Thoughts on the Market

Morgan Stanley analysts project an 80GW AI power deficit by 2029, driving demand for on-site generation and battery storage.

10 min Listen to the original
AI InfrastructureEnergy StorageUtility GridsElectric VehiclesRenewable EnergyIndustrial Demand

Summary

Morgan Stanley analysts Andrew Prococo and Dave Raccaro discuss the growing AI power bottleneck, projecting an 80 gigawatt deficit by 2029. They argue that the primary constraint is not power generation but delays in utility transmission and distribution networks. Consequently, hyperscalers are shifting from temporary on-site power solutions to long-term 15-20 year contracts for distributed energy. The episode highlights the rapid growth of battery storage demand from data centers, which is expected to outpace grid demand by 2030. Additionally, automakers like Tesla and Ford are leveraging EV manufacturing capacity for energy storage, while Vehicle-to-Grid technology is identified as a potential future supply source.

Episode sections

  1. 0:00 – 1:07

    Introduction and AI Power Shortfall Estimates

    Hosts Andrew Prococo and Dave Raccaro from Morgan Stanley introduce the topic of the AI power bottleneck. Raccaro notes that estimates for the power shortfall have grown, now projecting an 80 gigawatt deficit by 2029.

  2. 1:07 – 2:35

    Shift to On-Site Distributed Power

    Raccaro explains that data centers are increasingly adopting on-site power solutions to bypass slow utility grid connections. Hyperscalers are moving from viewing these as short-term bridges to signing 15-20 year contracts for long-term reliability.

  3. 2:35 – 3:59

    Transmission and Distribution Bottlenecks

    The discussion identifies that the primary bottleneck is not just power generation, but the utility transmission and distribution networks. Utilities are struggling to keep up with the unprecedented demand from data centers, leading to significant delays in grid expansion.

  4. 3:59 – 4:53

    Criteria for On-Site Power Technologies

    Developers are prioritizing speed, but increasingly also considering community support, air emissions, water use, and noise. Technologies like fuel cells, engines, and turbines are being pursued to meet these combined criteria.

  5. 4:53 – 6:31

    Industrial Demand and Energy Storage Growth

    Raccaro estimates 10 gigawatts of incremental annual power demand from U.S. reshoring and reindustrialization. He highlights that data centers are ramping loads instantaneously, making battery storage a critical solution, with data center demand for batteries expected to exceed grid demand by 2030.

  6. 6:31 – 8:15

    Auto Industry Pivot to Energy Storage

    Prococo notes that automakers like Tesla and Ford are leveraging underutilized EV manufacturing capacity to enter the energy storage market. Suppliers are also repurposing automotive-grade technology for turbine generators and microgrid inverters.

  7. 8:15 – 9:57

    Vehicle-to-Grid (V2G) Potential

    Raccaro discusses Vehicle-to-Grid technology, noting that California currently dispatches 500-1,000 megawatts during peaks. He projects this could become a significant power supply source by 2030+ as EV penetration increases across the U.S.

Assets discussed

AssetKindStanceContext
TeslaTSLAequityneutralObviously, Tesla, a very obvious player in the market that has a very large energy storage business in the U.S.

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

The AI power shortfall is estimated to be upwards of 80 gigawatts by 2029.

Dave Raccaro ·1:12

How to verify: Check Morgan Stanley research reports or press releases from late 2025/early 2026 regarding AI infrastructure power demand forecasts.

Data center developers are signing 15, 18, or 20-year contracts for on-site power solutions.

Dave Raccaro ·2:00

How to verify: Review public announcements or contract disclosures from major hyperscalers (e.g., Microsoft, Amazon, Google) regarding long-term power purchase agreements (PPAs) for distributed generation.

U.S. teams estimated 10 gigawatts of incremental annual power demand from U.S. reshoring and reindustrialization.

Dave Raccaro ·5:02

How to verify: Search for Morgan Stanley collaborative reports on industrial reshoring impact on energy demand.

Data center demand for battery storage in 2030 will exceed grid demand for battery storage in terms of incremental new projects.

Dave Raccaro ·6:06

How to verify: Compare market forecasts from energy research firms (e.g., BloombergNEF, Wood Mackenzie) for data center vs. grid-scale battery installations in 2030.

