What recent changes in the technology style are worth paying attention to?

In the past two weeks, we mainly discussed the seasonal effects of the dividend style ("How to view the 12.5% probability of the CSI Dividend Index outperforming the entire A-share market in June?") and the potential catalysts for subsequent index-level market trends ("Trigger to break free from the anxious oscillation range"). Entering June, we further observed several important changes in the technology sector that are worth noting, and this week we provide further analysis. Important Change One: Dividend assets release liquidity, potential incremental funds for the technology sector Starting in mid-June, dividend stocks entered a concentrated annual report dividend window period. Historically, after the ex-dividend date, dividend assets face realization pressure, and the frequency of dividends is basically negatively synchronized with the performance of the CSI Dividend Index. Most representative dividend leaders have shown a decline in excess probability (relative to the Wind All A) after the ex-dividend date. As the dividend sector adjusts, funds are flowing out of defensive assets. Under the premise that there is no significant downside risk for the index, this creates a favorable funding environment for other styles, bringing potential incremental funds to the technology sector. Indeed, upon review, there is a seesaw effect between technology and dividends in mid to late June. Especially after mid-June, when dividends turned downward, the trend of excess returns in the technology style became more pronounced. From the perspective of financing balance, the current position is relatively low. Compared to the significant increases in financing balances in February 2024, the end of 2024, and February 2025, the financing balance has remained at the bottom this year since the sharp rebound in April due to the equal tariff, with no signs of recovery, which also provides the possibility of incremental funds for subsequent market trends. Important Change Two: Marginal changes overseas, 232 investigation and overseas technology progress Based on the statements made by the U.S. Secretary of Commerce during an interview in April, the Trump administration's initiation of a "232 investigation" into the semiconductor sector may land in mid-June, which can be considered a suppressive factor for the TMT sector as the "shoe drops." Additionally, since May, the fundamentals of U.S. tech stocks have been continuously accumulating positive factors, and the market has returned to being performance-driven. The investment logic for overseas tech stocks has returned to CSP maintaining CAPEX investment, exploring higher model capabilities and reasoning abilities through training like Blackwell, and the stock prices of representative overseas companies have returned to new highs Previously, there was a clear "U.S. stock mapping" pattern in the domestic AI sector, where the stock prices of leading companies mirrored the performance of their U.S. counterparts. However, this round has shown a significant lag in growth, with stock price performance diverging, mainly influenced by recent factors (such as the aforementioned 232 investigation, the pace of domestic technology industry catalysts being slightly below expectations, and the diversion of funds to other strong sectors). Currently, the transaction concentration ratio in the TMT sector has reached the lower bound of the AI narrative range for 2023. With changes in the capital landscape, a easing of uncertainties between China and the U.S., and an acceleration of technological catalysts, the "overseas mapping" effect on Chinese tech stocks is expected to become apparent. Most companies are currently valued below 20X, and this week, overseas AI chain companies in the A-share market have started to perform. Important Change Three: Intensive Technology Catalysts in June June saw a concentration of catalytic events in the technology sector, including the nearing conclusion of the 232 investigation, and β-level stimulating events such as the Lujiazui Forum, as well as significant marginal changes across various fields including consumer electronics, domestic AI large models, computing power both overseas and domestically, and commercial aerospace. The intensive release of major companies in June could be very crucial. Finally, based on the relative performance of stock prices, we have ranked the popular technology sectors: ① High cost-performance ratio, with attention: This includes directions that have seen market activity since the beginning of the year but adjusted in March (such as domestic AI chains), as well as sectors that have historically been active but have not traded β this year (such as commercial aerospace); ② Medium cost-performance ratio, with general attention: Mainly focused on directions affected by trade friction, with recent positive news being dulled. If short-term negative news is exhausted, there may be a possibility of a rebound (such as consumer electronics); ③ Low cost-performance ratio, with high attention: Directions that have not adjusted or have just adjusted in the past month (such as robotics). Based on this relative cost-performance thinking, the technology sector is recommended to focus