Photovoltaic Industry Outlook for 2025: Destruction Before Construction

Photovoltaic components, especially double-glass photovoltaic components, are typical durable industrial products Their power generation principle and packaging form determine that photovoltaic components often have a product life of 30 to 40 years; the attributes of industrial products also mean that people will not easily replace existing photovoltaic panels due to technological upgrades or style updates. The price of photovoltaic components has fallen by two-thirds in the past two years, which further strengthens my belief that photovoltaic power generation will account for 50% by 2050 and become the mainstream energy source for humanity. However, even based on the most optimistic assumption that photovoltaic power will become the mainstream energy source for humanity, the global market's annual demand for photovoltaic components is unlikely to consistently exceed 1000 GW, as I mentioned in last year's annual outlook, "To the End," we will inevitably hesitate and wander in the TW era hesitation and wandering 。

The ceiling of 1000 GW of photovoltaic demand is there, whether you see it or not; and the photovoltaic industry chain capacity of 1500 GW has already been formed, whether you see it or not. How I wish I could travel back to May 11, 2023, when I visited Beijing Zoo on that bright day, with my current memories, to stop this unprecedented frenzy of production capacity instead of becoming a part of it. On the one hand, there is the ceiling of demand, and on the other hand, there is huge production capacity. The only way to reconcile this contradiction is to clear the production capacity, so when writing the 2025 annual outlook, the theme I most want to write about is “Destruction Before Construction” 。

I. "Price Prediction" and "Supply and Demand Analysis"

Let's first look at some basic facts and situations in 2025. Since I started writing annual outlooks in 2017, I have maintained a high degree of accuracy, and many annual outlooks can even be used as annual reviews. Being in the industry, from an industry researcher to an industry practitioner, it is very easy to obtain sufficient predictive information. Especially at this point in time, looking at the whole year of 2025, many outlooks are already clear

1. Global component shipments are expected to reach 600 GW in 2025, a slight increase from last year

Following China's photovoltaic installation capacity reaching 216.88 GW in 2023, it further reached a massive 277 GW in 2024. This is a significant impact on the power grid, after all, since the birth of photovoltaics in China, the cumulative installed capacity of photovoltaics in China is only 890 GW (end of 2024), and the installation capacity in 2023 and 2024 alone accounts for 55.52% of the historical total, which is incredible! This led to two policies from the NDRC and the National Energy Administration at the beginning of this year. One announced the end of the full-on-grid era for distributed photovoltaics (effective April 30); the other announced the arrival of the competitive bidding on-grid era for new energy (effective May 31). Not allowing full on-grid will cause projects such as fish pond photovoltaics, aquaculture photovoltaics, highway photovoltaics, and useless warehouse roof photovoltaics to leave the stage, thus reducing the scenarios for distributed photovoltaic projects and making development more difficult; while the pilot experience of competitive bidding tells us that competitive bidding will inevitably make the on-grid price of photovoltaics lower, or even far lower than the benchmark electricity price of desulfurized coal, and the reason is easy to explain: as an unstable power source, the long-term development of photovoltaics must establish the following equation:

Photovoltaics + Energy Storage = Desulfurized Coal Electricity Price

When the proportion of photovoltaics is small, the instability of photovoltaic electricity is harmonized by grid dispatching, but now, photovoltaics need to make its electricity price better than desulfurized coal electricity price, creating a profit space for energy storage development, in order to be eligible to compete with stable power sources.

Discussions with industry experts suggest that the negative impact of these two policies on photovoltaic development this year is relatively large, and the demand in the second half of 2025 is unclear. There is still a two-month rush installation window before the policy node, and I believe that with the diligence of the Chinese people, I predict that China's annual photovoltaic installation capacity will be over 230 GW, and China's demand is expected to decline by more than 15%, while the overseas market is expected to grow by 20-30%. Overall, global DC-side photovoltaic component shipments in 2025 are expected to exceed 600 GW, a slight increase compared to 580 GW in 2024.

2. Prices have passed the valley of despair, but the rebound is limited

In the third quarter of 2024, the prices of the photovoltaic industry chain rapidly slid towards the valley of despair: polysilicon at 36 yuan/kg; silicon wafers at 0.98 yuan/piece; battery cells at 0.265 yuan/watt; modules at 0.62 yuan/watt; photovoltaic glass at 11 yuan/square meter. Those familiar with the photovoltaic industry chain will know that the above prices are all below the cash cost, leading to a rapid wave of production capacity shutdowns, and the operating rate of some industry segments fell to 30%. The product prices at this time are completely irrational venting under panic emotions, and customers should not expect the above prices to reappear in the short term.

