Enabling AI: Geopolitical implications from an investment perspective
Portfolio Managers Richard Clode and Guy Barnard discuss how geopolitical factors have a significant impact on technology and property, two sectors that play a major role in enabling AI.
8 minute read
Key takeaways:
- Sovereigns have become a new customer base in the AI era, while generative AI has exponentially raised the cybersecurity threat level.
- Data centres, the backbone of the digital economy, have become ‘AI factories’ with strategic implications for countries.
- Investors need to be cognisant of how geopolitical factors are impacting the sectors and companies they invest in, among them implications for supply chains, costs, competition and sustainability.
Geopolitical factors are an increasingly important driver and risk for markets. As explored previously, it is important for investors to actively navigate this new era of deglobalisation, political and military conflict, climate change, and a focus on national security. In our first post on the subject, we discussed how technology and property companies are coming together to enable and benefit from the demand for artificial intelligence (AI). Here we discuss how, as one of our three key macro drivers, geopolitics is impacting the technology and property sectors and companies that are enabling AI.
The technology view (Richard)
AI is the new arms race – AI capability is now being viewed as a priority for countries and regions, being strategic to national security, cybersecurity and economic productivity. Consequently, unlike in the internet era, we are seeing countries becoming meaningful customers in AI infrastructure as they build their own data centres. This is creating a new customer base for the likes of nVIDIA, with demand for its GPU chips from sovereigns likened to being as large as the three largest hyperscalers combined. These sovereigns include new entrants to the technological geographic footprint, notably in the Middle East where AI is a strategic priority in both Saudi Arabia and the UAE. We have also seen the sovereign nature of AI play out in geopolitics, notably the US restrictions on exporting the latest nVIDIA GPUs.
The soft power of AI and its intertwining with geopolitics is evident in Microsoft’s recent US$1.5 billion investment in G42, the dedicated AI firm. Based in UAE, G42 will run on Microsoft’s Azure cloud platform and use its chips.1 The agreement involved a wider discussion with the US government around ongoing access to nVIDIA GPUs related to where their customers were based.
Another example is US attempts to control TikTok’s AI algorithm, with the potential to influence the US electorate leading to legislation being passed to force US investors in TikTok to divest, and suggesting the creation of a US-owned entity.
Chip wars: reshoring and nearshoring
2024 will see more than half of the world’s population heading to the polls, including the US. But we do not expect the interweaving of AI and geopolitics to unravel regardless of who wins the US presidential election. Access to the latest nVIDIA GPUs is key in the new AI arms race, and that is playing out both in terms of export restrictions as well as creating strategic resiliency in the supply chain of these chips.
Aiming to reduce risk and reverse decades of concentration of semiconductor manufacturing in Taiwan and South Korea away from the US, Europe and Japan, taxpayers globally are indirectly funding tens of billions of dollars in subsidies, and enabling tax breaks and favourable loan terms to encourage leading edge semiconductor manufacturers to build manufacturing facilities (fabs) locally. The current restrictions on the latest nVIDIA GPUs and ASML EUV (extreme ultraviolet lithography used to print chip patterns) tools being exported to certain countries have significant implications today, but its impact will only grow over time as Moore’s Law creates a wider chasm over time in terms of AI capabilities. It remains to be seen whether these restrictions are absolute or will prove to be relative and moved forward over time. Also, how willing are countries to accept this status quo and growing AI deficiency.
Growing cyberthreats
Geopolitics is playing out not just in the physical, but also the digital world with rising cyber threats that can only become more potent in an AI world. The ability of generative AI to create deep fakes at scale and at low cost, enables bad actors to engage in cybercrime using more sophisticated attacks with a greater armoury of threat vectors to gain access to computer systems and networks. Insurer UnitedHealth recently reported the first billion-dollar hack, causing a huge theft of Americans’ private healthcare data. Cybersecurity’s Ventures Cybercrime Report noted by 2025 cybercrime could cost US$10.5 trillion compared to US$3 trillion in 2014. Meanwhile, according to the World Economic Forum’s 2023 Global Security Outlook report, 74% of organisations surveyed believed global geopolitical instability influenced their cyber strategy. The recent outbreak of land wars has been accompanied by a major escalation in cyber warfare. To combat this escalating threat driven by geopolitical instability as well as AI, companies as well as countries are having to enhance their defence capabilities by embracing AI technologies.
The property view (Guy)
Data centres used to be large rooms housing mainframe computers. Today they have evolved into sophisticated facilities, requiring specialist management teams tasked with maintaining the security of servers, data and power supply, all whilst operating in spaces with consistent optimal temperatures and humidity levels as well as guaranteed 100% uptime.
As the backbone of the digital economy, data centres enable digital transformation, cloud computing, big data and AI technologies such as generative AI. The broadening demand for digital services and connectivity from consumers, corporates, and governments is a significant tailwind for this property type.