on:【Commercial Aerospace, Autonomous Driving, Domestic Computing Power/DeepSeek Chain】>【Consumer Electronics, Overseas Computing Power】>【Domestic Robotics, Solid-State Batteries】. Report Body 1. Topic of this issue: What recent changes in the technology sector are worth paying attention to? **In the past two weeks' weekly reports, we discussed the seasonal effect of dividend style (《How to view the 12.5% probability of the CSI Dividend Index outperforming the entire A-share market in June? (1) Main Influencing Variables in the June Market After the disturbances from the April earnings season clear, A-shares usually experience a round of risk appetite recovery in early May, with the market generally continuing until mid to late May. The subsequent weak oscillation phase typically lasts about one month until the end of June. In addition to the limited sustainability of purely risk-driven market movements, there are also some liquidity disturbance factors in the macro context in the later part of the second quarter that indirectly affect market valuation levels: 1. Window Period for Changes in Top-Level Macro Policies: After the Politburo meeting at the end of April, the next window for adjusting economic policies at the central level will not be until the end of July. In the later part of the second quarter, economic policies are more about local and departmental implementation of the spirit of the April Politburo meeting, and this factor is usually priced in by the end of April to early May, so the market typically does not face upward risks from the macroeconomic level. 2. Crowding Out of Real Demand: The second and third quarters are a seasonally active phase for the real economy, with vigorous production activities crowding out the space for funds to "shift from the real to the virtual," thus constraining the liquidity of financial assets. 3. Disturbance from Mid-Year Bank Assessments: At the end of June, banks face semi-annual assessments of both assets and liabilities, where the incremental credit supply is absorbed by the active demand from the real sector. Completing savings targets theoretically leads to a temporary redistribution of residents' asset structures, potentially having a negative effect on equity asset allocation scale. 4. Corporate Tax Payments: The end of May is the deadline for corporate income tax settlement and payment, requiring corporate sectors to gather funds to cope with potentially large one-time tax payments, which leads to a concentration of corporate assets more towards cash-like liquid assets, also temporarily affecting the allocation scale of equities and other assets. Other disturbance factors include mismatches in mid-year fiscal revenue and expenditure rhythms, accelerated government bond issuance siphoning, and concentrated annual report dividends from listed companies, all of which are temporarily unfavorable to macro liquidity and equity liquidity. This results in a typically weak oscillation in the A-share market in June. Since 2010, the main indices have generally followed a rhythm of "oscillation at the beginning of the month → adjustment in the middle of the month → recovery at the end of the month" in June. The ChiNext and small-cap indices, which are less correlated with macro liquidity, perform slightly better than the main board and mid-to-large caps. (2) Important Change in the Technology Sector: Dividend Assets Release Liquidity, Potential Incremental Funds for Technology Assets Although the index is weak in June, the performance of defensive dividend assets does not significantly outperform. From 2009 to the present, the CSI Dividend Index has performed weakly in June, with probabilities of outperforming the CSI 300 and Wind All A being only 25% and 12.5%, respectively, and an increase probability of only 37.5%, which is a noticeable decline compared to May. From the performance of major dividend sectors, the probability of rising and outperforming in mainstream dividend sectors such as banking, transportation, utilities, and coal is also not high. Among them, the transportation sector had a less than 20% chance of outperforming in June, with only the white goods sector performing relatively well. If we further observe the situation since 2021 (when the effectiveness of dividend strategies has been further strengthened), the overall performance in June remains weak. ** Specifically in June, dividend assets actually performed well at the beginning of the month, with the adjustment window usually starting in mid-June. Since 2010, the China Securities Dividend Index has averaged weak fluctuations in the first half of June, turning downward towards the end of the second trading week, with adjustments almost throughout the entire late June, only experiencing some rebounds on individual trading days at the end of the month (but the magnitude is also weaker than the index gains mentioned earlier). Annual report dividends are an important suppressing factor for the seasonal weakness of dividends starting in June (especially in the latter half of the month). June and July are intensive time windows for the annual report dividend equity registration dates (or ex-dividend dates) of dividend assets. From 2009 to now, the proportion of dividends in June for the China Securities