Infra Dig - A Green Street Podcast

PwC forecasts $31.6T global data center CapEx through 2050, driven by AI and power constraints

27 min Listen to the original
Data CentersInfrastructureAI Capital ExpenditurePower GridAsia Pacific MarketsData SovereigntyReal Estate Investment

Summary

Clara Cutajar of PwC outlines a forecast of $31.6 trillion in global data center capital expenditure through 2050, potentially reaching $50 trillion with accelerated AI adoption. She notes that ICT equipment will comprise 93% of CapEx by 2050, up from 70% in 2026, due to the short 4-6 year lifecycle of IT assets. Power availability is identified as the primary determinant for investment flows, with grid constraints driving interest in off-grid solutions like gas and batteries. Data sovereignty regulations are shifting capital from regional hubs like Singapore to domestic markets in India, Vietnam, and Indonesia. Financing models must adapt to hybrid risk profiles combining property, utilities, and semiconductor assets.

Key takeaways

  1. 1
    PwC forecasts global data center CapEx at $31.6 trillion through 2050, with a high-end scenario of $50 trillion if AI adoption accelerates significantly.

    Clara Cutajar · 0:27

  2. 2
    report, do continue to climb. We start with an annual amount of 800 billion from 2026, rises to 1.1 trillion in 2030, and then nearly doubles to 1.8 in 2050.

    Clara Cutajar · 2:55

  3. 3
    ICT equipment is estimated to comprise 70% of total CapEx in 2026, rising to 93% by 2050, fundamentally changing the asset's risk profile.

    Clara Cutajar · 5:10

  4. 4
    Your report identifies power availability as the most important factor determining where future data center capital expenditure will flow.

    Unidentified speaker (speaker_1) · 6:14

  5. 5
    India is attracting investment due to significant domestic demand, tax concessions, and streamlined planning approvals.

    Clara Cutajar · 10:52

Episode sections

  1. 0:07 – 2:15

    Introduction and PwC Data Center Outlook

    Clara Cutajar introduces the PwC forecast of $31.6 trillion in global data center CapEx through 2050, noting uneven distribution driven by power, chips, and regulation.

  2. 2:15 – 6:14

    Investment Cycle and Asset Lifecycles

    Cutajar explains that ICT equipment has a short 4-6 year life compared to buildings, with ICT CapEx rising from 70% in 2026 to 93% in 2050.

  3. 6:14 – 9:42

    Power Constraints and Off-Grid Solutions

    Power availability is the primary driver of capital flows. Grid constraints are leading investors to explore off-grid gas, fuel cells, and batteries in the US and Ireland.

  4. 9:42 – 14:10

    Asia Pacific Market Dynamics

    Asia Pacific CapEx is forecast at $8.2 trillion through 2050. India benefits from domestic demand and policy, while Singapore faces constraints, pushing investment to Malaysia and Indonesia.

  5. 14:10 – 17:33

    Data Sovereignty and Geopolitics

    Data sovereignty laws, with 30% of workloads currently required locally, are shifting investment from regional hubs to domestic markets in India, Vietnam, and Indonesia.

  6. 17:33 – 23:03

    Financing and Risk Profiles

    Financing models are evolving to handle the hybrid nature of data centers. Traditional infrastructure lenders may struggle with the short-life technology risks.

  7. 23:03 – 27:04

    Bankability and Future Watchlist

    Securing power and robust contracts is key to bankability. Risks include chip supply constraints, export controls, and regulatory uncertainty in jurisdictions like Saudi Arabia and Singapore.

Claims worth verifying

Global data center capital expenditure could reach $31.6 trillion through to 2050, potentially approaching $50 trillion if AI adoption accelerates.

Clara Cutajar ·0:27

How to verify: Check PwC's 'Global Data Centre Outlook 2026-50' report published in 2026. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Annual data center CapEx starts at $800 billion in 2026, rises to $1.1 trillion in 2030, and nearly doubles to $1.8 trillion in 2050.

Clara Cutajar ·2:55

How to verify: Verify against the annual breakdown table in PwC's Global Data Centre Outlook 2026-50. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

ICT equipment is estimated to be 70% of CapEx in 2026 and rises to 93% in 2050.

Clara Cutajar ·5:10

How to verify: Cross-reference the asset composition assumptions in PwC's report. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Servers and GPUs require replacement every four to six years.