Tesla has a very large energy storage business in the U.S.

Andrew Prococo ·6:53

How to verify: Review Tesla's quarterly earnings reports (10-Q/10-K) specifically the Energy Generation and Storage segment revenue.

Ford is building an energy storage business in the U.S. and has started to build a rather large backlog.

Andrew Prococo ·6:59

How to verify: Review Ford Motor Company press releases or investor presentations regarding their energy storage initiatives and order backlogs.

California dispatches between 500 to 1,000 megawatts of power to the grid during peak periods via vehicle-to-grid or home-to-grid systems.

Dave Raccaro ·8:33

How to verify: Check California Energy Commission reports or utility (e.g., PG&E, SCE) data on demand response and V2G program capacity.

Do your own research

  • Verify Morgan Stanley's 80GW power deficit projection by reviewing their latest energy or technology sector research reports.
  • Check EIA (U.S. Energy Information Administration) data on current utility transmission and distribution interconnection queues and wait times.
  • Review Tesla (TSLA) and Ford (F) quarterly earnings calls or 10-K filings for specific revenue and capacity figures related to their energy storage businesses.
  • Analyze CAISO (California Independent System Operator) reports on current Vehicle-to-Grid (V2G) dispatch volumes and pilot program results.
  • Research the specific 15-20 year contract terms and pricing structures of recent on-site power agreements signed by major hyperscalers.
  • Assess the regulatory landscape for V2G technology expansion beyond California, including state-level incentives and grid integration standards.
Full transcriptexpand
0:00
Unidentified speaker: Welcome to Thoughts on the Market.
0:01
Unidentified speaker: I'm Andrew Prococo, Morgan Stanley's U.S.
0:04
Unidentified speaker: Autos and Shared Mobility Analyst.
0:06
Unidentified speaker: And I'm Dave Raccaro, Morgan Stanley's U.S.
0:08
Unidentified speaker: Power, Utilities, and Cleantech Analyst.
0:11
Unidentified speaker: Today, why the AI Power Crunch Could Transform How We Generate, Store, and Consticity.
0:18
Unidentified speaker: It's Tuesday, November 3rd at 10 a.m.
0:20
Unidentified speaker: in New York.
0:23
Unidentified speaker: So, Dave, the whole AI power shortfall and how data centers is creating a power crunch is not necessarily a new topic.
0:30
Unidentified speaker: I think we've we've had many conversations.
0:33
Unidentified speaker: Many people listening have probably heard this is kind of being a bottleneck for AI deployment.
0:38
Unidentified speaker: But I'm curious, like, how have your views around the role of distributed power changed in terms of its role in solving some of the shortfalls in the market?
0:50
Unidentified speaker: I think there's a lot of debate in the industry around whether or not these solutions are quote unquote bridge or whether they're structural and they can be truly long-term sources of supply to the market.
1:01
Unidentified speaker: So maybe just help us think about how our views have really changed as it relates to distributed power specifically.
1:07
Unidentified speaker: One of the things that we've seen is that the shortfall has grown in our estimates over the last several iterations.
1:12
Unidentified speaker: We're now looking at upwards of an 80 gigawatt shortfall by 2029.
1:17
Unidentified speaker: And how do we solve that?
1:18
Unidentified speaker: I'd say we've leaned more into on-site power, bringing your own power.
1:23
Unidentified speaker: What is that?
1:24
Unidentified speaker: You know, if you are still waiting for years for a grid connection from the utility, what can you do as an alternative?
1:31
Unidentified speaker: You can bring your own power.
1:32
Unidentified speaker: You could bring small-scale power equipment to get the high reliability in a much faster time frame than you would waiting for the grid in comparison.
1:40
Unidentified speaker: And so we're seeing that increasing extremely rapidly in popularity among hyperscalers and data center developers.
1:48
Unidentified speaker: And we're seeing them much more being willing to consider it on a long-term basis.
1:52
Unidentified speaker: What used to be, hey, I need to cover a few years just a bridge until I have line of sight to getting a connection to my utility.
2:00
Unidentified speaker: Now we're seeing actually 15, 18, 20 year contracts for these, you know, what used to be considered just smaller scale, less attractive solutions.
2:10
Unidentified speaker: They're now long-term, what looks like much more permanent considerations for highly reliable data center AI power.
2:18
Unidentified speaker: I think there's a preconceived notion in the market that the grid's going to catch up and that we won't need these solutions longer term, or that we're building too much of it and there's too much manufacturing capacity coming to the market.
