Alibaba shares rise as AI drives 34% cloud sales jump

Alibaba reported fiscal Q2 revenue of 247.8 billion yuan, up 5% year‑on‑year, with cloud computing revenue accelerating 34% to 39.8 billion yuan (above consensus 37.9 billion) and cloud EBITA rising 35% to 3.6 billion yuan. Overall adjusted EBITA fell 78% to 9.1 billion yuan, weighed by heavy investment in quick commerce, though China e‑commerce revenue grew 16% to 132.6 billion yuan and quick commerce revenue jumped 60% year‑on‑year. Management said AI demand remains very strong, disclosed about 120 billion yuan of AI/cloud capex spent in the last four quarters against a 380 billion yuan three‑year target and signaled it may increase that capex, while its Qwen app reached 10 million downloads in one week.

Analysis

Market structure: Alibaba’s step-up in AI capex and 34% cloud growth make cloud providers, GPU vendors (NVDA) and memory/chip suppliers the primary beneficiaries as enterprise AI spend re-accelerates; incumbent retail players taking heavy quick‑commerce losses are the short-term losers. Pricing power will shift toward AI-infrastructure suppliers over the next 12–36 months as Eddie Wu expects supply to remain tight; expect higher realized prices for GPUs and memory and stronger gross margins for infrastructure vendors. Cross-asset: tighter GPU/memory supply supports semiconductor equities and commodity memory prices, puts mild upward pressure on risk premia in EM credit and could modestly strengthen CNY on service export strength; bond spreads for Chinese tech investment‑grade names may widen if capex forces more cash burn.

Risk assessment: Tail risks include abrupt China regulatory action on data/AI models or new US export controls on GPUs — both could cut revenue by >20% in 6–12 months. Short-term (days–weeks) volatility will hinge on guidance and export headlines; medium-term (1–6 months) risks are execution of large capex and margin recovery; long-term (3+ years) upside depends on model adoption and sustaining triple‑digit AI product growth. Hidden dependency: Alibaba’s AI roadmap is materially dependent on continued access to high-end GPUs and third‑party semiconductor supply chains. Catalysts: quarterly cloud growth prints, NVIDIA inventory/guidance, and Chinese AI regulation in the next 30–90 days.

Trade implications: Primary direct play is BABA equity and 6–12 month call spreads to capture cloud re‑rating while limiting exposure to quick‑commerce margin drag; pair trade: long BABA vs short China consumer discretionary ETF (KWEB) to isolate AI/cloud exposure. For NVDA, use 3–6 month call spreads to express continued GPU tightness but avoid naked long volatility; rotate into semiconductor suppliers and memory names on any supply‑tightness confirmation. Time entries over 2–6 weeks, add on any post‑earnings dip >8%, and trim if cloud yoy growth falls below 20% or regulatory guidance tightens.

Contrarian angles: The market underestimates Alibaba’s ability to scale AI monetization — cloud AI product triple‑digit growth for nine quarters suggests sustainable ARPU expansion, not just sales mix. Conversely, capex acceleration could produce multi‑quarter margin compression and cash burn larger than guided (Eddie Wu hinting 380bn CNY may be “small”), so upside is likely 12–36 months while near‑term EPS may disappoint. Historical parallel: cloud leaders (AWS/GOOGL) traded through profit troughs before rerating; here the executing risk is supply-chain and geopolitical constraints that could derail a similar outcome. Unintended consequence: aggressive capex could force Alibaba to slow buybacks/dividends, pressuring short‑term returns despite long‑term structural gains.

Key Decisions for Investors

• Establish a 2–3% portfolio long position in BABA ADRs (scale in over 2–4 weeks). Target a 30% upside over 9–12 months driven by cloud re‑rating; place a stop‑loss at 12% below entry or exit if quarterly cloud revenue growth falls below 20% YoY.
• Buy a 9–12 month BABA call spread to leverage upside while capping downside: long 1x ~25% OTM call and sell 1x ~50% OTM call, size 1–2% notional. Close if adjusted EBITA margins recover unexpectedly or if China issues new restrictive AI/data guidance within 60 days.
• Take a tactical 1–2% notional position in NVDA via a 3–6 month 10–20% OTM call spread to play continued GPU tightness; unwind if NVDA guidance cuts or implied volatility compresses >40% from current levels.
• Reduce China consumer discretionary/quick‑commerce exposure by 2–3% (trim KWEB or similar ETF) and redeploy proceeds into cloud/AI infrastructure names. Revisit re‑allocation after the next two China quarterly prints or if quick‑commerce unit economics fail to improve by Q2 next year.