Unlocking value in growing digital infrastructure

This episode explores where private equity can create value in today’s digital economy.

The digital economy is growing exponentially, driven by tech innovations like artificial intelligence and e-commerce. Behind this transformation lies a capital-intensive infrastructure buildout. This creates timely, strategic opportunities for private capital. 

In this episode, Luke Pais, EY-Parthenon Asia-Pacific Private Equity Leader, along with guests, Joongshik Wang, EY-Parthenon Asia-Pacific Strategy and Execution Leader, and Gilles Pascual, EY-Parthenon Asean Power and Utilities Leader, discuss where private equity can create value in today’s digital economy.

Key takeaways: 

• Demand for data centers remains strong, driven by cloud consumption and artificial intelligence (AI), but supply constraints on power and land persist, creating a structural gap 
• Data center operators and renewable energy developers should conduct a comprehensive strategy review, whether to internalize capabilities, externalize them, or pursue partnerships. 
• Strategic partnerships such as backward integration by data centers or forward integration by power and utilities providers can unlock investment and financing opportunities, especially for private equity players.

Luke Pais

Welcome to Money Multiple. The digital economy has experienced exponential growth over the past decade, driven by rapid advancements in technology and the increasing integration of digital solutions into everyday life. This transformation has been fueled by the proliferation of internet connectivity, the rise of e-commerce and the widespread adoption of digital services across various sectors of the economy. 

Artificial intelligence is playing a pivotal role in this digital revolution, enhancing the capabilities of digital platforms and services. 

The integration of AI into various applications has further accelerated the growth of the digital economy, creating new opportunities and efficiencies. As businesses and consumers increasingly rely on digital platforms for transactions, communication and information, the demand for robust digital infrastructure has surged. 

High-speed internet ensures seamless connectivity and access to digital services, while data centers and cloud computing provide the necessary storage and processing power to handle vast amounts of data generated by digital activities. Additionally, robust cybersecurity measures are essential to protect sensitive information and maintain trust in digital systems. 

Building out all this digital infrastructure is highly capital-intensive and presents a unique investment opportunity for private capital investors. Joining me to discuss key trends and opportunities in the space, we have Joongshik Wang, EY-Parthenon Asia-Pacific Strategy and Execution Leader, and Gilles Pascual, EY-Parthenon Asean Power and Utilities Leader. Welcome to the show. 

Joongshik Wang

Thank you for having us. 

Luke Pais

Joongshik, let me start with you. The global data center market is expected to grow significantly over the next decade. In fact, the global data center market is projected to show an 8% growth rate through to 2029 and result in a market size of about US$625 billion by 2029. I understand you’ve been keeping tabs on the sector over the years. Tell us what are you seeing in the landscape, specifically how you think the digital infrastructure landscape will evolve over the next five years and what does that mean for data center operators and investors? 

Joongshik Wang

Thank you, Luke.  

Interestingly, many things have changed. What has happened? AI chip controls by the US. How do these impact our data center landscape? 

Every five years, data center sector analysts try to look at the market’s fundamental drivers and changes. However, that has now become every three years. 

But one thing that’s consistent is the demand for data centers. What comprises future demand? 50% or up to 70% cloud consumption, and the rest largely driven by AI. 

Demand is solid. However, supply is limited even though there is capital. But power, land and (AI) chips? We are quite sure there will be a supply-demand gap in the market. In general, there will be over demand, lesser supply. 

Now, what’s the implication to us? In US, hyperscalers generally own 50% of their data centers. In Asia-Pacific, because of supply-demand gap, hyperscalers tend to outsource to third-party operators rather than own (data centers). That’s one trend that we see. 

Another trend is looking at Asia-Pacific’s top 10 or top 20 cities, data centers are well-developed – in Singapore, Tokyo, Seoul – but outside of those tier one cities in developed countries, because of supply-demand gap, we see a lot of data center development happening in South Asia or developing countries, and tier two and tier three cities because tier one cities have a fundamental constraint to supply new data centers. That’s the reason why recently, a lot of data center projects are happening around Malaysia and Thailand. In Indonesia, interestingly, Jakarta itself is well-developed and well-penetrated. However, outside of Jakarta, for example, Batam, would be the next future development for data centers. 

