Last week I opened my portfolio after work to find BABA down again. The 52-week high was $192. Now it's $95. Halved. My first instinct was to chalk it up to broad China stock weakness, but tracking the news story by story revealed something different. This isn't simply "cheap because it's China."
25% in One Month: This Was a Compound Shock
BABA was hovering around $130 in early June. By June 25 it had slid to $95 β a 25% decline in roughly a month.
A lot of coverage tries to explain this with a single cause. The reality is three separate shocks hit within the same window.
First, the US Department of Defense added Alibaba to its Chinese Military Company list under Section 1260H. There are no immediate sanctions, but once this label is attached, large institutions start trimming positions for compliance reasons. The institutional selling precedes any actual enforcement.
Second, Anthropic formally accused Alibaba's Qwen AI lab of creating 25,000 fake accounts to generate 28.8 million interactions with Claude models. Anthropic submitted this accusation directly to US senators and the White House in an official letter. This is not a company-to-company dispute β it has the potential to become a Congressional agenda item.
Third, Nomura's estimate of the June 18 shopping festival (618) showed core e-commerce revenue down 8% year-over-year. The market had priced in flat.
All three arrived simultaneously. That's why the stock moved like this.
Cloud Grew 36% β So Why Is the Stock Halved?
This is the sharpest contradiction surrounding BABA right now.
CEO Eddie Wu has said AI-related revenue has grown triple digits for ten consecutive quarters. Cloud revenue was up 36% year-over-year. On the surface, this looks like a growth company.
But the Q3 FY2026 results (reported March 2026) tell a different story. EPS came in at $6.96 against a consensus of $11.88. Net income fell 66% year-over-year. EBITDA fell 57%. Cloud revenue rose 36% while earnings evaporated by more than half.
The structure explains it. Alibaba is in the middle of a massive AI infrastructure buildout. That capital expenditure is consuming the current profit base. Spending for future growth isn't inherently bad β the problem is that the market has zero visibility into when that investment converts to positive free cash flow. Right now FCF is negative.
At P/E of 15.6x the stock looks cheap versus its five-year average of 20.9x. That discount is real. But P/E is nearly meaningless when FCF is negative. Measuring a multiple against earnings that don't exist is an optical illusion.
Why Geopolitical Risk Drags On This Long
The Pentagon blacklist story first appeared in February. Back then the listing was briefly floated and then withdrawn the same day. Hong Kong shares fell 3% before confirmation. Price moved before fact.
In June it came back. Alibaba has filed suit in federal court.
The relevant question isn't the lawsuit outcome. It's what happens while the label sticks. Large pension funds, sovereign wealth funds, and institutional managers often cannot hold securities classified as military-affiliated companies under their internal compliance policies. Supply deteriorates even without actual sanctions. Even if the lawsuit reverses the listing, the process takes quarters.
This has created a pattern for BABA: every technical bounce gets sold. As long as the institutional exit structure holds, rallies function as selling opportunities rather than recoveries.
The Anthropic Complaint Hits the Core of Alibaba's AI Thesis
The most recent trigger for BABA's decline was Anthropic's formal accusation.
If true, the issue isn't simple IP theft. It becomes a credibility problem for the Qwen model itself β the centerpiece of Alibaba's AI strategy. Those headline numbers (cloud up 36%, AI revenue triple digits for ten quarters) rest on an assumption that the underlying capability is genuinely proprietary. That assumption is now contested.
Alibaba has denied the allegations and signaled legal action. The back-and-forth will continue. But the fact that Anthropic sent formal letters to the Senate and the White House means this issue has the potential to become a new front in the US-China technology war. If it escalates to Congressional hearings, expanded export controls, or ADR-related legislation, the risk profile shifts to an entirely different order of magnitude.
Whether Anthropic's specific claims hold up is hard to judge from here. What's clear is that this may not end as a company-level dispute.
So Is BABA Cheap or Dangerous?
37 analysts have buy ratings with an average price target of $187. From $95, that's nearly a double. Valuation metrics alone make it look like an opportunity.
But consider the structure. Alibaba is currently fighting on three fronts simultaneously: profit destruction from AI investment costs, a federal lawsuit over the Pentagon blacklist, and the Qwen credibility problem opened by Anthropic's complaint. Each alone would be digestible. The problem is they're amplifying each other.
Because earnings visibility is low, the argument "wait through geopolitical risk" weakens. Because geopolitical risk is high, the argument "trust the AI growth" weakens. Both counterarguments are neutralized at the same time.
"Cheap" may be accurate. But cheap doesn't mean it goes up. For it to move, either FCF conversion needs to be confirmed, or the blacklist lawsuit needs to be reversed. Bounces before either of those happen are technical bounces β and the current pattern treats those as distribution events.
Can Alibaba recover eventually? Entirely plausible. Whether this is the floor is unknown. Until one of those two catalysts materializes, the honest question to ask yourself is: is this investing, or is this waiting?
#Alibaba #BABA #AlibabaStock #ChinaStocks #USChinaTension #ChinaAI #AlibabaCloud #BABAAnalysis #Qwen
Investment Disclaimer: This post is for informational purposes only and does not constitute investment advice. All investment decisions and their outcomes are the sole responsibility of the reader. Nothing in this post should be construed as a recommendation to buy or sell any specific security.