Polysilicon capacity acquisition: The Emperor’s New Clothes

Since first reports surfaced in mid-May 2025, the planned fund of major Chinese polysilicon manufacturers to buy up and shut down excess production capacity has spurred fantasies about reduced oversupply and rising prices. Now a Chinese author, writing under the pseudonym of “Polysilicon Industry Information,” has thoroughly picked the plan apart in a contribution on the Chinese platform WeChat. Although Bernreuter Research does not agree with all points of the analysis, we find it worth to publish a slightly shortened English translation.

By Polysilicon Industry Information

The photovoltaic industry’s polysilicon sector is mired in a controversy surrounding “stockpiling.” On the one hand, the industry is facing the challenges of overcapacity and price pressure; on the other, rumors of a “stockpiling rescue” plan are swirling, even catching the international media.

This supposed “salvation,” so highly anticipated by some, is in reality a classic case of the “Emperor's New Clothes" in the polysilicon acquisition sector – a seemingly glamorous “industry self-rescue” cloaked in emptiness, with its underlying logic failing to even support basic logic.

I. Stockpiling: Seemingly a rescue measure, but in reality a shell company in new clothes

On the surface, the logic behind stockpiling polysilicon seems sound – leading companies acquire outdated production capacity, thereby reducing market supply and alleviating downward price pressure. However, a deeper analysis reveals that this plan fundamentally violates market principles and is far from a lifeline for the industry. Rather, it resembles the superficial Emperor's New Clothes, a hype-fueled shell without a practical core.

First, it runs counter to the very essence of supply-side reform. The core of every successful supply-side reform in history has been eliminating backwardness and supporting advanced industries, eliminating inefficient assets and concentrating resources on high-quality enterprises.

However, stockpiling polysilicon requires leading companies with advanced technology and controllable costs (such as Tongwei) to provide a transfusion to inefficient, near-obsolete P-type polysilicon production capacity. This is tantamount to asking top students to pay for the mistakes of underperforming students. It’s essentially moral coercion, failing to optimize the industrial structure and instead solidifying excess capacity and delaying industry restructuring.

Secondly, the current state of the industry doesn’t require “stockpiling” to support it. Current polysilicon production capacity data is already skewed: The nominal capacity is 3.8 million tons, including planned production and idled capacity; actual year-round operating capacity is only 2.3 million tons, with an overall operating rate of 56.5%.

Domestic polysilicon demand is approximately 1.4 million tons annually, and the majority of the 900,000 tons of excess capacity is inefficient, with outdated technology and high costs. This capacity should naturally be eliminated through market competition, not prolonged by stockpiling. Pushing for stockpiling is merely a way to forcefully dress up outdated capacity, masking its inevitable market elimination.

More crucially, stockpiling lacks a sustainable business logic. It creates no real value; it merely transfers production capacity from inefficient enterprises to stockpiling entities, requiring continuous capital injections to maintain. If funding runs out, not only will the stockpiled capacity become a burden, but it could also trigger new industry turmoil, completely contradicting the original purpose of the “rescue” measure. This “new garment” lacks even the most basic fabric durability, destined to be a short-lived hype.

II. Stockpiling rumors: The hype logic of “new clothes” under capital manipulation

Once the guise of a “market rescue” is removed, the capital-trap nature of the polysilicon stockpiling rumors becomes increasingly clear, revealing the underlying manipulation of capital. The so-called “industry self-rescue” is little more than a gimmick fabricated by capital for arbitrage. Inferring from the perspective of profit, this farce is more likely the result of the meticulous planning of two forces.

The first category involves photovoltaic companies eager to divest and cash out. With the PV industry’s performance generally under pressure in the second quarter of 2025, some shareholders faced significant pressure to reduce their holdings. The “silicon stockpiling” incident was a compelling “new coat” – as long as rumors spread, it could temporarily boost stock prices, creating conditions for shareholders to flee at high levels.

The most direct evidence is that after shareholders of companies like Canadian Solar completed their divestment plans in mid-September, the previously widespread rumors of stockpiling quickly vanished, a highly consistent timing. Once the purpose of the “dressing” (divestment and cashing out) was achieved, the “new coat” was naturally discarded, without even a trace of attachment.

The second category involves polysilicon futures investors. With the launch of polysilicon futures on the Guangzhou Futures Exchange, some investors began artificially manufacturing market expectations to amplify price fluctuations and profit. Rumors of stockpiling were precisely their “new clothes” for the market – spreading rumors of capacity reductions and price increases to lure investors into following suit, and then profiting by buying low and selling high in the futures market.

When polysilicon futures prices fell to around 50,000 yuan per ton in mid-to-late September and capital outflows continued for four consecutive days, the rumors of stockpiling subsided, further confirming the evidence of capital manipulation. For investors, the value of these “new clothes” lasts only during the hype period; once they become unprofitable, they are quickly discarded without hesitation.

It’s worth noting that this rumor has long been debunked by authoritative sources: Tongwei explicitly warned investors not to believe rumors, and the China Photovoltaic Industry Association publicly clarified in July that the polysilicon stockpiling allegations were completely fabricated. Even so, the rumor has continued to garner global attention, demonstrating the power of the capital behind it and exposing the importance of the “new clothes” to certain stakeholders. Even knowing the rumors are false, they still use them to generate momentum for short-term profit.