Last but not least, people are also concerned about pricing because it’s a direct impact to the revenue model of data centers. Previously, when cloud consumption was a major driver of the growth of data centers, people were acceptable to planned pricing (models). Now, because of supply-demand gap, we see the market willing to accept 10%, 15%, 20% higher pricing depending on demand, depending on commercial model. A lot of pricing dynamics is happening because of supply-demand gap. 

Luke Pais

Clearly, a lot of investment opportunities. As we look at digital infrastructure, data centers have been in large focus but if you can just help to unpack, what are the different categories of assets that investors should look at across the whole digital infrastructure landscape? 

Joongshik Wang

10-15 years ago, digital infrastructure was largely driven by the tower business, satellites and subsea cables. Recent digital infrastructure focuses on data centers. Data centers require very strong fiber capability, so the major driver of the growth of digital infrastructure is largely from data centers and new investments in fiber. 

Meanwhile, for satellites and tower, it’s more about how (companies) optimize balance sheets, so they divested their tower business, subsea cable business, for example, for (business) optimization. However, more digital infrastructure demand is now largely coming from data centers, fiber and also subsea cables, because of the regionalization of data center platforms. 

Luke Pais

Thanks, Joongshik.  

Gilles, let me turn to you. Clearly, all of the development that Joongshik mentioned creates a tremendous demand for power and utilities. I understand that data centers, AI and cryptocurrencies currently consume about 2% of global electricity demand and this is expected to approximately double in the next five years, so that means a lot of investment in power and utilities, both fossil based as well as renewables. At the same time, countries, of course, have to take care of their own population demands for power, plus all these goals that they’ve set to get to net zero. Can you take us through your thoughts on the investment opportunity ahead and how the sustainability agenda gets embedded into digital infrastructure development? 

Gilles Pascual

Thanks, Luke.  

I think as you pointed out and Joongshik just covered, governments around the region are generally keen to attract investments in digital infrastructure. It’s about attracting FDI (foreign direct investment). It’s about creating jobs. It’s about future-proofing your economy and your workforce. And frankly, there’s also a little bit of a “cool factor” when you can announce billion-dollar investments from the likes of hyperscalers, which are all household names, so it’s about developing your economy. However, data centers are resource-hungry – electricity, water – and the pressure is on the utility providers, the electricity companies around the region to cope with this additional demand. 

Now, if we zoom in to the electricity sector and challenges faced – with a few exceptions, like Singapore and the Philippines – the electricity sector in Southeast Asia is managed by a single electricity company that handles everything from generation, transmission to distribution. As you pointed out, population is growing, the economy is growing, so fundamentally, these utilities (providers) are dealing with the need to increase generation capacity, to constantly invest in the grid. There are billions of dollars of investment to serve natural demand. Additional demand from data centers further increases those massive investments needed. 

On top of it, utilities (providers) are faced with the need to green the grid because governments have made commitments to reduce CO2 emissions. This is a key component. Electricity companies face triple pressure points – they need additional power in the system, they need this to be green, they need to adapt to intermittent renewable energy. The utilities (providers) are under pressure, and they’re essentially pushing back to the government, saying, “Well, I can’t quite be as fast to meet the demands from data centers.” 

If you remember a few years ago, Singapore increased sustainability standards. They said no more investments in data centers unless the data center (DC) meets certain minimum requirements. That was driven by land pressure, but also the electricity sector. Singapore is very keen to green the electricity sector. 

Now, Malaysia and Thailand recently introduced regulatory changes to allow data centers to procure electricity directly from renewable energy developers, as opposed to buying it from the grid. That’s to relieve the pressure on utilities (providers) and allow data centers to contract directly and to work with renewable energy developers. 

