First Solar (FSLR.US), a leading US solar company, saw its shares plummet more than 17% in pre-market trading on Wednesday. While the AI infrastructure boom is fueling a surge in demand for green power, positioning renewable energy systems like solar as potential winners in the unprecedented expansion cycle of AI data centers, the company's annual sales forecast fell significantly short of Wall Street analysts' consensus expectations. This outlook was provided against a backdrop of uncertainty in US renewable energy and tariff policies, coupled with widespread permitting delays reminiscent of the prior administration.
Some analysts view First Solar's growth narrative as a "2027 story," with the current market showing little confidence in the company's near-term growth expectations. The largest US solar panel manufacturer's latest earnings report projected full-year 2026 net sales to be in the range of $4.9 billion to $5.2 billion. According to data compiled by LSEG, Wall Street analysts had on average expected approximately $6 billion, making the new guidance substantially lower than market consensus.
Under the current policy agenda supportive of traditional energy sources like oil and gas, the solar industry and other renewables face tariff pressures and a freeze on permits for large-scale renewable projects. The core energy agenda focuses on oil, natural gas, coal, and nuclear power, marking a departure from the green energy policies advocated by the previous administration.
First Solar executives stated during an earnings call on Tuesday that the company expects the total tariff impact this year to be between $125 million and $135 million, far exceeding market expectations. First Solar is a US-based manufacturer of photovoltaic modules and solar technology. Its key differentiation from other solar technology companies lies in its cadmium telluride (CdTe) thin-film modules, as opposed to the mainstream crystalline silicon approach, and it continues to expand its domestic US solar cell manufacturing capacity.
Solar energy is considered a core beneficiary of the "AI infrastructure super-cycle," but the benefits are more medium to long-term and require a "systems engineering" approach to be fully realized. Rising electricity demand from AI data centers significantly increases the certainty of new power generation and grid investments. Goldman Sachs Research concludes that data center power demand by 2030 could increase by approximately 220% compared to 2023 levels, substantially boosting demand for new power generation, including solar. Large AI data centers, similar to projects like Stargate, represent 24/7 high-reliability loads. Solar power's intermittency means it typically enters the power supply structure as a combination of "solar PPAs + energy storage/peak-shaving power sources + grid redundancy," making it a "critical piece of the puzzle" for data center power supply.
The growth story pointing to 2027 was highlighted by RBC Capital Markets analyst Christopher Dendrinos, who suggested that the 2026 outlook fell short of expectations due to curtailment activities. However, he characterized this as a "clearing event" or a potential catalyst for a rebound. Assuming no additional tariffs are imposed, this could position the company for a strong sales recovery next year. The company added in its earnings statement that demand for its Series 6 solar modules—designed for data centers or utility-scale solar plants and produced in large volumes in Malaysia and Vietnam—remains significantly constrained recently.
In response, First Solar will establish a new US-based precision processing line in South Carolina, expected to begin operations in the fourth quarter. This production line will utilize some front-end capacity from its Southeast Asian factories. The company anticipates this move will significantly optimize freight costs, tariffs, and the proportion of domestic US production, enabling the continued large-scale sale of incremental products to the US domestic market.
Citigroup analyst Vikram Bagri noted, "The market broadly understands that First Solar is a 2027 earnings growth story, with several positive catalysts expected along the way."
The AI infrastructure wave has already ignited a surge in green power demand. A cost report from IRENA cited by Morgan Stanley shows that the vast majority of new renewable projects in 2024 have lower overall costs than comparable fossil fuel solutions. This is especially true for solar and onshore wind, whose average levelized cost of energy continues to decline. Consequently, as AI significantly increases global data center power demand, lower-cost renewable energy sources like wind and solar are naturally prioritized.
In terms of new capacity additions and cost per unit of electricity, renewables may soon become the "first choice" for power system expansion in the AI era, supplemented by storage and nuclear power to balance the supply curve. In the AI age, a clean energy supply is becoming increasingly critical. The exceptionally strong demand for clean energy from large data center operators like Google and Microsoft stems from the global decarbonization trend, focusing on high efficiency, low cost, and the zero-emission, clean attributes of renewable resources like wind and geothermal, which are poised to be the most important sources for the global AI power ecosystem.
Wall Street giant Morgan Stanley recently published a report stating that the unprecedented expansion cycle of AI data centers is increasingly linked directly to clean power supply and grid integration based on renewables. However, First Solar's stock plunged primarily because its "near-term cash flow/sales visibility" was disrupted by policy and project execution risks. The market is more inclined to price it as a "2027 recovery story" rather than a pro-cyclical renewable energy stock that benefits immediately from the AI power trend.
The company's 2026 net sales guidance of $4.9–$5.2 billion was significantly below consensus, with reasons cited including US policy uncertainty and widespread permitting/project advancement delays. Concurrently, the company expects a tariff impact of approximately $125–$135 million, and demand/deliveries for its US-market-bound Series 6 modules produced in Malaysia and Vietnam remain constrained by capacity.
Morgan Stanley's report referenced the case of Google's parent company, Alphabet, acquiring Intersect Power, pointing out that tech giants need to partner with renewables to meet 24/7 clean electricity demands. Google plans to mobilize $20 billion in renewable energy investments by 2030, with Amazon and Microsoft having similar plans. Google also plans to invest approximately $40 billion by 2027 to build three new hyperscale AI data center campuses in Texas. Morgan Stanley specifically highlighted that one campus is a co-located "solar + battery storage" project developed in partnership with Intersect.
Morgan Stanley's analyst team stated that structural incremental power demand from AI data centers is shifting "clean power + storage + grid integration" from a thematic investment to an asset allocation issue. This is particularly true for infrastructure-type renewable assets with contracted cash flows at their core, which are better evaluated using a "target price + dividend + total return" framework to depict the risk-reward ratio. Consequently, such assets are more likely to experience valuation recovery during periods of interest rate cuts and improving risk appetite.