Market sentiment experienced a sharp downturn on July 17th, with the Shanghai Composite Index declining 3.05% and the ChiNext Index falling 7.15%. The total turnover for all A-shares reached 2.67 trillion yuan, representing a 10% increase from the previous trading session. Most sectors closed lower, with defensive dividend-oriented sectors showing relative resilience. Banking (+0.90%), oil & petrochemicals (+0.40%), and utilities (+0.11%) managed gains, while communications (-9.84%), electronics (-8.19%), and machinery & equipment (-6.61%) led the declines. In terms of style, dividend-focused indices like the SSE Dividend Index (-0.47%) were relatively resistant, whereas growth-oriented styles, exemplified by the STAR 200 Index (-9.53%), were at the forefront of the sell-off.
The continued slump in overseas computing power stocks is further dampening risk appetite, with short-term irrational panic emerging in the domestic equity market. Since late June, the market decline has accelerated, exhibiting signs of irrational panic. From the end of June, the market has seen persistent heavy losses, with the Shanghai Composite's maximum drop reaching 9.6%. The ChiNext Index, growth-style stocks, and key industrial sectors like communications and electronics have suffered maximum declines of 21.6%, 22.5%, 27.4%, and 27.9%, respectively. The pace of decline this week has been notably steeper than in the first week of July, indicating a typical acceleration in the sell-off. While recent market disturbances have been numerous—including renewed tensions in US-Iran relations, a slowdown in China's Q2 GDP growth to 4.3%, and the upcoming Changxin IPO potentially straining micro-liquidity—market expectations had already largely priced in these events. Today's sharp decline lacked a specific, sudden negative catalyst, highlighting the market's short-term irrationality and trend-accelerating characteristics.
The ongoing correction in the overseas computing power sector is suppressing risk sentiment for domestic tech stocks. Recent collective pullbacks in overseas computing power stocks have significantly dampened risk appetite for related domestic sectors. South Korea officially introduced restrictive policies on July 16th, substantially raising the threshold for single-stock leveraged ETFs. On that day, SK Hynix and Samsung Electronics fell sharply by 12% and 9%, respectively. The Philadelphia Semiconductor Index has retreated up to 19% since late June. On July 16th, key US computing power stocks like Micron Technology, AMD, Marvell Technology, Google, and Oracle dropped 6%, 5%, 9%, 4%, and 6% in a single session. The sharp decline in overseas tech sectors has transmitted overnight risk to the A-share market.
Reiteration of the Second-Phase Earnings-Driven Rally
The second phase of the current industrial cycle rally is not over. The AI industry's current upcycle has entered its second phase of growth. Historical analysis suggests that such second-phase rallies typically last around eight months, with the ChiNext Index potentially gaining about 60%, growth-style stocks rising 40-50%, and leading industrial sectors advancing 70% or more. Since the rally's inception on April 7th, the ChiNext Index, growth-style stocks, communications, and electronics have achieved maximum gains of 39%, 39%, 65%, and 91%, respectively, over approximately three months. Overall, compared to historical patterns, the current second-phase rally still has room to run in terms of both duration and magnitude.
The recent rapid market adjustment has already exceeded the upper limit of similar historical declines. Historically, following the initial rapid ascent in the second phase of a growth industry cycle, adjustments for growth-style stocks and leading industrial sectors typically range around 10% and 10-20%, respectively. The market correction since late June has surpassed the upper bounds of such post-rally adjustments, making a further deep short-term correction less probable. With the mid-year earnings previews disclosed in mid-July, some computing power companies have reported strong results, continuing to provide robust support for the sector's high growth narrative. Should a rebound materialize, historical precedent suggests growth-style stocks and leading sectors could see gains of 10-20% and outperform during the recovery period. There is a reasonable probability that growth-style stocks and leading sectors could rebound to new highs; even if they fail to surpass previous peaks, they are likely to approach them closely. More importantly, further breakthroughs beyond previous highs remain plausible in the accelerated final stage of the second phase. Given the significant oversold condition in growth sectors since late June, we believe it is not the time for continued pessimism.
Short-Term Outlook and Investment Strategy
The market's irrational decline this week is pronounced. Reviewing similar historical declines suggests the market is already oversold in the short term, making further substantial losses less likely. As tech company earnings continue to validate high growth and expectations for large IPOs are realized, the market is poised for a potential rebound. The primary investment theme should remain focused on the upstream and midstream segments of the AI industry, a structural trend expected to persist until year-end. Regarding allocation strategy, the logic of prioritizing growth sectors with strong industrial cycle earnings, amid a still-weak economic cycle, remains clear. The second-phase earnings-driven rally is ongoing, and the significant correction in growth and tech sectors since late June has created more upside potential. Therefore, investors can continue to concentrate on increasing exposure to the AI industry trend, with computing power centers and supporting hardware in the upstream/midstream being the most indispensable areas. This segment boasts a sufficiently long industrial chain and investment capacity and is currently in a mid-phase of significant institutional concentration. Although some areas may show signs of short-term trading overheating, the sector is broad enough to allow for rotation within it ("changing horses midstream") without altering the medium-term uptrend. When trading cools and turnover ratios reach low levels, it presents a buying opportunity. Overall, investors should focus on the following two investment themes and their timing:
The first core theme is the AI industrial chain, with particular emphasis on upstream and midstream computing power hardware. The second-phase rally of the AI industry cycle is being validated and is set to continue. Based on the timing, magnitude, and structural patterns of second-phase rallies, upstream and midstream AI segments offer significant potential for outperformance with considerable room for gains. Monitoring turnover ratios can help refine timing; while medium-term opportunities remain intact, the communications sector, with its currently lower turnover ratio, may offer better value relative to electronics in the short term. It is advisable to position for a potential rebound in computing power center and supporting infrastructure areas likely to outperform, such as optical modules, fiber optics, liquid cooling, data centers, and AI data centers (AIDC).
The second theme involves sectors likely to benefit from the diffusion and catalysis of the AI chain, primarily including machinery & equipment, robotics, gaming, and software. These sectors often represent extensions of the current AI industrial chain's upstream and midstream. For instance, machinery & equipment benefits from both an upcycle in construction machinery and thematic spillovers from robotics, commercial aerospace, liquid cooling, and high-end manufacturing. Robotics, gaming, and software are the most promising downstream directions to receive thematic diffusion from AI computing power.
Key Risk Factors to Monitor
Potential risk transmission from concerns over valuation digestion following sharp rallies in US chip stocks; renewed volatility in US-Iran tensions; a rapid economic slowdown in China echoing past calendar-year effects; persistent market disturbances from expectations of Federal Reserve monetary policy tightening; and limitations or deviations in learning from historical patterns of growth-style stock performance.
Comments