III. Stockpiling failed: The September agreement was broken, the new clothes finally showed holes

Since July, rumors have circulated in the market that a purchase agreement would be signed in mid-September and a reorganized company would be established by the end of September. Now, there has been no concrete progress on the purchase. This broken promise is no accident; the “Emperor's New Clothes” of the polysilicon acquisition ultimately revealed flaws. It lacked any real fabric, merely a facade propped up by hype. Over time, five core contradictions have erupted, leaving the holes in the new clothes growing bigger and bigger.

1. The deadlock in the game of interests: Even the new clothes’ stitches can’t be sewn together

The core of stockpiling is pricing, but the expectations of the acquirers and the companies to be acquired are so clashing that even the basic stitching of the new garment is out of sync. Leading companies believe that outdated production capacity should be acquired at a 50% to 70% discount on net asset value – after all, this capacity primarily consists of P-type polysilicon, which has become obsolete due to technological advancements, and its value has significantly decreased after depreciation.

Meanwhile, the second- and third-tier companies to be acquired insist on offering a 30% to 40% discount; some even assert: “The stockpiling is the result of the leading companies begging us to exit, so why should we lower the price?" This assertion is unwavering. Especially after the spot price of polysilicon rebounded to 50,000 yuan/ton in September , small and medium-sized factories, feeling they still had room for profit, were unwilling to sell at a low price. Negotiations deteriorated to a stalemate, with even the most basic stitching unable to be completed.

2. The funding gap is difficult to fill: Even the fabric for the new clothes is hard to come by

The scale of stockpiling and the capital requirements far exceeded expectations. Industry estimates suggest that achieving supply-demand balance requires the withdrawal of 1.5 million tons of production capacity. Even at a 50% discount, the associated costs for equipment, land, and personnel placement would total 50 - 70 billion yuan [7 - 10 billion US$]. Even with leverage (e.g., bank loans), leading companies would still need to raise 15 - 20 billion yuan [2.1 - 2.8 billion US$].

However, polysilicon companies currently have high debt-to-asset ratios (some exceeding 70%), and there is an unresolved debate over whether to allocate capital based on production capacity or sales quotas. Tongwei argues that its high market share stems from its technology and efficiency and that costs should not be allocated excessively, while GCL advocates that capital allocation based on production capacity is fairer.

No company is willing to shoulder the enormous costs alone. Even if GCL and other companies accumulated some capital through private placements, it would be far from sufficient to cover the overall shortfall. The “new garment” project is simply a matter of time, with no implementation possible.

3. Policy shift to substitution: Finding that new clothes are unnecessary

On September 16, the Standardization Administration of China (SAC) released a draft for public comment on the “Energy Consumption Limits per Unit of Polysilicon and Germanium Product.” This directly negates the logic of “stockpiling to prevent backwardness” and makes it clear to the market that this stockpiling policy is completely unnecessary.

The new regulations clarify that existing enterprises must meet an energy consumption standard of 6.4 kilograms coal equivalent per kilogram (kgce/kg [52.1 kWh/kg])  during the transition period; failure to do so will result in closure. Newly established enterprises must meet the stricter standard of 5.5 kgce/kg [44.8 kWh/kg]. According to estimates by the Silicon Industry Association, the implementation of the new regulations will directly eliminate 31.4% of inefficient production capacity.

Combined with natural market clearance, this will significantly alleviate the overcapacity problem. The policy provides a clear path for forced clearance, eliminating the need for costly and inefficient capacity consolidation through stockpiling. Consequently, leading enterprises have lost the incentive to promote stockpiling.

4. The hype cycle ends: Those who wore the new clothes have already left

The cooling of stockpiling rumors coincided closely with the cycle of capital speculation. Those who had “dressed up in new clothes” to create hype had long since departed with their profits. In mid-September, shareholder divestments at companies like Canadian Solar and Trina Solar expired – announcements indicate that some shareholders cashed out over 1 billion yuan.

During the same period, open interest in the main polysilicon contract in the futures market decreased by 14,000 lots month-over-month, with net capital outflows for four consecutive days, and speculation plummeted. For investors, the “stockpiling expectation” had fulfilled its mission of boosting stock prices and amplifying futures volatility. With the core demands met, the rumors naturally ceased to fuel the rumors.

5. Fundamentals improve: Even the need for new clothes has disappeared

Industry data from September showed that polysilicon inventories had dropped from 280,000 tons at the end of June to 204,000 tons, demonstrating significant reductions. Meanwhile, overseas demand surged (installations in Europe and Latin America increased by over 30% year-on-year), with leading companies securing over 20GW of module orders, indirectly absorbing some of the excess polysilicon production capacity

This easing of supply and demand imbalances eliminated the rationale for stockpiling to save the market. When the market regulates itself, the need for human intervention vanishes. This “Emperor’s New Clothes” has completely lost its meaning.

