The recent correction in China's A-share market is more of a recalibration of overcrowded trades rather than a Korean-style deleveraging shock, according to analysis. This conclusion is supported by three key factors: overall leverage remains safe, with the number of rising stocks in July nearing half of all A-shares, significantly surpassing June's figures; the current decline in margin financing is modest compared to historical deleveraging events globally; and the ETF market has seen sustained capital inflows, with technology-focused ETFs providing ample liquidity support. However, localized liquidity pressures remain, particularly among non-core AI stocks, which have impacted portfolio adjustments within the tech sector, causing temporary pricing inefficiencies in core holdings. It is believed this effect has now largely dissipated.
The probability of a broad recovery in August is increasing, but it is not simply a case of a bounce-back from oversold conditions. The negative narratives surrounding non-AI sectors are showing marginal improvement, and the funding environment is conducive to a degree of recovery. For portfolio positioning, it is recommended to increase allocations to energy and chemicals, non-ferrous metals, non-bank financials, and innovative drugs. Within the technology sector, investors should use any rebound to focus more on core holdings.
The current adjustment is a correction of overcrowded trades, not a Korean-style deleveraging shock
The overall leverage situation is relatively safe. As of July 30, the average collateral ratio for margin trading across the market stood at 264.5%. Although this is down from 296.4% at the end of June, the lowest point in July was still 261.6%, indicating a sufficient safety cushion for margin accounts. In terms of market breadth, 2,546 A-shares achieved monthly gains in July, significantly higher than the 1,419 in June. The proportion of rising stocks increased from 25.7% to 46.0%. Among these, non-tech sectors saw 2,291 stocks rise, accounting for 50.6% of the total, suggesting the market trend is not a simple contraction but a rotation from the previously highly concentrated tech trades into more non-tech industries. The current market correction is a process of de-congestion within tech and structural rebalancing, not a systemic reversal of the bull market logic. Even if the Shanghai Composite Index were to correct by about 10% from its recent cyclical peak, this would be relatively mild compared to other major global equity markets.
Compared with typical historical deleveraging scenarios globally, the current decline in margin financing is not substantial. During this adjustment, the balance of margin financing fell from a peak of 3.01 trillion yuan on June 25 to 2.59 trillion yuan on July 31, a cumulative decline of roughly 14%. Historically, the most severe event was in mid-2015, with a 60% drop from peak to trough. In similar sharp short-term corrections globally, margin balance declines have ranged from -60% to -14%, with an average of -32%. The current decline in A-shares is at the lowest end of this spectrum. However, within the TMT sector, margin balances have fallen 20.2% from their highs, exceeding the declines seen in March-April 2022 and early 2024, though still less severe than the 56.1% drop in 2025 and the 25.3% decline in early 2026.
The ETF market has experienced sustained capital inflows, with tech-oriented ETFs providing liquidity support. From June 25 to July 30, the ChiNext Index fell 25.8% during the period. Concurrently, tech-oriented ETFs saw cumulative net subscriptions of approximately 1,655 billion yuan. In 13 out of 15 down days, net subscriptions were recorded, totaling about 1,159 billion yuan on those declining sessions. This persistent buying demand indicates that the current adjustment is not due to a disappearance of market absorption capacity. Instead, it results from a combination of the unwinding of previously crowded trades, active reduction of portfolio concentration by institutions, and ETF funds buying against the trend. This is distinctly different from historical leverage-driven shocks where buying demand vanishes and liquidity dries up.
Liquidity pressure on non-core AI stocks caused temporary pricing failures in tech, now largely resolved
The stocks that have truly faced short-term liquidity pressure during this correction share specific characteristics: delayed price rallies (with main gains this year, not in previous years), non-institutional heavy holdings, high margin financing ratios, and high entry costs for early buyers. According to calculations based on a sample of 37 such pan-AI tech stocks under greater liquidity pressure, they saw an average price gain of 114% in the second quarter, far exceeding the 49% in the first quarter. Many of these stocks only joined the AI rally this year, with gains concentrated from April to June. These stocks are predominantly from the AI upstream pricing chain. In terms of participants, these stocks were not among the top 30 heavy holdings of active public funds in the second quarter. They are mainly non-institutional heavy holdings with significant involvement from hot money, private equity, and industrial capital, and have a relatively high proportion of margin buying. Buyers who chased gains in May and June are now facing significant losses, with average paper losses on weighted costs from that period estimated at around 34%.
During the early stages of the correction, these second-tier or even third-tier AI stocks fell more sharply than core institutional tech holdings. This caused the weight of core holdings to increase passively, making it difficult for institutional funds to control portfolio drawdowns by simply switching from fringe to core positions. Core stocks were also dragged down. Only after the fringe stocks had completed their de-positioning and entered a phase of supplementary decline could tech-focused funds actively adjust their portfolios, thereby releasing overall liquidity pressure in the sector. Once this process is complete, differentiation within the tech sector will re-emerge, with fundamental pricing regaining influence, ending the negative feedback loop of declines.
