CITIC SEC: The probability of a general recovery in August is still on the rise, and the configuration strategy should gradually shift from trading on oversold conditions to balance.

CITIC SEC released a research report stating that the current adjustment in the A-share market is more of a correction due to crowded trading, rather than a deleveraging shock similar to that in South Korea. The firm believes that localized liquidity pressures remain, especially for certain non-core AI stocks. Adjustments in these stocks have impacted the internal positioning adjustments within the technology sector, causing temporary pricing inefficiencies for core stocks, but it believes that this impact has largely dissipated as of now. The probability of a general recovery in August is increasing, but this is not merely a simple rebound after excessive declines. The negative narrative surrounding non-AI sectors shows marginal improvement, and the funding environment also supports appropriate recovery. In terms of allocation, it recommends increasing exposure to energy and chemical industries, non-ferrous metals, non-banking financials, and innovative pharmaceuticals, while technology investments need to focus more on core holdings during the rebound. Key points from CITIC SEC are as follows: The current adjustment is more a correction of crowded trading than a deleveraging shock akin to that in South Korea. 1) The overall leverage situation is relatively safe, with the number of rising stocks in July actually exceeding that in June. As of July 30, the average guarantee ratio for the entire markets margin trading was 264.5%. Although this is a decrease from 296.4% at the end of June, it still reached a low of 261.6% in July. Financing accounts exhibit a sufficient safety margin overall. In terms of market breadth, 2,546 A-shares recorded monthly increases in July, significantly higher than the 1,419 in June; the proportion of rising stocks increased from 25.7% to 46.0%. Among these, the number of rising non-tech stocks reached 2,291, accounting for 50.6%, indicating that the market trend is not merely contracting but is expanding from the previously highly concentrated tech trades to more non-tech sectors. The current market adjustment reflects a decongestion within the tech sector and a restructuring of market dynamics, rather than a systemic reversal of the bull market logic. Even when considering the nearly 10% decline from this round's high in the Shanghai Composite Index, it remains at a relatively moderate level when compared globally (KOSPI -39%, Nikkei 225 -16%, S&P 500 -4%). Overall, the index still possesses strong endogenous stability, and the adjustment in the tech sector does not signify the end of this bull market cycle. 2) Compared to typical deleveraging markets in global history, the current decline in financing is not substantial. During this round of adjustment, the financing balance fell from a peak of 3.01 trillion yuan on June 25 to 2.59 trillion yuan on July 31, representing a cumulative decline of approximately 14%. Historically, the most dramatic decline in financing balance occurred in mid-2015, where the depth of the decline was around 60% from peak to trough. In past similar short-term dramatic adjustments globally, the financing balance typically receded between -60% and -14%, with an average decline of -32%. The current decline in A-share financing is at a relatively low level. However, the financing balance in the TMT sector has dropped by 20.2% from its peak, surpassing the declines of 13.2% in March-April 2022 and 13.8% in early 2024, only following the declines of 56.1% in 2025 and 25.3% in early 2026. 3) The ETF market is experiencing sustained inflows, and the inflows into technology ETFs provide liquidity support. From June 25 to July 30, the ChiNext index saw a temporary decline of 25.8%, and during the same period, technology ETFs had a cumulative net subscription of approximately 165.5 billion yuan. Among the 15 days of decline, there were net subscriptions recorded on 13 days, with a total net subscription of around 115.9 billion yuan on down days. If we only consider the positive subscription amounts during the down days, the bottom-fishing funds amount to about 121.3 billion yuan, with the maximum single-day net subscription reaching 23.6 billion yuan. The continuous presence of incremental buy orders indicates that this round of adjustment has not been due to a loss of market absorption capacity, but rather is closer to the dissolution of previously crowded trading, proactive reduction in institutional concentration, and the countertrend absorption of ETF funds. This clearly differs from historical instances where leverage shocks led to a disappearance of buying power and liquidity exhaustion. Liquidity pressure from non-core AI stocks has caused temporary pricing inefficiencies in technology stocks, but this impact has now been mostly cleared. During this round of adjustment, the stocks genuinely affected by short-term liquidity pressure primarily have characteristics such as delayed stock price increases (the main rise occurred this year, with little gain in previous years), low institutional holdings, high financing proportions, and high costs of chasing the uptrend previously. According to our calculations for 37 AI tech stocks with significant liquidity pressure: 1) On average, their stock prices rose by an astonishing 114% in Q2 2026, well above the 49% recorded in Q1 2026, with many stocks only catching up to the AI trend this year. The main gains were concentrated in April to June of this year, and structurally, most of these are upstream stocks in the AI price increase chain; 2) In terms of participating entities, none of these stocks were among the top 30 holdings of actively managed public funds in Q2 2026; they are mainly non-institutional holdings, with significant participation from retail investors, private equity, and industrial funds, which have high proportions of financing purchases; 3) The chasing funds in May-June incurred significant losses. According to our estimates, by July 31, the weighted average floating loss of the buying costs from May to June was about 34%. These second or even third-tier AI-related stocks significantly outperformed the core institutional holdings in the technology sector during the initial phase of the adjustment, leading to a passive increase in the weight of core holdings. During this period, institutional funds found it difficult to manage portfolio drawdowns through high-cut-low strategies of selling the margins, buying the cores, which instead could lead to core stocks being affected. Only after the marginal stocks complete their chip clearing and move into the phase of core stocks experiencing downturns can technology funds have the space to actively adjust positions, thus releasing the overall liquidity pressure within the sector. Once this process is completed, the differentiation within the technology sector will resume, with the weighting of fundamental pricing increasing, thus terminating the previous negative feedback of declines. From the quantitative tracking indicators, we use the aforementioned 37 AI tech stocks with relative liquidity pressure as a marginal tech holding portfolio and 10 core technology stocks represented by light communication leaders, wafer manufacturers, and semiconductor equipment as a core tech holding portfolio. By judging the excess return difference between the two, we can identify the bottom signals for the sector adjustment. Based on this indicator, the excess return of marginal tech holdings relative to core tech holdings plummeted from 31% at the end of June to -26% by July 21, before fluctuating back to around -14% by July 31, suggesting that this indicator is currently showing signs of bottoming out, and the overall liquidity shock in the technology sector has essentially concluded, with differentiation set to begin. The probability of a widespread recovery in August is still increasing, but it is not just a simple rebound from excessive declines. 1) Those that dropped significantly are often the ones that rebound in early recovery, particularly stocks that faced liquidity pressure before. We calculate that, after past severe declines resulting from similar liquidity shocks, the effects of excessive rebounds are typically fulfilled within 5-10 trading days after the lows, while the market rebound structure after 10 trading days tends to gradually decouple from the previous decline structure. Selling pressure following a tech rebound might mainly stem from funds that prematurely bottom-fished during the decline, as well as from institutional funds that had previously overly concentrated holdings adjusting their positions. Taking semiconductor equipment ETFs as a typical example, net subscriptions mostly occurred during the initial days after the related indices began their downward adjustments, and this portion of holdings could become selling pressure during the rebound process. For actively managed products, many showed significant increases in technology holdings in June, with 237 out of 1,731 flexible allocation funds in June raising their Beta relative to the ChiNext 50 Index by over 0.05, corresponding to a scale of 208.8 billion yuan; this part of the funds still had a median Beta of 0.44 in July, indicating that their net value still exhibited a high tech exposure during the technology adjustment phase. These products may adjust their positions to reduce portfolio volatility during future technology rebounds due to excessive deviations, which is also a factor restricting the strength of the technology rebound. Therefore, even if the overall technology sector recovers in August, internal performances may still reflect significant differentiation, and not all stocks that experienced greater previous declines will rebound stronger. To fundamentally improve the current chip structure, we need significant breakthroughs in the industry to sufficiently open the sector's imaginative horizons, attracting incremental funds to take on the pressure from the realization of existing chips. 