2:33
Unidentified speaker: So maybe we can segment those into two separate questions.
2:35
Unidentified speaker: Like one, what is actually the bottleneck?
2:38
Unidentified speaker: I think everyone thinks of it as a power generation bottleneck, but I think you've done some interesting work around how there's also transmission and distribution bottlenecks that don't get solved as quickly as the power generation.
2:47
Unidentified speaker: So maybe you can talk to that.
2:48
Unidentified speaker: But then secondly, when we think about like the long-term case for distributed power, are there certain types of technologies that maybe screen better or more favorably, or certain characteristics of certain technologies that screen more favorably when you think about this being a structural source of power rather than maybe a short-term fix?
3:13
Unidentified speaker: Yeah.
3:14
Unidentified speaker: The bottleneck is very often in the utility space.
3:18
Unidentified speaker: Utilities are seeing such an unprecedented demand for electricity from the data center industry.
3:23
Unidentified speaker: They're scrambling to keep up with it.
3:25
Unidentified speaker: They're not used to growing this fast.
3:26
Unidentified speaker: And it's often their electric transmission network that is falling behind.
3:31
Unidentified speaker: It's the long-distance power lines and substation equipment that the utilities are really trying to react and to grow alongside the data center industry, but we expect them to fall short and pretty meaningfully short.
3:43
Unidentified speaker: You know, this is large-scale infrastructure that takes years to develop and build.
3:48
Unidentified speaker: So then when it comes to on-site power, as you're growing alongside that constraint or trying to avoid that constraint, we're seeing that data centers, they're very much prioritizing speed.
3:59
Unidentified speaker: The quicker they can get the power, the better.
4:01
Unidentified speaker: But more, you know, more and more we're seeing other factors come into play.
4:04
Unidentified speaker: How, how supportive is the community for different applications?
4:08
Unidentified speaker: You know, what's the air emissions profile, the water use, the noise consideration?
4:12
Unidentified speaker: What are the overall emissions that you're considering here?
4:15
Unidentified speaker: And then also what are the operational characteristics of that equipment?
4:19
Unidentified speaker: So we've looked carefully at those, and we come down to these technologies, fuel cells, engines, turbines.
4:25
Unidentified speaker: The industry is really pursuing all of them in this chase for quick time to power and to get around that bottleneck.
4:32
Unidentified speaker: And to that point, it seems like what matters to maybe the developers is changing.
4:37
Unidentified speaker: It used to just be cost, then it was cost plus time, and now it's cost plus time plus noise and emissions, because obviously there's a lot of political pushback.
4:46
Unidentified speaker: Take that maybe one step further and say it's not just the data center market that is looking at these technologies now.
4:53
Unidentified speaker: It's also commercial and industrial applications where obviously there's a big trend to onshore manufacturing, a reindustrialization of the U.S.
5:02
Unidentified speaker: And our teams have collaboratively estimated as I think 10 gigawatts a year of incremental demand for power just from industrialization of or reshoring of the U.S.
5:14
Unidentified speaker: But one technology we haven't talked about yet is energy storage.
5:16
Unidentified speaker: And this is part of the, when I think about the auto industry, this is obviously where they have a they play a somewhat critical role, which I'll talk about in a second.
5:23
Unidentified speaker: But Dave, just help us understand where is energy storage being used and how big is that market for data centers specifically?
5:31
Unidentified speaker: Yeah, storage has a number of applications out there, and and one of the fastest growing applications is going to be with data centers.
5:39
Unidentified speaker: You know, they consume power in such a new way that data centers are ramping up and down instantaneously, massive quantities of electricity load going up and down instantaneously, which is not what the grid is capable of handling.
5:54
Unidentified speaker: And even some of this on-site power equipment isn't ready to accommodate such quick changes in load.
6:00
Unidentified speaker: And so we're seeing battery storage as the solution in a growing proportion of data centers around the country.
6:06
