The four phases of the AI trade: How to invest in the new market super cycle

Phase 1 – The core enablers

Everything begins with chips. Nvidia, AMD, TSMC, and ASML manufacture the processors and lithography equipment that allow AI models to run. These companies form the base layer of the ecosystem, supplying the computing power for every AI application in use today.

The market treated them as growth stocks in 2023 and 2024, but they are also capital-intensive businesses tied to cyclical supply chains. Taiwan and South Korea dominate advanced chip fabrication, while the U.S. and Europe are racing to secure their own capacity. For investors, this means opportunity mixed with geopolitical risk.

Even after rapid share-price gains, demand remains firm. Each new model requires more processing power, and the training cycles are getting longer. Investors can gain diversified exposure through semiconductor ETFs or focus on key suppliers if they can tolerate volatility. Hardware is cyclical, but the need for computing power is structural.

Phase 2 – The infrastructure builders

Once the chips exist, they need power and data capacity. Cloud giants Microsoft, Amazon, and Alphabet are spending tens of billions on new data centres. Energy producers and grid operators are quietly becoming the next beneficiaries of the AI boom. Utilities like RWE, NextEra Energy, and Iberdrola, as well as data-centre REITs Equinix and Digital Realty, are seeing record demand.

Power has become the true bottleneck. A single large training run can consume as much electricity as a small city. That pressure has triggered new investment in renewables, battery storage, and transmission upgrades. Analysts expect data-centre power usage to double this decade.

Investors who want a steadier entry point into the AI theme can use infrastructure or utility ETFs. These provide exposure to the physical backbone of the digital economy without the valuations attached to pure tech stocks.

Phase 3 – The software monetisers

In this stage, AI moves from cost to profit. Companies such as Adobe, Salesforce, Palantir, and ServiceNow are embedding AI into their platforms to boost productivity for clients and pricing power for themselves. The winners here are those converting innovation into repeatable revenue.

Markets are beginning to separate noise from evidence. The best signal is customer adoption—are firms willing to pay more for AI-enhanced tools? Early data suggests yes. Enterprise clients are upgrading subscriptions and reporting efficiency gains of 10–20 percent in routine tasks. That kind of proof attracts long-term capital.

Broad software ETFs focused on automation and analytics can capture this theme, spreading risk across companies with different approaches and product cycles.

Phase 4 – The productivity adopters

The fourth phase extends AI’s reach into the wider economy. Banks, retailers, and healthcare groups are integrating AI into daily operations. JPMorgan uses it for fraud detection, Walmart for inventory logistics, and UnitedHealth for patient analytics. These gains are incremental but lasting.

This stage is where AI shifts from novelty to necessity. The productivity lift feeds directly into earnings. Research from McKinsey suggests that AI could add more than one percentage point to annual global productivity growth through the next decade. For investors, that means broader participation across sectors, not just technology.

Exposure here comes through diversified equity funds or thematic ETFs focused on efficiency and automation. The story moves from building to using.

Market drivers behind the cycle

Valuation and liquidity cycles are shaping the speed of each phase. After the sharp rally in 2023, markets spent 2024 digesting gains while corporate investment in AI infrastructure kept climbing. Global tech capex is now growing faster than overall GDP, with roughly half of it linked to AI.

Investor sentiment remains strong but more selective. The market now rewards proof of earnings rather than announcements. Regulation is tightening across regions, and interest-rate expectations continue to guide sector rotations. AI has moved from concept to competition, and that change will determine which firms stay ahead.

Building an AI portfolio that lasts

AI is no longer a trade—it’s a theme that will evolve over many years. To stay invested through the full cycle, investors can layer exposure across the four phases.

Mix early- and late-stage beneficiaries

Hardware and infrastructure provide cyclical upside, while software and adopters smooth returns as the theme matures.

Diversify by geography

The U.S. leads in innovation, Asia in chip manufacturing, and Europe in energy infrastructure. A global approach captures different drivers of return.

Manage concentration risk

The largest technology stocks now represent an outsized share of global indices. Balanced exposure helps avoid over-reliance on a handful of mega-caps.

Use ETFs strategically

Thematic and sector ETFs offer targeted access while maintaining liquidity. Combining semiconductor, infrastructure, and automation ETFs can create a simple multi-phase structure.

Revisit allocations regularly

The AI cycle moves quickly. Hardware booms may pause while adoption accelerates elsewhere. Periodic rebalancing keeps portfolios aligned with where growth is shifting.

Outlook for the next stage

The market is entering a transition period: the first wave of investment built capacity, and the next will measure return on that spend. Energy, infrastructure, and productivity metrics will start to matter as much as model innovation. Countries able to secure power and chips will have a strategic advantage.

Long-term investors can expect volatility but also compounding growth. AI is feeding into capital formation, corporate margins, and policy decisions. It is no longer a side story—it’s part of the market’s core narrative.

Closing thoughts

AI is now embedded in global business, from code to construction. The opportunity is not confined to Silicon Valley; it stretches across supply chains, utilities, and service sectors. Investors who think in phases (hardware, infrastructure, software, adoption) can stay positioned through the full supercycle and avoid chasing short bursts of excitement.

The trade is not about guessing the next breakout stock. It’s about owning the companies building, powering, and using the technology that will define this decade’s growth.