Unidentified speaker (speaker_1) ·4:41

How to verify: Verify against industry standards for server lifecycle and PwC's report assumptions. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Asia Pacific cumulative data center capital expenditure is forecast at $8.2 trillion through 2050.

Unidentified speaker (speaker_1) ·9:42

How to verify: Check the regional breakdown in PwC's Global Data Centre Outlook 2026-50. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Data sovereignty could raise Asia Pacific data center investment by 7% above the central forecast.

Unidentified speaker (speaker_1) ·14:15

How to verify: Review the sensitivity analysis in PwC's report regarding sovereignty scenarios. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Approximately 30% of workloads today have to be hosted locally.

Clara Cutajar ·15:21

How to verify: Verify against industry reports on data residency requirements and PwC's market data. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Asia Pacific CapEx drops from $8.2 trillion to $6.4 trillion under the chip export control scenario.

Clara Cutajar ·25:16

How to verify: Verify the specific numerical impact in PwC's scenario analysis. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify PwC's 'Global Data Centre Outlook 2026-50' report for methodology and assumptions regarding the $31.6T and $50T CapEx forecasts.
  • Check grid connection waitlists and power availability data for key markets like the US, Ireland, India, and Singapore.
  • Review specific data sovereignty laws in India, Vietnam, Indonesia, and the Philippines to assess the 30% local hosting mandate and its projected growth.
  • Analyze the capital stack structures for hybrid data center projects that combine real estate, utility, and technology assets.
  • Monitor chip supply chain constraints and export control regulations that could impact the projected ICT equipment CapEx share.
  • Assess the bankability of off-grid power solutions (gas, fuel cells, batteries) in target jurisdictions.
  • Compare land and regulatory costs in emerging markets like Malaysia and Indonesia against established hubs like Singapore.
  • Review historical IT equipment lifecycle data to validate the 4-6 year renewal cycle assumption.
Full transcriptexpand

0:07Clara Cutajar: Hello and welcome back to the InfraDig podcast. I'm Mila Patel, Senior Editor at Green Street News Infrastructure, formerly IJ Global, and today I'm delighted to be joined by Clara Koutejar, partner and global infrastructure leader at PWC. We're going to be looking at one of the biggest infrastructure investment opportunities of our generation, the global expansion of data centers and the infrastructure needed to support the AI revolution. PWC's latest global data center outlook forecasts that global data center capital expenditure could reach U31.6 trillion US dollars through to 2050, with the potential to approach $50 trillion if AI adoption accelerates.

0:47Unidentified speaker (UNKNOWN): But

0:47Clara Cutajar: the report also highlights that this investment opportunity won't be even evenly distributed. Power availability, access to advanced chips, data sovereignty, and regulatory certainty will all influence work capital flows. So today we're going to explore what this means for infrastructure investors, lenders, and developers with a particular focus on Asia Pacific. Clara, welcome to the InfraDeg podcast. Please briefly introduce yourself to our listeners before we get into it. Hello, everyone. It's an absolute pleasure to be here today. I am Clara Cutter, I'm a partner with PWC. I have the privilege of leading our infrastructure business globally, a sector that I am hugely passionate about, a sector that is super exciting because we enable the world's global somatics. And today we're going to be talking about. OLAS. We are enabling the move into the digital infrastructure world. So super exciting. So great to be here. Wonderful. Thanks, Clara. Right. So

1:48Unidentified speaker (speaker_1): let's start with the headline numbers. As I said earlier, PWC is forecasting $31.6 trillion of data center capital expenditure globally through to 2050, potentially rising to almost $50 trillion.

2:01Unidentified speaker (UNKNOWN): That's

2:02Unidentified speaker (speaker_1): an extraordinary level. of investment. What has fundamentally changed in the data center market to create an opportunity of this scale? And how confident are you that demand will sustain this level of investment over the coming decades?

2:15Unidentified speaker (UNKNOWN): Yeah,

2:15Clara Cutajar: so the numbers are very large. 31.6 trillion is a very hard number to get your head around. If we look back into history and how in the infrastructure market has developed, this is a very different cycle to infrastructure investments of the past. If you think about infrastructure investments of the past, like building out the railroads or electrification or building the internet, there was a significant amount of capital that was required up front. But then it was about maintaining the assets and of course deploying deploy more capital when you wanted to expand the capacity of those assets. This is very different. The numbers, as you can see in our report, do continue to climb. We start with an annual amount of 800 billion from 2026, rises to 1.1 trillion in 2030, and then nearly doubles to 1.8 in 2050. And that is because all the the building building itself. itself lasts for a certain amount of time, but then you have all the equipment inside the building and the chits themselves, which have a much shorter useful life. And therefore, there's this constant renewing, renewing cycle of capital that you sign up for when you build a Dart

3:31Center.