The listing of polysilicon futures at 42,000 yuan in the fourth quarter of 2024 provided the industry with correct price guidance, causing the spot price of polysilicon to quickly rise to over 40,000 yuan, and simultaneously causing the price of 182.2*183.75mm silicon wafers to rise to over 1.1 yuan/piece. After that, due to the arrival of the Lunar New Year, there were no more fluctuations in the price trend.

The two policies issued by the National Energy Administration before and after the Lunar New Year triggered the current rush installation situation, and it is expected that this wave will last at least until early May 2025. Because the market situation in recent years has been too extreme, we cannot predict how high the industry chain prices will rebound during the rush installation period. I will only talk about a reasonable price for the industry chain: polysilicon needs to recover to at least 55 yuan/kg, M10 silicon wafers to 1.55 yuan or more, batteries to 0.39 or more, and modules to 0.8 or more, so that related companies can hope to gradually recover their investment costs through long-term operations. Correspondingly, even though the domestic photovoltaic industry chain prices have experienced a surge after the Lunar New Year, the current prices are still far from recovering the industry's costs.

3. The Externalities of Photovoltaics and the Hesitation and Wandering Fate of the TW Era

China's photovoltaic industry is extremely external The characteristics of the industry, where by "externalities" I mean: our Chinese photovoltaic industry transfers benefits to the whole world, while retaining very little profit for itself. The electricity generated over the entire life cycle of a standard container of photovoltaic components is equivalent to that generated by burning 6220 tons of 5500 kcal coal, which would require 180 trucks to transport. The high stability, long lifespan, and zero marginal cost of photovoltaic components mean that power plant owners will always make a fortune over the long asset holding period. Therefore, the world trade of photovoltaic components is tantamount to a huge energy trade. However, while people are fighting over oil due to its "resource attribute", they scoff at the "manufacturing attribute" of photovoltaics. Although photovoltaics have energy-like attributes, they do not and cannot enjoy resource benefits. Our Chinese entrepreneurs sell components at extremely low prices, which has not earned the respect of the world market, but rather accusations of dumping and forced labor, and a strong response in the form of tariffs. Once upon a time, I was proud and self-moved by the unique external (Here, externalities refer to the characteristics of value spillover in the photovoltaic industry, where the industry itself gains little but brings great value to other fields) but now, under the reality of extreme involution and the suffering of industry practitioners, I have begun to reflect: should we stop the continued proliferation and spread of this "externality"?!

The fundamental reason for the existence of externalities in the photovoltaic industry is determined by the ultra-long lifespan of photovoltaic components. The sales profit of photovoltaic components can only be realized once, and the power generation revenue during their subsequent lifespan of at least 30 years belongs to the power plant owners and has nothing to do with the photovoltaic manufacturing industry. To help everyone understand why the photovoltaic manufacturing industry will inevitably hesitate and wander in the TW era, I will present last year's formula again for everyone to hear:

If 1 TW of photovoltaic components are installed annually and the component lifespan continues for 30 years, then after 30 years, the existing power plants will reach 30 TW. According to the average annual power generation hours of 1100 hours for photovoltaic power plants, after 30 years, the accumulated photovoltaic power plants can output 33 trillion kWh of clean energy to the world each year, which is more than the current global total electricity consumption of 32 trillion kWh. Even considering the increase in electricity consumption brought about by energy substitution and artificial intelligence, it is not enough to change our fate of wandering and hesitating in the TW era, because the unstable and unbalanced characteristics of photovoltaic power generation in day and night, seasons, and climate are destined to make it difficult to achieve 100% popularization.

However, our entire industry chain manufacturing capacity has exceeded TW, so how can this contradiction be resolved? The externalities determined by the long lifespan of photovoltaics cannot be changed, so we can only break the deadlock by optimizing the competitive landscape and obtaining reasonable profits through one-time sales. I heard that in the European market, 1 watt of photovoltaic components only sells for 0.7 yuan, but the photovoltaic system EPC price is as high as 7 yuan. It is not that European customers do not cherish our products and use photovoltaic components as fences, but that iron fences in Europe are really more expensive than our photovoltaic components! If the current situation of mutual slaughter and extreme involution among our manufacturers does not change, the outcome waiting for us will be a complete defeat.

II. Failure to change will lead to complete defeat

1. A question that needs to be clarified for everyone

The value of an asset depends on its profitability, not its historical investment. In more standard economic terms: Asset price = Present value of future cash flow This means that when an asset with huge investment suddenly loses its profitability, it will instantly become worthless. This is completely different from the ten-year straight-line depreciation method commonly used in our financial statements!

Although the discounted cash flow method is the most accurate method for expressing asset prices, the assessment of future cash inflow capacity involves many assumptions, making it difficult to apply in accounting practice. It can only be reflected in asset transactions through the judgment of the buyer and seller on the asset's profitability. Therefore, the discounted cash flow method is mainly used in the "transaction" process of assets, while in the daily accounting practice of financial statements, we still have to use the less accurate straight-line depreciation method. At this time, we should pay attention to the difference in the methodology for assessing asset prices in transaction practice and accounting practice, which may cause a potentially huge deviation between the tradable price of an asset and the price presented on the statement. Obviously, if the current situation of overall losses and extreme involution in the photovoltaic industry continues, the value of assets reflected in the financial statements will be far higher than their actual tradable price. The existence of this deviation may blindfold current photovoltaic industry decision-makers, making them unaware of the severity of the current problem and leading to strategic misjudgment.

2. Revaluation of asset value under deteriorating profitability

Let's use the example of Topcon cell production capacity to illustrate the problem of "revaluation of asset value under deteriorating profitability." Topcon cells, with their unique passivated tunnel oxide layer process characteristics, have a relative efficiency advantage of more than 5% over Perc cells. They can bring low cost and high premiums, representing a technological revolution that can further enhance the competitiveness of photovoltaic power generation. However, the problem lies in the high consistency of everyone's judgment of technological value and excessive investment enthusiasm. The popularization of the technology from 1% to 100% was too fast. At the beginning of 2023, its popularity rate just exceeded 15%, but by the third quarter of 2024, it reached more than 80%. In the two years from 2023 to 2024, the total effective Topcon production capacity formed in the market reached 700 GW, and the capacity under construction exceeded 1300 GW, far exceeding global demand.

In 2023, when investment enthusiasm for Topcon was high, the equipment procurement price for 1 GW of production capacity was as high as 170 million yuan. However, when the tide receded, equipment manufacturers had equipment backlog, and the price of new equipment for 1 GW fell to around 80 million yuan. Some manufacturers were unwilling to continue investing and sold new, unused second-hand equipment at a price as low as 50 million yuan per GW. Those manufacturers who entered the market in 2023 with heavy investment at 170 million yuan, even after two years of depreciation, still have a remaining asset value of 136 million yuan on their balance sheets. The deviation between the market price and the remaining value on the balance sheet of Topcon cell equipment production capacity is a typical example of asset illusion. The fundamental reason for the deviation is that the current Topcon cell production capacity lacks profitability and even suffers losses. Investors buy assets to make money, not to lose money. According to the profitability of Topcon, calculated using the discounted cash flow method, the price of 50 million yuan/GW is already high.

The existence of leverage further exacerbates the severity of the problem. If the 170 million yuan is all equity investment, then even if the equipment price falls to 50 million yuan in a short period of time, the company's net assets will still be positive. However, if during the investment frenzy in 2023, everyone entered the market with high leverage, such as only 50 million yuan of equity capital for a 170 million yuan investment, with the rest being bank loans and supplier arrears, then the large fluctuations in asset prices will easily cause the net assets of the production capacity investor to become zero, resulting in insolvency. The use of leverage tools is a common phenomenon among current photovoltaic industry practitioners. Even the debt ratios of giant companies in the photovoltaic industry are generally high. If we re-evaluate asset prices based on today's profitability using the discounted cash flow method, many companies will become insolvent The situation. This is the important reality of the industry today, and it is especially worthy of our high vigilance.

3. Orderly capacity reduction actually increases asset value

The purpose of explaining these problems is not to continue to create a pessimistic atmosphere, but to encourage everyone to have a common awareness of the danger! And to find a way out under this common awareness: When looking at the history of the American Industrial Revolution, there is one segment that impressed me deeply. Rockefeller had just acquired an oil refinery and gave it to his subordinate. The subordinate asked Rockefeller what he needed him to do, Rockefeller replied: Shut it down . The solution to today's predicament is also clear: Under the condition of overcapacity, orderly shutdown of some production capacity, return to supply-demand balance, restore profitability, can actually increase the asset value of the remaining production capacity. It is precisely because of the orderly shutdown of some production capacity that the entire industry has been saved, and the asset value of everyone can be preserved, so as not to lose everything.

