I have been observing investment opportunities in energy for 2026, and I have found that the market logic has completely changed. In the past few years, discussions were still about electric vehicle subsidies and solar overcapacity, but now the focus has shifted entirely to AI power demand and grid upgrades.

The core change is this: AI data centers have become the true drivers of electricity demand. According to forecasts from the IEA and Goldman Sachs, global data center electricity consumption will surge from 460 TWh in 2022 to 1,050 TWh in 2026, with more than half contributed by AI-related activities. The power consumption of training a large AI model can reach several thousand MWh, equivalent to the annual electricity use of tens of thousands of households. This is not hype; it is a rigid demand.

Traditional wind and solar energy face a critical problem: intermittency. AI data centers require 24/7 uninterrupted power supply, so tech giants like Microsoft, Amazon, and Google have begun massive investments in nuclear energy starting in 2025. Amazon plans to deploy 12 small modular nuclear reactors, and Google has committed to tripling its nuclear capacity by 2030. The logic behind this is clear—the power companies with nuclear and natural gas assets will receive the highest premiums.

But that’s not all. The real bottleneck lies on the transmission side, not the generation side. The global power grid is severely aging, with high-voltage transformers now taking 2-3 years for delivery, and the supply-demand gap is expected to persist until at least 2027. This directly drives power companies’ revenue growth from 1% to 4-6%.

There are also many opportunities in Taiwan’s renewable energy concept stocks. Delta Electronics, as a leader in power electronics, saw explosive growth in AI server orders by 2025, and electric vehicle electronics are also taking shape. Huasheng Electric, a long-term partner of Taipower, benefits directly from the NT$564.5 billion grid upgrade plan and is leading the charging station industry. Companies like United Renewable Energy and Yuanjing, solar panel manufacturers that experienced overcapacity, are now entering a phase of cost reduction and demand recovery. Wuxi’s wind turbine blade materials hold a high market share, with order backlog exceeding NT$10 billion in 2026.

On the US stock side, Constellation Energy is the largest nuclear power operator in the US, having signed a 20-year contract with Microsoft, and its data center projects are expected to expand significantly in 2026. Oklo’s micro nuclear reactor technology is low-cost and quick to deploy, with explosive potential amid AI power shortages. Eaton and GE Vernova are core beneficiaries of grid upgrades, with transformer demand directly driven, high gross margins, and long order visibility. NextEra Energy, as the largest renewable energy company in the US, leads in wind and solar capacity, with stable dividends, making it a core defensive play in green energy.

In terms of investment logic, it is recommended that AI power stocks constitute 50-60% of the portfolio (high growth, high volatility), traditional energy stocks account for 30-40% (stable defense), and the remaining cash or bonds serve as buffers. Because renewable energy stocks are volatile, avoid chasing highs; look for short-term dips within the long-term upward trend as buying opportunities.

The key is to monitor leading indicators: tech giants’ AI capital expenditures, scale of grid investments, order backlogs, and technological iteration progress. Renewable energy concept stocks are not just hype but are driven by order certainty and rigid demand. From 2026 to 2030, in the context of the AI era and net-zero transition, renewable energy is the most promising structural opportunity worth deep cultivation.