IV. Historical lessons: Beware of transforming “new clothes” into “monopoly clothes”

When discussing industry integration plans such as polysilicon stockpiling, the Trust model of the United States in the late 19th century is a historical mirror that cannot be avoided. It is highly similar to the logic of polysilicon stockpiling, seemingly effective but actually harmful. It is also like “new clothes for the old times” that once confused the market, but was eventually abandoned because of stifling innovation. This also sounds the alarm for the current polysilicon stockpiling: If it is forced through, the Emperor's New Clothes may become a monopoly cloak.

At the time, leading American companies in industries like oil and steel established trusts to entrust the production capacity of small and medium-sized enterprises to them for unified management. This model did increase industry concentration in the short term, avoid vicious price wars, and stabilize market supply and demand – just like the initial rumors of stockpiling made some people feel that it was “effective.”

However, in the long run, the trust model, under the guise of stability, actually practiced monopoly: Leading companies achieved excessive monopoly profits by controlling production capacity and prices, but lost the motivation to iterate on technology – why invest in R&D when they could make money without doing anything? This “innovation inertia” directly led to the gradual loss of global competitiveness of related American industries in the early 20th century.

It is precisely for this reason that the United States later enacted the Sherman Antitrust Act, which rigorously broke up and regulated trust companies with the goal of breaking up monopolies and revitalizing market innovation. This history demonstrates that industry stability should not come at the expense of innovation, and consolidation cannot serve as a cover for monopoly.

If the polysilicon industry were to implement stockpiling, even if it stabilized prices in the short term, in the long term, it would, like the trusts, allow outdated production capacity to linger, depriving leading companies of the motivation to reduce costs and increase efficiency, and ultimately hindering technological advancement across the entire PV industry chain. After all, the widespread adoption of N-type polysilicon and breakthroughs in plasma processing were never achieved by protecting backwardness, but rather by the forced “survival of the fittest.”

V. Breakthrough: How big is Tongwei? It relies on hard power, not new clothes

Amidst the clamor of stockpiling rumors, the true path to success in the polysilicon industry lies in the industry landscape surrounding Tongwei’s dominance. However, it must be made clear that Tongwei’s dominant position is fundamentally different from monopolies and polysilicon stockpiling. It is the result of market competition, not administrative intervention or capital manipulation. It is certainly not a case of the Emperor’s New Clothes, but rather a “hard power barrier” built on technology, scale, and business models.

From the perspective of hard power, Tongwei’s advantages are almost impossible to replicate:

• Technically, more than 90% of its polysilicon meets the requirements for N-type. Black technologies, such as a plasma method, are in pilot production, and the speed of technological iteration far exceeds that of its peers.
• In terms of scale, the production capacity reaches 970,000 tons (global market share of 30%), and it has 30% production capacity flexibility – the Baoshan and Leshan bases can take advantage of overproduction and cost reduction during the flood season; the Baotou base is in full production all year round and can flexibly respond to market fluctuations.
• In terms of business model, “shareholder underwriting + super integration” locks in 545,600 tons of sales (accounting for 41.9% of the market), while self-digesting 250,000 tons of polysilicon, building a closed loop of production-sales and having strong risk resistance.

More importantly, Tongwei’s dominance isn’t a monopoly, but rather a cornerstone of industry stability. Its advantages stem from technological leadership and cost control, not from stockpiling to protect lagging behind. Tongwei’s operating rate is over 80%, far exceeding the industry average of 56.5%. This is essentially the market-driven result of efficient production replacing inefficient capacity.

Without Tongwei’s price guarantee (proposing a minimum full-cost price for polysilicon) and rule-making (participating in the design of polysilicon futures), the polysilicon industry would have long since fallen into a game of worst-case scenario. The price plunge in 2023/2024 and the failure rate in random module quality inspections reaching 16% (2025 data) are the consequences of disorderly competition.

VI. The Future: Dual drive from regulation and technology to end the new clothes fantasy

The mystery of the polysilicon industry will eventually be completely broken by policy supervision and technological iteration. The survival space for non-market means such as stockpiling will become smaller and smaller, and the Emperor's New Clothes will no longer be possible to be sewn.

On the policy front, the draft energy consumption quota released in September for public comment, though not officially implemented until the end of 2026, clearly sets the direction for mandatory phase-out. According to calculations by the Silicon Industry Association, domestic polysilicon production capacity will drop to 2.4 million tons after the new regulations are implemented, significantly alleviating oversupply and eliminating the need for stockpiling.

On the technical level, the penetration rate of N-type polysilicon continues to increase, while the value of P-type polysilicon is depreciating rapidly. Tongwei is developing new processes, such as granular silicon and plasma methods. Costs will continue to fall in the future.

Editor’s note: On qq.com, you will find the original Chinese article (published on September 30). The author obviously sugarcoats the role of market leader Tongwei, which is the largest contributor to the polysilicon overcapacity in China: Around one third of the capacity expansion between 2020 and 2024 was caused by Tongwei alone. The company’s utilization rate in the first half of 2025 was 39%, not 80%. Moreover, the author withholds the fact that ingot/wafer producers have another 200,000 tons of polysilicon inventories in store. Total inventories are rising, not decreasing.