But the seemingly simple orderly shutdown is the most difficult to achieve, because there is a prisoner's dilemma: The industry does not have only one competitive entity. For example: If companies A, B, and C reach a consensus on production reduction, but company D expands under the profitable market conditions after production reduction, it means that A, B, and C will lose market share. Everyone hopes that it is others rather than themselves who shut down, so things will be difficult to reach a consensus between "beneficial to the group" and "beneficial to the individual", and will be forced to accept the predicament of group blood loss.

III. The way out lies in the transformation of the competition paradigm

Life always finds a way out, and even the most difficult predicament will always have a "solution". The way out lies in: Using silicon materials as a consensus basis to promote the transformation of the competition paradigm.

As we know from the previous discussion: At present, the photovoltaic industry is in a severe situation of absolute overcapacity, demand hitting the ceiling, loss of profitability leading to a reassessment of huge asset value, and years of accumulated net assets of enterprises facing zeroing or even insolvency. An obvious way out is: Through industry consensus to achieve orderly capacity exit, achieve supply-demand balance or even slightly tight supply, restore industry profitability, and thus save the entire photovoltaic industry. However, the difficulty of this beautiful picture of the industry lies precisely in the two words " consensus ". We cannot pursue the consensus of all companies and all industrial links; that is too difficult and simply impossible! We can only pursue a partial consensus in a certain link, while this link has a global impact on the entire industry. This link is the source of the photovoltaic industry, the mother of the photovoltaic industry cycle, and has many characteristics such as a long production cycle, low production elasticity, high capital density, and extremely high approval difficulty. Silicon! Material! Industry link. Because it is the source of the industry, the consensus on silicon material reduction is equivalent to the consensus on the reduction of the entire industry; because it is the mother of the cycle, the restoration of the profitability of silicon materials means the improvement of the profitability of the entire industry; because there are very high entry barriers, only six valuable companies remain in this link, namely Tongwei Co., Ltd., Xinyi Solar, Daquan Energy, New Energy, Oriental Hope, and Asia Silicon Industry. These six companies account for more than 85% of the industry's effective output, and their current capacity is sufficient to meet 150% of the current global demand; because there are only six valuable companies left, I say that we should take silicon materials as the basis for industry consensus.

The consensus of the silicon material industry link, or more precisely, the consensus of the above six companies, should include:

1. Consensus on orderly production reduction

The global effective demand for photovoltaic power in 2025 is expected to be 600GW. According to the equation that 10,000 tons of photovoltaic silicon materials can produce 4.5GW of photovoltaic components, our demand for silicon materials in 2025 is only 600÷4.5=133.33 million tons. However, there has always been a puzzling contradiction in this equation, that is, the number of components calculated from the annual polysilicon output is always greater than the total demand statistically reported annually. The largest deviation is in 2024. In 2024, China's polysilicon output was 1.83 million tons, and the estimated total output of overseas polysilicon, represented by Wacker, was 100,000 tons, totaling 1.93 million tons. According to the equation of 10,000 tons = 4.4GW photovoltaic components in 2024 (note: the equation of 10,000 tons of silicon materials producing 4.4GW of photovoltaic components has fully considered the problem of production loss in all links), it is enough to produce 849GW of corresponding components, while the official statistics show that the global component installation demand in 2024 is only 520GW. The deviation is enormous. There are some obvious reasons for these deviations: 1. Silicon material inventory. 2024 is a stage with a huge accumulation of photovoltaic silicon material inventory. Combining industry research, I estimate that 300,000 tons of the total 1.93 million tons of silicon material output in 2024 has become inventory. After inventory correction, 1.53 million tons of silicon material correspond to 688GW of components; 2. Inventory at various links. In 2024, under the situation of overall oversupply, inventory at various links will increase, especially the inventory of silicon wafers and components will increase significantly, and the increase in inventory of battery cells is due to the increase in turnover inventory due to increased production. I estimate that this reason can explain the overproduction of more than 100GW of silicon materials, and after correction again, it is 588GW; 3. The actual global component demand may be greater than the official statistical values! This is a problem that various research institutions have never pointed out, and I believe that this problem has existed since the explosive application of photovoltaics in 170 countries around the world in 2018, so I need to spend more chapters discussing how much the actual global demand is, because this number is related to the calculation of the silicon material output balance point, and the silicon material output balance point is related to where the industry self-discipline target is set.

