Thoughts on the Market

Asia Power Demand Surges Beyond AI to Include Cooling and Industry

5 min Listen to the original
Power GenerationAsia InfrastructureData CentersAir ConditioningElectricity DemandGrid InvestmentSemiconductorsEnergy Markets

Summary

Mayank Maheshwari argues that global electricity demand is underestimated, projecting a 36 trillion unit total by 2030, driven by AI, industrial growth, and air conditioning. He notes that Asia's demand will grow 20% faster than the US, with non-China Asia reaching 23 trillion units. AI accounts for only a quarter of incremental demand, while cooling and industrial projects are major drivers. US grid constraints may shift data center and semiconductor investments to Asia, particularly Thailand, Singapore, India, and Japan. Power prices rose 20% in 2026, with margins expected to widen by 2028, favoring generation and grid infrastructure owners.

Key takeaways

  1. 1
    Global power prices rose approximately 20% year-on-year in 2026, and the spread between electricity prices and generation costs is expected to widen by around 15% by 2028, suggesting improving margins for generators.

    Mayank Maheshwari · 3:27

Episode sections

  1. 0:00 – 0:30

    Introduction and Broad Electrification Thesis

    Mayank Maheshwari introduces the theme that power demand is driven by more than just AI, citing a broader electrification super cycle. He argues that market consensus underestimates future power needs globally and in Asia.

  2. 0:30 – 0:48

    Global and Regional Demand Forecasts

    The speaker forecasts global electricity use reaching 36 trillion units by 2030, 15% above current estimates. Asia's demand is expected to grow 20% faster than the US, with non-China Asia reaching 23 trillion units.

  3. 0:48 – 1:23

    Drivers of Incremental Demand: AI, Industry, and Cooling

    While AI-related demand is rising, it accounts for only a quarter of incremental power needs. New industrial projects and air conditioning are significant additional drivers, with AI agents requiring more computing power than conventional chatbots.

  4. 1:23 – 1:46

    AI Agents and Computing Power

    AI agents require significantly more computing power than conventional chatbots because they search for information, use software, and check results, potentially driving a 1.6-fold increase in global power consumption by 2030.

  5. 1:46 – 2:13

    Data Center and Semiconductor Investment Shifts to Asia

    Investments in data centers and semiconductor factories may shift to Asia due to US grid constraints. Asia could absorb 15 GW of additional demand linked to the US, with Thailand seeing $25 billion in commitments and Singapore, India, and Japan showing strong demand.

  6. 2:13 – 2:23

    Thailand and Regional Winners

    Thailand has already secured $25 billion in data center investment commitments, while Singapore, India, and Japan are experiencing stronger demand from US hyperscalers, indicating specific national winners in the infrastructure build-out.

  7. 2:23 – 2:47

    Air Conditioning as a Major Power Driver

    Cooling demand is identified as a major source of electricity growth, potentially requiring as much power as data centers by 2030. In India, cooling accounts for up to 40% of peak summer power use.

  8. 2:47 – 3:27

    Power Generation Mix and Grid Infrastructure Needs

    About 60% of Asia's data center capacity currently runs on coal and natural gas. Gas generation outside China is expected to increase by 19 GW by 2030, supporting 20 million tons of annual LNG imports. Grid investments could reach $1 trillion through 2030.

  9. 3:27 – 4:00

    Power Price Trends and Investment Implications

    Global power prices rose 20% year-on-year in 2026, with the gap between prices and generation costs expected to widen by 15% by 2028. The speaker suggests investors may be underestimating demand, creating value for reliable power generation and grid infrastructure owners over equipment suppliers.

  10. 4:00 – 4:33

    Top Power Markets and Conclusion

    The most attractive power markets are identified as India, Malaysia, Australia, and the US. The segment concludes with a standard disclaimer and call to action for podcast reviews.

Claims worth verifying

Global electricity use could reach nearly 36 trillion units by 2030, 15% above current market estimates.

Unidentified speaker (UNKNOWN) ·0:30

How to verify: Check Morgan Stanley research reports from 2026-08-12 to 2026-11-12 titled 'Global Electricity Demand Forecast' or similar; compare with IEA World Energy Outlook 2026 or BloombergNEF forecasts. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Outside China, Asia's electricity demand could reach 23 trillion units by 2030, with Asia's demand growing 20% faster than in the US.

Unidentified speaker (UNKNOWN) ·0:30

How to verify: Verify against IEA Asia Pacific Energy Outlook 2026 or national grid operator forecasts for major Asian economies (India, Japan, etc.) and US EIA data. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

AI-related demand is only about a quarter of incremental power demand.

Mayank Maheshwari ·0:48

How to verify: Review Morgan Stanley's detailed breakdown of power demand drivers in their Asia Energy research; cross-reference with data center industry reports (e.g., from Uptime Institute or Equinix) on AI vs. non-AI load growth. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

AI agents could drive a 1.6-fold increase in global power consumption by 2030.

