$Alibaba(BABA.US) reported earnings. The numbers were solid, yet the stock fell, which did not feel great.

Dolphin Research has explained the reasons both on the Longbridge community and in its earnings take. Faster cloud growth plus e-com losses bottoming were already known and priced in ahead of the print.

The new delta was RMB 67 bn in CAPEX and negative FCF of RMB 45 bn. Mgmt said the CAPEX spike reflects procurement timing, meaning chips are available in 2H, and cautioned against linearly extrapolating this spend.

Still, the market seeing such a large CAPEX number prompted investors to lock in gains. Then an overnight follow-on was launched over the weekend, which again hurt sentiment.

Dolphin Research had flagged during its earnings read that single-quarter profit of RMB 30–50 bn before the food-delivery war implies future quarterly investment of RMB 30–50 bn. That means either a hard exit from the delivery war or financing.

Between the two, a straight exit from the delivery war is almost impossible. More likely, they will fight and retrench, keeping e-com steady while channeling the rest into a bold AI bet to build a new Alibaba.

In this process, e-com still consumes some firepower while AI requires investment. With a timing mismatch, financing becomes necessary. That said, Dolphin Research expected debt first to lever the balance sheet, or off-balance-sheet structures like compute ABS, and only then equity if the gap persisted.

This is mainly because Alibaba’s net cash is RMB 208 bn. That includes cash and equivalents, short-term investments, U.S. Treasuries, and other unrestricted equity stakes, net of bank loans, preferreds, convertibles, and other interest-bearing debt.

In reality, this past weekend $BABA-W(09988.HK) executed a HK placement raising HK$80 bn. It was an overnight bookbuild with T+3 settlement and quick cash-in.

Thinking it through, Dolphin Research sees a few likely reasons. The first:

a. Issuing USD bonds offshore pits Alibaba against U.S. CSP majors for capital. Its credit rating is lower, making offshore funding costlier, and tapping the China USD bond market for tens of billions could take time.

b. For equity, non-USD demand can be tapped without a long process. Also, the stock had recovered, making issuance more feasible.

Now to pricing. First:

a. HK$112.7 per share vs. Friday’s HK$123 close implies a ~9% discount. The stock gapped down at the open to reflect that.

b. The deal issued 710 mn new shares vs. a prior base of 19.175 bn. That equates to dilution of 3.7%.

c. In 2025, buybacks will barely offset SBC. With this issuance, pushing the stock back into an upward channel leaves little room for those focused on shareholder returns.

From a public-market investor’s lens, without a delivery war in 2025, this scale of dilution might have been avoided. Whether the war truly reshapes the domestic food-delivery competitive landscape will take 2–3 years to judge, once subsidies fade and scale plus ops discipline determine who gains share.

Channeling funds into AI, rather than delivery, is a long-overdue capital allocation correction. Because it is late, this HK$80 bn raise might have been unnecessary, or at least not this early.

Beyond short-term expectations and capital moves, Alibaba has hit a clear strategic pivot: from ‘full-stack AI + one-stop consumption’ to going all-in on full-stack AI. E-com just needs to stay broadly stable to fund AI.

Relying on a single e-com leg has already cost Alibaba dearly. The e-com battlefield is now a multi-front war of attrition, and throwing in more capital only defends the perimeter with little chance of a knockout.

Fortunately, the AI era offers Alibaba another leg to grow and strengthen. This AI venture capital bet is almost unavoidable, and returns from prior equity investments—such as into domestic model leaders and CXMT—already outstrip the e-com war burn and the near write-offs from new retail offline bets.

On relative ROE and capital allocation, Dolphin Research maintains its view that this financing likely marks a stock low. Further compute ABS may still come, but this should enhance Alibaba’s upside optionality.

<End here>

Dolphin Research on Alibaba (recent):

《阿里（纪要）：AI CAPEX 3 年回本》

《阿里狂投，只为 ‘中国版谷歌’？》

《业绩后，阿里小会交流信息》

本文的风险披露与声明：海豚研究免责声明及一般披露