What is the global demand? After China's 531 policy in 2018, the price of photovoltaic components rapidly dropped by 40%, coupled with the efficiency improvement brought about by the technological revolution of monocrystalline + Perc, photovoltaic electricity entered the era of grid parity, and photovoltaic demand exploded in over 170 countries worldwide. In the past, when photovoltaics were expensive, global demand was easy to judge because the mainstream market was concentrated in Europe, America, Japan, and policy-stimulated China. The market was concentrated, and power stations needed to receive subsidies, so all construction projects would actively apply for them. Research institutions could easily obtain accurate global demand figures through official data, so before 2018, the judgment of global demand by official institutions was relatively accurate. However, after 2018, market demand became widespread and diverse, and traditional research methodologies (referring to official statistical bulletins) gradually became inaccurate. Japan saw the emergence of 1.7:1 ultra-high capacity ratio power stations to achieve better power generation curves; Africa, South America, and the Middle East saw the emergence of a large number of off-grid power stations in economically underdeveloped and war-torn regions, and with technological innovation, there was also a large demand for replacing photovoltaic versions of already grid-connected photovoltaic power stations, but this demand would not appear in the demand statistics of official bulletins, but it is a real demand in the manufacturing process. Therefore, there is an un-statistical demand correction parameter, and with photovoltaic components becoming cheaper and more widely used, penetrating every corner of the world, this un-statistical demand correction parameter will continue to increase year by year. I personally estimate that 8% is reasonable in 2024, and we can take 10% in 2025, because only after this correction can it be consistent with glass production data, film shipment data, and polysilicon production data, forming a logically self-consistent conclusion.

Now, let's return to the topic of this article. According to the correction value of 10% of un-statistical demand in 2025, the calculated polysilicon production to meet the supply-demand balance is 1.333 million tons × 1.1 = 1.4666 million tons. This is the polysilicon production limit target to be achieved through industry self-discipline. After the target is set, the next question is how to achieve it? More than ever, we need a stronger photovoltaic industry association. Today's problems in the photovoltaic industry are no longer the problems of a single enterprise. If the current problems continue, after enterprises lose their net assets, ordinary creditors, banks, and employees will be affected more widely. At this time, the problem is no longer a problem of a single enterprise itself, but will spread to suppliers, banks, and employees, affecting countless entities. This is the legal basis for an institution with a certain official background to intervene in today's industry management! Due to the widespread and negative nature of potential problems, target enterprises should have the self-awareness of collaboration and cooperation. Obviously, the achievement of a production limit consensus is conducive to maintaining common interests. When the industry association can overcome the prisoner's dilemma, target enterprises should naturally have the initiative to collaborate and cooperate. The production limit target should not be an "appeal" from a certain institution, but a "clause" with certain mandatory enforcement requirements, normative nature, and regulatory measures. Clause ”。

2. Target Price Management

After experiencing so many cycles, I finally understand: prices that are too low or too high are harmful to the industry. The disadvantages of low prices are obvious, but what are the disadvantages of high prices? The depth of our current cycle is the evil result of the greed that is difficult to restrain under the peak of prices. I always remember this sentence: To destroy it, one must make it crazy. If there are no measures to manage the upper and lower limits of prices, then the industry will always be in disorder. Those who benefited from high prices in the past will also become victims because of greed. In the last silicon material cycle, except for Daquan New Energy, which was able to restrain its production capacity and raised funds through listing to accumulate huge cash, several other silicon material giants invested the profits they earned with leverage into new production capacity, thus bearing a heavier burden.

The lower limit of target price management is the industry's average cost line. Therefore, the current market price of polysilicon of 40 yuan/kg is obviously inappropriate, because this is a price at which even the industry's most cost-leading giant company will lose money, especially considering that the current industry's operating rate is only 35%. The market price of 40 yuan/kg for polysilicon is a cruel reality that must be changed. Based on my long-term understanding of the polysilicon industry, the target price for the first phase of polysilicon should be 60 yuan/kg. Only at this price or above is it possible for enterprises to afford loans and recover their investments in the long run.

