Let's start with the semiconductor industry tracking:
SMIC Performance Exchange Meeting: SMIC held a performance exchange meeting in the morning, and I listened to it, and no one consulted about 7nm during the whole process (it is expected that the company said hello in advance, probably for confidentiality);SMIC's performance is expected to be low, and it does not reflect the performance of foundry Kirin chips at all, and the company's performance exchange meeting does not mention it, and no one is allowed to ask. However, SMIC's performance disclosure this time the market only saw that the capital expenditure exceeded expectations: the good targets were Yaxiang Integration and Shenghui IntegrationDue to yesterday's Nvidia's H20 card, a painful blow to domestic chips, the market continues to pay attention to advanced packaging and testing, because there is no advanced packaging and testing, the progress of domestic AI chips has a bottleneck, so it continues to speculate on Wenyi Technology, Qiangqiang New Materials, two equipment and material stocks.
Storage**Weekly Tracking:
The recovery rhythm of DRAM modules has been adjusted, and the demand for NAND AI has been strongDRAM: The spot price has narrowed, and the recovery rhythm of module manufacturers has been adjusted. According to DRAMexchange, last week (1031-1106) the spot prices of 18 categories of DRAM were -29%~+4.2%, with an average width of +05% vs. +11%。Last week, 11 of the 18 parts were opened, and the DDR4 was significantly better than DDR3 and DDR5. NAND: The spot price has dropped slightly, and the demand is dominated by AI products. Last week (1031-1106), the spot price of NAND particles was -4 month-on-month6%~+0.4%, and the average overlap of 31 categories was -07% vs. +05%。Among them, 6 models are ** up, 14 models ** are flat, and 11 models ** are down. Last week, NAND's medium and large capacity part numbers** generally stabilized or increased, while the lower part numbers were mainly concentrated in small capacity parts such as 1GB and 2GB. Last week, spot demand was still dominated by AI-related products, and some spot inventories were adjusted along the trend**, but there was no significant pull-up. Due to the large previous stacking (price reduction of more than 70% at the high point), which is seriously lower than the cash cost price, large manufacturers are bound to open above the breakeven line (there is still a large gap at present), and the price is expected to continue in the fourth quarter. Pan Jiancheng, CEO of Phison, a NAND flash control IC manufacturer, said on the 7th that with the successive production reductions of major memory chip factories and the recovery of some end markets, the NAND market has begun to strengthenRecently, customers have been more active in pulling goods, but the market supply continues to shrink due to the production reduction of large manufacturers, and it is expected that there will be a shortage of goods next year. Some consultants are optimistic that the global storage market size will increase by 38% year-on-year (2023e: -42%) to $110.5 billion in 24. Under the cycle reversal, the order of attention is module and peak measurement, supporting chips, and li-based storage design. It is expected that after the 24-year cycle is upward, the module and peak measurement link will be the first to benefit, and the normalization of the inventory level of the module link will drive the demand for supporting chips and li-based storage design.
The double bottom of PV has appeared, and the expectation has been pessimistic to the extreme, and the future will be better and better.
Those who understand whether they understand photovoltaic are overcapacity, who dares to blindly expand production, many planned production capacity may be stranded, and the overall overcapacity is seriously expanded, because in the end there is not so much real implementation, and the future stable and high-speed growth demand is selectively ignored when the negative is everywhere, and the wave of photovoltaic emotions is really everywhere. The bears are over-rendered, and after the bears are exhausted, there will be more and more benefits. First, controlling IPO and refinancing is the most beneficial to the photovoltaic industry, and controlling production capacity from the source has not already been solved. Second, new entrants will naturally stop when polysilicon, wafers, cells, modules, auxiliary materials and other links are not profitable, and in the future, we will often see the postponement or suspension of production of various projects. Third, the Fed stopped raising interest rates and even started the interest rate cut cycle, which is the most beneficial two industries, that is, photovoltaic and innovative drugs, and high interest rates have seriously hindered the installed capacity of photovoltaics in European and American countries. Fourth, at the San Francisco summit next week, the small essay on the United States' relaxation of photovoltaic imports has been circulating for many days, and there is no wind and no waves. Fifth, after the sector is cut in half and then cut in half, the valuation stock price is at the lowest position in the history of extremely undervalued, and the extreme low and extreme undervaluation itself is the biggest benefit.
It is the nature of the company to attract traffic and attract people's attention, and it is the norm to exaggerate and exaggerate over-exaggerated. Let's talk about photovoltaic involution and excess, it is indeed but far from being as exaggerated and scary as advertised. They confuse planned production capacity, nominal production capacity and effective production capacity, completely ignoring the elimination of backward production capacity and the still high-growth industry demand. There is definitely a surplus of polysilicon, but when Tongwei Daquan is not very profitable, will the PPT capacity of other new entrants still be put into production?The backward production capacity of silicon wafers and cells has two or three hundred GW, and the first thing to be eliminated after the war starts is these backward production capacity, such as the widely criticized TOP cell planning capacity of thousands of GW, how much can be landed by the end of the year, except for Junda Co., Ltd., most of the others are the head integrated companies Jinko JA Solar, Trina, Canadian Solar, etc., and how many new entrants across the bank will land and when will they be put into production?The bucket effect supply should not look at the most but the least link, which is part of the reason why Jinko said that there is a shortage of backward overcapacity, high-quality and efficient production capacity.
Many people like to compare the photovoltaic and pig industry, they only focus on the similarity of the supply side and ignore the huge difference in the demand side, the demand for pork is stable every year, almost no growth, photovoltaic will maintain double-digit rapid growth for many years in the future. This year, the global installed capacity is about 400GW, and the market is expected to grow by 20-30% next year, taking the median value of 25% growth rate, that is, 500GW demand, according to 1The demand for 3 times the module is 650GW, and it is likely that it will continue to grow at a rate of 15-20% in the next few years. In this way, it can be quantified to see if there is less worry and worry about overcapacity.
Let's put it this way, after a period of clearing and washing in the future, the photovoltaic industry is still one of the most worthy of investment. How long is the future period of time, maybe March or May, or one or two years, but the stock price will bottom out in advance, and the double bottom of the photovoltaic ETF a few days ago is not a real historical bottom, let the market prove it.
Article**: