous year, fueled by expectations of a substantial increase in demand from energy-intensive data centers necessary for the expansion of Big Tech’s artificial intelligence technologies.
However, the introduction of AI models like the one created by DeepSeek, which claims to be more cost-effective and efficient than existing models used by U.S. companies, has raised concerns about a potential decrease in overall electricity demand. This could result in a smaller power build-out, according to analysts and economists.
Analysts at Evercore ISI noted, “If proven true, the efficiencies utilized within DeepSeek’s open-source model could be adopted by hyperscalers, leading to a more moderate demand.”
Big Tech firms, also known as hyperscaling data center developers, have invested billions of dollars in AI data center development over the past year. In the U.S., data centers consumed approximately 4.4% of electricity in 2023, with projections indicating that this figure could rise to 6.7% to 12% by 2028.
The impact of DeepSeek’s advancements was immediately felt in the market, with independent power provider Constellation Energy seeing a 20% decline in its stock price on Monday. Other companies like Vistra and Talen Energy Corp also experienced significant drops.
Furthermore, DeepSeek’s AI technology poses a threat to current AI leaders based in Silicon Valley and could potentially slow down their deployment of data centers. DeepSeek’s AI assistance even surpassed U.S. rival ChatGPT in downloads on Apple’s app store.
While the adoption of more energy-efficient AI models could lead to a surge in power demand, energy economist Ed Hirs from the University of Houston cautioned against a hasty sell-off of power stocks. He emphasized that embracing new technologies could result in greater and faster development.
Despite the potential benefits of AI technology, electricity companies and related feedstock producers are under pressure. Constellation’s recent acquisition of natural gas producer Calpine Energy for $16.4 billion signaled a growing expectation of increased gas demand as a generation source for AI.
Shares of publicly-traded natural gas producers, including EQT Corp and midstream operator Energy Transfer, also saw declines following DeepSeek’s advancements.
In conclusion, the introduction of DeepSeek’s AI technology has caused disruption in the U.S. power industry, prompting a reevaluation of projected electricity demand and tech spending. Companies are now faced with the challenge of adapting to these new developments in order to remain competitive in the evolving landscape of AI technology.
