Alibaba

Alibaba's Q3 fiscal 2026 (calendar Q4 2025) print confirmed the AI pivot is no longer aspirational — Alibaba Cloud revenue grew 34% YoY driven by AI training and inference workloads from external customers. Cumulative external Cloud revenue surpassed RMB 100 billion (~$14 billion) through February of fiscal 2026, with management explicitly framing FY2026 as the inflection point where Cloud profitability scales materially. Q4 fiscal 2026 (calendar Q1 2026) earnings are scheduled for May 2026 release.

What makes Alibaba structurally distinctive in 2026 is the convergence of three strategic pivots under CEO Eddie Wu (in role since September 2023). First, Qwen 3 was released as open source in April 2026, joining a family of over 200 models with more than 300 million cumulative downloads worldwide and over 100,000 derivative models built by third parties — making Qwen the world's largest open-source LLM family by adoption. Second, Wu created a new business group called Alibaba Token Hub (ATH) in early 2026, consolidating Tongyi Lab, the Model-as-a-Service (MaaS) business line, the Qwen unit, the Wukong unit (video model), and an AI innovation business unit. Third, the e-commerce business has stabilized after multiple years of competitive pressure from PDD/Pinduoduo, with Taobao and Tmall growing low-single-digits as merchant subsidies have been calibrated against PDD's pricing.

BABA is the highest-conviction smart-money name we track outside of U.S. tech: David Tepper's Appaloosa Management initiated a position in Q3 2024 at average cost in the $80s and has held continuously through Q1 2026 — the position is now Tepper's third-largest holding and was sized at approximately $1.4 billion as of the most recent 13F. Tepper has been publicly vocal about his thesis: BABA is undervalued on Cloud-revenue growth alone, and the AI optionality is essentially free.

Michael Burry's Scion Asset Management held BABA through 2024 (notably one of his higher-conviction positions before Scion deregistered in November 2025); since deregistration, Burry has not publicly disclosed a current BABA position, but his last Substack commentary in March 2026 reiterated bullish framing on Chinese tech valuations broadly. Stanley Druckenmiller cycled out of BABA in mid-2025 but has discussed it positively in subsequent public commentary. Mohnish Pabrai initiated a small position in late 2025 according to his Q1 2026 13F. The combined smart-money positioning is the strongest among non-U.S. mega-caps and signals that the China-tech regulatory overhang has compressed materially since the 2021-2022 crackdown peak.

Alibaba ADRs trade at $135 per share (HK shares at HK$132 equivalent), roughly 13× consensus FY2027 (March-end) EPS of $10.40 — meaningfully cheaper than Tencent at 18× and PDD at 14×. EV/Cloud-revenue sits at approximately 4× versus Alphabet (GCP) at 9× and Microsoft (Azure) at 12× — making BABA materially the cheapest hyperscaler-AI exposure on global public markets. Wall Street consensus across 36 covering firms averages $185 (JPMorgan $200, Goldman $190, Morgan Stanley $180, HSBC $165 as the bear), implying ~37% upside. Berkshire Hathaway does not hold BABA (Buffett has historically avoided China). The Tepper-led smart-money conviction provides the institutional anchor instead.

1. May 2026: Q4 fiscal 2026 earnings (calendar Q1 2026) — first full quarter to confirm whether Cloud growth holds above 30% as comparables tighten, and the first explicit ATH (Alibaba Token Hub) revenue disclosure
2. Mid-2026: Qwen 4 release and Alibaba Token Hub commercial monetization disclosure — first quarter where Wu's strategic restructuring should produce visible revenue contribution from AI-Model-as-a-Service rather than embedded in core Cloud line
3. Throughout 2026: U.S.-China audit-cooperation negotiations under HFCAA — any meaningful framework breakdown re-triggers ADR delisting risk; conversely, formal cooperation extension de-risks the multi-year regulatory overhang

Alibaba is the most asymmetric mega-cap I track outside the U.S. and arguably the cheapest hyperscaler-AI exposure on global public markets. You get Tepper as smart-money anchor at $1.4 billion (third-largest Appaloosa position), Cloud growth at 34% YoY with positive segment margins, Qwen as the world's largest open-source LLM family by adoption, and a 13× forward multiple that prices in the China-tech regulatory tail risk plus the PDD competitive headwind. My add-trigger is any quarter where Cloud growth holds above 30% AND ATH (Alibaba Token Hub) discloses commercial AI-MaaS revenue separately — that combination would invalidate the bear case that AI optionality is non-monetizable. I would not chase BABA above $160; I am building the position aggressively at any pullback below $125 where the multiple compresses below 12× and Tepper's average cost remains above water. The thesis breaks if the ADR delisting framework re-activates with regulatory teeth or if Cloud margins plateau below 15% for two consecutive quarters.