Recently exploring the new energy sector, I found that the investment logic for 2026 is completely different from the past few years. Previously, new energy stocks relied on policy subsidies and capacity expansion; now, it has become a demand driven by AI.

It may sound a bit exaggerated, but the data is there. AI training and data center electricity consumption are expected to surge from 460 TWh in 2022 to 1,050 TWh in 2026, with more than half contributed by AI-related parts. The electricity used to train a large AI model is equivalent to the annual electricity consumption of tens of thousands of households. This is not hype; it’s a real power demand.

The key is that traditional wind and solar energy cannot provide stable 24-hour power, so tech giants are competing for nuclear energy resources. Microsoft, Amazon, and Google are heavily investing in nuclear power and small modular reactors between 2025 and 2026. Goldman Sachs predicts that by 2030, data centers’ demand for nuclear energy will reach several tens of gigawatts.

But there is a bottleneck many overlook—power generation is easy, but transmission is difficult. The global power grid is severely aging, with delivery times for high-voltage transformers and switchgear reaching 2-3 years. Major manufacturers have already invested billions of dollars to expand capacity, yet supply still falls short. This is actually the best investment opportunity because manufacturers of grid equipment have high profit margins and long order visibility.

Furthermore, the long-term goal of global net-zero emissions remains unchanged. Renewable energy will still account for nearly 50% of global electricity by 2030. So, the investment logic for new energy now runs on a dual track—on one side, the short-term explosive growth driven by AI electricity demand; on the other, the long-term steady growth from net-zero transition.

In Taiwan stocks, I favor Delta Electronics, Huacheng Electric, and United Renewable Energy. Delta is a leader in power electronics, with high-power-density AI servers directly boosting orders. Huacheng is a long-term partner of Taiwan Power, benefiting from the NT$5.645 trillion grid upgrade plan, and is also a leader in the charging station market. United Renewable Energy is a solar cell leader, benefiting from anti-dumping tariffs and technological upgrades in Europe and the US in 2026. Shangwei’s wind turbine blade material backlog exceeds NT$564.5B, with an expected revenue growth of 18%. Yuanjing is a major solar module manufacturer, known for good cost control and stable dividend policy.

On the US stock side, Constellation Energy is the largest nuclear power operator in the US, with a 20-year contract with Microsoft, offering stable cash flow and attractive dividends. Oklo is a pioneer in micro nuclear reactors, supported by Sam Altman, with NRC approval progress leading into 2026. Eaton is a leader in grid automation, with orders surging and delivery times extended to 24 months. GE Vernova, spun off from GE, is in the power grid and generation business, with order backlog reaching new highs. NextEra Energy is the largest renewable energy company in the US, with stable dividends and strong long-term defensive qualities.

Investing in new energy requires patience. It is recommended that AI power stocks make up 50-60% of your portfolio, aiming for high growth but accepting volatility; traditional energy stocks account for 30-40% for steady defense; the remaining 10% in cash or bonds as a buffer. Since new energy stocks are volatile, it’s best to add on dips during long-term upward trends rather than chasing highs. Key indicators to monitor include AI capital expenditure, grid investment scale, and order backlog—these are leading indicators.

From 2026 to 2030 is the most promising structural opportunity window for new energy. Every bear market low has historically been the start of a long-term bull run, and this time is no different.