The upper limit of target price management is the impulse of new marginal suppliers to resume production. If the price recovers to a point where those un-commissioned and suspended production capacities have a strong desire to resume production, it is a warning sign, because this new supply capacity will quickly bring the price down. It should be noted that the price ceiling at this stage is not a fixed value. When the polysilicon sector has sufficient consensus and sufficient measures to restrict the resumption of production capacity, the price ceiling will be higher, until it recovers to a point where some enterprises have the impulse to add new production capacity. Based on the current polysilicon production technology and cost control, I believe this price is around 80 yuan. Therefore, the conclusion is that as industry consensus gradually strengthens, especially when a considerable degree of mergers and acquisitions and restructuring has been achieved, and the industry is in a relatively orderly state, the price ceiling is expected to eventually reach 80 yuan/kg. However, if the price is higher, potential new entrants will come in and disrupt the consensus that has been formed. In the long run, the price consensus ceiling under industry self-discipline should be controlled within 80 yuan.

Under industry consensus, the target price for polysilicon in the first phase of price management is 60 yuan. As consensus deepens and some production capacity no longer has the ability to resume production, the target price can be gradually increased appropriately, but it must ultimately be limited to within 80 yuan to suppress potential industry unrest. The goal is beautiful, but how to achieve it? What measures should we take?

3. Industry Stabilization Inventory

Price formation is based on continuous supply and demand signals. Behind the management of target prices is supply and demand management. Here, I propose the concept of "industry stabilization inventory" for polysilicon. The "standard" here refers to the target price. Due to the continuous fluctuation of industry demand, supply and demand often deviate from our target, so a certain amount of inventory adjustment is needed. When supply exceeds demand, inventory is absorbed; when demand exceeds supply, inventory is released, thus smoothing fluctuations and adjusting prices to approach our target price. This is the concept of stabilization inventory.

Polysilicon is an industry segment with a long production cycle and low production capacity elasticity. Due to its characteristics as a large chemical industry, even for already built production capacity, the time it takes for "cold" capacity to become quality-qualified "hot" capacity, even under the excellent management of a mature team, is at least 2-3 months. Based on these industry characteristics, I believe that the industry stabilization inventory should match this cycle, and the high value density, small space occupancy, stable quality, and easy long-term storage of polysilicon inventory also support the long-term storage of inventory for a 2-3 month cycle. Based on the 2025 annual demand of 1.46 million tons, each month requires 120,000 tons, so it is reasonable to establish an industry stabilization inventory of 300,000 tons. The current extremely low price of 40,000 yuan/ton is precisely the stage where crazy storage is needed.

Heaven helps those who help themselves; Buddha saves those with predestined affinity. We need to save our industry ourselves. The industry-related companies can initiate and establish the industry fund needed for inventory equalization. For 300,000 tons of long-term inventory, based on the current price of 40,000 yuan per ton, the industry fund needed for inventory equalization is 12 billion yuan. The contribution amount of each company to the equalization fund can be determined or referenced according to its inventory scale. For example, if Company A, a polysilicon manufacturer, has 60,000 tons of polysilicon backlog inventory, it can contribute 2.4 billion yuan, accounting for 20% of the equalization inventory fund. The equalization fund will then purchase 60,000 tons of inventory from Company A. The 2.4 billion yuan investment will eventually be recovered through the sale of the backlog products, and the funds will be in a balanced state. The only difference is that the inventory has been digested. After the equalization fund purchases the inventory, it will be sealed and disposed of, and will not be sold until the target price is reached.

As long as the industry reaches a consensus, we can backlog inventory into equalization inventory within a day. Polysilicon industry-related companies can quickly clear their polysilicon inventory, while the equalization inventory will be managed under the target price agreed upon by the industry, and will not become a driving force for disorderly price wars, but a guardian of industry order. After the inventory is cleared, the price of polysilicon products can quickly recover to 60 yuan/kg. The equalization inventory should also set 60 yuan as the first-stage target price management, and will not release a single gram of inventory until this target price is reached. The price of 60 yuan will not incentivize the construction of new production capacity, but it will make existing production capacity eager to operate. After 300,000 tons of inventory becomes equalization inventory, whether the price of 60 yuan/kg can be maintained long-term depends on the large amount of existing but un-commissioned production capacity, and these capacities are distributed among numerous entities. I believe that industry mergers and acquisitions are the optimal solution to this problem.

4. Industry M&A Fund

Inventory equalization is only a low-level consensus. If the consensus can be further consolidated, an industry M&A fund can be established to merge and completely shut down some production capacity, reduce the number of industry competitors, and thus make the industry more orderly.

