Abstract

AI, energy, defence, and supply-chain re-architecture are not separate trends — they are the interlocking load-bearing columns of Asia’s industrial supercycle. Each deserves a close look.

The argument for Asia’s industrial supercycle rests not on a single thesis but on four distinct capital expenditure waves, each with its own investment logic and timeline, all converging on the same geography at the same moment in history. The risk of treating them as a single monolithic trend is that it obscures their individual dynamics — and their individual vulnerabilities. The risk of treating them in isolation is that it misses the systemic amplification effects that make the cycle so powerful. What follows is an attempt to do both: to examine each pillar closely, and then to understand how they reinforce one another.

Pillar I — Technology Infrastructure

The AI Compute Imperative

The race to build AI infrastructure is the largest voluntary capital expenditure cycle in corporate history — and Asia holds most of the keys.

• $700B+ Global AI capex announced 2025–27
• 92% Advanced logic chips from TSMC
• 74% HBM memory from Korea
• 38GW New data centre power demand by 2028

When the four American hyperscalers — Microsoft, Google, Amazon, and Meta — announced a combined capital expenditure budget exceeding $320 billion for fiscal year 2026, the technology press focused on the dollar figure. The more revealing story was in the supply chain. Every advanced GPU in those data centres was built on TSMC silicon. Every high-bandwidth memory stack was produced by Samsung or SK Hynix. The silicon wafers came from Japanese suppliers. The specialty gases and photoresists came from chemical companies in Japan and South Korea. The advanced packaging — the technology that stacks chips together to create the AI accelerators — was done in Taiwan.

The TSMC effect TSMC’s capital expenditure alone exceeded $40 billion in 2025. Its expansion into Japan (Kumamoto) and the United States (Arizona) is creating secondary demand for Japanese and Korean equipment makers — companies like Tokyo Electron, Shin-Etsu, and ASML’s Korean partner base.

This is not a transient dependency. The process of building a leading-edge semiconductor manufacturing ecosystem takes, at minimum, fifteen years and tens of billions of dollars in sustained investment. South Korea’s Samsung and SK Hynix have been building their DRAM and NAND expertise for four decades. TSMC’s manufacturing processes embody thirty years of continuous refinement. The idea that this capability can be replicated at speed outside Asia is the most consequential piece of industrial magical thinking of our time.

The data centre land rush

The AI infrastructure buildout is not only about chips. Data centres require land, power, water, and fibre connectivity — and Asia’s governments have been faster than their Western counterparts to designate and permit the industrial zones, transmission corridors, and subsea cable landing rights that make large-scale data centre clusters possible. Singapore, despite its land constraints, remains Asia’s primary data centre hub. Malaysia’s Johor corridor has attracted over $60 billion in commitments from American hyperscalers. Japan’s Osaka and Tokyo regions are seeing data centre construction at their fastest rates in twenty years.

The power requirement is the binding constraint. A modern large-language-model training cluster running continuously draws between 100 and 500 megawatts — comparable to a small city. This is driving an intimate connection between AI infrastructure capex and energy infrastructure capex, the two pillars reinforcing each other through the simple physics of electricity consumption.

“Every AI model trained is a vote for more copper wire, more transformer steel, more silicon carbide power electronics — almost all of which is made in Asia.”— Dr. Chen Wei, Head of Technology Equity Research, Nomura Securities

Pillar II — Energy Infrastructure

The Energy Transition as Industrial Policy

Asia is not merely participating in the global energy transition — it is manufacturing it. And in doing so, it is generating a self-sustaining loop of industrial demand.

• 80% Global solar manufacturing in Asia
• $620B Asian grid investment 2025–30
• 54 Nuclear reactors under construction in Asia
• 65% Global EV battery capacity

Asia’s energy story operates on two levels that are easy to conflate but important to distinguish. The first is the manufacturing-for-export story: China produces roughly 80 percent of the world’s solar panels, 70 percent of its wind turbines, and 60 percent of its lithium-ion batteries. These industries generate employment, trade surpluses, and industrial learning-by-doing — but they are increasingly subject to protectionist measures from Western governments that limit their export markets.

The second story is more structurally important: Asia’s domestic energy transition. The continent’s electricity demand is growing at approximately three to four percent annually — nearly three times the rate of Western Europe and the United States. Meeting this demand while reducing dependence on imported fossil fuels is an existential strategic priority for every significant Asian economy. The investment required is staggering.

The nuclear revival

Nuclear resurgence Japan has approved the restart of 14 nuclear reactors since 2023, with more expected. South Korea has reversed its nuclear phase-out entirely, commissioning new APR1400 units. India has 22 reactors under construction. Each reactor requires 50,000+ tonnes of specialised steel and years of precision engineering.

Perhaps the most underappreciated energy story in Asia is the nuclear revival. After Fukushima, the conventional wisdom was that Asia would trend away from nuclear power. The opposite has happened. Japan has approved the restart of more reactors in the past two years than in the previous decade. South Korea has formally abandoned its nuclear phase-out and contracted for new builds. China has 22 reactors under construction, with plans for dozens more. India’s nuclear programme is accelerating. Each reactor represents a capital investment of $7–12 billion, a construction timeline of eight to twelve years, and sustained demand for specialised steel, zirconium alloys, control systems, and precision-engineered components — almost all of which are made in Asia.