Luke Pais

Thanks, Gilles.  

And that is a good segue for us to get into some of the key challenges. Joongshik, can you talk us through the key challenges faced by digital infrastructure operators and how investors are influencing these operators to incorporate the sustainability agenda. 

Joongshik Wang

Sustainability is a very luxurious topic, right? Because today’s supply-demand gap is coming from constraint of land, constraint of power; before even talking about whether it’s green power or not. Like Gilles mentioned, Singapore has already announced the moratorium on data centers. It’s an order in decarbonization, zero carbon, that the country wants to target, which limits any new supply in Singapore. Same thing in Seoul. For example, in South Korea, there is no available power in the tier one city, Seoul. That’s the one city where hyperscalers want to install more data centers, but no one can get power. Sometimes, residents protest against data centers because data centers are not necessarily welcomed in the residential area and the government also needs to listen to those residential complaints. 

So, before we (address) any sustainability matter, the data center itself is a fundamental issue around providing enough supply. Outside of power (constraints), land (constraints) and residential complaints, another key issue is you cannot find a good contractor. There are not enough EPC (engineering, procurement and construction), good general contractors in the market. 

Sustainability of energy is very, very important. However, the fundamental challenge is, it’s not easy to source green energy. Now, I will ask Gilles: Why is it very difficult to source green energy? 

And from the data center operator perspective, to manage that challenge, they’re making efforts to improve energy efficiency. Everybody has heard about direct liquid cooling. AI data center requires almost 10 times higher cooling mechanism. Existing cloud consumption of the data center can rely on air-condition-based cooling. However, the AI data center itself, because it requires 50 kilowatts, 150 kilowatts per rack, the heat control requires not just liquid cooling but also very AI-specialized chip design. The data center operator is making a lot of continuous effort to improve energy efficiency. However, the fundamental challenge is, how can we source green energy? 

Gilles Pascual

In many countries around the region, it is possible to procure renewable energy. There are various forms, either from the main utility company because they have a green tariff or directly from the private sector, from a renewable energy developer. And for that, you have virtual solutions, where the contract is not to buy and sell electrons but to buy and sell certificates, or there are physical solutions, like the ones I mentioned were recently introduced in Malaysia and Thailand, where it’s a real purchase of electrons going through the grid. Each one of these options has a different price point. And I will throw a question back to you later: What is the hyperscaler’s appetite to pay more to gather 100% green electricity? All the hyperscalers have ambitious decarbonization targets, but for now, they have been managing this at the global level, and they can offset their global carbon footprint through the purchase of carbon offsets or renewable energy certificates. 

There is a trend – regulatory trend – to push decarbonization targets to the country level. If you are a global operation, you either manage at the global level or you push this to the country level, or even at some point, to the facility level. Today, you have solutions, but they have different price points. If I just buy solar or wind, I can procure electricity cheaper than if I were to buy it from the grid, but you know that this is intermittent. If, as a DC, I need 24/7 power, I cannot afford any interruption, so I need to combine solar with battery storage. And that solution, solar and battery, today is more expensive than the electricity – not green – that I can buy from the grid. So eventually, it goes back to the hyperscaler: Are you willing to pay more to get a green DC? 

Joongshik Wang

I can’t represent the hyperscalers but let me put it this way, there are pricing dynamics and there is some room for a hyperscaler to pay some premium. However, energy cost is already 40 to 60% of the OPEX (operating expenditure). It’s already very high. It’s a very, very critical and sensitive expense, so I would look at the more fundamental change of infrastructure, rather than paying the very spot price to comply with decarbonization. 

For example, co-location between data center and power plant. Why not plan to source for green energy, especially for the AI data center, which requires a minimum 100 megawatts. It’s a larger scale project that is worthwhile (to consider), and then discuss with the energy service provider, earlier than later. 