Unidentified speaker: We're expecting even by 2030, it's such a fast-growing end market, a fast-growing TAM, that we would expect the data center demand for battery storage in 2030 to even exceed the grid demand for battery storage in that year in terms of the incremental new projects that are added in that year.
6:23
Unidentified speaker: So it absolutely is a very fast-growing opportunity for lithium-ion battery storage being added to data centers.
6:29
Unidentified speaker: Yeah, I mean, that that makes a lot of sense.
6:31
Unidentified speaker: And I think that corroborates what we're hearing and seeing in the auto industry where you are seeing a number of companies explore energy storage as a new end market to essentially reutilize their low-utilized capacity or underutilized capacity as it relates to EVs, uh, just given the lower than expected demand there.
6:53
Unidentified speaker: Obviously, Tesla, a very obvious player in the market that has a very large energy storage business in the U.S.
6:59
Unidentified speaker: Ford is now building an energy storage business in the U.S.
7:03
Unidentified speaker: and started to build a rather large backlog.
7:06
Unidentified speaker: So I think there's some obvious ways in which the auto industry is essentially taking the opportunity where EV demand has been lower to reutilize that in the energy storage market, which is faster growth and in potentially some cases higher margin than they've seen in the EV industry.
7:22
Unidentified speaker: So that's to me is like the most obvious opportunity for the auto industry in terms of how they can participate in the growth in AI power infrastructure.
7:31
Unidentified speaker: The second derivative would be, you know, power generation itself.
7:36
Unidentified speaker: And that we've seen some suppliers, auto suppliers essentially take their automotive grade technology and repurpose it into turbine generators, into energy storage systems, into microgrid inverters to essentially serve this market with existing technology that they manufacture today for vehicles, just repurposing it for the data center industry.
8:00
Unidentified speaker: So, you know, it's an auto industry that hasn't really seen much growth over the last five years.
8:05
Unidentified speaker: So it's a very interesting opportunity for these companies to, with very little incremental capital investment, find areas of growth outside of their traditional subsectors.
8:15
Unidentified speaker: And as we're looking for, you know, new creative sources of power, electric vehicles are an area that some are exploring to become a source of power rather than just consuming it.
8:25
Unidentified speaker: I guess how meaningful could vehicle-to-grid technology become for the broader energy system?
8:30
Unidentified speaker: Aaron Powell I think over time it could be fairly significant.
8:33
Unidentified speaker: I mean, today if you look at California, there's some existing vehicle-to-grid or home-to-grid systems or programs in place, and they dispatch between 500 to 1,000 megawatts of power to the California grid during peak periods of time.
8:50
Unidentified speaker: So it is fairly significant.
8:51
Unidentified speaker: It's fairly isolated to California because there's a very high degree of EV penetration and also residential solar with battery storage.
8:58
Unidentified speaker: So it makes it a somewhat of a unique market to be able to provide that.
9:02
Unidentified speaker: But over time, as we start thinking about, you know, 20, 30, 40% of annual vehicle sales being EVs across the US, that's when that could start to become a much more significant contribution to the overall power supply.
9:15
Unidentified speaker: So it's probably more of a 2030 plus conversation, if I'm going to be honest in terms of when it starts to move the needle.
9:20
Unidentified speaker: But in places like California, you are already starting to see it, you know, contribute fairly significantly during peak periods of time.
9:26
Unidentified speaker: All right.
9:26
Unidentified speaker: Well, I think we can wrap it there.
9:28
Unidentified speaker: Um, Dave, thanks for taking the time to talk.
9:30
Unidentified speaker: Great speaking with you, Andrew.
9:33
Unidentified speaker: And thanks for listening.
9:34
Unidentified speaker: If you enjoy Thoughts on the market, please leave us a review wherever you listen and share the podcast with a friend or colleague today.
9:42
Unidentified speaker: The preceding content is informational only and based on information available when created.
9:47
Unidentified speaker: It is not an offer or solicitation, nor is it tax or legal advice.
9:51
Unidentified speaker: It does not consider your financial circumstances and objectives and may not be suitable for you.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3 and interpreted by Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL. 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