2) There has been marginal improvement in the negative narrative surrounding non-AI sectors, and the funding environment also supports recovery. The Federal Reserve's July monetary policy meeting lacked significant hawkish measures, noting that the rise in long-term interest rates had led to some tightening of financial conditions, further indicating that the Fed does not need to rush to raise interest rates. The subsequent release of U.S. Q2 GDP and June PCE data also fell short of expectations, further weakening the short-term rate hike narrative. Over the past few months, the strong dollar and interest rate expectations have been significant macro factors suppressing demand expectations in non-AI sectors and exacerbating market K-shaped differentiation; the marginal erosion of this narrative is conducive to a certain recovery in non-AI sectors. This week's Politburo meeting had a more cautious stance on the economy, shifting from emphasizing "strong start in economic growth and major indicators exceeding expectations" in April to stressing "high importance to difficulties and challenges in economic operations" this time, further emphasizing the need for timely planning of incremental policies and enhancing counter-cyclical adjustment efforts. Unlike the downward adjustment of relatively high growth expectations seen in the first half of the year, the market in the second half will gradually wait for the marginal changes brought about by accelerated fiscal spending, monetary tool adjustments, and the rollout of incremental policies, starting from a more cautious expectation. The change in anticipated directional runs suggests that the market structure in the second half may no longer merely favor the strongest stocks; policy support and the recovery of demand expectations could promote a migration in market sentiment from single high prosperity to more sectors. Furthermore, the funding ecology in the second half of the year will also be favorable for the recovery of non-AI sectors. In the first half of the year, the cumulative net redemption of A-share ETFs reached 1.63 trillion yuan. However, as we entered July, the redemption direction of ETFs saw a clear reversal, with a cumulative net subscription of 478.4 billion yuan for the entire month. In 23 trading days, there were net inflows on 19 days, recovering about 29% of the cumulative net redemptions from the first half. For non-AI sectors, which already have low active institutional holdings, the return of broad-based ETFs creates a more stable and balanced passive absorption; coupled with the earlier easing of crowded trades, the pricing constraints for non-AI sectors are expected to lessen, laying the financial foundation for a balance of styles and widespread recovery in August. It is advised to increase exposure to energy chemicals, non-ferrous metals, non-bank financials, and innovative pharmaceuticals, while focusing technology holdings during the rebound. The short-term market rebound is likely to exhibit more characteristics of excessive correction recovery; stocks that have significantly dropped and have seen thorough chip clearing are expected to perform particularly well, especially those non-core AI varieties previously impacted by deleveraging and liquidity shocks. Such rebounds mainly stem from improvements in liquidity, risk appetite recovery, and short covering, and do not imply that the existing industrial logic and valuation system have been reestablished. As market liquidity and the price discovery mechanism return to normal, the allocation strategy in August should shift from trading excessive declines to achieving balance. Utilize the rebound to optimize the holding structure, returning to pricing logic based on fundamentals, industrial position, and mid to long-term profitability. We maintain a mid-term judgment of "three convergences": 1) Within the AI industry chain, the relative excess returns of upstream hardware and price-increasing varieties are tending to converge with those of downstream platforms, cloud services, and core application links; 2) The valuation discount of non-AI industrial sectors relative to comparable overseas companies is expected to recover in stages; 3) The extreme differentiation between technology and non-technology sectors is tending to converge. Within the technology sector, it is recommended to utilize the rebound in marginal varieties to timely adjust towards core assets (such as light communication leaders, wafer manufacturing platforms, and semiconductor equipment, etc.). For non-technology sectors, a focus on increasing exposure to energy chemicals, non-ferrous metals, non-bank financials, and innovative pharmaceuticals is advised. Risk factors include: Increased friction between China and the U.S. in technology, trade, and finance; domestic policy intensity, implementation effects, or economic recovery may fall short of expectations; macro liquidity tightening exceeds expectations both domestically and internationally; further escalation of conflicts in regions like Russia-Ukraine, the Middle East; the digestion of housing inventory in China is less than anticipated.