By laila kearney and liz hampton
New York/Houston (Reuters) -The stocks of US power, utility and natural gas companies were sold in some of the most recorded daily drops on Monday. US electric demand and technology spending.
Power Producers was one of the biggest winners, depending on the expectation of the expansion of energy processing data centers needed to expand Big Tech’s artificial intelligence technology last year.
The wider adoption of the AI model, such as DeepSeek, is built in two months, and if it is cheaper than the model used by the US company, the overall demand for electricity can be reduced and the power construction can be smaller. Analysts and economists said.
Analysts of EVERCORE ISI said, “If the truth is proven, the efficiency used within the open source model of DeepSeek can be applied to the model by HyperScalers, which can lead to more relaxed demand,” Evercore ISI said.
Known as a HyperScaling Data Center developer, a large company has invested billions of dollars in developing AI data centers last year.
According to a report from the Lawrence Berkeley National Laboratory, the US data center consumed about 4.4%of electricity in 2023, but is expected to use 6.7%to 12%of all power by 2028.
In 2024, an independent power supplier, Constellation Energy, with an increase of about 100% per share for the ability of stocks to sell nuclear and gas combustion power to US data centers, increased about 20% in Monday trading on Monday’s development.
VISTRA decreased 30% and Talen Energy Corp decreased by 22%.
DeepSeek AI can threaten the dominance of the current AI leader based on Silicon Valley and slow down data center deployment. DeepSeek’s AI support was downloaded from Apple’s App Store on Monday and overturned American rivals.
But Houston’s energy economist, Ed Hirs, said that even if AI has been adopted more extensively, power demand can increase anywhere, even if there is more energy model. He warned that the selling of power stocks could be shortsighted and short -term.
“In this case, DeepSeek turns out to be what everyone wants, and it means faster and faster development if you change the algorithm to adopt an algorithm,” HIRS said.
Nevertheless, even the power producing country related to the power generation and the development of the electricity company was under pressure.
