AI, energy, and the new supercycle of sustainable infrastructure investments

Despite the dynamic investments made by hyperscalers, investors should bear in mind that this heavy dependence on just a few players also entails risks. Strategy changes, shifts in CAPEX budgets or regulatory interventions can slow down investment cycles – as was evident from 2022 to spring 2025, when higher interest rates, political uncertainties, and supply chain disruptions led to a significant underperformance across the entire green tech segment. Even though the environment has since improved structurally, macroeconomic conditions and the willingness of major cloud providers to invest remain important factors influencing the short-term development of the industry.

The Hyperion Data Center is a mega data centre developed by Meta and financial investor Blue Owl Capital in Richland Parish, Louisiana, and, at an investment volume of about USD 27bn, is one of the largest and most ambitious AI infrastructure projects worldwide. The project comprises a hyperscale campus, which is scheduled for completion by 2030 and will be capable of delivering several gigawatts of IT power (source: CNBC).

Solar, wind, and geothermal energy are the beneficiaries

For AI systems to grow, energy must be scaled at a historically unprecedented rate. Renewable forms of energy play a key role here because they are the only technologies that can be deployed quickly enough to actually meet the growing demand. Whereas conventional large-scale power plants require several years of planning and approval cycles, solar and wind farms can be connected to the grid in a much shorter time – often within a few quarters to a few years. This time-to-power is the key competitive advantage in the AI age.

In addition, there is predictability: long-term power purchase agreements (PPAs) offer reliable revenue structures over terms of up to twenty years. This gives data centre operators stable, predictable electricity prices – a critical factor when loads and computing capacities are increasing on an annual basis. At the same time, the expansion of renewables contributes to decarbonisation targets that are firmly anchored in regulations on both sides of the Atlantic. Geothermal energy complements the system with base-load-capable, weather-independent generation, while battery storage increasingly provides hourly flexibility and grid services. The result is an energy mix that combines speed, cost transparency, and sustainability – exactly what the AI economy needs.

The new infrastructure supercycle: grids, dissipation, and electrification

We should like to point out that energy generation is only half of the equation. At least as important is the infrastructure that reliably delivers electricity to data centres – and distributes and dissipates it efficiently there. The necessary grid expansion includes higher transmission capacities, additional high-voltage lines, and significantly more precise load control.

At the same time, demand is growing for transformers, switchgear, power electronics, high-performance cables, and highly efficient cooling systems.

Companies in the electrification and grid technology sector – such as Schneider Electric, ABB, Quanta Services, MYR Group, Hammond Power Solutions and others – are becoming indispensable partners in AI infrastructure. In the United States, the transmitted grid capacity will likely have to be effectively doubled in some areas to cope with the projected loads. The result is a broadly based infrastructure supercycle that extends far beyond the electricity sector and is reshaping entire industrial clusters.

Attractive valuations in the green-tech sector

It is striking that this robust, long-term growth is not yet fully reflected in the valuations of many green tech shares. While major tech indices have seen valuation premiums, numerous renewable energy, grid, and infrastructure companies continue to trade at significant discounts to the overall market. According to our analysis, the P/E ratios of relevant environmental and green tech strategies are in the range of 19x–20x, while the S&P 500 is closer to about 25x.

The fundamental environment has improved: for institutional investors, this results in a rare combination of moderate multiples, visible cash flows, and predictable growth in end demand – a set-up that opens up both relative and absolute outperformance potential. Please note: investing in securities involves risks as well as opportunities.

Risks: politics, commodity prices, and financing costs

Investors should bear in mind that this potential also involves certain risks. Changes in political support mechanisms, delays in approvals or grid connections, and fluctuations in commodity prices and financing costs can influence individual project calculations. Similarly, technological leaps – such as a significant increase in the energy efficiency of data centres or alternative computing architectures – can change the timing of demand trends. Climate risks, supply chain bottlenecks, and increasing competition in the infrastructure sector also remain factors that should be considered. That being said, in the current environment, the structural appeal outweighs these factors, provided they are carefully managed and integrated into the portfolio structure.

Political environment and takeover momentum as additional driving forces

The political environment, especially in the United States, is a key anchor of stability. Despite recurring debates, the regulatory framework – from safe harbour rules to the subsidy mechanisms of the Inflation Reduction Act – remains reliable in its basic features. Even in traditionally conservative states, economic rationality now prevails: “Economics trumps ideology” – wherever renewables are the cheapest and fastest option, they are being built.

Also, M&A activity is increasing noticeably: pension funds, infrastructure managers, and private equity firms are taking advantage of the valuation environment to make strategic acquisitions – from developer platforms to operator portfolios. Such transactions confirm the attractiveness of the asset class and attract further capital. Where institutional capital is invested, economies of scale arise in procurement, construction, and operation – a self-reinforcing mechanism that further accelerates the expansion of renewables and grid infrastructure.

Conclusion: AI and energy constitute the investment engine for the coming decade

The transition to an AI-driven economy is triggering an infrastructural revolution. Energy generation, grids, storage, electrification, and data centre peripherals are the pillars of this transition. Institutional investors are thus faced with an exceptionally clear long-term trend: renewable energies and green tech companies are not just followers, but a prerequisite for economic growth in the AI age.

The combination of structural demand growth, political tailwinds, attractive valuations, and increasing M&A support offers a compelling risk/reward profile: for broadly diversified, sustainable, and growth-oriented portfolios, Erste Asset Management's environmental and green tech strategies represent a strategically relevant exposure to one of the most significant megatrends of the coming decade.

For explanations of technical terms, please visit our Fund Glossary.