Polysilicon investment has high capital intensity, a long investment cycle, and much industry know-how, so there are relatively few competitors. However, even so, the current polysilicon industry is still in a 6+N industry structure. The 6 refers to the six leading companies: Tongwei Co., Ltd. (820,000 tons of capacity), Xinyi Energy (480,000 tons), Daquan Energy (300,000 tons), Xin’te Energy (300,000 tons), Oriental Hope (255,000 tons), and Asia Silicon Industry (220,000 tons), with a total of 2.36 million tons of silicon material production capacity, enough to meet global demand. The effective delivery volume also accounts for at least 75% of industry demand; N represents all the remaining companies in the silicon material industry segment, numerous in number but accounting for a relatively small share of the total market. Some representative companies include Qinghai Lihao, Hesheng Silicon Industry, Xinjiang Jingnuo, Xinjiang Qiya, Baofeng Energy, Ningxia Runyang, Guyuan Oriental Sunrise, Guyuan Hongyuan Green Energy, Tianhong Ruike, and Qinghai Nanbo. The number of entities is still relatively large, and mergers and acquisitions are needed to reduce the number of competing entities in order to consolidate a higher-level consensus. However, there are two questions that need to be answered: How much money is needed for the M&A fund? Who will provide the money?

The money for the M&A fund is essentially provided by the market. Mergers and acquisitions and the closure of production capacity make the market orderly. The first-stage goal is to ensure that the price of silicon materials can remain stable above 60 yuan; the second-stage goal is to consolidate a higher-level consensus to gradually restore the price of silicon materials to above 80 yuan. Then the answer is simple: whoever wants their silicon material production capacity to continue operating should contribute to the M&A fund. For example, if Company A has 100,000 tons of silicon material production capacity that it wants to continue operating, and the price per ton increases from the current 40,000 yuan/ton to the expected 80,000 yuan/ton due to the orderly market, with an increase of 40,000 yuan per ton, then the company has an obligation to contribute 4 billion yuan; assuming that the industry needs to maintain a total production capacity of 1.5 million tons, then the companies with quotas will jointly contribute 60 billion yuan to establish an industry M&A fund to acquire and close some polysilicon production capacity, reduce the number of industry entities, and optimize the industry structure. This investment will eventually be recovered through higher product prices in the future continuous operation of the business, so I said that the money for the M&A fund is essentially provided by the market.

As mentioned earlier, the current silicon material industry is in a 6+N industry structure. The six leading companies have a total of 2.36 million tons of completed production capacity, while the numerous tail-end N companies have about 1.5 million tons of completed production capacity of various types. At an investment intensity of 70,000 yuan per 10,000 tons, the tail-end N companies have a total capital expenditure of over 100 billion yuan for the construction of production capacity. So, can a 60 billion yuan merger and reorganization fund complete the industry restructuring? The answer is yes, mainly because: 1. The 100 billion yuan investment is not all capital expenditure; on average, at least 50% is various bank loans, so 60 billion yuan in M&A funds is sufficient to complete the acquisition of equity; 2. For the equity capital portion, there should be a discount, on the one hand because the operation of polysilicon production capacity in the past two years has suffered serious losses, and on the other hand, many companies that rushed into the industry at the peak of the cycle have realized their mistakes and are willing to exit at a loss after bearing some losses.

IV. Mergers and Reorganizations, Not a Life-or-Death Struggle

Let's recall this scene from the history of the American Industrial Revolution: Rockefeller had just acquired an oil refinery and assigned it to his subordinate, who asked Rockefeller what he needed him to do, Rockefeller replied: Shut it down . When I first saw this clip as a young man, I had an inexplicable admiration for Rockefeller, but I also wondered: Since he never intended to use this production capacity, why did he spend a huge sum to acquire it?

Now, having witnessed and experienced so many business stories, I can fully understand Mr. Rockefeller's actions. And as a reader, if you have carefully read and savored this article, I believe you will also have the answer: The value of an asset depends on its future profitability, not its historical investment. Optimizing the competitive landscape, raising product prices, and improving profitability can significantly increase the value of existing assets of related companies, and vice versa. The loss of profitability can instantly turn assets worth hundreds of billions into scrap metal.