Tracking this process using quantitative indicators, the excess return of fringe tech holdings relative to core tech holdings has rapidly fallen from 31% at the end of June to -26% on July 21, before recovering to around -14% by July 31. This indicator now shows signs of bottoming out, suggesting the liquidity shock in the tech sector is largely over, and differentiation will follow.
Probability of broad August recovery rising, but not just a simple oversold bounce
Stocks that have fallen the most typically rebound most sharply initially, especially those that experienced liquidity pressures. Based on analysis of past similar corrections driven by liquidity shocks, the oversold bounce effect tends to be concentrated within 5 to 10 trading days after the low point. After 10 trading days, the market's rebound structure gradually decouples from the prior decline structure. Selling pressure during the tech rebound may primarily come from funds that bought the dip too early during the decline and from institutional funds with previously concentrated holdings looking to rebalance. For active funds, some products significantly increased their tech exposure in June. Among 1,731 flexible allocation funds, 237 saw their beta relative to the STAR 50 index increase by more than 0.05 in June, representing assets under management of 2,088 billion yuan. The median beta for these funds in July remained at 0.44, indicating high tech exposure even during the correction. These products may rebalance during a future tech rebound to reduce portfolio volatility, limiting the strength of the tech rally. Therefore, even if the tech sector recovers broadly in August, internal performance is likely to be clearly differentiated, with the biggest decliners not necessarily rebounding the strongest. To fundamentally improve the current ownership structure, a major breakthrough in the industrial sector is still needed to open up the imagination space for the sector and attract incremental funds to absorb the realization pressure from existing holdings.
The negative narratives surrounding non-AI sectors are marginally improving, and the funding environment supports a recovery. The Federal Reserve's July meeting lacked substantive hawkish measures, mentioning that the rise in long-term interest rates has already brought some tightening of financial conditions, further suggesting the Fed sees no urgent need to raise rates. Subsequent weaker-than-expected US Q2 GDP and June PCE data further weakened the narrative of imminent rate hikes. In recent months, a strong US dollar and rate hike expectations were key macro factors suppressing demand expectations for non-AI sectors and exacerbating the K-shaped market divergence. A marginal shift in this narrative should facilitate some recovery in non-AI sectors. The Politburo meeting this week also adopted a more cautious tone regarding the economy, emphasizing difficulties and challenges and reiterating the need for timely counter-cyclical adjustments. Unlike the first half of the year, when the market moved from high growth expectations to constant downward revisions, the second half will see the market start from relatively cautious expectations, awaiting marginal changes from faster fiscal spending, monetary policy adjustments, and the implementation of incremental policies. This shift in the direction of expectations means the market structure in the second half may no longer be simply 'the strong getting stronger'. Policy support and the restoration of demand expectations should help drive the market's rotation from a single high-growth sector to more industries.
Furthermore, the market funding ecosystem in the second half is also favorable for a recovery in non-AI sectors. In the first half, A-share ETFs saw cumulative net redemptions of 1.63 trillion yuan. However, entering July, the direction of ETF subscriptions and redemptions clearly reversed, with cumulative net subscriptions reaching 4,784 billion yuan for the full month, with inflows on 19 out of 23 trading days, recouping about 29% of the first half's net redemptions. For non-AI sectors where active institutional holdings are already low, the return of broad-based ETF flows provides a more stable and balanced form of passive support. Combined with the loosening of previously crowded trades, pricing constraints on non-AI sectors are expected to ease, creating a capital base for style equalization and broad recovery in August.
Recommended allocation: increase exposure to energy/chemicals, non-ferrous metals, non-bank financials, and innovative drugs; within tech, use the rebound to focus holdings
Short-term market rebounds may primarily exhibit characteristics of an oversold correction. Stocks that have fallen the most and seen a thorough cleanup of positions are expected to perform more prominently, especially non-core AI stocks that previously suffered from deleveraging and liquidity shocks. Such rebounds primarily stem from improved liquidity, a recovery in risk appetite, and short covering, and do not signify the re-establishment of original industry logic and valuation systems. As market liquidity and price discovery mechanisms return to normal, the investment approach for August should shift from trading on oversold conditions towards equalization. Use the rebound to optimize portfolio structure, refocusing on fundamentals, industry standing, and long-term profitability as the core pricing logic. The medium-term judgment of "three convergences" remains: the excess return of upstream hardware and price-rise stocks relative to downstream platforms, cloud services, and core applications within the AI chain tends to converge; the valuation discount of non-AI industrial sectors relative to their overseas peers is expected to narrow temporarily; and the extreme divergence between tech and non-tech sectors tends to converge. Within the tech sector, it is advised to use the rebound in fringe stocks to promptly rotate into core assets. For non-tech sectors, key allocations should focus on increasing exposure to energy and chemicals, non-ferrous metals, non-bank financials, and innovative drugs.
Risk factors
Risks include: intensified friction between China and the US in technology, trade, and finance; domestic policy strength, implementation effects, or economic recovery falling short of expectations; unexpected tightening of macro liquidity at home and abroad; further escalation of regional conflicts; and slower-than-expected digestion of China's real estate inventory.
Comments