AI-Powered Energy Stocks Rebound Despite DeepSeek Model Concerns

Key Takeaways:

* AI-exposed energy stocks surged on Tuesday, reversing previous losses.
* DeepSeek's competitive and cost-effective AI model raises questions about future energy demand.
* Experts acknowledge uncertainties but anticipate accelerated electricity consumption driven by AI.
* Data center deployment remains intact in the medium term, with long-term projections under scrutiny.
* Wall Street analysts emphasize the speculative nature of energy usage estimates for AI.

Details:

On Tuesday, AI-exposed power stocks experienced a rebound after facing a significant downturn the previous day due to a tech rout triggered by the release of a new AI model from DeepSeek, a Chinese startup. Constellation Energy (CEG), the top nuclear plant operator in the US, gained over 1% after a 21% drop in the prior session. Electricity and power generator Vistra Corp (VST) jumped 9% after a record 28% decline. Power equipment maker and servicer GE Vernova (GEV) also saw a 7% increase following a 21% slide.

DeepSeek's AI model, introduced in late January, has gained traction, surpassing OpenAI's ChatGPT in US iPhone downloads and receiving acclaim. This has sparked concerns among investors who believed in the demand for data centers and energy due to growing AI utilization. However, questions are now emerging about the long-term strength of energy demand for AI-related purposes.

Analysts acknowledge that data center deployment remains on track in the medium term. However, they question if the trend will accelerate or stabilize, leading to moderated demand growth in the latter part of the decade. Wall Street analysts emphasize the speculative nature of data center energy usage estimates as AI technology evolves and becomes more efficient.

Despite the uncertainties, data shows that US electricity consumption has already increased by 2% in 2024, with continued growth projected for the coming years. This is occurring simultaneously with utility companies' expansion of electric grids and infrastructure.

As a result, analysts emphasize that the transition to data-driven infrastructure is a long-term process. While DeepSeek's AI model raises questions, the broader trend of increasing energy needs appears to be intact for the foreseeable future.