The dramatic shift toward electrification has moved beyond a simple AI narrative, and a wealth-creation bonanza worth as much as $500 billion is accelerating across global power markets, while Wall Street pricing models generally lag behind reality. Companies with reliable baseload power and control of core grid nodes are entering their strongest cash-flow supercycle since industrialization in the last century.
According to Morgan Stanley's global utilities team in a latest research report on November 2, the signal is clear: global power tightness will significantly exceed expectations. The next major leg of the electrification supercycle has fully upgraded from simply "powering AI" to "powering everything."
Analysts including Mayank Maheshwari and Stephen C Byrd noted in the latest report that the drivers of the electricity supercycle have expanded well beyond AI data centers, with industrial electricity use and household cooling joining AI to form a "troika" of demand growth, a trend that the stock market has not yet fully priced in.
Morgan Stanley raised its global electricity consumption forecast again by 10%, projecting that global power use will reach about 36 trillion kWh by 2030, with a compound annual growth rate of about 4.2%. The report highlighted four major "surprises":
Demand sources expanding from AI to "everything"; fossil fuels regaining share; a structural shift in the electricity pricing landscape; and pricing power shifting from equipment suppliers to power infrastructure operators.
Morgan Stanley estimates these trends will create about $500 billion in value globally and drive upward revisions of about 14% to 2028 earnings per share forecasts for related companies. At present, Asian power "AI beneficiary stocks" have generally seen valuation compression this year, but actual profit forecasts are being raised, and the divergence between the two creates a clear investment opportunity.
In terms of regional distribution, Asia will absorb about 64% of global incremental electricity demand from 2025 to 2030. The United States, Southeast Asia and India are leading in both demand growth and power market tightness, making them the core investment regions highlighted in the report.
Demand explodes: the "3A" engine drives power consumption far beyond expectations
Electricity demand growth has far exceeded the single category of data centers. This is the first major surprise in the report. According to the report, global power consumption expectations have been raised by 10%, and by 2030 they will surge to about 36 trillion kWh at a 4.2% compound annual growth rate.
Morgan Stanley summarizes the core drivers of this round of power demand expansion as the "three As" - Agentic AI, Allshoring and Aircons. These three factors are extending the foundation of power demand growth from data centers to broader industrial and civilian scenarios.
On the AI dimension, the report notes that Asia's AI adoption curve is one of the fastest in the world. Although the computing cost per token continues to fall, agentic workloads have sharply increased token consumption per task, keeping total computing demand climbing, and electricity has thus become a key bottleneck constraining AI expansion.
Data center electricity use in the Asia-Pacific region is expected to grow at a compound annual rate of 27%, reaching about 133 billion kWh by 2030, requiring 84 GW of data center power supply capacity, about four times the 2025 level and roughly in line with the U.S. growth rate of 26%.
On the industrial relocation dimension, the report lists it as the biggest driver of accelerating industrial electricity growth in Asia. Morgan Stanley expects Asia's industrial electricity consumption to grow at a compound annual rate of about 4%, accounting for 43% of all incremental power demand, about 2.7 times the growth rate of data center electricity use. Data show that since the start of 2025, year-on-year industrial electricity growth has accelerated notably in several markets including India and Southeast Asia.
On the cooling demand dimension, the report characterizes it as an "underestimated source of demand." By 2030, the incremental scale of space cooling electricity use will approach the increase from data centers and contribute an additional about 40 basis points to Asia's compound annual power demand growth, while cooling's share of total global electricity use will rise from about 10% now to 12%.
Global power forecasts raised sharply, with Asia at the core of demand growth
Morgan Stanley raised its 2026-2030 power demand growth forecasts for almost all major regions. The three markets with the largest upward revisions were India (+454 basis points), the United States (+169 basis points) and Australia (+141 basis points); Europe and Japan also showed signs of modest growth again after years of stagnant or shrinking electricity use.
In absolute terms, Asia is expected to consume about 64% of global incremental electricity demand from 2025 to 2030. Morgan Stanley estimates Asia's power supply needs to grow at a compound annual rate of about 5%, about 140 basis points faster than the current average.
Fossil fuels regain share, "behind-the-meter" demand multiplies
The second "surprise" in the report is that fossil fuels' role in the power mix is rising rather than falling. Morgan Stanley expects fossil fuels to still account for about half of the global power mix by 2030, while 60% of Asia's data center power supply comes from coal and natural gas, a clear divergence from some investors' expectations of accelerated clean energy substitution.
The report notes that as power markets tighten, reliable fossil fuel generation is gaining a scarcity premium, and natural gas and coal are increasingly being paired with battery storage and renewables. For price-sensitive Asian markets, coal combined with storage can provide a highly competitive all-day generation cost of $50 to $60 per MWh.
At the same time, the power shortfall facing U.S. data center developers is spilling over into Asia. Morgan Stanley estimates that beyond existing solutions, U.S. data center developers will face a 43 GW power supply gap by 2028, widening to 80 GW by 2029, driving a large amount of data center capacity to migrate overseas.
About 15 GW of that gap may spill over into Asia, mainly Southeast Asia (about 8 GW) and India. This trend is already emerging - U.S. and Chinese hyperscalers and emerging cloud providers have locked in about 10.6 GW of data center capacity through contracts in Asia, with more than 70% concentrated in India and Southeast Asia.
"Behind-the-meter" demand is also expanding rapidly. Against the backdrop of continued strain on power grids, Morgan Stanley expects U.S. behind-the-meter demand to rise to about 47 GW by 2030, while Asia's will be about 8 GW, creating significant business opportunities for on-site power suppliers.
Power prices return to an upward path, pricing power shifts to generators
The third and fourth "surprises" identified in the report both point to a structural reshaping of the electricity pricing system. Against the backdrop of persistently tightening supply and demand, wholesale power prices in major markets rose about 20% year on year in 2026; Morgan Stanley expects commercial electricity price spreads to widen further by about 15% by 2028.
From a market perspective, after a brief pullback following the peak of the 2022 energy crisis, wholesale power prices have strengthened again in several markets including India, China, Malaysia, Singapore, Japan and Europe, supporting an expansion in generators' profitability.
On the change in pricing power, the report notes that power infrastructure operators are reclaiming pricing leadership from equipment manufacturers. Over the past 18 months, equipment manufacturers' prices climbed rapidly, and that trend is now beginning to pass through to generators, whose selling prices have been continuously rising.
The report also notes that hyperscalers are relatively less sensitive to power price fluctuations - estimates show that a 20% change in electricity prices affects AI hyperscaler profits by only about 1% - giving power companies room to seek higher electricity pricing.
At the same time, the "duck curve" of renewable energy is gradually evolving into a "double-humped camel curve":
Morning and nighttime power demand rises sharply around the daylight period, expanding profit opportunities for flexible generation assets and storage, while returns for pure renewable generators come under pressure.
Comments