Grid infrastructure: the overlooked capex story

The most unglamorous but arguably most important energy investment is in transmission and distribution grids. Renewable energy is intermittent and often located far from demand centres; connecting it to consumers requires the largest expansion of high-voltage direct current transmission infrastructure in history. This is a copper-intensive, transformer-intensive, civil-engineering-intensive undertaking. Asia’s grid investment over the next five years is estimated at $620 billion, and the suppliers of the equipment — transformer manufacturers like Hitachi Energy (Japan), ABB’s Asian operations, and a host of Chinese electrical equipment makers — are running multi-year order backlogs.

Pillar III — Defence & Security

The Reluctant Rearmament

Asia’s strategic environment has changed irreversibly. Governments that spent decades suppressing defence spending are now racing to rebuild — or build from scratch — sovereign industrial defence capabilities.

• $640B Asia-Pacific defence spending 2025
• 2.0% Japan GDP defence target (2027)
• 19 Asian nations raising defence budgets
• $14B Korea arms exports 2025

The defence capex cycle in Asia has a different character from its counterparts in Europe. European rearmament is largely a procurement story — governments buying platforms and munitions that are primarily manufactured in the United States or, at best, in large European primes. Asian rearmament is a manufacturing story. The strategic ambition driving it is explicitly about developing sovereign capability — not merely buying weapons, but building the industrial base to design, produce, and maintain them.

Japan’s industrial transformation

Japan deserves particular attention because its rearmament represents the reversal of a strategic posture maintained for eighty years. The doubling of Japan’s defence budget to two percent of GDP is not a marginal adjustment — it is a fundamental transformation of one of the world’s largest economies’ relationship with its own military-industrial base. Japan’s shipbuilding, aerospace, and electronics industries, which have maintained dormant dual-use capabilities throughout the post-war period, are now being explicitly mobilised for defence production. Mitsubishi Heavy Industries, Kawasaki Heavy Industries, IHI Corporation, and Fujitsu Defence are all expanding at a pace not seen since the 1970s.

“South Korea’s defence industry in 2026 resembles South Korea’s consumer electronics industry in 1986 — technologically competitive, cost-disciplined, and on the cusp of global dominance.”— Marcus Tanner, Senior Defence Analyst, IISS

Pillar IV — Supply Chain Architecture

The Great Diversification

The post-pandemic redesign of global supply chains is the most sustained greenfield industrial investment cycle in Asia since the original China manufacturing boom of the 1990s.

• $85B FDI into Vietnam 2024
• $47B Apple India manufacturing by 2027
• +220% Indonesia FDI growth 2021–25
• 340M India working-age population by 2035

The phrase “China plus one” entered business vocabulary around 2020. By 2026 it is more accurately “China plus three or four” — a systematic diversification of production footprint across a complex web of Asian manufacturing locations, each developing specialised niches within the broader supply chain. This is not a simple story of production migrating from China to cheaper alternatives; it is a more complex story of supply chain architecture — of deliberate design choices about where to place different manufacturing steps, based on cost, capability, political risk, and proximity to end markets.

The India factor

India’s emergence as a serious manufacturing destination is the most consequential industrial story of the mid-2020s. The Production Linked Incentive scheme — which provides financial incentives to manufacturers who achieve specified production thresholds — has attracted commitments across fourteen sectors, from semiconductors to pharmaceuticals to electronics assembly. Apple’s decision to assemble a growing share of its products in India is not merely a supply-chain hedge; it is a signal to the entire contract manufacturing ecosystem that India is open for serious industrial business.

The structural underpinning of India’s manufacturing ambition is demographic. India’s working-age population will reach 340 million by 2035, making it the world’s largest reservoir of young, trainable industrial workers. Combined with increasingly reliable power infrastructure, improving logistics connectivity, and a legal system that, while slow, provides property rights protections that some competing locations cannot, India’s long-term manufacturing proposition is compelling.

Southeast Asia’s diversification

Vietnam has absorbed the largest single share of manufacturing displaced from China, becoming a major hub for electronics, footwear, and garment production. Its proximity to China’s supply chains, combined with its relatively low wages and stable government, has made it the default second choice for many Chinese supply-chain-dependent industries. Indonesia is pursuing a different strategy: leveraging its extraordinary natural resource endowment — the world’s largest nickel reserves — to position itself as the centre of the global battery supply chain, insisting that raw materials be processed domestically before export. Malaysia has emerged as a sophisticated semiconductor back-end and advanced packaging hub, attracting investment from Intel, Infineon, and a host of OSAT (outsourced semiconductor assembly and test) companies.

How the Four Pillars Interact

The most important analytical point about these four pillars is not their individual scale — though each is historically significant — but the way in which they interact and amplify one another. Understanding these interactions is the key to understanding why this cycle has the characteristics of a genuine supercycle rather than a conventional boom.

AI data centres require electricity, which requires grid investment, which requires copper, transformers, and power electronics. Defence electronics require advanced semiconductors, which require investment in chip manufacturing, which requires the same precision chemicals and equipment that support AI chip production. Supply chain diversification across Southeast Asia requires infrastructure — roads, ports, power grids, industrial zones — which requires construction equipment, steel, and engineering services. Every loop feeds back into the others, sustaining demand across the entire industrial complex for longer than any single investment cycle would on its own.

This is the defining characteristic of a supercycle: not the scale of any individual investment decision, but the self-sustaining nature of the demand it creates. Asia’s four pillars are now generating precisely this kind of structural, mutually reinforcing demand. The next article examines what this means for investors — where the opportunities are greatest, and where the risks lie hidden.

Data cited represents analyst consensus estimates and publicly disclosed figures as of Q1 2026. Article II of III in the Asia Industrial Supercycle series. ← Article I: The Big Picture Article II — The Four Pillars Article III: Investing the Cycle →

Other People are Reading