ReadingResearchRadarInvestment framework
Sign in / Sign up
← Back to reading
NewsMaterial2 min readDisconfirmAI Editorial

Janus Henderson Portfolio Manager Denny Fish: Stricter Data Center Regulation Will Benefit Diversified Giants and Weaken New Cloud Startups

Janus Henderson portfolio manager Denny Fish argues that regulatory risks for data centers stemming from Democratic state-level victories in the 2026 U.S. midterm elections are more likely to benefit hyperscale cloud service providers with diversified business lines and ample cash flow, while weakening AI new-cloud startups reliant on single-product pricing advantages and smaller new entrants. He notes that AI infrastructure construction will not stop, but site selection and the pace of advancement will be readjusted.

2026-11-06 18:41 GMT+8

On November 5, 2026, the NASDAQ Composite Index recorded its worst single-day performance since April 2025. Denny Fish, a portfolio manager and research analyst at Janus Henderson, pointed out that the true catalyst for this sell-off was not the return of a divided government in Washington, D.C., but changes in the state-level political landscape—in states such as Texas, Georgia, Michigan, and Ohio, newly elected Democratic governors had all publicly opposed data center construction.

This distinction is crucial because the regulatory, permitting, and zoning frameworks for U.S. data center projects primarily operate at the state and local government levels. Fish believes the election results have cast a thicker veil of policy uncertainty over the entire AI infrastructure value chain. However, he emphasized that AI construction will not cease; rather, questions have been raised regarding which regions will host these projects and at what pace they will proceed.

Fish's core judgment is that stricter regulation will reshape the competitive landscape within the AI ecosystem. Large technology companies typically possess sufficient financial strength and personnel to handle more stringent compliance requirements. Hyperscale cloud service providers, backed by robust cash flows from software sales, cloud services, and digital advertising, will hold an advantage over new cloud providers focused solely on AI computing power. He believes that single-product companies that achieved profit increases only through pricing power during periods of supply-demand imbalance, smaller new entrants, and companies with lower energy efficiency (tokens produced per watt) also face unfavorable situations.

Sign in / Sign up
ReadingResearchRadarInvestment framework

Over a longer time horizon, Fish expects gigawatts of AI computing capacity scheduled to come online in 2027 to largely proceed according to plan. However, projects encountering new regulatory hurdles may shift to more friendly jurisdictions or even migrate to emerging market countries that previously had little exposure to AI construction but are willing to accept it for economic and geopolitical reasons. The Trump administration still views AI dominance as one of the highest national priorities, and hyperscale cloud service providers view the continued advancement of AI as fundamental to their businesses, providing underlying support for the entire construction cycle.

Fish concluded that AI construction will unfold across multiple election cycles, with inevitable ups and downs in policy, technology, and financing paths. However, he believes that as the transformative potential of AI in productivity and knowledge advancement becomes increasingly evident, this process will not reverse. While shifts in the regulatory regime will affect winners and losers, the willingness of companies to drive innovation and transform their business models to integrate AI will determine outcomes to a much greater extent. In his view, this election itself has proven that the landscape and ultimate winners can change rapidly.

Sources

  • janushenderson.comUS Advisor↗
Continue researching →