Luke Pais

That’s an interesting observation. Let me address this to both of you. Given the direct correlation between digital infrastructure demand and the need for a stable, high-quality supply of power and utilities, do you see opportunities for collaboration and convergence and maybe even the emergence of new business and financing models? 

Gilles Pascual

Things change really fast on the electricity side. Prices are coming down; they’ve been coming down over the last few years, and it’ll continue to do so. For any DC operator, I would recommend a strategic review to look at the solutions, the potential business model when it comes to greening your electricity. You can bring the development of renewable energy in-house and start doing your own projects, attached or linked to your DC. You can externalize it; go and talk to renewable energy developers and see the developments happening in each country and whether you can sign a single contract or explore partnership models, whereby for the whole of APAC, partner with an existing large or small-scale renewable energy developer, a regional party, a domestic dominant player. 

I would even say likewise to the renewable energy developers. The skills to develop renewable energy – it’s selecting a site, it’s managing permits, managing the construction of an asset – the two-asset class, renewable energy investments – solar project, wind project – and DC, they’re both considered infrastructure assets, meaning from investors’ point of view, they have a similar risk and return consideration. Typically, DC investment will require a slightly higher threshold because there’s a little bit more uncertainties on the cash flows than your typical renewable energy project. But fundamentally, it’s similar capital, infrastructure, long-term risk-return consideration. So there is a key question, whether we will see a convergence of the platforms, the developers, renewable energy players getting into data centers, DC operators getting into renewable energy and those are the strategy questions that we need to solve with clients. 

Joongshik Wang

That’s a great suggestion, Gilles. Let me put it this way, they are two different infrastructures. The data center itself has a little bit more infrastructure risk. However, commercially (the market opportunity) is very much bigger than renewable energy infrastructure investment. So that’s the reason many of the capital market (players) are coming on now, because the cost of capital can be lower. However, the returns can be much higher. 

Going back to your question, how can those two assets be converged? Because there are benefits to converge the assets. Why? Because everybody needs green energy but there are technical challenges, which require a long-term investment period, as well as upstream and upfront collaborative planning, with a large CAPEX (capital expenditure). 

Today, it has been driven by the data center business, but the energy service provider can (also) come up with a debt plan, combine projects from the finance perspective, technology perspective. I think if the project has a debt angle, whether it’s driven by the renewable energy operator or the data center operator or through a partnership, I think that’s a fantastic idea. I mean, this idea itself can be a competitive advantage. 

Luke Pais

That’s fascinating. What you are basically saying is the first step could be a strategic partnership business model but as you take this forward, you would actually see convergence, and that could be backward integration of the data center, or forward integration of the P&U (power and utilities) providers, or there are opportunities to split the infrastructure layer from the operating layer. And all of that creates investment and financing opportunities for our private capital investors. 

Joongshik Wang

Yes, indeed.

Luke Pais

Joongshik, Gilles, I’d like to thank you both for joining us. Before we close, given the fascinating investment opportunity the space presents, I’d like to ask each of you to share a final thought with our audience.

Gilles Pascual

Well, from my end, Luke, really, if I’m a DC operator today, I would undertake a comprehensive review of the strategy when it comes to renewable energy procurement, whether to internalize, whether to externalize, whether to create partnerships and with whom. And if I’m a renewable energy developer, I would do exactly the same: How do I get into the DC market? Internalize, externalize or partnership. It’s all about strategy. 

Joongshik Wang

I really enjoyed today’s discussion, especially because the capital market is excited about this over demand cycle, but beyond the super demand cycle, I think the conversation that we had today, it’s very much a cutting-edge business idea, which requires maybe more than five years of preparation in terms of the technology, commercial planning. Convergence generally creates new idea, innovation and solution, so we should have a more open ecosystem approach rather than just (focusing) on your own segment.

Thank you, Luke. 

Luke Pais

Thank you very much. 

Gilles Pascual

Thank you. 

Luke Pais

You have been listening to Money Multiple. If you liked what you’ve heard, subscribe to our show on Apple Podcasts, Spotify or wherever you listen.