3:33Unidentified speaker (UNKNOWN): In

3:33Clara Cutajar: terms of the confidence around the numbers, of course, it is a very long-term forecast. Forecasting out 25 years is no main feat,

3:42Unidentified speaker (UNKNOWN): and

3:42Clara Cutajar: we've had to make a certain amount of assumptions. And there are a number of assumptions, and the one big assumption that underpins these numbers is what is the uptake in AI. We have put some scenarios in our report. What if the AI uptake is higher or what if the AI uptake is lower? And of course, that does impact the numbers pretty significantly. And the range goes from 22 trillion on the low side to 50 trillion on the upside. So it's a significant swing, all due to the uncertainty around the uptake of AI into the future.

4:21Unidentified speaker (speaker_1): Um One of the particularly interesting findings in your report um is that this isn't a traditional. infrastructure investment cycle. Yeah. You estimate that for every one dollar spent on data center construction, there could be around $12 in subsequent um IT uh equipment expenditure with servers and GPUs requiring replacement every four to six years. How should infrastructure investors think about this different investment profile? Does it fundamentally change how data centers should be valued, financed, under underwritten?

4:51Clara Cutajar: Yeah, so as you mentioned, the life cycle of assets that you put into infrastructure investments into data centers is quite different. And because of the different useful lives of the assets, you do have an ecosystem of assets that come together. As we move into the future, the ICT equipment actually becomes. a much more significant component of the CapEx. It's roughly, we've estimated to be 70% in 2026, and that rises to 93% in 2050. So that does make the asset quite different as you move into the future. Investors absolutely need to consider the different components that make up the complete asset and where the risk risks actually. actually lie in those pieces. And as you note, the risks differ across the different components of infrastructure. Capital, development, leasing, and counterparty risk all behave differently. And so investors do need to consider that

5:56Unidentified speaker (UNKNOWN): in

5:56Clara Cutajar: terms of the different risks and, of course, opportunities that come with that. And that all impacts on the ultimate valuation of the asset. and the type of financing that you put into the asset.

6:05Unidentified speaker (speaker_1): Thanks, Clara. So let's turn to uh what is arguably the biggest challenge facing the sector. Yeah. Power.

6:13Unidentified speaker (UNKNOWN): Yeah.

6:14Unidentified speaker (speaker_1): Your report identifies power availability as the most important factor determining where future data center capital expenditure will flow. We're already seeing developers face grid constraints, lengthy connections. connection timelines, and shortages of critical equipment.

6:29Unidentified speaker (UNKNOWN): So,

6:30Unidentified speaker (speaker_1): how significant is this constraint? And are we reaching a point where access to reliable electricity will determine which markets and developers win the race for AI infrastructure?

6:41Clara Cutajar: Yes, absolutely. So the availability of power absolutely impacts where capital flows into this sector, and to a certain extent, it already is. If you sort of think about one of the critical aspects of this market, it's speed to market. It's the speed at which you can build out the data center, satisfy the demand of the customers, and of course connect to power. And I guess that power question: if you ignore data centers for a minute, we were already on a world uh a world where we were electrifying and having to build out transmission distribution networks, having to build out the grid. So we're already having a bit of a challenge around how we build out the grid network. And of course with the additional load and demand from data centers, that has accelerated that need. Now there are some countries that are doing well. If you look at um places in the Nordics, they've really pushed along in terms of building out their renewable energy, and they are moving fast in terms of building out the grid. But then other places around the world, across Europe, parts of Asia Pacific, there are significant delays in building out the grid, and that is impacting how quickly

8:00data centers can connect. And so investors then and data center operators are looking to other ways to get to market quicker. And one of those potentially is to look at repurposing old industrial sites that are already powered and using those to build data centers. So you know, investors are looking at other ways is to get power quicker.

8:22Unidentified speaker (speaker_1): Okay. Are we likely to see more investment in behind the meter general? generation, dedicated power assets and integrated energy solutions as developers look to secure their own electricity supply?

8:34Clara Cutajar: Yeah, absolutely. Because you know the automatic assumption cannot be that every data center has to be reliant on the grid.

8:42Unidentified speaker (UNKNOWN): Yeah.

