The Endgame of Polycrystalline Silicon

Carbon Rush, Reconstructing the Value of the New Electric Industry

Recently, Carbon Rush returned to his hometown to pay respects to ancestors. He heard that the local farmers had already switched industries. After inquiring further, he learned that at the rural market, pork costs only five yuan per jin, and each pig farmer loses 500 yuan per pig raised. Carbon Rush couldn’t quite understand why, in Shanghai supermarkets, a jin of pork sells for 20 yuan—this is a typical “Shanghai currency,” just like Zhang Longgen’s overseas silicon materials in Oman. Regardless, the pig cycle has reached a freezing point, marked by the culling of breeding sows, piglet prices plummeting, and farmers’ willingness to restock drying up.

The situation with polysilicon is similar. Carbon Rush was somewhat surprised that silicon material prices could return to 30k yuan/ton, which is actually cheaper than pork in Shanghai supermarkets. Cutting capacity in silicon materials is like culling breeding sows; eliminating “outdated” capacity requires a swift and decisive action. Even if production can continue, the excess must be killed off—no matter how painful, there’s no other way. This is a clear signal.

Without the support of industry self-discipline like production limits and price caps, and lacking the encouragement of silicon stockpiling expectations, silicon material prices naturally fall like a waterfall. Market-driven capacity reduction is, of course, not a bad thing—it should be so—but Carbon Rush still has some words he can’t hold back.

Based on current polysilicon prices, the entire industry should be selling below cost, which means the whole industry is violating price laws. If industry associations’ self-regulatory measures like production and price limits are suspected of violating anti-monopoly laws and are halted, then selling polysilicon below cost is also suspected of violating price laws and should, in theory, be stopped. Moreover, relevant departments had previously held meetings and clarified opinions on this.

However, until today, there seems to be no concrete law enforcement action against selling below cost. The entire industry is selling below cost—why is no one managing this? Ultimately, it’s because this is difficult to operate; it’s not as simple as stopping production limits and price caps. Similarly, pork has been sold below farmers’ costs for a long time. Who can regulate such blatant violations of price laws? And how should they be regulated? Is banning pig trading even an option?

In the era of silicon dominance, silicon material was once a rising star, soaring on the wind; in the capacity reduction era, silicon material is the pig falling from the sky. When pigs fly, it seems no one cares; when they fall, debates about whether to slaughter or bury, whether to make red-braised pork or steamed pork, have raged for nearly two years. I remember in July 2024, silicon material prices even dipped above 30k yuan. If it had been clarified then that market-based clearing was necessary, the photovoltaic industry might have already emerged from the trough by now.

To say that there is no big picture, no understanding of the overall situation—after two years of anti-involution efforts in photovoltaics, the current feeling is one of chaos. Is this grand campaign truly ending in failure? Who should be responsible for today’s results?

Let’s not dwell on the past; let tomorrow continue. Where is the future of photovoltaics?

The weekly data released by the Silicon Industry Branch on April 1 shows that the average price of n-type re-investment material is 36.5k yuan/ton, while InfoLink’s weekly report at the same time shows the average price of n-type dense material is 37k yuan/ton.

In the report from the Silicon Industry Branch the previous week, it was described as:

Market prices continue to decline, mainly because upstream and downstream companies are relatively pessimistic about the future. Downstream procurement is extremely low, and price pressure is very strong; silicon material companies, under inventory and capital pressure, passively accept low prices to seek cash flow, forming a negative cycle of “the more they fall, the less they buy”… Currently, the polysilicon industry is facing severe tests, and the market clearing process may be more complicated than expected.

The Silicon Industry Branch urges:

Rational price cuts will only further erode the overall profit margins of the industry. Companies should strengthen market information exchange, enhance consensus on supply and demand, stabilize market expectations, and reduce misjudgments and malicious price suppression caused by information asymmetry.

In one sentence, Carbon Rush summarizes: polysilicon has reached a critical, difficult, risky, and heavy moment.

Like a battle, the final stage is often more bloody and violent than the frontal confrontation, similar to the scenes in war movies like “Hacksaw Ridge,” where it’s a one-sided slaughter of the defending enemy, whether in resistance or surrender.

