Alibaba: Buybacks Overpower Anthropic Spat

Alibaba shares sold off sharply, with additional pressure after Anthropic alleged the company was distilling Claude models to improve its own AI capabilities. The stock was already declining ahead of the allegation, driven by rising memory costs, a large Chinese government data center buildout, and an earnings miss. Despite the downturn, Alibaba is now the cheapest since early 2025 and its AI business is still growing at triple digits, leaving the near-term outlook mixed and cautious.

Analysis

The market is treating this as an AI credibility event, but the deeper issue is unit economics: if memory inputs stay tight and China’s data-center buildout remains policy-driven, Alibaba’s AI revenue can compound fast while contribution margin stays weak. That usually keeps the stock’s multiple anchored even when reported growth looks impressive, because investors eventually pay for free cash flow, not usage metrics. The spillover is broader than BABA: other China internet/cloud names with AI ambitions but limited pricing power should trade with the same “capex before monetization” discount.

The allegation from Anthropic matters mainly as a catalyst for policy risk. Over the next few sessions it can pressure sentiment and invite more scrutiny around access to frontier models, but it is not yet an earnings event unless it forces real product restrictions or customer churn. Over 1-3 months, the key question is whether this turns into a broader U.S.-China AI containment narrative; if not, the headline should fade faster than the valuation reset. The risk to the bearish case is that the market has already priced in a lot of bad news and is now discounting a permanently impaired AI franchise that may still be growing much faster than the rest of the business.

Contrarian view: the selloff may be overdone because the market is conflating reputational noise with durable economics. If Alibaba’s AI business is still compounding at triple digits, the more important variable is how much of that growth converts into cloud ARPU and operating leverage once memory costs normalize. The thesis would be falsified by a stabilization in gross margin or any indication that AI demand is taking share without incremental capex, which would force a sharp re-rating from “story stock under attack” to “cheap platform with optionality.”

Key Decisions for Investors

• Do not chase the short immediately; wait 1-3 trading sessions for the headline to either produce follow-on policy risk or fade. If no new regulatory escalation emerges, look for a tactical long BABA on stabilization, targeting a 10-15% mean reversion over 1-2 months with a tight stop below the post-selloff low.
• For higher-conviction expression, use a BABA call spread 3-6 months out rather than outright stock. Risk/reward is favorable if the market is over-discounting AI reputational risk, but the structure limits downside if China AI capex/memory costs keep compressing margins.
• Relative-value pair: long BABA / short KWEB or short a weaker China AI proxy on any relief rally. The thesis is that Alibaba’s valuation has detached more than fundamentals, while sector-wide sentiment remains vulnerable to the same input-cost and policy pressures.
• Set an alert on any further U.S. policy or model-access restrictions tied to China AI usage. That would convert this from a one-off headline into a 1-3 month derating event and would invalidate a tactical long.
• Watch next earnings for AI-to-cloud monetization and operating margin commentary. If management cannot show that triple-digit AI growth is translating into better cash economics, the stock can stay cheap for another 6-18 months despite headline growth.