The US Polysilicon Paradox: Reshoring Capacity Does Not Guarantee Domestic Demand

Semiconductor demand is not large enough by itself

Polysilicon sits unusually far upstream from the companies investors normally associate with semiconductors. Electronic-grade material is converted into monocrystalline ingots and semiconductor wafers before reaching companies such as TSMC, Samsung Electronics, Intel and Micron. Solar-grade polysilicon follows a parallel path through ingots, wafers, cells and modules.

That distinction matters because the semiconductor market is the technologically demanding part of the business, but not the volume anchor. The White House's August polysilicon proclamation said semiconductor-grade material represents only 2.4% of global polysilicon production. It explicitly concluded that producers increasingly need larger solar-grade volumes to maintain competitive unit costs across their operations. [2]

This creates the paradox. Washington may value domestic polysilicon primarily because advanced chip production requires extremely pure material, but the plants making that material can still depend economically on solar customers. A semiconductor reshoring boom can therefore coexist with weak economics at a domestic polysilicon facility if the plant cannot secure enough solar volume.

The new trade regime does not fully eliminate that problem. Beginning in December, the US plans a minimum import price of $21 per kilogram for polysilicon, $100 per kilogram for ingots and wafers, $0.22 per watt for solar cells and $0.38 per watt for modules, along with tariffs on downstream derivatives. But the rules do not simply require imported wafers, cells or modules to contain US polysilicon. [2]

That distinction changes the competitive map. A foreign wafer producer can still source foreign polysilicon. A solar manufacturer can still build its upstream chain outside the United States. The policy raises the economics of importing the finished or intermediate product, but it does not automatically redirect the underlying polysilicon purchase to Tennessee or Michigan.

The buyer structure matters more than the capacity number

Hemlock Semiconductor shows what a more protected domestic position looks like. Corning has vertically expanded from its Hemlock polysilicon operation into US solar ingot and wafer production. In 2025 Corning increased the investment in its Michigan solar manufacturing expansion to $1.5 billion, explicitly building that operation around Hemlock's domestic polysilicon base. [3]

Reuters noted that this internal downstream buyer makes Hemlock more insulated than Wacker from the current policy problem. [1] Economically, that is the critical difference. Hemlock does not have to rely entirely on independent wafer manufacturers choosing its material over lower-cost foreign alternatives. Part of the customer relationship has been internalized.

Wacker has a different structure. Its Charleston plant competes inside a global Wacker production network that also includes major German capacity. Foreign polysilicon can also reach US demand through wafers and other downstream products manufactured abroad. Reuters identified Chinese competition, Wacker's own German plants and Oman-based United Solar Polysilicon as parts of the alternative supply picture. [1]

The exact identities of Charleston's two recently lost customers were not confirmed publicly by Reuters, so they should not be guessed. What matters for investors is the purchasing layer those customers occupy. Polysilicon producers sell into wafer manufacturing, not directly into a TSMC fab or an Nvidia accelerator. A reshored semiconductor fab can increase US chip output without creating an equivalent increase in demand for US polysilicon if its silicon wafers are still produced abroad.

Altsets relationships illustrate how diversified the next layers of the chain can become. Shin-Etsu Chemical supplies TSMC, Samsung Electronics, Intel and Micron. TSMC represents 4.02% of Shin-Etsu revenue while Shin-Etsu represents 1.33% of TSMC COGS. Samsung represents 2.43% of Shin-Etsu revenue and Shin-Etsu represents 0.29% of Samsung COGS. Intel represents 1.79% of Shin-Etsu revenue and Micron 1.83%.

These are company-level relationships and do not prove that a particular batch of polysilicon flows through each connection. They do show something economically important about the downstream market: a major silicon materials supplier can serve several large chipmakers without any one of them dominating its revenue base.

That diversification gives the wafer and materials layer procurement flexibility. For a merchant polysilicon producer, the dependency can run in the opposite direction. Losing only a small number of qualified wafer customers can materially reduce plant utilization even when semiconductor demand downstream remains strong.

The second-order investment exposure

For Wacker Chemie, the US semiconductor buildout is therefore not automatically a bullish demand signal for Charleston. The relevant variables are customer qualification, plant utilization, solar-grade volume and whether future trade rules create an explicit economic advantage for polysilicon made in the United States.

Corning has a structurally different exposure. Its ownership of Hemlock is increasingly connected to its own domestic solar wafer manufacturing. That vertical integration means the company is not only a seller of a strategic upstream material. It is also becoming one of that material's downstream buyers.

The same framework matters when evaluating wafer manufacturers such as Shin-Etsu, SUMCO, GlobalWafers and Siltronic. These companies sit between polysilicon producers and semiconductor fabs. If qualified foreign polysilicon remains available, they can preserve sourcing alternatives even while the United States spends heavily to reshore chip fabrication.

That is the US polysilicon paradox. The material is strategically indispensable, domestic semiconductor investment is rising, and policymakers are explicitly trying to protect the industry. Yet an American polysilicon plant can still struggle because the economic bottleneck sits one layer downstream.

For investors, domestic production capacity is the wrong endpoint to track. The more useful question is whether domestic polysilicon has a committed domestic route into ingots and wafers. Where that route exists, as Corning is building around Hemlock, policy support can translate into durable demand. Where it does not, foreign alternatives and mobile wafer buyers can leave even strategically important US capacity underutilized.

For Altsets research conventions and interpretation limits, see the research methodology.

Sources

1. "Trump's bid to shield chip supply chain could backfire in Tennessee," Reuters, September 4, 2026. https://www.reuters.com/business/trumps-bid-shield-chip-supply-chain-could-backfire-tennessee-2026-09-04/

2. "Adjusting Imports of Polysilicon and its Derivatives into the United States," The White House, August 6, 2026. https://www.whitehouse.gov/presidential-actions/2026/08/adjusting-imports-of-polysilicon-and-its-derivatives-into-the-united-states/

3. "Corning Expands Advanced Manufacturing Capacity To Meet Increased Demand for U.S.-Made Solar Products," Corning Incorporated, April 29, 2025. https://www.corning.com/worldwide/en/about-us/news-events/news-releases/2025/04/corning-expands-advanced-manufacturing-capacity-to-meet-increased-demand-for-US-made-solar-products.html

How to Cite This

For research inquiries or data access: press@altsets.com

Go Deeper