What’s terrifying is that although silicon material prices are now at historic lows, they may still continue to fall—soon the southwest flood season will arrive, electricity prices will further decrease, and industrial silicon prices are on a downward trend. Leading companies might abandon their “shock therapy” of full shutdown and inventory reduction, and resume production. If that happens, by the second quarter of this year, silicon material prices could drop to 30k yuan.

The collapse of polysilicon is deadly.

On one hand, silicon material prices have already fallen below the full costs of all companies; on the other hand, low operating rates or even shutdowns have caused depreciation and capital costs to skyrocket.

Carbon Rush roughly estimated that when the industry’s operating rate drops below 20%, the average full cost will exceed 60k yuan/ton. Even leading companies like Tongwei, GCL-Poly, and Daqo, under low utilization, will face full costs of around 50k yuan/ton. Is there any survival in this collapse? Who can survive this stampede?

Based on the debt, cash reserves, self-owned power plants, N-material ratio, and inventory devaluation of silicon companies, we assessed the survival red lines for various firms, including Tongwei, GCL-Poly, Daqo, Xintech, Orient Hope, Red Lion, Lihua, Hongyuan, He Sheng, Jinnuo, Qiya, Runyang, Xinjiang Gones, and others.

0

Silicon Material Avalanche: No Bottom in the “Floor Price”

When the photovoltaic industry’s joint capacity reduction was halted and the expectations for silicon stockpiling were completely dashed, silicon prices lost their last psychological defense and began an unstoppable avalanche.

In January 2026, regulators explicitly denied the behavior of market coordination through capacity and price agreements in the photovoltaic industry, and the six-month “anti-involution” joint capacity reduction campaign was fully suspended. This meant the market would fully revert to market-based clearing logic. The previous support from “policy backing” and “silicon stockpiling” expectations was completely shattered, becoming the main trigger for this round of price collapse.

Looking at the price trend, silicon material prices had already fallen beyond the scope of cyclical adjustment, entering an irrational stampede. The cause of the plunge was unprecedented supply-demand imbalance and inventory backlog.

According to estimates from the Silicon Industry Branch, Galaxy Futures, and others, by the end of March 2026, China’s total polysilicon inventory had risen to 518.7k tons, with available days of inventory approaching five months—far exceeding the healthy industry level of 1-2 months. Among these, in-house finished goods inventory at production plants reached 330k tons, the core source of low-priced spot market dumping.

Demand’s complete slowdown further exacerbated the supply-demand contradiction.

Data from the National Energy Administration shows that in January-February 2026, the country’s new photovoltaic installed capacity was 32.48 GW, a year-on-year decrease of 17.71%, marking the first decline in recent years. Due to project returns and grid absorption limits, project startup progress fell far short of expectations. Last year, domestic new PV capacity was 315 GW, and the China Photovoltaic Industry Association’s forecast for 2026 is 180-240 GW.

Overseas markets are also grim.

Despite the energy crisis triggered by the US-Iran conflict sweeping the globe, overseas demand for photovoltaics also suffered a blow. On April 1, the export VAT rebate policy for PV products was officially canceled, directly increasing the export costs of modules. The “export rush” market in March had already front-loaded orders for Q2, and after the policy was implemented, overseas orders sharply declined. Production plans for modules, cells, and wafers were significantly cut, with all three segments producing less than 50 GW. Downstream demand for silicon material continued to shrink.

What’s more alarming is that silicon material prices still have room to fall.

First, the southwest region will enter the flood season in May-June, with water prices in Sichuan and Yunnan expected to decrease by 0.05-0.1 yuan/kWh compared to dry season, which will reduce silicon production costs by 3,000-6,000 yuan/ton, fueling expectations of a lower cost center and further suppressing spot prices.

Second, industrial silicon prices continue to decline. According to Antaike’s April 2 pricing, the nationwide industrial silicon price averaged 8,958 yuan/ton, down 52 yuan/ton from the previous week.

Third, the negative spiral of “the more it falls, the less they buy” has formed. Downstream wafer companies generally keep raw material inventories at just 3-7 days’ demand, maintaining small orders, low prices, and just-in-time procurement strategies. Upstream companies, to free cash flow, can only continue lowering prices to clear inventory, with transactions at ultra-low prices of 33k yuan/ton already occurring.

