JPMorgan Maintains Bullish Stance on Chinese Equities, Recommends AI + CXO and High-Dividend Barbell Strategy
nashnova research
J.P. Morgan's October strategy report holds its CSI 300 year-end target at 5,200, backed by consensus 19% forward EPS growth; with liquidity thin and AI names under pressure, the house recommends a barbell of high-dividend defence and AI growth — arguing that reasonably valued AI sub-sectors still offer downside cushioning.
What holds the 5,200 target together?
J.P. Morgan's base-case year-end target for the CSI 300 is 5,200, anchored by consensus next-twelve-month EPS growth of 19%.
This means → the call rests on earnings carrying valuation, not on a sentiment turn.
But the report flags that A-share liquidity metrics — turnover, margin balances, ETF flows — are all depressed, with no sign of large-scale "national team" buying. In plain terms = the money has not shown up yet.
The barbell — what goes on each end?
J.P. Morgan recommends a "barbell" — splitting exposure between two extremes: defensive high-dividend names (banks, hydro utilities) and growth plays in AI infrastructure plus CXO / innovative pharma.
This means → the house does not see a one-directional long setup; it wants steady cash flow in one hand, growth optionality in the other.
This reflects a tension at the heart of the call: bullish on direction, but acknowledging that near-term liquidity and geopolitical risk remain unresolved.
How far has AI fallen — and why hasn't it bottomed?
Over the past five trading days, AI-linked themes fell almost across the board: optical modules −16.8%, optical chips −15.5%, GPUs −14.2%, versus a CSI 300 decline of just 4.6%.
The direct trigger is geopolitical: fears of tighter U.S. FCC rules, broader import restrictions on Chinese products, and the MATCH Act — a bill aiming to coordinate multilateral chip-export controls on China — pressuring the Chinese AI supply chain.
J.P. Morgan judges that A-share AI positioning has not yet bottomed. This means → selling pressure may not be over, capping the size of any rebound.
Where is the valuation cushion — which AI sub-sectors are still cheap?
J.P. Morgan highlights two verticals: memory and server / switch ODMs/OEMs — both trading below 5× forward price-to-sales with solid revenue growth. The top pick is Hua Qin Technology (华勤技术).
IDC / telecom operators, cabling, OSAT (outsourced semiconductor assembly and test), foundry, thermal management, and power supply segments also trade below 5× forward P/S.
In plain terms = J.P. Morgan's logic is simple: AI is selling off broadly, but sub-sectors that are cheap on revenue and still growing have a floor underneath them.
Do the bulls' three arguments hold up?
Bullish participants cite three points: ① CSI 300 IT-sector forward EPS growth of 53%; ② Chinese firms hold roughly 70% of the U.S. optical-module market, hard to replace quickly; ③ China has export-control leverage over indium phosphide — a key laser material.
J.P. Morgan's view: all three arguments are valid, but none is sufficient — none fully resolves market concerns about sales and earnings for the relevant names beyond 2028/29.
This reflects the house's core stance: near-term cushion exists, but long-run geopolitical risk cannot be offset by fundamental arguments alone.
What do the policy signal and global AI fundamentals say?
On October 9, China released a policy document on "developing new quality productive forces." Article 10 specifically lays out an "AI+" action plan covering compute, algorithms, data supply, industry pilots, and safety governance. J.P. Morgan reads this as a clear policy-priority signal and a likely driver of the afternoon A-share bounce that day.
However, the document names no specific funding figures; budgets are deferred to subsequent annual plans. This means → the signal is on the table, but the money has not landed; a subsequent budget that explicitly raises AI capex targets would be the real catalyst.
Globally, J.P. Morgan's Asia equity strategy team sees AI hardware supply-demand tightness lasting at least through 2028, with Asian companies holding significant pricing power and ecosystem net profits potentially exceeding $1 trillion. Under a U.S. Treasury bear-flattening curve (elevated long-end rates), CSI 300 P/E tends to compress — but the IT sector, with 53% forward EPS growth, is the segment most likely to offset valuation compression through earnings momentum.
市场有风险，内容仅供研究参考，不构成投资建议。