8:42Clara Cutajar: And therefore the off the grid solutions do need to develop. In some countries they are rapidly developing and have evolved pretty significantly. And the US is one of those markets where there's multi-hundred megawatts and gigawatts of off-the-grid solutions that have been developed in terms of gas generation, fuel cells, batteries. Ireland is another one where, as part of the grid sort of connection policies, they are looking at the regulation and ensuring that it's part of connecting a data center that you have generation and storage as part of that. So now there are some countries where that's developed, but if you look to Australia, which is you know where I'm from, we're still pretty reliant on the grid for data centers, having that conversation about off-grid quite a lot. And some operators are thinking about it, but we haven't quite got these. there yet. But you know, it it does need to evolve.

9:42Unidentified speaker (speaker_1): Yeah. Okay, thanks, Clara. Right, so let's turn specifically to Asia Pacific. PwC forecasts $8.2 trillion in cumulative data center capital expenditure across the region through 2050. Again, so looking at that longer, longer period. So China and India are identified as major sources of incremental demand. but the report also highlights significant differences in the outlook for uh individual markets. Which markets do you believe are best positioned to capture the next wave of investment?

10:15Unidentified speaker (UNKNOWN): Um,

10:15Unidentified speaker (speaker_1): and what do you think will differentiate the winners from those that struggle to attract capital?

10:20Clara Cutajar: Yeah, absolutely. So, as we highlight in the report, capital is mobile and it's looking to be deployed quickly in the countries that can offer the most reasonable return.

10:32Unidentified speaker (UNKNOWN): And

10:32Clara Cutajar: as we look across the Asia Pacific, there's quite a spectrum of where countries are at in terms of regulation and policy and guardrails in relation to data centres, and that obviously affects where capital will flow. Mentioned India. India has put in place some very attractive policies to attract data center investment, significant tax concessions, cutting red tape, making approvals and planning much quicker. Of course, all of this underpinned by significant domestic demand. So that is a very, very attractive market for a number of investors. Across Asia Pacific, if you look at one of our uh one of our scenarios, which was the AI adoption being higher or lower than our base case, that does significantly impact where capital will flow as well. So the markets that have the larger domestic demand,

11:32their countries like Japan and Australia, are more resilient in that scenario. Whereas markets that rely on AI workloads are more exposed because that demand can slow to other countries.

11:46Unidentified speaker (speaker_1): But how should investors be thinking about established hubs such as Singapore compared with emerging markets like Malaysia, Indonesia, Vietnam, and India?

11:57Clara Cutajar: Yeah, so when investors are looking at deploying capital, you know, they're obviously looking to diversify. their risk. Singapore is an attractive market. It leaves in terms of policy coordination, although it does have constraints in terms of land availability and in relation to the availability of renewable energy.

12:19Unidentified speaker (UNKNOWN): The

12:19Clara Cutajar: Malaysian market is very attractive. It is and there is a significant amount of data center pipeline being built there. It has the demand in Singapore, it has land, power, good connectivities, Singapore, and projects are being built in a much quicker timeframe. So becoming very attractive. Indonesia has a very underpenetrated local domestic market. And therefore, in a scenario where sovereignty becomes key, that will become quite an attractive market. And then if you look at where investors are looking to deploy capital, I mean, some markets are very hot in terms of the amount of capital that are flowing into them. And if you're looking for an earlier stage opportunity, certainly in Vietnam is at the earlier stage in terms of domestic demand is still. lower, but does have advantages in terms of uh speed to market and lower build costs. And then we obviously talked about India being one of the leading markets across Asia Pacific, which is supported by significant domestic demand, which is driven by large domestic economy, plus

13:36Unidentified speaker (UNKNOWN): all

13:36Clara Cutajar: of the services that get undertaken in India. So, you know, investors, depending on where they're at in their life cycle of considering these investments, will be attracted to some more established markets or potentially the newer markets, depending on the return and the risk profile they're looking for.

13:56Unidentified speaker (speaker_1): Thanks, Claris. And let's talk about data sovereignty and geopolitics. One of the more interesting aspects of the report. is the impact of data sovereignty on where investment could be deployed. Yeah. Your modeling suggests that if governments increasingly require critical workloads to be hosted domestically, Asia Pacific could see data center investment rise by seven percent above the central forecast. However, some established hubs, including Singapore, could see lower investment under that scenario. How significant could this shift be and could data sovereignty fundamentally change the regional data center landscape over the next decade?