Dividend Index component stocks is close to 40%, and the proportion in July is around 30%. In the past two years, the proportion of dividends in June for the China Securities Dividend Index has further increased. In terms of rhythm, the frequency of dividends is basically negatively synchronized with the performance of the China Securities Dividend Index. The ex-dividend dates (the day after the equity registration) of the China Securities Dividend Index component stocks remained low in the first two weeks of June, surged in the third week of June, and then decreased at the end of June, reaching a second peak of dividend equity registration in mid-July, followed by a weekly decline. The most direct logic is that as the equity registration date approaches, the demand for cashing out dividend stocks is temporarily suppressed and will be concentrated after the registration date, thus suppressing performance after the ex-dividend date. Observing representative dividend leaders, most individual stocks have shown a decrease in excess probability (relative to the Wind All A) after the dividend ex-date. With the adjustment of the dividend sector, funds are being withdrawn from defensive assets. Under the premise that there is no significant downward risk in the index, this creates a favorable funding environment for other styles. Theoretically, the opposite of defensive assets, the more aggressive technology sector, will benefit. This is indeed the case. A review shows that in mid to late June, there was a seesaw effect between technology and dividends. Since 2010, the Wind Technology Major Index has performed relatively well in June, with an average return of 1% and a win rate of 67%. Moreover, June has generally maintained positive returns since 2019. Particularly, after mid-June, when dividends turned downward, the trend of excess returns in the technology style became more pronounced. However, there are two points worth noting regarding the response strategy for the technology style: ① There may be a resonant adjustment. A review shows that the adjustment of dividend weights in mid-month can drag down index performance, indirectly causing the technology style to experience a certain pullback in mid-month from a β perspective. From the perspective of precise timing, theoretically, the optimal return point for technology is after the resonant adjustment day of dividends and technology. ② Structural characteristics of the market. The upward trend of the technology major index has a large amplitude, which means that the technology market is likely to have structural and rotational pulse characteristics, requiring more attention to marginal catalysts and segments with superior cost-effectiveness. From the perspective of financing balance, the current position is relatively low, with potential for future increases. Compared to the significant increases in financing balance in February 2024, the end of 2024, and February 2025, we see that since the sharp rebound from the equal tariffs in April, the financing balance has remained at the bottom for this year without any recovery, which also provides the possibility of incremental funds for the subsequent market. (3) Important Change in the Technology Sector: 232 Investigation and Overseas Technology Progress 1. 232 Investigation: According to an interview with the U.S. Secretary of Commerce in April, there is a possibility that the semiconductor "232 tariffs" will be implemented in mid-June. On April 14, the Trump administration initiated a "232 investigation" in the semiconductor field. The products under investigation include semiconductor substrates and bare chips, traditional chips, advanced chips, microelectronic components, and downstream products containing semiconductors, basically covering parts of the "equal tariffs" products that were exempted on April 11. In the same month, U.S. Secretary of Commerce Gina Raimondo stated in an interview with ABC that the tariff exemptions for technology products such as mobile phones and computers were only "temporary." Raimondo also mentioned that these products would soon be included in the so-called "semiconductor tariffs," which are expected to be implemented "within a month or two." **If, as Raimondo said, the "232 tariffs" may be implemented in mid-June, this could also be a "shoe dropping" factor for TMT suppression, and the pressure on sectors such as electronics is expected to ease 2. New highs for US tech stocks and the "overseas mapping" transmission to A-shares This year, the market for US tech stocks has experienced a shift from "macro narrative" to "fundamental pricing" — (1) A significant adjustment in Q1, mainly due to ① the impact of DeepSeek's computing power deflation expectations; ② NV's short-term capacity and yield pressures; ③ macro pressures in the US stock market. (2) Entering April, reciprocal tariffs once again formed an impact, and the H20 ban led NV to recognize a $5.5 billion asset impairment. By April, the macro narrative of DeepSeek and tariffs had a profound impact on tech stocks. (3) Starting in May, the fundamentals of US tech stocks continued to accumulate positive factors, and the market returned to being performance-driven. ① At the beginning of May, the four major CSPs disclosed their financial