UBS: A-Share Tech Sector Crowding Has Significantly Eased, Downside Risk Is Limited
nashnova research
UBS strategist Lei Meng's team said on October 14 that crowded positioning in A-share tech has been systematically cleared out, with turnover concentration back to early-April levels; combined with sustained ETF inflows and intact earnings forecasts, further downside looks limited.
How crowded was tech positioning before the sell-off?
Mutual-fund overweight in tech hit a record 18.5% in Q2 2026, surpassing the prior peak of 16.2% in Q4 2015 and approaching the all-time extreme of 22.8% set by consumer stocks in Q3 2010.
Actively managed tech-tracking funds reached 27.5% of total active-fund assets — also a record.
This means → before the correction began, virtually everyone was in the same trade. Extreme concentration made a stampede almost inevitable once sentiment turned.
How aggressive was the deleveraging?
A-share margin balances fell by RMB 408 billion in July 2026 alone — the second-largest single-month drop on record, behind only the RMB 709.2 billion decline in July 2015.
That drop exceeded every other historical stress event: the August 2015 crash, the January 2016 circuit-breaker episode, the April 2022 Covid shock, and the April 2025 US-China trade escalation.
In plain terms = leveraged money exited faster than in any prior sell-off except the 2015 peak — the "squeeze" has been rapid and thorough.
Where does turnover concentration stand now?
Weekly turnover in the broad tech sector (electronics, telecom, computers, defense) fell from a record 49.7% of total A-share turnover in early July to 37.2% currently — essentially back to early-April levels.
The share of turnover held by the top 1%, top 5%, and top 10 stocks has also retreated to early-April readings.
This means → concentrated trading and position crowding within tech have been fully digested — the crowding risk indicator has materially receded.
What signal are ETF flows sending?
Since October 10, trading volumes in major broad-based A-share ETFs have surged. Historically, every time the Shanghai Composite broke below a key round number — 3,000 in 2024, 3,200 in 2025, 3,800 in 2026 — ETFs recorded sustained net inflows.
This time the flow pattern differs from 2024–2025: instead of concentrating in CSI 300 ETFs, inflows are spread more evenly across CSI 300, CSI 500, CSI 1000, ChiNext, and STAR 50 products.
This reflects a more diversified positioning by market participants, creating a broader support base than in previous episodes.
Have fundamentals actually deteriorated?
All-A earnings growth accelerated to 29.4% year-on-year in Q2, well above Q1's 9.3%, delivering 19.5% growth for the first half; UBS forecasts full-year All-A earnings growth of 15%.
The key signal: despite the sharp price correction, consensus earnings estimates have not been revised down. The 2026 earnings-growth consensus for both ChiNext and STAR 50 remains stable.
In plain terms = "prices fell but earnings expectations didn't" — historically, that divergence indicates the correction is sentiment-driven, not fundamentally driven.
What near-term risk should investors watch?
UBS flags mutual-fund redemption pressure as the main short-term disruption. Some investors who bought funds during the May–June rally saw their positions return to break-even after the August rebound and are motivated to lock in — a "break-even, then redeem" effect.
This effect may create selling pressure in the near term. UBS sees the pace at which redemption pressure is absorbed as the key validation point for whether tech can truly stabilize.
This means → the biggest uncertainty has shifted from "will fundamentals worsen?" to "how long until retail redemption sentiment is fully flushed out?" — a different kind of risk, and a milder one.
市场有风险，内容仅供研究参考，不构成投资建议。