14:37Clara Cutajar: Yes, so sovereignty will certainly have an impact. And yeah, as the report outlines, it it doesn't significantly impact the overall quantum of spend, but of course does impact. where the money gets spent. In certain parts of Asia, India, Vietnam, Indonesia, Philippines, and Thailand, they all gain due to sovereignty, due to the requirements to build more in those countries because of the domestic to meet the domestic demand that they already have. And then other countries that are currently regional hubs. potentially are losers in that scenario. And that's places like Hong Kong, Singapore, potentially Australia, they will potentially see less investment in a central case. We are already seeing this happening in terms of this move because of sovereignty. About 30% of workloads today have to be hosted locally, and this share is growing much faster as governments.

15:39put in some regulations and mandate the a certain percentage of workloads needed to be run in certain countries. Inference, day-to-day running of AR models is drawn towards where the users are. Of course, training can still go wherever the power and the chips are. So it it is changing the landscape already. And then depending on how regulation. regulations evolve in countries and how restrictive those regulations are, it could change significantly there.

16:10Unidentified speaker (speaker_1): And just how should investors balance the opportunity presented by domestic demand against established advantages of regional data center hubs?

16:19Clara Cutajar: Yeah. So investors, I think, should look at those two not as a competition between each other or a substitute, but as a balancing of risk between the two and a balancing of opportunity, because they do serve different parts of the data center demand and they offer different risk characteristics. And investors would want to balance those risk characteristics to end up with you know an overall balanced portfolio. In terms of how investors should think about those, they obviously need to ask themselves some questions about what is the workload in those in those countries, not the geography. Consider the domestic demand, and the domestic demand is effectively a flaw in terms of the utilization in those countries, and then work through the impacts of sovereignty, because that does. provide some diversification depending on how that sovereignty plays out. And then, of course, balance that all off in terms of how much exposure they want from a risk perspective and a growth perspective. So it's a bit of an art in terms of working through it and

17:31balancing the risk versus the reward.

17:33Unidentified speaker (speaker_1): Okay. Um And just talking about financing the infrastructure build-out, build out, the scale of investment requires. required raises quite important questions about how these projects will be financed. So and we're seeing increasing interest in private credit, debt capital markets, and project finance, while renewable energy PPAs are also becoming quite important in supporting bank or power infrastructure. Do you think we'll see a fundamental shift? in how data centers are financed, particularly as projects become increasingly integrated with power generation and transmission assets?

18:10Clara Cutajar: Yeah, and the financing models are evolving and will continue to evolve.

18:17Unidentified speaker (UNKNOWN): You

18:17Clara Cutajar: know, if we think about this asset class, it is still relatively immature in the context of infrastructure assets, because sort of a 15, 15-20-year 20 year thing versus other asset classes that have been around for over 100 years. You know, historically, data center projects were financed on a relatively independent basis, um, you know, secured land, grid connection, some of customer leases. But as projects become much more complex, then the model of financing. does need to be reconsidered because you're not just financing one asset, you're financing an ecosystem of assets which have different useful lives, different risk characteristics, different cash flow profiles. And then if you think about Up the financing of those, there's not one stack of capital that can finance all that. You are going to need different layers of capital to come in and finance that. And that's effectively what's happening now. Different capital stacks are coming in, depending on which part of the asset

19:28you're looking to finance, and matched with the type of risk that that capital is looking to finance. But it will it it's you

19:38Unidentified speaker (UNKNOWN): know

19:38Clara Cutajar: it's gonna continue. going to to evolve.

19:40Unidentified speaker (speaker_1): Okay. And and do you do you think that traditional infrastructure lenders and investors are sufficiently, I mean, are they sufficiently equipped to assess the technology and equipment risks associated with these assets, particularly given the relatively short replacement cycles for GPUs and servers?

19:56Clara Cutajar: Yeah, no, in terms of uh infrastructure investing, it has evolved pretty significantly over the years. It's become much more sophisticated. And of course, what we call infrastructure becomes broader and broader and broader, which you know, for those that work in the sector is great because we get to work on more and more things. And and and that that data center asset class is no different.

20:21Unidentified speaker (speaker_1): Yeah.

20:22Clara Cutajar: It it is becoming more and more complex, more and more broad. and still relatively immature in terms of the profile of its life. Infrastructure investors and the lenders into these assets, of course, are very comfortable assessing assets with long, useful lives, long cash flows, contracted assets versus then comparing it to what's in. inside the asset, the technology inside the asset becomes more complicated to consider. And that's where we're seeing these different types of financing structures, because you have different investors coming in to finance that part of the asset versus the long-style cash flows.