reports, maintaining high CAPEX investments, with META unexpectedly raising its full-year capital expenditure to $64-72 billion (the guidance for Q4 2024 was $60-65 billion), alleviating concerns about CSP's reduction in computing power investment; ② From May to June, the earnings season began, with impressive performances from Nvidia, Credo, Broadcom, and others; ③ Continuous positive factors from the industry side, including TSMC raising CoWoS capacity, improvements in GB200 capacity, upward revisions of 800G demand for 2026, and potential launches of H20 alternatives. Thus, the logic of US tech stocks returned to CSP maintaining CAPEX investments, training through Blackwell, and exploring higher model capabilities and reasoning abilities. In terms of A-shares, how do US tech stocks influence A-share investments? Historically, the main trading logic of A-share's overseas AI chain is the mapping of US stocks, but this round of A-shares has clearly underperformed US stocks. Under concerns of chip suppression, the A-share tech sector still shows a gap. As shown in the figure below, the market synchronization between Nvidia and Xinyise is extremely high, and the price turning points of Zhongji Xuchuang, Industrial Fulian, and others also correspond closely. Theoretically, the rise of US tech stocks in this round should drive A-shares. However, comparing the performance of A-shares and US stocks: since the end of April, the stock price performance of A-share overseas AI chain companies has been relatively average. Except for core companies like Xinyise, which were directly boosted by ASIC, most companies' stock price performances have significantly lagged behind US stocks, creating a gap with US tech performance. This week, A-share overseas AI chain companies began to perform (for example, Shenghong Technology has risen 22% since the end of April, with a 16% increase this week), and currently, most companies are valued below 20X. Under the conditions of the style rotation mentioned earlier in June, the US tech market is expected to continue to map to A-shares, and the overseas AI chain may become a "trigger" for the rebound in TMT trading activity 。 (4) Important Change Three: Concentrated Catalysis of the Subsequent Technology Industry Chain In June, the heavy catalysis of the technology industry chain is relatively dense, including the nearing end of the Section 232 investigation, events like the Lujiazui Forum as β-level stimulating events, and significant marginal changes in various fields such as consumer electronics, domestic AI large models, overseas and domestic computing power, and commercial aerospace. Considering the structural market performance mentioned earlier, it is expected that these directions will have at least pulse-level opportunities. (5) Recommended Directions for the Technology Sector Therefore, at the industry comparison level, the decision-making weight of relative cost-effectiveness is temporarily increasing, and it will have a greater odds advantage under structural market conditions. In selecting screening indicators, considering that June and July are performance window periods, the sensitivity of funds to valuation will be weakened. We classify popular technology sectors into three categories based on the relative performance of stock prices: ① High cost-effectiveness with attention: This includes directions that have had market performance since the beginning of the year but adjusted in March (such as domestic AI chains), as well as sectors that have historically been active but have not traded β this year (such as commercial aerospace). ② Medium cost-effectiveness with general attention: Located near or below the gap from early April, usually with a certain proportion of overseas revenue and involving risks from trade frictions (such as consumer electronics). Attention is dragged down by the uncertainty of the U.S. imposing tariffs, reflected in the recent stock prices' relatively dull feedback to marginal benefits, but the current position is adequately pricing the friction risks. If there is a catalyst for the phase-out of negative news (such as the end of the Section 232 investigation), there is potential for further repair and rebound. ③ Low cost-effectiveness with high attention: Mainly sectors that started adjusting in May-June or have not yet adjusted (such as robotics). Although there may be marginal catalysts, due to the relatively high position, funds realized from other non-technology sectors may have concerns about odds when entering, which may affect elasticity. Based on this relative cost-effectiveness approach, it is recommended to pay attention to: [Commercial Aerospace, Autonomous Driving, Domestic Computing Power/DeepSeek Chain] > [Consumer Electronics, Overseas Computing Power] > [Domestic Robotics, Solid-State Batteries]. Authors: Liu Chenming, Zheng Kai, Yi Xin, Yang Zezhen, Source: Chenming's Strategic Deep Thinking, Original Title: "What Changes in Technology Style Are Worth Paying Attention to Recently?" Risk Warning and Disclaimer The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not consider individual users' specific investment goals, financial conditions, or needs. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investing based on this is at one's own risk