During periods of market panic, it is crucial to step back and analyze the data rationally. Recent sharp declines have been driven by a liquidity feedback loop and the concentrated release of pessimistic sentiment. The underlying fundamentals have not undergone significant substantive deterioration. For instance, despite Taiwan Semiconductor Manufacturing Company reporting record profits and significantly raising its capital expenditure this week, it failed to lift market spirits. This highlights how liquidity-driven feedback has amplified market volatility and negative sentiment, turning even positive news into a negative catalyst.
Why Data Matters During Downturns
First, the sentiment indicators we track have fallen to bottoming zones, suggesting the phase of adjustment driven purely by liquidity and sentiment may be nearing its end.
The RSI indicator, a measure of market sentiment, has dropped into the oversold territory below 30. Historically, periods when it reached this level (September 2024, January 2025, April 2025, December 2025, March 2026) have presented significant bottom-fishing opportunities.
The 'margin financing purchase ratio' indicator, which gauges leveraged investor sentiment, has fallen to its lowest level since July of last year. Following recent concentrated outflows, market risks have been substantially released.
This is coupled with renewed substantial net inflows from market-stabilizing funds, which helps anchor market expectations and mitigate liquidity-related feedback risks.
Assessing the AI Sector Correction
Second, the magnitude of the current correction in the AI thematic sector is approaching or has even exceeded historical extremes for thematic pullbacks.
For the US market, the Philadelphia Semiconductor Index has fallen over 20% since its peak on June 23. This decline is second only to the mid-2024 correction driven by weaker-than-expected fundamentals and the early-2025 correction triggered by trade war concerns.
For the A-share market, the TMT index has declined 23.72% since July 1, exceeding the scale of all previous AI-related corrections since 2023. Historically, this magnitude is only surpassed by the early-2020 correction in the new energy sector caused by the pandemic.
Given that this correction is not due to systemic risk or a substantive slowdown in fundamentals, the extent of the current decline appears relatively sufficient.
Understanding Sharp Drops Within Bull Markets
Finally, sharp single-day declines have sparked fears among some investors that the bull market may be ending. However, historical analysis shows such sharp corrections are not uncommon during bull markets and do not typically signal their termination. Following a sharp drop, the probability of a short-term oversold rebound actually increases. The long-term trajectory of the market ultimately still depends on fundamentals.
A review of performance following single-day drops exceeding 5% in the ChiNext Index during bull markets since 2010 reveals two key points. First, the likelihood of a short-term oversold rebound increases. On average, the index rose 1.2%, 3.9%, and 4.68% over the subsequent 1, 5, and 20 trading days, respectively. Second, 'sharp corrections are common in bull markets.' A single-day plunge does not signal the end of the trend; the index's long-term direction still hinges on fundamentals. Except for the 2015 period which lacked fundamental support and led to prolonged weakness, short-term sharp corrections have often provided buying opportunities for subsequent rallies.
Examining the Fundamental Backdrop
Regarding fundamentals, there is insufficient evidence to suggest a slowdown in AI industry prosperity.
For A-shares, mid-year earnings previews confirm that AI-related demand remains the most significant driver of corporate performance, with its positive impact continuing to spread across upstream and downstream sectors.
Overseas, the profit expectations for the 25 key AI companies we track, along with the long-term capital expenditure expectations of Hyperscalers, are still being revised upward against the market downtrend.
The next crucial validation point will be the US earnings season in late July. Following the recent sentiment volatility, the market is poised to return to rational fundamental verification. Communication from North American tech giants will provide clearer right-side signals regarding global computing power demand and the sustainability of AI hardware company performance.
Furthermore, late July will also bring policy meetings from major global central banks and China's Politburo meeting. Once key signals regarding AI industry demand, the global monetary policy path, and domestic policy direction become clearer, it will be much easier to make informed decisions on market direction and sector allocation.
Identifying Oversold Sectors with Potential
Following the recent panic-driven selloff, the technology and growth sectors now offer improved value. First, our tracked crowding metrics show a shift from 'new to old' and 'high to low': sectors like pharmaceuticals, beauty & personal care, and animal husbandry within the low-positioned consumer space have seen crowding rise to relatively high levels, while crowding in most technology sectors has fallen to lower levels. Second, crowding in core segments of the AI industry chain has retreated to moderate or even偏低 levels. Notably, crowding in key North American computing power chain segments, represented by optical modules, fiber optic cables, and PCBs, has fallen to historical lows.
Further, we have identified potentially 'oversold' sectors by screening based on four key metrics: the magnitude of this correction (performance since July), the degree of being oversold (RSI indicator), crowding (short-term trading拥挤度), and the strength of earnings revisions (earnings revision trends since July).
The screening results highlight sectors including: TMT (optical modules, fiber optic cables, memory, display panels, IDC), manufacturing (power grids, lithium battery supply chain, photovoltaic辅材), and cyclical sectors (rare earths, plastics, chemical fibers).
Risk factors include fluctuations in economic data, policy easing falling short of expectations, the Federal Reserve's interest rate cuts being less aggressive than anticipated, and an escalation in geopolitical tensions.
Comments