My China Stocks: Ground-Level Reality vs What the Market Is Pricing
I was in Beijing in early April — walked around Xidan and Wangfujing, watched how people spent money, or didn’t. Two weeks later, China’s Q1 2026 GDP came in at 5.0%, beating every forecast. The number and what I saw on the ground didn’t quite tell the same story.
That gap is worth unpacking, because about a third of my portfolio is in China-centric names. If you haven’t read my March 2026 portfolio performance review yet, that’s the broader context — this post goes deeper on the macro China backdrop, runs a proper benchmark comparison using Hong Kong indexes, and looks honestly at what the drag numbers look like.
What Q1 China GDP Actually Said
The headline 5.0% beat the 4.5–4.8% consensus. But pull it apart and it’s a familiar picture.
The strong side: industrial output up 6.1%, high-tech manufacturing up 12.5%, exports up 14.7%. China’s factory sector ran hot, partly driven by front-loading shipments ahead of Iran war trade disruption.
The weak side: retail sales up just 2.4% for the quarter, decelerating to 1.7% in March alone. Property investment still contracting.
It’s still a two-speed economy — advanced manufacturing and exports pulling hard, consumer spending and property dragging. The K-shaped dynamic is very much intact.
My time in Beijing confirmed it at street level. Consumer activity felt muted — not dead, but not the energy you’d expect from a 5% headline. EV showrooms were competing aggressively on price with nobody buying. Property overhang visible in half-occupied developments on the outskirts. Too much supply chasing not enough demand. Beijing’s policy response has been to support producers and manufacturers, not directly boost household spending. Until that changes, the consumer recovery stays theoretical regardless of what the headline GDP says.
For H2 2026, the IMF has full-year GDP at around 4.6%. Export front-loading reverses, Iran war energy costs feed through more visibly, retail sales stay weak without structural reform. The Q1 beat was real — but it was front-loaded.
That two-speed story shows up in Hong Kong equity markets. The HSCI sector indexes as at March 31, 2026:
| Sector | Ticker | YTD | 1 Year | 3 Year | 5 Year |
|---|---|---|---|---|---|
| Energy | HSCIE | +24.1% | +48.95% | +82.40% | +156.10% |
| Conglomerates | HSCIG | +9.1% | +28.28% | +21.67% | +6.09% |
| Industrials | HSCII | +7.6% | +27.21% | +10.41% | –34.28% |
| Properties & Construction | HSCIPC | +4.0% | +13.61% | –25.84% | –59.25% |
| Utilities | HSCIU | +3.4% | +13.58% | +8.26% | –17.75% |
| Consumer Staples | HSCIS | +2.1% | +5.84% | –23.64% | –41.87% |
| Financials | HSCIF | +1.2% | +28.52% | +56.95% | +31.84% |
| Healthcare | HSCIH | +0.5% | +25.08% | +2.09% | –47.09% |
| Materials | HSCIM | –0.4% | +115.30% | +157.10% | +137.30% |
| Telecommunications | HSCIT | –3.0% | –4.52% | +23.21% | +44.22% |
| Consumer Discretionary | HSCID | –12.5% | –12.93% | –9.81% | –34.67% |
| Information Technology | HSCIIT | –18.6% | –8.58% | +22.28% | –32.79% |
| HSCI (broad) | HSCI | –4.8% | +8.75% | +21.12% | –16.34% |
Source: Hang Seng Indexes Company Limited March 2026 factsheet, all data as at 31 Mar 2026.
Energy leads — oil above $100 from the Iran war. Industrials positive — infrastructure spend, old-economy construction names. Consumer Discretionary and Information Technology at the bottom — weak household spending, US-China tech decoupling. Healthcare flat — BIOSECURE noise suppressing the sector despite decent underlying fundamentals. The market is pricing the same two-speed economy I saw on the ground in Beijing.
Benchmarking Against the Right Index
In my March post I used US sector ETFs as benchmarks. For HK-listed names, that’s a rough proxy at best. The right benchmark is the Hang Seng Composite Industry Indexes — the HSCI sector sub-indexes. I’ve switched to those here, using the official HSI March 2026 factsheet with all data as at March 31.
| Sector | YTD Portfolio Impact | YTD HSCI Benchmark | Alpha |
|---|---|---|---|
| Consumer Discr. (BABA + XPEV) | –5.0% | HSCID –12.5% | +7.5% |
| Industrials (PONY + WRD) | –4.1% | HSCII +7.6% | –11.7% |
| Healthcare (WuXi) | +0.3% | HSCIH +0.5% | –0.2% |
| Total China | –8.8% | HSCI –4.8% | –4.0% |
One sector is outperforming. Two are underperforming. But the story is more nuanced than it looks.
