CITIC Securities: Seize the Last Offensive Window for A-Shares This Year — Non-Institutionally Held Stocks Offer Greater Upside
nashnova research
CITIC Securities argues A-shares are in the late stage of an industrial super-cycle, where institutional heavy-holdings have peaked in alpha — historical playback shows non-institutional small-caps deliver significantly more upside at this juncture, and the year's final offensive window is approaching.
What is the "late stage of an industrial super-cycle" — and what happened before?
An industrial super-cycle — a multi-year bull phase driven by one dominant sector, such as AI — eventually reaches a point where the leading institutional heavy-holdings have exhausted their excess returns.
CITIC reviewed three comparable episodes: 2009, Q4 2015, and Q2 2022. Each time, after institutional favorites peaked, a wave of non-institutional small-caps hit new highs, forming a "double-top" structure.
In the post-crash divergence of Q3 2015: from September 15 to year-end, the low-ownership small-cap basket rallied 83.2%, far outpacing the institutional high-ownership basket at 54.1%. This means → in the cycle's late stage, stocks *not* crowded by institutions run harder.
What is holding institutional favorites down right now?
Factor one: AI monetization expectations are resetting. Capex hasn't slowed, but the market now sees ceiling risk — Ramp data shows roughly 80% of enterprise revenue for OpenAI and Anthropic comes from the top 1% of clients, and per-capita AI spending by those heavy users fell 9.7% month-over-month in August. In plain terms = the biggest spenders are already pulling back.
Factor two: earnings growth is nearing its peak. CITIC expects the year-over-year earnings growth peak for all-A ex-financials to arrive around Q4 2026, after which high tech base effects and slowing commodity prices drag growth lower. This means → institutional investors are reluctant to add positions now because the outlook for next year is uncertain.
Factor three: a hawkish Fed. As of September 18, the implied probability of another rate hike this year reached 90.1%, with the market pricing a cumulative 100 basis points of hikes through Q2 2027. Tight macro liquidity directly caps the rebound room for institutional names.
How depressed is sentiment — and why is that actually good?
Per CITIC's channel survey, as of September 11 the sampled active private-fund position stood at 73.0%, below the historical median of 76.4% and at roughly the 29th percentile since late 2020.
September has already priced in Fed hikes, AI deceleration, and high oil prices. This means → most of the bad news is "in the price," and the low point in the sentiment cycle provides a launchpad for a short-term offensive.
When does the window open — and what to buy?
CITIC expects the offensive window to open around Q3 earnings season in October, after the Mid-Autumn and National Day holidays. Sentiment could warm enough to front-run next year's optimism — an early valuation switch.
In tech, two focus areas: first, names riding rising manufacturing complexity — optical-communication new tech, PCB, advanced packaging; second, those with clear volume-growth logic — wafer fabrication, gas turbines. Among these, non-institutional heavy-holdings offer greater upside, and the North America supply chain may outperform.
Outside tech, focus stays on energy-chemical plays and leading brokerages with overseas expansion potential. This reflects CITIC's core thesis: under high rates, K-shaped divergence widens, and capital re-concentrates along the AI and going-global themes.
How do we verify this framework — what are the checkpoints?
CITIC names two explicit tests: whether Q3 earnings catalyze an effective second offensive, and whether non-institutional favorites break above their prior highs.
In plain terms = if small-caps actually rally after October's earnings reports and surpass the previous peak, the "late-cycle, small-caps outperform" thesis is confirmed; if not, the framework needs reassessment.
市场有风险，内容仅供研究参考，不构成投资建议。