Ed Ludlow's segment for Bloomberg Tech landed at a perfect moment of tension. Alibaba just told the market it needs more money to build AI infrastructure — and it is going to shareholders for it. The HK$80 billion placement is a record, but the real story is the economics of this AI buildout.

What Alibaba Just Did

The mechanics are straightforward, but the scale is historic. Alibaba launched an HK$80 billion (USD 10.2 billion) share placement on August 23, 2026, selling 710 million new ordinary shares at HK$112.70 each. That makes it the largest-ever primary follow-on offering by a Hong Kong-listed company and the world's third-largest this year, trailing only Alphabet and Intel.

The deal was oversubscribed, which forced the size up. Sovereign wealth funds were among the buyers; Morgan Stanley, HSBC, UBS, and CICC ran the books. One critical detail: the shares were not registered under US securities laws, so US investors were not eligible. That is a structural choice.

The part to focus on: 100% of the net proceeds are earmarked for Alibaba's "full stack" AI capabilities. Not debt repayment. Not buybacks. Every dollar is going into compute, models, and infrastructure.

The 10.2 Billion Question: Why Equity, Not Debt

When a company with Alibaba's balance sheet chooses equity over debt, you have to ask why. UBP senior equity advisor Vey-Sern Ling put it bluntly: this was unexpected, and it could signal that Alibaba needs more AI funding than investors realized, and wants to move faster than cash flow alone would allow. Equity dilution is the more expensive form of capital. It signals urgency: they are not waiting for operating cash flow to catch up.

Michael Burry is on the other side of the trade. He argues that Alibaba's return on invested capital could keep declining, and he will not rebuild his position at current levels. The numbers support him. June-quarter capital expenditure rose 75% to 67.7 billion yuan, roughly USD 9.5 to 10 billion. Net profit plunged 75% year over year. Alibaba has already burned through nearly half of its three-year capex plan of at least 380 billion yuan.

CEO Eddie Wu framed it as a prerequisite: "In order to be able to capture that future growth, we first need to make these capex investments to build out the necessary compute capacity." That is the bull case. The bear case is that you are spending ahead of returns, and the market is punishing you for it. The stock is down roughly 40% from its January high, and shares fell about 7% on the earnings miss. Meanwhile, Alibaba Cloud AI revenue surged 45%. Growth is real; the question is whether it is profitable.

The Circular Financing Machine

This is where the segment gets uncomfortable, and rightly so. Alibaba is not operating in a vacuum. It is participating in a global system where the biggest AI players are financing each other in a loop that looks increasingly self-referential.

Start with Microsoft: its 2023 investment made OpenAI a giant Azure customer. Amazon and Alphabet invest in Anthropic, which buys AWS, Google Cloud, and chips. Nvidia has taken big stakes — roughly USD 30 billion in OpenAI's USD 110 billion round in February 2026 — and is reported to be in talks to guarantee nearly USD 250 billion for OpenAI data center projects. Nvidia also funds neoclouds like CoreWeave and Nebius, then buys cloud from them. OpenAI has struck close to USD 1 trillion in deals with Microsoft and AMD.

The numbers are staggering: hyperscalers — Microsoft, Amazon, Alphabet, Meta — are expected to spend roughly USD 725 billion in capex in 2026. Three-quarters of a trillion dollars from four companies, largely to serve AI labs those same four companies fund.

ING's Jan Frederik Slijkerman argues strong balance sheets mitigate the risk, and he is not wrong — these are not junk-rated companies playing with leverage. But he also acknowledges the domino risk: if the AI bubble bursts, everyone falls together. The lender is the customer is the investor is the supplier.

Meta's Quiet Corner

The segment also highlighted something that gets less attention: Meta has quietly become one of the biggest forces in AI infrastructure financing. They do not get the same headlines as Microsoft or Nvidia, but they are deploying enormous capital into the ecosystem, often in ways that are not obvious.

Meta is not just building data centers; it is funding and guaranteeing capacity across the supply chain with the discipline of a company that has been through a brutal efficiency reset. The lesson: you do not need to be the most visible player to be the most consequential. You need to control the capital flows.

What This Means: The System Is Working Exactly as Designed

Here is my take, and it is not the one you will hear from the cheerleaders: this is what a capital-intensive platform shift looks like. Every major technology transition — railroads, electricity, the internet — required a period where capital was deployed ahead of returns. The winners were the ones that could survive the gap between spending and revenue.

The circularity is a feature until it is not. Right now the system functions because everyone believes the other guy will keep paying. Microsoft believes OpenAI will keep buying Azure. Nvidia believes OpenAI will keep buying chips. OpenAI believes the hyperscalers will keep funding them. Alibaba believes the market will reward its capex with future AI revenue.

That works until it does not. The risk is not that any single company fails. The risk is that the loop tightens simultaneously. If one major player pulls back, the others feel it immediately — the dominoes are wired into the financing structure.

For operators and investors, the metric to watch is not revenue growth. It is cash conversion. Alibaba's AI Cloud revenue is growing 45%, but net profit is down 75%. That gap is the whole story — whether these companies can turn compute into profit before the loop tightens.

What Comes Next

The immediate question is whether the placement closes clean. Oversubscription is a good sign, but the real test is where the stock trades in the weeks after. If the shares hold above the placement price, the market is endorsing the strategy. If they drift lower, the market is saying the dilution is not worth the bet.

Watch capex guidance at the next earnings call. They have already spent nearly half of their three-year plan. Raising the ceiling means all-in; holding the line means feeling the pressure from the profit collapse.

The bigger question is whether other Chinese hyperscalers follow. Tencent and Baidu are watching. Success will tempt them to do the same; struggle will send them looking for other ways to fund AI ambitions.

And the systemic question remains: what breaks first if financing loops tighten? Probably the weakest balance sheet in the chain — not the smallest company, but the one with the least cash flow to cover its commitments. Alibaba just bet shareholders will fund the gap. The market is watching whether it pays off.

— Allan Ali, Sylt.ing