CITIC Securities: "K-Shaped Convergence" May Be the Most Important A-Share Market Feature by 2027
nashnova research
CITIC Securities argues that A-share corrections were front-loaded in 2026; in 2027 AI stock momentum fades while non-AI sectors improve at the margin — "K-shaped convergence and broadening market breadth" will be the most important shift, meaning opportunity spreads beyond a handful of AI leaders.
Why was the market so split in the past year?
Global equities showed extreme "K-shaped divergence" in 2026: across China, the US, Japan and South Korea, AI-core portfolios outperformed non-AI-core by roughly 30% through October — and as much as 150% at the peak.
The AI hardware boom generated heavy negative externalities — it is worth spelling out what that means here: one sector's surge pushed up raw-material costs for consumer goods, siphoned capital from the bond market and drove long-end rates higher, repeatedly crushing non-AI sectors.
This means → the split far exceeded start-of-year expectations — the market didn't move as one; half was on fire, half was in an ice bath.
Can AI stocks keep leading?
CITIC's view: AI industry trends are still rising, but stock-level momentum is likely fading — the odds of AI hardware being the market's main thread for a fifth straight year are falling.
The bank estimates that if global AI capex reaches $1.6 trillion by 2030, A-share and Hong Kong-listed AI hardware names priced at 15× 2030 earnings would sit only about 51% above today's levels — aggregate compound return is no longer compelling.
In plain terms = the pie is still growing, but the share price has already eaten most of the gain upfront — what's left requires individual stock-picking.
Will pressure on non-AI sectors ease?
CITIC flags three sources of marginal improvement: ① North American AI investment slows → private-sector debt issuance peaks → long-end rates may retreat; ② RMB appreciation decelerates → FX losses shrink (the drag on all-A ex-financials net-profit growth hit 7.3% in H1 2026, projected to drop to 0.5%–6.2% at a 6.35–6.65 USD/CNY range in 2027); ③ overseas-expansion valuations are near historical extremes, and EU-China trade friction is unlikely to block top manufacturers long-term.
On domestic demand, 2026 was a rare year of year-on-year fiscal deficit contraction — tax revenue beat expectations while spending lagged; 2027 fiscal execution may adjust, leaving room for stimulus.
This means → non-AI sectors face easing signals from rates, FX and fiscal policy simultaneously.
Where do earnings and valuations stand?
CITIC forecasts 2027 all-A net profit growth at 9.5% (ex-financials 14.7%) — a clear step down from 2026's 15.5% / 20.1%, with the high-base effect sharpest in tech (from 101.9% to 43.0%).
Valuation compression is largely done: the 15 core AI stocks corrected over 77 trading days from July, falling as much as 44.5%; the top-200 institutional-heavy names dropped as much as 35.8% — exceeding the peak drawdown during the 2021–22 new-energy crowded-trade unwind.
Non-AI, non-financial leaders trade at just 15.6× P/E, the 21st percentile historically; consumer, manufacturing and pharma leaders sit at the 12th, 1st and 3rd percentiles respectively. In plain terms = these sectors are priced near the floor of their historical range.
What strategy fits 2027 best?
The capital-flow picture is not yet fully cleared: retail inflows are slowing, active equity funds are trapped in a "NAV rebounds → redemptions spike" feedback loop; thematic ETFs, quant hedge funds and long-horizon capital will be the main incremental sources.
CITIC argues that when demand signals are unclear and momentum strategies weaken, PB-ROE — price-to-book versus return-on-equity, a framework measuring "how much return per unit of capital" — style steady-return strategies outperform significantly. Adding a distressed-turnaround overlay makes the strategy even more stable.
Two clusters fit: high dividend + steady earnings names in white goods, textile & apparel, food & seasoning, thermal power; and supply-clearing potential plays in tanker shipping, express delivery, top property developers, select lithium-battery materials and midstream petrochemicals.
Which four threads shape sector allocation?
Steady returns: energy-chemicals, tanker shipping, toll roads, logistics, white goods, pharmacy retail, insurance, overseas-facing brokers, financially sound property leaders.
Top exporters: shipbuilding, diesel engines, two-wheelers, CXO (contract research & development outsourcing for pharma), power tools, electrical equipment.
Domestic-demand recovery (contingent on stronger central fiscal support): catering supply chains, airport duty-free, health supplements and medical-device leaders.
AI new tech (emphasis on earnings delivery): passive optical components, PCB (printed circuit boards), memory, semiconductor equipment & materials, robotics, on-device AI.
This reflects CITIC's core logic: diversify moderately, stop betting on a single thread. A-share shareholder returns are also improving — combined dividends and buybacks now run 3.4× equity financing; all-A net cash return rose from −1.5% in 2016 to 2.5% in 2024, the mirror image of the US trend declining from 4.6% to 2.5%.
市场有风险，内容仅供研究参考，不构成投资建议。