Key facts
- Morgan Stanley analysts believe demand for AI compute will significantly exceed supply for years.
- Constraints such as limited power supply, labor shortages for data center construction, and regulatory barriers are slowing AI infrastructure buildout.
- Morgan Stanley expects a 'barbell' outperformance over the next 6-12 months between AI adopters and infrastructure providers.
- The bank's analysts rate these 12 stocks as 'Overweight,' equivalent to a 'Buy' rating.
Morgan Stanley analysts are recommending a basket of 12 US stocks that they believe are critical to addressing the infrastructure bottlenecks in the artificial intelligence boom. The bank's analysts, led by Stephen Byrd, stated in a September 24 note that demand for computing power is expected to significantly outstrip supply for the foreseeable future. They pointed to several factors contributing to this bottleneck, including limitations in power supply, a scarcity of skilled labor for data center construction, and regulatory hurdles. Unlike previous technology cycles where infrastructure trades typically underperform as adoption grows, Morgan Stanley anticipates continued strength in this sector due to persistent demand and buildout constraints. The bank forecasts a "barbell" market leadership scenario over the next six to twelve months, with both AI adopters and the underlying infrastructure providers experiencing outperformance. The identified stocks are rated "Overweight" by Morgan Stanley's analysts and are projected to have at least 40% upside potential to their price targets. Among the highlighted companies are NextEra Energy (NEE), GE Vernova (GEV), Cummins (CMI), Williams Companies (WMB), and Liberty Energy (LBRT), with specific upside potentials and price targets provided for each.