Addressing sustainability challenges
However, data centres are power hungry; they constantly consume a huge amount of resources, from the energy to operate them, as well as either energy or water to cool down the highly temperature-sensitive equipment. The International Energy Agency says that by 2026 data centres could globally consume more than 1,000 terawatt-hours of electricity, more than double 2022 levels, roughly equal to Japan’s total electricity usage.2 In response to this, data centres are already being forced to relocate or find alternative locations. Singapore, Frankfurt and Amsterdam have seen power shortages and restrictions on new data centres. Indeed, almost 20% of Ireland’s power consumption today is used in data centres. This is forecast to rise to 32% by 2026.3
This is however, creating opportunities for other markets. For example, due to Singapore’s current moratorium on data centres, neighbours Malaysia and Indonesia are benefiting as the building of new data centres switches to these alternative locations. The potential geopolitical risks from such actions are clear given the importance and requirement for power.
An additional consideration is the source of power itself. There are currently many sources, however most of these produce large amounts of CO2. The lack of renewable energy and grid capacity presents challenges for data centre operators. Over time they will be forced to decarbonise and provide alternative sources of power. Ultimately, to become ‘green data centres’ companies will have to use power sources such as hydrogen, solar, wind and nuclear. Here there is good news, as wind and solar generation costs are down by almost 70% and 90% respectively since 2010 and are now the cheapest form of power generation for more than 80% of global energy demand.4
However, besides the immense power needed to operate data centres, cooling is an equally important aspect, particularly with AI requiring more power to run superchips such as those from nVIDIA. Data centres require enhanced cooling capacity, with liquid cooling likely to become the standard going forward. Access to water is also not a given. About a fifth of data centres are located in water-stressed regions as they tend to provide better access to wind and solar power.5 The Middle East has the largest expected economic losses from climate-related water scarcity according to the World Bank.6 Given the expected growth in AI this places further stress on already scarce resources and is shaping geopolitical challenges and influencing regional power dynamics.
With every new seismic shift in technology there are winners and losers, opportunities and threats. Indeed, while today’s high returns on data centre development will attract more capital to the space and result in more supply in the future, the current hurdles to development are considerable. These include bottlenecks in power transmission, data centre equipment, labour, as well as increasing prevalence of anti-data centre zoning in key markets. All this means that the data centre business is likely to remain landlord-friendly for several more years to come, and open only to those with access to capital and the experience to execute to the required exacting standards.
Summary
AI is not just a theme. This latest wave of technology has immense potential across multiple areas, including dynamic sectors like property and tech. From an investment perspective, geopolitics has huge implications for companies and sectors, including those that are enabling and meeting the strong demand for AI; these financially-material impacts can affect company valuations and their long-term growth prospects. While investors need to be mindful of these considerations, the flip side is that as companies navigate these challenges, there is also the potential for innovation, collaboration and partnerships. Active management and engagement can help identify the companies that seize these opportunities, and are more likely to be the leaders and winners in their respective spaces.
2,3 International Energy Agency as at 20 May 2024. https://www.iea.org/reports/electricity-2024/executive-summary
4 IRENA, Macquarie Equities Research, Data Centre Power Crunch, April 2024.
5 NBC News, 19 June 2021, Drought-stricken communities push back against data centers.
GPU: a graphics processing unit performs complex mathematical and geometric calculations that are necessary for graphics rendering and are also used in gaming, content creation and machine learning.
Hyperscalers: companies that provide infrastructure for cloud, networking, and internet services at scale. Examples include Google Cloud, Microsoft Azure, Facebook Infrastructure, Alibaba Cloud, and Amazon Web Services.
Moore’s Law: predicts that the number of transistors that can fit onto a microchip will roughly double every two years, therefore decreasing the relative cost and increasing performance.
There is no guarantee that past trends will continue, or forecasts will be realised.
Technology industries can be significantly affected by obsolescence of existing technology, short product cycles, falling prices and profits, competition from new market entrants, and general economic conditions. A concentrated investment in a single industry could be more volatile than the performance of less concentrated investments and the market.
REITs or Real Estate Investment Trusts invest in real estate, through direct ownership of property assets, property shares or mortgages. As they are listed on a stock exchange, REITs are usually highly liquid and trade like shares.
Real estate securities, including Real Estate Investment Trusts (REITs) may be subject to additional risks, including interest rate, management, tax, economic, environmental and concentration risks.
These are the views of the author at the time of publication and may differ from the views of other individuals/teams at Janus Henderson Investors. References made to individual securities do not constitute a recommendation to buy, sell or hold any security, investment strategy or market sector, and should not be assumed to be profitable. Janus Henderson Investors, its affiliated advisor, or its employees, may have a position in the securities mentioned.
Past performance does not predict future returns. The value of an investment and the income from it can fall as well as rise and you may not get back the amount originally invested.
The information in this article does not qualify as an investment recommendation.
Marketing Communication.