China Weekly 24.8.2026
brief · 2026-08-24 · China internet, HK equities, semis · 2-4 week horizon
§01 · Macro snapshot
The US screens that price offshore China risk.
| Series | Value | Prior | Δ | As of |
|---|---|---|---|---|
| 10Y-2Y Treasury spread | 0.50 | 0.51 | -0.01 pp | 2026-08-21 |
| 10Y-3M Treasury spread | 0.86 | 0.82 | +0.04 pp | 2026-08-21 |
| High Yield OAS | 2.75 | 2.71 | +0.04 pp | 2026-08-20 |
| US Unemployment (U3) | 4.1% | 4.2% | -0.1 pp | 2026-07 |
| Initial jobless claims | 206k | 212k | -6k | 2026-08-15 |
| Sahm rule | -0.03 | 0.07 | -0.10 | 2026-07 |
HY OAS at 2.75% and a positive 10Y-3M at +0.86 pp are not risk-off levels; U3 stepped down to 4.1% and Sahm went negative. The offshore-risk backdrop this week is benign; anything pushing China ADRs is coming from the names themselves.
§02 · Themes of the week
Alibaba's $10.2B placement resets the China AI capex read
Alibaba priced a $10.2 billion share placement on 2026-08-24 to fund AI infrastructure, and the ADR traded down 3.6% on the print. YMTC is separately raising into the same window. The size of the capital call implies management believes further AI spending still pays off, and is willing to take a tape hit to secure the balance sheet for it. That reads across the China internet complex: if Alibaba is spending at this size, peers face a harder time holding the line on their own AI capex.
Counter-thesis: this is straight dilution, and the "capex signal" reading is post-hoc. Wrong if BABA closes below $110 through end-September without a specific AI product or contract catalyst — that would mean the placement is being absorbed as pure dilution, not signal.
Sources: Bloomberg, LiveMint.
Burry rotates from Alibaba to JD.com
Michael Burry disclosed exiting his Alibaba position and building a JD.com stake, framed publicly around valuation and preference for JD's model. The direction is what matters — an investor known for concentrated contra-consensus bets is signalling a relative-value view within China internet, not a call on the sector as a whole. JD's tape absorbed the flow (+1.07% week) while BABA broke lower (-3.61% week).
Counter-thesis: one manager's rotation is a data point, not a trend, and JD's own RSI-14 at 25.1 says the tape is not confirming the thesis in real time. The rotation could be idiosyncratic (Burry-specific portfolio construction) rather than a read others echo. Wrong if by end-September JD trades flat-to-down against BABA on a total-return basis — that would mean the disclosed rotation drew no follow-through and the pair has already reset.
Falsifiability: track the BABA/JD relative-return spread through September; if the spread narrows against JD (BABA outperforming), the thesis is failing.
Sources: Business Insider, GuruFocus.
Baidu capitulation reads as AI-narrative displacement
Baidu is down 10.1% for the week, 13.2% for the month, 28.9% over three months. RSI-14 sits at 14.5 — deep enough that the tape is being sold into every bid, not just drifting. The July 2026 dual-primary Hong Kong listing has not stemmed the drawdown. The clean read: Baidu is losing the China AI leadership story to Alibaba, and the price is closing the gap. A Polymarket contract on "Will Baidu have the #1 AI model at end of August 2026" is open into that same question.
Counter-thesis: RSI 14.5 with $86 as the 52-week floor is a mean-reversion setup, not a fundamental read. A single AI product win or a dual-listing inflow window could bounce the name 15% in a week. Wrong if BIDU closes above $110 by mid-September without a Baidu-specific fundamental catalyst — that would mean the sell-off was technical exhaustion, not narrative loss.
Sources: Financial Times, South China Morning Post, Polymarket.
Shein's $27bn Hong Kong IPO reasserts the venue
Shein is targeting a $27 billion valuation for a Hong Kong listing, with HK Chief Executive John Lee simultaneously pitching the venue as the "golden gateway" for start-ups. Read as one signal: large China-adjacent float is routing toward HKEX, and Hong Kong is actively courting it. If Shein books at or near $27bn, expect the next tier of China-adjacent private companies to route their IPO decisions through Hong Kong first.
Counter-thesis: one deal at a particular size is not a venue trend, and Shein's own status as a China-adjacent — not pure China — issuer weakens the read-across. Wrong if Shein prices below $22bn or pulls the listing before Q4 — that would signal even Hong Kong cannot absorb the largest China-adjacent floats at target.
Sources: Financial Times, South China Morning Post.
§03 · Companies of interest
Research surface — not investment advice.
| Ticker | Exchange | Name / sector | Theme link | Technical snapshot |
|---|---|---|---|---|
| BABA | NYSE | Alibaba Group Holding · Internet Retail | $10.2B placement (2026-08-24) is the primary frame — capital action, not a clean tech-cycle read | $119.34 · 27th %ile 52w · RSI 40 · below 200d & 30w · flat 1y · industry DEGRADING |
| 9988.HK | HKEX | Alibaba HK line | Monday convergence trade against ADR after placement broke post-HK close | HK$123.00 · 35th %ile · RSI 47 · below 200d & 30w · +2.59% wk |
| JD | Nasdaq | JD.com · Internet Retail | Burry rotation drives tape this week; positioning event, not fundamental | $29.37 · 39th %ile · RSI 25 · above 200d & 30w · +7.15% 6m · industry DEGRADING |
| BIDU | Nasdaq | Baidu · Internet Content & Information | AI-narrative displacement thesis operates against a July 2026 dual-primary HK listing that has not stemmed the drawdown | $93.21 · 9th %ile · RSI 14.5 · below 200d & 30w · -28.94% 3m · industry DEGRADING |
| 0700.HK | HKEX | — | Read-across from the BABA capex signal — deep oversold on the tape (RSI 27) | HK$439.80 · 11th %ile · RSI 27 · below 200d & 30w · -24.95% 1y |
| NIO | NYSE | NIO · Auto Manufacturers | China EV small-cap; industry cohort DEGRADING | $4.63 · 7th %ile · RSI 42 · below 200d & 30w · -8.68% 1y · industry DEGRADING |
| 1211.HK | HKEX | — | China auto cohort pair against NIO | HK$93.05 · 44th %ile · RSI 49 · below 200d & 30w · -19.09% 1y |
| 1398.HK | HKEX | — | The one structural bid on the list, and the only name above both moving averages | HK$7.48 · 92nd %ile · RSI 67 · above 200d & 30w · +24.5% 1y |
Technicals: market data, computed 2026-08-24T06:00 UTC. HK ticker sector/name fields not populated in this cycle's data.
§04 · Calendar ahead
- 2026-08-25 — Denmark Q2 GDP (Statistics Denmark). Nordic release, no direct China read; noted for completeness.
- 2026-08-28 — Norway Q2 GDP (SSB). Same category as above; peripheral to the China thesis.
- 2026-09-04 — US Non-Farm Payrolls. Prints the US dollar and the front of the Treasury curve, both of which reprice offshore China ADRs. Any read that shifts the September FOMC path shows up in BABA and BIDU beta first.
§05 · Methodology + disclaimer
Compiled from public macroeconomic data, financial press, and proprietary analytical tools.
This is research material, not investment advice.
2026-08-24 · brief v1