Consumer Discretionary at –5.0% portfolio impact beat HSCID which was down 12.5% — a +7.5% alpha. The sector got hammered broadly, and on a fund-weighted basis the combined BABA and XPEV positions held up relatively well. Healthcare at +0.3% is essentially on par with HSCIH at +0.5% — WuXi is doing its job.
The real underperformance is in Industrials. –4.1% against HSCII’s +7.6% is an –11.7% gap, and that’s where the honest problem sits. But HSCII is dominated by old-economy names like Cosco Shipping, China Railway Construction, and Sinotruk — PONY and WRD don’t trade against those businesses. The classification is technically correct, but it’s the wrong peer group for an AV commercialisation thesis.
Are They Dragging My Portfolio or Pulling It Up?
Dragging — significantly.
My overall portfolio ended Q1 at +0.1% YTD. Here’s what my China positions contributed:
| Position | YTD Portfolio Impact |
|---|---|
| BABA | –4.6% |
| XPEV | –0.4% |
| PONY | –4.1% |
| WRD | 0.0% |
| WuXi | +0.3% |
| Total China | –8.8% |
YTD portfolio impact = position P/L ÷ total active portfolio cost basis.
The damage is concentrated in BABA and PONY — together –8.7 percentage points. Without them, the China book would be roughly flat. That’s a useful distinction heading into the next earnings cycle.
BABA (9988.HK) — A New Wrinkle in the Thesis
I wrote about the BABA rotation thesis in March — the margin collapse, free cash flow turning negative, the food delivery subsidy war. The trimming rationale stands. What’s changed is my read on the AI angle.
I have been treating Qwen and cloud acceleration as the catalyst to close the gap between BABA’s valuation and its underlying business quality. A Bloomberg piece this week gave me pause. China now processes 140 trillion tokens daily — up from just 100 billion at the start of 2024. Chinese AI models have surpassed US models on OpenRouter. The token economy is real. But the market isn’t rewarding BABA for it.
MiniMax (0100.HK) — in which Alibaba holds a stake — is up over 460% from its IPO price since its January Hong Kong debut. Zhipu AI (2513.HK) is up over 700%. Both are now worth more than Baidu. Meanwhile BABA’s HK-listed shares are down about 8% YTD.
The re-rating is going to pure-play model developers, not to the platforms sitting behind them. The market is pricing BABA as an e-commerce conglomerate with AI as a footnote — and that framing may be right for longer than I assumed. The trimming continues.
Pony.ai (2026.HK) — Still the Contrarian Bet
I covered PONY’s Q4 miss and the April recovery in the March performance review post. The –36.5% from cost is ugly, but this is a contrarian position in pre-profit AV commercialisation — short-term price action isn’t the signal I’m watching.
What I’m watching: a Counterpoint Research report out this week projects the global robotaxi market at $168 billion by 2035, with China leading on deployment speed, government backing, and cost structure. PONY is named directly alongside Waymo, Baidu Apollo Go, and Tesla as the companies expected to define this market. The milestones — unit economics breakeven in Guangzhou and Shenzhen, Zagreb live with Uber, Singapore regulatory approval with ComfortDelGro — are tracking.
PONY has recovered meaningfully in April. I’m less worried now than I was at end of March. May earnings are the next real data point.
WuXi AppTec (2359.HK) — Doing What It Should
WuXi is the only China position beating its benchmark. Up 5.1% from cost against HSCIH at +0.5% — a sector weighed down by BIOSECURE Act noise all year. WuXi’s own FY2025 results were strong — continuing operations revenue up 21.4%, 2026 guidance of +18–22% growth. The market is pricing regulatory risk that hasn’t materialised, and in the meantime the fundamentals are compounding quietly.
The GDP context helps here — high-tech manufacturing up 12.5%, R&D spending prioritised in the 15th Five-Year Plan. WuXi sits at the intersection of all of that. The position is smaller than the conviction warrants, which is the one thing I’d change in hindsight.
The Takeaway
China’s Q1 GDP beat looked good on paper. The ground reality in Beijing told a more cautious story — and the equity markets reflected that more honestly than the headline number did. My China positions returned –8.8% against HSCI’s –4.8%, underperforming by 4 percentage points. Some of that is the persistent China discount — the market has been slow to re-rate Chinese names regardless of fundamentals, and that’s not changing in the near term.
Most of that gap sits in two names: BABA, where the AI re-rating thesis is playing out through channels that don’t directly benefit the stock, and PONY, where the market hasn’t caught up to the operational progress yet. WuXi is the one quietly doing what it’s supposed to.
BABA is the one I’m watching most carefully going into the next quarter — the risk isn’t timing, it’s whether the AI catalyst I was counting on lands at the platform level or bypasses it entirely. May earnings for PONY and BABA’s March quarter print will tell me a lot.
Disclaimer: This is a review of my personal portfolio and performance for learning purposes. It is not investment advice. Always consult with a qualified financial advisor before making investment decisions.