
"While Nvidia ( NASDAQ: NVDA | NVDA Price Prediction) stock surged 34.9% over the past year and the S&P 500 gained 12.2%, the infrastructure REITs powering AI's physical backbone have been left behind. These five companies house, cool, and connect the servers running ChatGPT and Gemini, yet their stock performance has lagged far behind Nvidia and the broader market. They offer dividend yields ranging from 3% to 5% alongside direct exposure to AI infrastructure demand."
"Crown Castle ( NYSE: CCI) operates 40,000 cell towers connecting data centers to end users, but it has struggled. Revenue declined 4.2% year-over-year in Q4 2025 after divesting its fiber business, and management slashed the dividend 32% from $1.565 to $1.0625 per share quarterly. Adjusted EBITDA fell to $718 million. Crown Castle remains the only pure-play U.S. tower REIT, with a 4.9% dividend yield."
"American Tower ( NYSE: AMT) owns 148,000+ communications sites plus a U.S. data center network. Q3 2025 revenue of $2.72 billion beat estimates, growing 7.7% year-over-year on strong hybrid-cloud and AI-related demand. Adjusted EBITDA reached $1.82 billion, up 7.6%. The company raised its quarterly dividend 4.9% to $1.70 per share and lifted full-year guidance. JPMorgan maintains an Overweight rating with a $245 price target, citing its global footprint as essential for AI inference at the network edge."
Nvidia stock surged 34.9% over the past year while the S&P 500 gained 12.2%. Infrastructure REITs that house, cool, and connect servers running ChatGPT and Gemini have lagged those gains. These REITs offer dividend yields ranging from 3% to 5% and provide direct exposure to AI infrastructure demand. Crown Castle operates 40,000 cell towers, saw revenue decline 4.2% year-over-year in Q4 2025, cut its quarterly dividend by 32% to $1.0625, and reported adjusted EBITDA of $718 million. American Tower owns 148,000+ sites, posted Q3 2025 revenue of $2.72 billion, and raised its quarterly dividend to $1.70.
Read at 24/7 Wall St.
Unable to calculate read time
Collection
[
|
...
]