J.P. Morgan: China Equities Saw Broad-Based Correction in October; Defensives Outperformed Growth
nashnova research
J.P. Morgan's October review shows the CSI 300 fell 8.5% as growth stocks plunged 16.2% while dividend plays posted gains — money rotated from the AI narrative into cash and defensives, repricing weak domestic demand and uncertain AI monetisation timelines.
How deep was the sell-off?
All three A-share benchmarks fell hard: the CSI 300 dropped 8.5%, the CSI 500 15.7%, and the CSI 1000 15.6% — small- and mid-caps lost nearly twice as much as large-caps.
Hong Kong held up better: the Hang Seng fell 5.4%, the HSCEI 4.4%. This means → offshore selling pressure was lighter than onshore; the A-share drawdown was driven more by domestic sentiment.
Average daily turnover hovered around RMB 1.5 trillion with turnover rates of 3%–4% — in plain terms = the market was stuck in wait-and-see mode, neither panicking nor committing.
How extreme was the style rotation?
The CSI 300 Growth index plunged 16.2%, while the CSI 300 Value index rose 1.6% and the Wind Dividend index gained 0.3%. This means → capital did not leave the market — it migrated from "betting on the future" to "collecting dividends."
Sector divergence was equally stark: energy +3.4%, consumer staples +2.0%, financials +0.5% — all defensive; information technology −24.5%, real estate −14.9%, materials −14.5% — all high-beta.
This reflects institutions raising cash and cutting risk exposure, driven by fears over rate-hike cycles in three major economies and the global AI monetisation outlook.
Are valuations cheap? What do earnings expectations say?
The CSI 300 trades at roughly 11.5× forward 12-month P/E, about 0.5 standard deviations below its 12.3× median since 2016. In plain terms = historically on the cheap side, but not at an extreme discount.
Consensus forward EPS growth is 19%, revised up 4.9% from Q2-end — sounds bullish, but the upgrades are concentrated: IT +26.8%, real estate +16.9%, healthcare +10.7%.
This means → if AI monetisation timelines slip, that 19% growth forecast faces significant downward revision, and the valuation "cushion" may not be thick enough.
What does the macro data say?
Production is strong: industrial output grew 5.5% y/y, with high-tech manufacturing at 15.3% — outpacing overall industry by 9.8 percentage points. Dollar-denominated exports surged 36.3% y/y, led by AI hardware.
Consumption is weak: retail sales rose just 1.1% y/y, dragged by a 21.2% drop in auto sales; excluding autos, retail grew 3.9%. Telecom equipment sales jumped 31.1%, driven by AI hardware and a smartphone upgrade cycle.
In plain terms = China's "external loop" — exports and manufacturing — is accelerating, but the "internal loop" — consumption and investment — has not caught up. That split caps any near-term rally.
Are investment and credit improving?
Fixed-asset investment fell 7.2% YTD y/y: real estate −20.6%, private investment −10.4%. September's seasonally adjusted m/m reading turned marginally positive (+0.07%), ending six months of decline — but barely.
New RMB loans reached RMB 1.08 trillion, yet y/y growth slowed to 4.8% — a record low. Corporate medium- and long-term loans contributed RMB 890 billion, consistent with PBOC targeted support; household medium- and long-term loans — mainly mortgages — were just RMB 87 billion, ending two months of contraction but still weak.
This reflects a key tension: policy is pushing credit supply, but households and private firms are reluctant to borrow — confidence is scarcer than liquidity.
What to watch next?
Inflation is mild but structurally split: headline CPI 0.9% y/y, core CPI 1.0% — limited demand-side improvement. PPI jumped from 3.8% to 5.0%, pulled by upstream coal (+35.3%) and oil-and-gas (+24.5%) — upstream price rises squeeze midstream and downstream margins.
USD/CNY closed at 6.65; the 10-year CGB yield edged down to around 1.62%; FX reserves fell $38.1 billion to $3.4003 trillion on dollar strength and global asset repricing.
This means → valuations sit in a historically low band, but two variables determine whether the market can stabilise: the pace of domestic demand recovery and the AI earnings-delivery timeline. Until both materialise, defence over offence remains the dominant playbook.
市场有风险，内容仅供研究参考，不构成投资建议。