Unidentified speaker (UNKNOWN) ·1:23

How to verify: Search for Morgan Stanley research citing '1.6-fold increase' in power consumption due to AI agents; compare with projections from other consultancies like McKinsey or BCG on AI energy intensity. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Asia could absorb around 15 gigawatts of additional demand from projects linked to the US.

Unidentified speaker (UNKNOWN) ·1:52

How to verify: Check Morgan Stanley's 'Asia Data Center Investment' report for 2026; verify against public announcements of data center projects in Asia linked to US hyperscalers. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Asia may need 84 GW of data center capacity by 2030, roughly four times that of 2025 levels.

Unidentified speaker (UNKNOWN) ·1:52

How to verify: Verify 2026 data center capacity figures from industry trackers like Synergy Research Group or Uptime Institute; compare with 2026 projections in Morgan Stanley's Asia Energy notes. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Thailand has already seen $25 billion in data center investment commitments.

Mayank Maheshwari ·2:13

How to verify: Check Thai government announcements or major developer press releases (e.g., from JLL, CBRE, or local developers) for data center CAPEX commitments in Thailand between 2026 and 2026. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Air conditioning could require as much electricity as data centers by 2030.

Mayank Maheshwari ·2:23

How to verify: Compare IEA or Morgan Stanley forecasts for cooling vs. data center electricity demand in 2030; verify against historical growth rates of HVAC adoption in Asia. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

In India, cooling can account for up to 40% of peak power use during the summer.

Unidentified speaker (UNKNOWN) ·2:35

How to verify: Check India's Central Electricity Authority (CEA) or state discom reports for summer peak load composition; verify against Morgan Stanley's India energy research. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

About 60% of Asia's data center capacity currently runs on coal and natural gas.

Mayank Maheshwari ·2:47

How to verify: Verify against data center operator disclosures (e.g., Equinix, Digital Realty) or regulatory filings in major Asian markets regarding energy mix; cross-reference with IEA data on data center energy sources. For recent events, verify 2026-08-12 through 2026-11-12. Use another period only when the speaker explicitly names it.

Primary source

Do your own research

  • Verify global electricity demand forecasts against IEA World Energy Outlook or EIA International Energy Outlook for 2030 projections.
  • Check US grid interconnection queue data from NREL or FERC to confirm capacity constraints and delays.
  • Review Thailand's BOI investment promotion data for the $25 billion data center commitment figure.
  • Analyze India's Central Electricity Authority reports for peak summer cooling load percentages.
  • Compare global power price trends against BloombergNEF or S&P Global Platts data for 2026 year-on-year changes.
  • Assess LNG import volumes for Asia outside China against IEA or Gas Exporting Countries Forum data for the 20 million tons projection.
  • Examine grid investment capital expenditure plans of major utilities in India, Malaysia, and Australia against their annual reports.
  • Verify the 1.6-fold increase in power consumption for AI agents by reviewing semiconductor industry power efficiency reports from NVIDIA or AMD.
Full transcriptexpand
0:00
Unidentified speaker (UNKNOWN): Welcome
0:00
Mayank Maheshwari: to Thoughts on the Market. I am Mayang Mahishuri Morgan Stanley's Asia Energy Analyst. Today, why Power Thematic is moving beyond AI. It's Thursday, November 12th at 10am in Singapore.
0:15
Mayank Maheshwari: Whenever I have a conversation on power, it is mostly thought as a story about AI, but the electrification super cycle is much broader than just AI. Consensus is underestimating how. much power the world will need by the end of the decade.
0:30
Unidentified speaker (UNKNOWN): We
0:30
Mayank Maheshwari: think global electricity use could reach nearly 36 trillion units by 2030, 15% above our current market estimates. Outside China, that could reach nearly 23 trillion units, with Asia's demand growing 20% faster than in the US.
0:47
Unidentified speaker (UNKNOWN): For
0:48
Mayank Maheshwari: years, the focus was on the electricity needed by. data centers and semiconductor factories. That need is still rising, but AI-related demand is only about a quarter of incremental power demand.
1:00
Unidentified speaker (UNKNOWN): New
1:01
Mayank Maheshwari: industrial projects and air conditioning are now adding to the pressure. Not surprisingly, AI adoption is a big reason that data center demand keeps making the problem especially acute.
1:12
Unidentified speaker (UNKNOWN): A
1:13
Mayank Maheshwari: conventional chatbot may answer a question. with one request.
1:16
Unidentified speaker (UNKNOWN): An
1:16
Mayank Maheshwari: agent can work through a task by searching for information using software and checking its results.
1:23
Unidentified speaker (UNKNOWN): That
1:23
Mayank Maheshwari: process requires more computing power, even as the cost of each unit of computing falls. AI agents could drive a 1.6-fold increase in global power consumption by 2030.
1:35
Unidentified speaker (UNKNOWN): That
1:35
Mayank Maheshwari: raises a practical question.
1:37
Unidentified speaker (UNKNOWN): Where will
1:38
Mayank Maheshwari: the new data centers and semiconductor factories be built? And how will they get the electricity they
1:44
Unidentified speaker (UNKNOWN): need?
1:46
Unidentified speaker (UNKNOWN): Some
1:46
Mayank Maheshwari: of that investments may shift to Asia because the US cannot add power capacity quickly enough.
1:52
Unidentified speaker (UNKNOWN): The
1:53
Mayank Maheshwari: shift could be sizable. We estimate that Asia could absorb around 15 gigawatts of additional demand from projects linked to the US. Asia may need 84 GW of data center capacity by 2030, roughly four times that of 2025 levels. Thailand has already seen $25 billion in data
2:13
Mayank Maheshwari: center investment commitments, while Singapore, India, and Japan are seeing stronger demand coming through for power across various US hyperscalers. Scalars.
2:22
Unidentified speaker (UNKNOWN): But
2:23
Mayank Maheshwari: AI is just the beginning of a global electricity super cycle. Cooling is another major source of demand. Air conditioning could require as much electricity as data centers by 2030.
2:35
Unidentified speaker (UNKNOWN): In
2:35
Mayank Maheshwari: India, cooling can account for up to 40% of peak power use during the summer. Taken together, that raises the bigger question:
2:44
Unidentified speaker (UNKNOWN): where
2:44
Mayank Maheshwari: will all this electricity comes come
2:46
Unidentified speaker (UNKNOWN): from?
2:47
Mayank Maheshwari: About 60% of Asia's data center capacity currently runs on coal and natural gas.
2:53
Unidentified speaker (UNKNOWN): We
2:53
Mayank Maheshwari: expect gas generation outside China to increase by around 19 gigawatts between 2025 and 2030, supporting roughly 20 million tons a year of LNG imports into Asia. Renewables and batteries will be part of the answer,
3:10
Unidentified speaker (UNKNOWN): but
3:11
Mayank Maheshwari: the grid will need to expand as
3:13
Unidentified speaker (UNKNOWN): well. Asia
3:14
Mayank Maheshwari: could require about a trillion dollars of new grid investments through 2030. In some markets, data centers are being encouraged to build their own power because a central grid cannot connect them quickly enough.
3:27
Unidentified speaker (UNKNOWN): The
3:27
Mayank Maheshwari: strain is already showing up in power prices. Global power prices rose about 20% year on year in 2026, and we expect the gap between electricity prices and generation cost to widen by around. 15% by 2028. Investors may be underestimating how much electricity Asia will need and how broadly that demand
3:47
Mayank Maheshwari: will spread beyond data centers. That could create more value for the owners of reliable power generation companies
3:54
Unidentified speaker (UNKNOWN): as
3:55
Mayank Maheshwari: well as grid infrastructure than for companies that simply supply equipment.
4:00
Unidentified speaker (UNKNOWN): We
4:00
Mayank Maheshwari: believe the most attractive power markets are. India, Malaysia, Australia, and the US. Thanks for listening. If you enjoy thoughts on the market, please leave us a review wherever you listen and share the podcast with a friend or a colleague today.
4:16
Unidentified speaker (UNKNOWN): The
4:16
Unidentified speaker (speaker_1): preceding content is informational only and based on information available when created. It is not an offer or solicitation, nor is it tax or legal advice. It does not consider your financial circumstances. and objectives and may not be suitable for
4:29
Unidentified speaker (UNKNOWN): you.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.

