By Tanja Faller
For much of the past three decades, investors were rewarded for believing that the world was becoming less physical.
Software displaced hardware. Platforms displaced assets. Globalisation reduced costs. Capital became lighter, faster, and increasingly detached from geography. The world’s most valuable firms were now digital networks. And no longer oil companies, industrial conglomerates, or utilities.
The assumption underpinning this era was that physical constraints mattered less than they once had.
That assumption is now being tested. Across energy, industry, technology, and geopolitics, the world is rediscovering an old truth: economic power ultimately rests on physical systems. Electricity still has to be generated. Just as coods still have to be manufactured. Data still has to be stored and water still has to be moved.And increasingly, these systems are struggling to keep pace with demand.
The defining economic challenge of the next decade may not be innovation itself. It may be the capacity to support it.
AI’s Inconvenient Reality
Artificial intelligence is often described as the next digital revolution.
In practice, it is also an infrastructure revolution. Every AI model depends upon vast quantities of electricity, computing power, cooling capacity, and data transmission. Behind every elegant interface lies an industrial system of data centres, transmission lines, substations, transformers, and power generation assets.
The physical footprint of the digital economy is expanding rapidly. According to the International Energy Agency, global electricity consumption from data centres is expected to more than double by 2030, reaching levels comparable to the current annual electricity consumption of Japan. In the United States, data centres are expected to account for nearly half of all growth in electricity demand during the remainder of the decade.
Utilities in Virginia, Texas, Ireland, and parts of Northern Europe are already reporting connection requests that exceed available grid capacity. Waiting times for new power connections are lengthening. In several markets, access to electricity is becoming a strategic determinant of economic growth. The race to dominate artificial intelligence increasingly looks less like a software competition and more like a contest over energy, infrastructure, and industrial capacity.
This should not be surprising. Every previous industrial transformation eventually encountered physical limits. Railways required steel. Automobiles required roads. Electrification required grids.
Artificial intelligence requires power. Lots of it.
The Return of Scarcity
For much of the globalisation era, scarcity appeared to be in retreat. Supply chains stretched across continents. Components arrived precisely when needed. Capital flowed freely. Energy was assumed to be abundant. Infrastructure was largely taken for granted.
Recent years have challenged those assumptions.The age ppf policrisis emerged. The pandemic exposed the fragility of highly optimised supply chains. Russia’s invasion of Ukraine reminded Europe that energy security cannot be taken for granted. Rising geopolitical tensions have highlighted dependencies on critical minerals, semiconductors, and advanced manufacturing capabilities.Meanwhile, climate-related disruptions have revealed vulnerabilities in water systems, transport infrastructure, and electricity networks.
What these events share is not their cause, however there is one lesson: Efficiency and resilience are not the same thing.
For decades, efficiency was rewarded. Inventories shrank. Redundancies disappeared. Production concentrated in the lowest-cost locations. The result was a system optimised for stability.For much of the post-Cold War period, economic policy operated under the assumption that capital was scarce and infrastructure abundant.
Global financial assets exceed half a quadrillion dollars. Yet the OECD estimates that the world must invest nearly $7 trillion annually in infrastructure through 2030 to support projected growth and meet development and climate objectives. The constraint is no longer capital availability, but rather the ability to deploy capital into financeable , scalable, real-economy projects.
Why Infrastructure Has Become Strategic
Infrastructure has traditionally been viewed as defensive. As investments, they mainly provided stable returns and a predictable cash flow. Limited excitement.
However, the balance is adjusting to what historically had been valid for centuries: Today, once again, infrastructure is determining national competitiveness. A modern economy’s ability to attract industry, support artificial intelligence, electrify transport, secure supply chains, or withstand climate shocks depends upon the quality of its physical systems.
The distinction between economic policy and infrastructure policy is narrowing.
The distinction between national security and energy policy is narrowing too.
Governments have already begun responding.
The United States committed approximately $370 billion through the Inflation Reduction Act to accelerate domestic industrial and energy investment. Across Europe, the Green Deal Industrial Plan and associated initiatives are mobilising hundreds of billions of euros towards energy systems, manufacturing, and strategic technologies. Germany recently announced a €500 billion infrastructure and competitiveness package.
These initiatives are often presented as climate policies. They are equally about competitiveness, energy security, industrial resilience, and geopolitical positioning. What appears at first glance to be a collection of unrelated investment themes is, in fact, a single phenomenon” The return of physical constraints.
The Investment Gap
Yet a curious mismatch exists between where capital is most needed and where it is often allocated.
Financial markets remain heavily geared towards software, consumer technology, and financial assets. Many of the technologies required to strengthen physical systems occupy an awkward middle ground: Grid technologies, industrial electrification, Advanced materials, energy storage, industrial automation.
These sectors are often too capital-intensive for traditional venture capital investors and too early-stage for conventional infrastructure funds. As a result, a significant financing gap has emerged precisely where many of the most important innovations are occurring.
The scale of investment required is substantial.McKinsey estimates that achieving global net-zero objectives would require approximately $9 trillion of annual investment by mid-century—roughly three times current levels. Yet much of the financing challenge lies not in mega-projects but in the “missing middle”: industrial technologies and infrastructure-enabling businesses that have moved beyond the laboratory but have not yet reached institutional scale.
Europe provides a particularly striking example. The continent possesses world-class engineering capabilities and industrial expertise. Yet many companies struggle to scale because growth capital remains scarce. It is a shortage of capital willing to embrace complexity.
Beyond ESG towards princi
For much of the past decade, sustainable investing has been framed through the language of ESG. That framework has generated valuable discipline around risk, governance, and disclosure. However, will ESG manages risks and benefits, as if today there has not been a significant shift in the capital allocation principles that could ensure that Capital allocates itself along these principles. As macroeconomists, we blame distortions: Natural monopoly, public good characters and the like. Yet, we are st searching for a more practical organizing principle
Resilience may prove to be that principle. Unlike many sustainability concepts, resilience is immediately measurable and priceable: Can a power system continue operating during disruption? Can a supply chain absorb shocks? Can a city withstand extreme heat? Can an industrial facility remain competitive despite volatile energy prices. Can critical infrastructure recover quickly after a cyberattack or natural disaster?
These questions move beyond disclosure and into capability. The shift is also reflected in the economics of risk. According to Munich Re, annual global losses from natural catastrophes now regularly exceed $250 billion, while geopolitical disruptions, cyber threats, and supply-chain interruptions have become material concerns for corporate boards and investors alike.
The Next Trillion Dollars
The coming decade is likely to witness one of the largest reallocations of capital towards physical systems in modern economic history. This reallocation Partly will be driven by energy transition, by artificial intelligence, geopolitical competition and by climate adaptation.
But beneath these trends lies a common reality that economic growth remains constrained by the capacity of physical systems. Bloomberg NEF estimates that global investment in energy transition exceeded $2 trillion for the first time in 2024. Yet investment in electricity grids—the backbone of electrification, industrial competitiveness, and artificial intelligence—continues to lag far behind what is required. Building sufficient physical capacity to support a transition remains the challenge
The digital economy rests upon the laws of physics. The biggest investment opportunity of the next decade may not be another digital platform. It is the rebuilding of the physical foundations upon which the digital age depends.
The challenge facing advanced economies is therefore not simply technological innovation. The investors who prosper in the years ahead may be those who ask different questions. Not where growth will happen, but also which infrastructure will be required to make that growth possible.
The next trillion dollars will not be allocated solely to software; they will flow toward the infrastructure of resilience.
Read more on thought leadership at From Boardrooms to Market Stalls: Different Worlds, Same Reputation Lesson
