Morgan Stanley Declares 'Transformer Supercycle' as AI-Driven Power Demand Surges, Straining US Supply

Morgan Stanley has identified a 'transformer supercycle' running through 2030, driven by the massive electricity demand from AI data centers.

This demand surge is creating a severe bottleneck, as the U.S. relies heavily on imported transformers and new domestic factories won't be operational until 2027.

The supply-demand imbalance creates a seller's market, significantly benefiting transformer manufacturers like GE Vernova and HD Hyundai Electric.

Morgan Stanley recently argued that the U.S. has entered a 'transformer supercycle' that is expected to last until 2030.

The primary driver behind this trend is the explosive growth in electricity demand, fueled by the rapid expansion of AI and data centers. We're seeing this play out in real-time. For instance, Meta recently committed to funding seven new natural gas power plants just to power a single data center campus in Louisiana. This move shows that major tech companies, or 'hyperscalers,' can't wait for the grid to catch up; they are securing their own power sources now, which in turn requires a massive amount of new electrical hardware, including large power transformers (LPTs).

This is where the problem lies—a significant supply and demand mismatch. The demand for transformers is skyrocketing, but the supply side is severely constrained. First, the U.S. is heavily dependent on imports, with over 85% of its large transformers coming from overseas. Second, building new domestic factories takes time. Major manufacturers like Hitachi Energy, Siemens Energy, and HD Hyundai Electric have announced new U.S. plants, but they aren't expected to contribute meaningful supply until 2027. This creates a multi-year bottleneck.

This structural imbalance creates a classic 'supplier's market.' With demand far outstripping supply, manufacturers hold significant pricing power. Companies like GE Vernova and HD Hyundai Electric are already seeing the benefits. GE Vernova reported a massive $150 billion backlog, and HD Hyundai Electric secured record orders from the U.S. Their stock prices and valuation multiples reflect this optimism, suggesting investors believe this profitable cycle will continue for years.

In essence, the supercycle thesis is built on a solid foundation. The surge in AI-driven power demand is no longer a forecast but a reality confirmed by corporate actions and official projections from bodies like the North American Electric Reliability Corporation (NERC). Combined with a structurally tight supply chain that won't see relief for several years, the stage is set for a prolonged period of high demand and strong profitability for key players in the transformer industry.

Large Power Transformer (LPT): A critical component of the electrical grid used to step up or step down voltage for long-distance transmission and distribution. They are large, expensive, and have very long manufacturing lead times.
Hyperscalers: Large cloud service providers and tech companies (like Amazon, Google, Meta) that operate massive data centers requiring enormous amounts of power.
NERC (North American Electric Reliability Corporation): A non-profit organization that develops and enforces reliability standards for the bulk power system of North America.

Morgan Stanley Declares 'Transformer Supercycle' as AI-Driven Power Demand Surges, Straining US Supply

Morgan Stanley has identified a 'transformer supercycle' running through 2030, driven by the massive electricity demand from AI data centers.

This demand surge is creating a severe bottleneck, as the U.S. relies heavily on imported transformers and new domestic factories won't be operational until 2027.

The supply-demand imbalance creates a seller's market, significantly benefiting transformer manufacturers like GE Vernova and HD Hyundai Electric.

Morgan Stanley recently argued that the U.S. has entered a 'transformer supercycle' that is expected to last until 2030.

Large Power Transformer (LPT): A critical component of the electrical grid used to step up or step down voltage for long-distance transmission and distribution. They are large, expensive, and have very long manufacturing lead times.
Hyperscalers: Large cloud service providers and tech companies (like Amazon, Google, Meta) that operate massive data centers requiring enormous amounts of power.
NERC (North American Electric Reliability Corporation): A non-profit organization that develops and enforces reliability standards for the bulk power system of North America.