21:09Unidentified speaker (speaker_1): Yeah. Okay.

21:10Clara Cutajar: I think those investors that understand the distinctions between the various assets. asset classes are going to be the ones that are better positioned overall to underwrite the asset class.

21:19Unidentified speaker (speaker_1): And your report makes an interesting point that investors need to look at data centers as hybrid assets, including property, utilities, and semiconductor exposure, each carrying different risk profiles.

21:31Clara Cutajar: Yes.

21:32Unidentified speaker (speaker_1): As more capital enters the sector, where do you think investors are more at risk of mispricing opportunities or overlooking? risks?

21:40Clara Cutajar: Yeah. And if we sort of maybe look at the renewable energy sector firstly and think about when that initially got going, there was, of course, a lot of capital that was deployed into the renewable energy sector across the world. Investors made certain assumptions about the return of those assets, made certain assumptions about how the market would move going forward. And as we know now, uh hindsight's always a wonderful thing. Um, the market, yes, in certain countries, has not evolved as investors thought. And some investors took on more risk than they thought, and returns were not what they thought they were. would. This, I think, is gonna is similar, probably go in a similar way, in that as the market evolves, you know, certain investors will have made certain assumptions about how these assets move into the future. Some of those assumptions may not prove to hold. And, you know, on on the flip side, there'll be other assumptions. that have provided opportunities as well. And and as with any market, it will start to normalize and then start to move

22:59forward. So you know, it it's an it's an evolving market.

23:03Unidentified speaker (speaker_1): And how important will long term contracts, credit worthy off takers, and the ability to secure power be in determining which projects are ultimately bankable?

23:13Unidentified speaker (UNKNOWN): Yeah,

23:14Clara Cutajar: I mean they're fundamental. Uh you know, when any when any uh financier comes along to look at a project, first question they're gonna ask you is, do you have power? And then second question is what's the nature and tenor of the contractual arrangements and what's the economics of the cash flow moving forward? So that is fundamental question that all financiers are going to ask. It it it's interesting in that we, because we're so early in this cycle, we haven't really seen a lot of churn in contracts as yet, and what the behavior will be once those contracts churn. And as we move into the future, once again, there'll be some learnings as to how that market moves into the future. But absolutely fundamental. for bankability is having power and having contractual arrangements that underpin the cash flows.

24:12Unidentified speaker (speaker_1): And just looking ahead, what happens next? So if we're looking over the next five years, what do you think will be the biggest development to watch in the digital infrastructure market? Do you think it will be the appeal ability to secure power, the emergence of new financing structures, the geographical redistribution of data center capacity, or something else entirely?

24:35Clara Cutajar: Yeah, there's so many things. Um I think one of the areas that isn't is a real area to watch is the supply of chips. Because ultimately if the supply of chips becomes impacted. in some way, shape, or form, and where those chips flow, that is obviously going to significantly impact where CapEx gets deployed. And you know, one of our scenarios in our report was if there were export controls around chips, what would that do to the market? And CapEx did fall significantly, it sort of halved effectively. And in terms of Asia Pacific, it did drop the overall amount of CapEx from 8.2, 6.4. So I think that piece around where chips flow and whether there are going to be some controls around those is an important consideration.

25:33Unidentified speaker (UNKNOWN): Mm-hmm.

25:34Unidentified speaker (speaker_1): And if you were advising an infrastructure investor looking to enter or expand. in this market today, where would you tell them to focus?

25:42Clara Cutajar: Yeah, it it's in terms of what investors want to see, it is certainty. Certainty at at the point at which they make decisions, they need certainty around policy, regulatory landscape, tax landscape. And so those were where I would encourage them to focus are the areas and the jurisdictions that can coordinate energy, the capital, planning and approvals, and get that to a single point of contact. And if that is coordinated in that way, then that makes the deployment of capital a much easier race and improves returns over overall. And there are some countries that are leading this, Saudi Arabia, Singapore have really coordinated their policies. And that's making them very sort of attractive markets to invest in.

26:39Unidentified speaker (speaker_1): Clara, thank you very much. It's been fascinating to hear your perspective on the scale and opportunity and the challenges that will determine where investment flows over the coming decades. And thank you to our listeners for joining us for this latest episode of Infradig. We'll see you next

26:55Unidentified speaker (UNKNOWN): time.

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.
All insights