The great wisdom of acquiring and closing down lies in the exit path of failed competitors . In the ups and downs of the business world, we inevitably bear too much personal responsibility for our enterprises: the leadership responsibility for failed investment attraction projects with Chinese characteristics, the bank responsibility signed under personal guarantees, and the social responsibility of countless suppliers and employees. The responsibility is too great, the problems are too many, and it is not easy to say that we want to quit, we can only persevere and not give up. On the other hand, the era of reform and opening up has brought about dividends, making the decision-makers of China's manufacturing industry today still the generation born in the 1960s. Their difficult upbringing has made them unafraid of life-or-death struggles, and it seems that for enterprises with competitive disadvantages, the only way is success or death The only way out is mergers and acquisitions, which were never considered by these predecessors. Various factors have led to the fact that in China, once an industry experiences overcapacity, it always takes an extremely brutal price war that lasts a long time to barely clear the inventory. At this time, even industry-leading companies with competitive advantages are severely weakened, which has formed the "involution" trend that we have been talking about, and it is difficult to extricate ourselves from it.

Acquiring and closing down, allowing failed competitors to exit relatively decently, and optimizing the market structure will significantly benefit large companies with huge production capacity. Ultimately, they will earn back the cost of regulating the industry order from the market. Therefore, from the perspective of responsibility, rights, and interests, mergers and acquisitions are not only an option for large factories but also a necessity. In today's situation where the industry structure is difficult to solve, where production capacity cannot be digested through demand growth, and where supply-demand balance must be achieved through production capacity clearance, the leading large factories naturally have the responsibility of mergers and acquisitions and will also enjoy the benefits of mergers and acquisitions. Mergers and acquisitions, rather than a life-or-death struggle, the fate of large factories will also be redeemed through mergers and acquisitions under consensus.

V. "Breakthrough to Rebirth" and "Turning Point"

I really like the Chinese term " crisis " crisis " which contains the spiritual core of both danger and opportunity. The more I experience cyclical baptism and the ups and downs of the industry, the more I like this term. Whenever intense contradictions in a field reach an irreconcilable level, it often means that a turning point is coming. Therefore, being in the current predicament of the photovoltaic industry, I see the following opportunities:

1. We need industry consensus more than ever before; crises can facilitate consensus.

2. A sufficiently large crisis can lead to a sufficiently deep industry consensus, thereby promoting a transformation of the industry competition paradigm.

3. The problems faced by the Chinese photovoltaic industry in competing in the global market will serve as an example for other fields to learn from.

4. The traditional photovoltaic giants led by the generation born in the 1960s have been severely weakened, but there is always hope after a crisis. A new batch of entrepreneurs will take to the historical stage and become the new decision-makers of the era.

Those who know me will know that I transitioned from the financial field to the real industry in 2019 to become part of photovoltaic manufacturing. Undoubtedly, I am also experiencing the pain that all photovoltaic people are experiencing. Therefore, some people are worried about my mental state. Here, I would like to express my gratitude to everyone who cares about me and tell everyone that I am in good spirits and am actively working to solve a series of complex problems and striving for a new turning point for the company. For my age, what I am experiencing today is not necessarily a bad thing; it is making me stronger than ever before. Although the photovoltaic industry is currently experiencing an unprecedented cycle, and although the ceiling of photovoltaic component demand is obvious, the great cause of human energy transformation under the photo-storage-charging system is still enough to last for another 100 years. There will still be countless opportunities to emerge, and countless cycles to experience. Let us maintain a youthful mentality and become more experienced veterans in the industry.

" Breakthrough to Rebirth " is the theme I set for the photovoltaic industry in 2025. The "break" here can be the break of bankruptcy or the break of breaking through. The difference between these two breaks is our current industry consensus. The general trend of the industry is unstoppable. Those who follow it will break through, and those who go against it will go bankrupt. Today, no company or individual can stop the general trend of industry consensus. We photovoltaic entrepreneurs should organize ourselves with a more positive attitude to deal with the current industry crisis and abandon the previous life-or-death struggle mentality. Only in this way can we turn the tide.

A Few Thoughts

This dry, lengthy, and abstract discussion may be difficult to read. If any reader has carefully read this word by word, I would like to sincerely thank you for your reading and company, because there are too few people who are truly worried about the industry and can think about it. The researchers who were active in the photovoltaic capital market in previous years are now nowhere to be found, and people in the industry generally suffer from the problem of being too close to the situation to see it clearly, so even fewer people can think deeply about the photovoltaic industry. Compared with the annual outlook in previous years, this year's article is the least like an outlook. It is more of a hope, my personal hope for the healthy development of the industry, a hope for a new pattern in the photovoltaic industry, and a hope that we can all successfully get out of the crisis and turn the tide.

Zhang Zhiyu

May 1, 2025