More from Thoughts on the Market

Thoughts on the Market

2027 Global Outlook: The Case for a Stronger Dollar, Longer AI Cycle

Episode discussion: Morgan Stanley's 2027 Outlook: Stronger Dollar, AI Cycle, and Global Equity Overweight

Speakers’ claims: Seth Carpenter and Serena Tang discuss Morgan Stanley's 2027 market outlook, with Tang identifying 'embracing risk' as the primary theme due to an intact AI CapEx cycle and robust growth. The firm has moved to a maximum overweight in global equities, particularly the US, Japan, and Europe, while maintaining a neutral stance on investment grade bonds. Tang highlights out-of-consensus views including a stronger US dollar driven by superior US growth and a constructive stance on UK gilts. Risks to this outlook include an abrupt end to the AI investment cycle, rising oil prices, or unrelated inflation reacceleration. Tang argues that US midterm election outcomes are unlikely to significantly impact data center construction, as projects are largely state-level and already permitted.

Global EquitiesAI CapExUS Dollar
7 min2 assets0 claims to verify
Thoughts on the Market

2027 Global Outlook: Powered by AI, Tested by Oil

Episode discussion: Morgan Stanley forecasts AI-driven US growth and two Fed hikes in 2027 amid oil risks

Speakers’ claims: Seth Carpenter and Serena Tang outline Morgan Stanley's 2027 global outlook, attributing resilient growth to AI investment cycles despite energy and geopolitical shocks. Carpenter forecasts US real GDP growth of 2.5% driven by AI capital expenditure, with Europe at 1.3% and China at 4.3%. He predicts two Federal Reserve rate hikes in December 2026 and March 2027, anticipating an inflation peak in Q1 2027. The analysts highlight that US growth expectations exceed consensus due to anticipated AI productivity gains, though they warn of downside risks if AI CapEx falls short.

Global GrowthFederal ReserveArtificial Intelligence
13 min1 assets10 claims to verify
All insights