July Outlook

In the second half of the year, focus on the core industry主线, clarify allocation priorities, and adhere to directions within each sector that have higher capital recognition and industry景气度. As the window for mid-report earnings disclosures approaches, I continue to place the storage track at the top of computing subdivisions, with allocation priority higher than optical communication. The core logic lies in the significant gap in market size between the two industries. Currently, short-term market capital is concentrated on the storage sector, with a clear tiered景气 structure within its subdivisions: storage颗粒 leads in景气度, followed by supporting equipment for颗粒 factory expansion, and then niche storage modules and distribution companies. Among these, DDR4 supply and demand remain tight, providing relatively outstanding industry profitability stability. The storage sector's rally has entered the mid-to-late stage, with sector heat gradually approaching its end. During the mid-report earnings disclosure period, storage companies still have short-term catalysts from earnings, but the sustainability of the rally is uncertain. Combining industry景气度 and valuation positions, the recommended allocation order for each track is: storage > optical communication > PCB > domestic substitution > domestic信创 > semiconductor equipment > liquid cooling. Understand the priority of tracks, achieve unity of cognition and operation, and once you understand the underlying industry logic, do not waver repeatedly. Avoid dispersing capital into niche, low-景气 support targets that lack attention.

Optical communication has a clear long-term growth logic, with vast industry development space. Its overall market size is similar to that of the PCB industry. Even under the global tech capex framework, the combined share of these two sectors remains low, providing ample medium- to long-term growth room. However, there are obvious current issues: short-term capital clustering is overheated, and many targets' stock prices have already priced in long-term industry expectations prematurely, making current chasing less cost-effective.

The power semiconductor track was previously outlined. At this stage, sector rallies are more driven by short-term event-based sentiment catalysts. Multiple industry catalysts have landed sequentially: Infineon raised prices on related products, some categories saw tightened supply; South Korea also proposed making power semiconductors a key development industry. These news have boosted short-term market sentiment. But investment should not be purely driven by short-term sentiment; we need to clarify the underlying industry logic: digital semiconductors handle digital logic signals; power semiconductors manage current transmission, directly determining core hardware performance such as fast charging, high voltage tolerance, and high-temperature operation. High-power new energy vehicle fast charging and AI data center high-load power supply scenarios cannot do without power device support. SiC, GaN, and silicon-based power devices are the medium- to long-term core growth drivers of the track. Currently, the sector's short-term sentiment speculation is relatively heavy and needs rational viewing. Be cautious of subsequent sector pullback volatility risks. Even if industry景气 remains upward, individual stocks still have volatility uncertainties; heavy positions in a single target are not recommended. The short-term rally dividend of storage is gradually ending, and market capital will continue to seek new allocation directions. The core certainty in the second half of the year lies in the entire domestic super node industry chain, with domestic信创 and semiconductor equipment likely to see independent structural rallies. The global trend of independent tech development is clear, and overseas high-end computing hardware quotes continue to rise. According to public industry information, affected by export control tightening, the quote for a single H200 has exceeded 4.1 million, and the B300 equipment quote has reached over 8 million, forcing the entire domestic chain to accelerate autonomous controllability. The premium circulation prices of such high-end computing equipment also indirectly accelerate the pace of domestic super node and computing hardware autonomous substitution. The independent development of the domestic tech industry chain is a long-term clear direction. Combined with continued policy support and gradual order fulfillment, related tracks have medium- to long-term allocation value that can withstand market turbulence. The current market environment is complex: traditional sector recovery pace is weak, consumer market复苏 is below expectations; AI tech-related sectors have generally high valuation levels, with much industry divergence and uncertainty, making overall operational profitability more difficult. At the operational level, I will continue to balance risk/reward, actively avoid chasing highs, steer clear of areas with high capital congestion and completely unanimous market views, and adhere to the core industry主线 logic. Prioritize mining targets with expectation gaps, lower stock price positions, and near the annual moving average range. Don't force huge short-term returns from a single target. Deploy in stages across different景气 cycle tracks, grasp波段 opportunities in phases, achieving higher operational safety. Reduce meaningless frequent换手, follow波段 operation rhythm, patiently layout and wait for industry realization. In the second half of the year, there is an opportunity to seize good structural opportunities.

The biggest taboo in trading is fickleness; frequently switching tracks only leads to getting hit from all sides. Large silicon wafers, indium phosphide substrates, pump EML, PPO resin, SiC, CMP, advanced packaging, de-Japan materials, liquid cooling, storage, optical communication, FT probe—the entire industry chain—once you identify them, hold them steadfastly from beginning to end. This is a typical structural bull market. After half a month of adjustment, one day can recover all lost ground. Mid-June was suitable for stock selection and positioning. Now it's late June, the positioning window has closed, and the only correct operation at this stage is to hold positions and wait for the rally to materialize.

Most people stare at the Shanghai Composite Index, while I have been staring at the科创 Chip ETF—the main line is all here. By mid-July, many retail investors were previously focused on watching events and not entering the market. Once the events end, this wave of sidelined capital rushes in collectively, easily forming a short-term top. Many people wait for the event to end before chasing highs; this is just a reasonable inference. Whether the market will be "poor in May, desperate in June, and turnaround in July," or "surge in May, attack in June, peak in July," or repeat last year's rally until National Day—the answer lies with capital. Every time the Shanghai Composite Index hits a new high, be more vigilant. Over the next decade, the global economy enters low growth, various black swan events will only become more frequent, and black swans have become the norm. Those who survive long-term cannot seek a safe haven; they must cultivate themselves like a ping-pong ball—the heavier the pressure, the stronger the bounce.

Currently, as stock prices move higher, the trend becomes clearer and win rates improve, but the risk/reward ratio continues to shrink. Positioning at low levels is more cost-effective. The higher the stock price, the less worthwhile the risk/reward. By the time everyone sees the upward trend and enters, there is no longer any positioning significance. The trading model must be replicable long-term; you cannot merely计较 gains and losses of one city or one territory. Risk/reward is always equal; only focus on high-odds low-price varieties, and resolutely avoid high-position themes. Set a firm rule for yourself: do not participate in any high-position heavily speculated targets; only ambush low-position subdivisions with expectation gaps. From beginning to end, the capital market earns entirely from expectation gaps. Continuously exceeding earnings expectations is an expectation gap; the gradual realization of industry景气 is also an expectation gap. Mundane, changeless, mediocre tracks with no variables do not offer excess returns and are not worth waiting for over a long time.