Carbon Rush believes that if no new policies or external shocks intervene, silicon material prices could further decline. From the trend:

• In the short term (Q2 2026): silicon prices will continue to fall, possibly reaching historic lows of 30k-32k yuan/ton;**
• Mid-term (Q4 2026 - Q1 2027): as high-cost capacity is cleared, inventories are depleted, and global PV demand recovers, prices may rebound to 38k-45k yuan/ton;**
• Long-term (after Q2 2027): supply and demand will balance, and prices could stabilize in the reasonable range of 45k-55k yuan/ton, corresponding to 8%-15% gross profit margins for leading companies operating at full capacity.

A senior executive from a leading silicon company told Carbon Rush that he considers the above forecast overly optimistic and that silicon prices don’t need to fall further; many companies can’t hold on much longer.

0

The Silicon Material Elimination Race Is Near Its End

The freezing point of the pig cycle is when farmers slaughter breeding sows to cut losses; the end of the polysilicon cycle is a comprehensive test around the survival red line. Using authoritative data from the Silicon Industry Branch, listed company announcements, and investor communications, we have conducted a comprehensive assessment and stratification of the survival capacity of mainstream industry companies from the following dimensions.

From Dongfang Hope’s official website

Dimension One: Cash costs and self-owned power plants—this is a life-and-death line at the production end.

Cash cost is the company’s cash flow survival line. Once spot prices fall below cash costs, production results in net cash outflow, and companies can only choose to halt production. The coverage of self-owned power plants directly determines the bottom line of cash costs.

• Safety tier (cash costs <30k yuan/ton): GCL-Poly (grain silicon as low as 24k yuan/ton), Tongwei (as low as 27k yuan/ton), Dongfang Hope (29k yuan/ton).
• Critical tier (cash costs 30k-37k yuan/ton): Daqo Energy (39k yuan/ton, note: Daqo’s cash costs are slightly higher, but with zero-interest debt and no interest pressure), Xintech (38k yuan/ton), Hongyuan Green Energy (35k yuan/ton). These three are close to the breakeven point, with pressure but still able to maintain production through low load operation.
• Death tier (cash costs >38k yuan/ton): Red Lion Silicon Materials (47k yuan/ton), Qinghai Lihua (45k yuan/ton), Runyang Shares (46k yuan/ton), Nanjing South Glass (49k yuan/ton), Jinnuo Technology (48k yuan/ton), Qiya Group (50k yuan/ton), Xinjiang Gones (47k yuan/ton). These companies have fully fallen below cash costs, producing at a loss, and can only halt all production, relying on inventory sales to recover cash flow.

Dimension Two: Interest-bearing debt and capital costs—cash flow survival line

High debt levels bring rigid interest expenses, which are the biggest cash flow black hole during downturns. Under industry-wide losses, a company’s resilience depends directly on the scale of interest-bearing debt and its cash reserves’ ability to cover interest payments.

Companies with coverage multiples below six months face immediate risk of cash flow rupture; while Daqo Energy, Dongfang Hope, and others, with very low debt and ample cash reserves, have a strong safety cushion to survive the cycle.

Note: Silicon companies’ interest-bearing debt includes current liabilities and project loans. For recent polysilicon capacity expansions, bank loans typically have a five-year repayment cycle, similar to residential mortgages, making extensions difficult. This puts pressure on companies like Qinghai Lihua, which face significant repayment burdens.

Dimension Three: Inventory and devaluation provisions—asset safety life-and-death line

As of March end, the industry’s massive inventory of 518.7k tons is not only the core source of price suppression but also an “invisible mine” on the balance sheet.

Current spot prices have fallen more than 40% below the average production costs of 2025. Whether inventory devaluation provisions are sufficient directly affects asset safety.

• Among leading companies, Tongwei had already made inventory devaluation provisions of 30k yuan in the first half of 2025, over 90% of which related to polysilicon inventory, indicating that inventory devaluation risk has been largely realized; GCL-Poly’s inventory turnover is about 10 days, with almost no inventory backlog, and negligible devaluation risk;

• Second-tier and smaller companies generally hold 10,000-30,000 tons of inventory, with turnover exceeding 90 days. Many have not fully provisioned for inventory devaluation. As prices continue to fall, they face huge asset impairment losses, further eroding net assets and increasing financial risks.

0

Final Speculation

The end of the pig cycle is marked by the large-scale elimination of inefficient breeding sows, leading to increased industry concentration among top breeders; the end of the polysilicon cycle has long been predetermined by cost, debt, and product competitiveness stratification. This is an irreversible market-based clearing process. Based on survival red line calculations, we have clarified the three-stage rhythm of capacity clearance and the final list of eliminations and survival patterns.

Of course, companies like Dongfang Hope, Lihua, Qiya, Jinnuo, etc., are non-listed firms. Most of Carbon Rush’s estimates for these companies come from industry interviews and may not be entirely accurate.

First Stage: Breaking through the cash cost line, early elimination (Q2-Q3 2026)

Trigger condition: Spot prices persistently below 38k yuan/ton, leading to immediate cash losses in production, rapidly depleting cash flow.

Elimination targets: Companies with cash costs >38,000 yuan/ton, no self-owned power plants, and N-material ratio <40%, i.e., the least efficient silicon material producers.

Representative companies: Qiya Group, Jinnuo Technology, Xinjiang Gones, Nanjing South Glass, Inner Mongolia Dunan, etc.

Capacity scale: About 400k tons, roughly 15% of current total capacity.

Core features: These companies have mostly halted production, relying on selling old inventory to recover funds. Cash reserves can only support a few months of interest payments. If prices do not rebound in Q2, their silicon business will likely face bankruptcy or capacity sale, exiting the market entirely.

Second Stage: Debt pressure triggers, concentrated clearance (Q4 2026 - Q1 2027)

Trigger condition: Spot prices persistently below 45k yuan/ton, ongoing losses, and high interest costs exhaust cash reserves.

Elimination targets: Companies with full costs >45,000 yuan/ton, debt-to-asset ratio >60%, and self-owned power coverage <30%.

Representative companies: Red Lion Silicon Materials, Qinghai Lihua, Runyang Shares.

Capacity scale: About 500k tons, roughly 20% of current capacity.

Core features: These companies are near breakeven but cannot withstand the industry winter due to high debt interest. Continued losses erode assets, and financing tightens. They can only achieve debt restructuring through capacity sales or bringing in strategic investors, or exit the market.

Third Stage: Lack of competitiveness, marginal elimination (post Q2 2027)

Trigger condition: Industry enters oligopoly, market share concentrates among top players, and small capacities lose survival space.

Elimination targets: Companies with N-material ratio <50%, no integrated layout, or weak technological iteration.

Representative companies: Some cross-industry silicon capacity.

Capacity scale: About 200k tons, roughly 8% of current capacity.

Core features: These companies have some cost advantages but lack product competitiveness, stable downstream clients, or long-term contracts. In the scale competition among giants, they gradually lose market share, eventually becoming supporting self-use capacity or completely exiting market competition.

Final Pattern: Oligopoly era

After this deep capacity clearance, by the end of 2027, China’s polysilicon industry will undergo a thorough restructuring:

• Effective capacity will shrink from over 3 million tons to 1.5-1.6 million tons, restoring the supply-demand ratio to a healthy 1.2:1, ending overcapacity;

• Industry concentration will significantly increase, with CR5 (Tongwei, GCL-Poly, Daqo, Xintech, Dongfang Hope) surpassing 85%, entering a low-profit oligopoly era;

• Industry price center will stabilize at 45,000-55k yuan/ton, with leading companies operating at full capacity achieving a reasonable gross profit margin of 8%-15%, ending the cycle of “super profits—sharp drops,” and entering a phase of steady development.

Postscript

The above are only some of Carbon Rush’s speculations, or more accurately, industry guesses. Perhaps even top companies may not all survive, and black swan events are not impossible.

From 300k yuan/ton to 30,000 yuan/ton, polysilicon has gone through a complete cycle in four years. This brutal price war is not the end of the industry but the end of the era of savage growth.

Just as the pig cycle hits a freezing point, a new price rise will inevitably follow; after the winter of polysilicon, surviving companies will benefit from long-term growth in global PV installations and will truly control the pricing power of the global PV industry chain.

For those companies that fell in the cold winter, their ending has already been fixed in this bloodbath-like slaughter on Steel Ridge.

Editor: Carbon Detective