On the evening of August 23rd, Alibaba (NYSE: BABA) announced its first discounted share placement since its Hong Kong listing seven years ago: issuing 710 million new shares at HK$112.70 each to raise HK$80 billion, with all proceeds earmarked for AI. The following day, Hong Kong-listed shares plunged over 9%, briefly dipping below the placement price during trading. On the surface, this is a fundraising move. But looking deeper, it reveals two simultaneous battles the company is fighting and a long-standing corporate trait that hasn't changed over the years.

Let's first clarify the nature of this capital raise. The 710 million new shares represent approximately 3.7% of the existing share capital, with a placement price of HK$112.70, a discount of about 9% to the average price over the previous five trading days. This marks the first new share placement since Alibaba's Hong Kong listing in 2019 and the largest primary follow-on offering in Hong Kong market history. The funds are 100% allocated to AI—computing power, large language models, and commercialization—not for debt repayment or e-commerce subsidies. Subscribers are primarily sovereign wealth funds from the Middle East, Europe, and Asia, with the offering oversubscribed within an hour. However, the heat in the primary market didn't translate to the secondary market: Hong Kong shares fell over 9% on August 24th, briefly touching HK$110.10, breaking below the HK$112.70 placement price. In other words, institutions that acquired shares at a 9% discount were already sitting on paper losses by day one. The market's message was clear through the sell-off.

The real question is: why is Alibaba asking shareholders for this HK$80 billion at this particular moment? The most direct reason is that its cash flow is no longer sufficient—not that the balance sheet is empty, but that operating cash flow can't keep pace with expenditures. Looking at recent fiscal years: in FY2025, free cash flow was still a positive inflow of RMB 73.9 billion; in FY2026, it reversed sharply to an outflow of RMB 46.6 billion; and in the single quarter ending June 2026, capital expenditure reached RMB 67.678 billion, which is 2.95 times the operating cash flow of RMB 22.945 billion for the same period. In other words, the company's operating earnings can't even cover a fraction of its capital spending. Annualized, capital expenditure stands at roughly RMB 270 billion against operating cash flow of about RMB 92 billion, leaving a gap of approximately RMB 180 billion. The balance sheet does hold RMB 474.5 billion in cash, but that cash isn't reserved solely for AI—it also needs to fund dividends (RMB 33.7 billion in FY2026), buybacks (cumulative USD 24.4 billion over the past two fiscal years), and maintain a safety buffer. If the company were to rely purely on internal funds to finance AI, the cash pool would deplete at a rate of RMB 150-200 billion annually, hitting bottom within two to three years. So this HK$80 billion isn't exactly "lifesaving money"—it's money raised because Alibaba doesn't want to exhaust its cash reserves to fund AI.

But here's the catch—AI isn't the only place where money is burning. Alibaba is simultaneously fighting two battles. It's currently pouring funds into two bottomless pits. The first is AI: a three-year investment plan of RMB 380 billion, with FY2026 capital expenditure of RMB 126 billion and quarterly capex up 75% year-over-year, which has directly slashed net profit by 75%. This is the main battleground where it's betting on the future. The second is instant retail. In April 2025, Ele.me was formally integrated into "Taobao Flash Purchase," with Alibaba setting its top priority for 2026 as "absolute No.1 market share," firmly increasing investment. The result: quarterly revenue surged to RMB 20 billion, up 57% year-over-year, and market share briefly overtook Meituan by 0.2 percentage points to claim the top spot—but at a cost. According to HSBC research estimates, losses from food delivery plus instant retail over the past year amount to tens of billions, with some media calculations putting the figure at over RMB 80 billion annually, consuming nearly half of full-year profit. And just after being summoned for regulatory talks (over forced price cuts and exclusive dealing), the company continues to burn cash. While the intense investment phase in instant retail has temporarily eased, the sector has already absorbed significant cash flow, and this battle is far from over.

On one side, there's AI, a long-cycle deep pit; on the other, food delivery, a high-intensity subsidy war. Both are money-burning battles, and neither can afford to be lost—losing AI means missing the next era, while losing food delivery means failing to defend the core e-commerce base. This is Alibaba's biggest problem right now: it isn't fighting one battle but two simultaneously, with only one pool of money to fund both. Moreover, this isn't the first or even second time the company has turned to shareholders for capital. Looking back: in 2007, it listed its B2B business separately on the Hong Kong Stock Exchange; in 2012, it privatized and delisted; in 2014, the group went public in the U.S. with a USD 25 billion IPO, then the world's largest; in 2019, it returned for a secondary listing in Hong Kong, raising HK$87.5 billion from new shares; in between, it also issued convertible bonds; and now, another HK$80 billion placement. Over the past two decades, Alibaba's fundraising history has been more密集 than most companies. Every fundraising round has been backed by a new vision—B2B, globalization, cloud computing, new retail, and now AI. Money keeps coming from shareholders, and the vision keeps changing.

But why does Alibaba always need so much money and so many new narratives? This brings us to its most fundamental nature. Charlie Munger bought Alibaba, used leverage, and eventually exited at a loss, leaving behind a now-frequently-quoted remark: "Investing in Alibaba was one of the worst mistakes I've ever made in my life. I was fooled by its dominant position in China's internet, but I failed to see that it's essentially a goddamn retailer." Internet retail is an extremely competitive arena where no one can easily coast to profits. Munger's words hit the mark: no matter how glamorous e-commerce appears, it's still retail at its core, and retail is a tough business. Alibaba has spent years trying to escape this "toughness." It has turned Taobao and Tmall into a "virtual Wanda Plaza"—Tmall as the upscale mall, Taobao as the wholesale market—essentially operating as a "traffic landlord and ecosystem tax collector." But the problem is that "prime locations" in the digital world lack physical scarcity. Pinduoduo opened a new entrance through WeChat virality and rock-bottom pricing, while Douyin shifted from "people looking for products" to "products finding people" through content recommendation. Taobao and Tmall's market share has steadily declined from over 60% in 2019 to 31-33% in 2025. The moat still exists but has grown shallower—degraded from a "monopolistic deep moat" to an "infrastructure-level medium moat." So Alibaba is trying to use high-frequency scenarios like food delivery and instant retail to pull users back, driving traffic to the lower-frequency transactions on Taobao and Tmall. This is the real motivation behind its food delivery war—not a belief in delivery itself, but a strategy to use high-frequency engagement to rescue low-frequency commerce. Yet this logic, at its core, is still "buying traffic with money," no different in essence from its old model a decade ago of "buying traffic to fill shelves and sell ads." It tried to escape the intense competition of retail, only to leap into an even more capital-intensive battleground.

If it were just the tough e-commerce business, the situation might not have reached this point. Alibaba's second trait has magnified its problems severalfold: its love for grand narratives. Among these grand narratives, Tmall, Taobao, and Alibaba Cloud have delivered results, but some ambitious visions have come with expensive tuition fees. Consider the entertainment division: Youku Tudou was fully acquired in 2016 for approximately USD 4.5 billion, and nearly a decade later it's still loss-making, rebranded as "Killer Whale Entertainment" in 2025 in an attempt to turn the page. Over those ten years, it told countless stories about "content ecosystems and full IP industry chains," but the jam never actually made it to the table. Then there's local life services (Ele.me), which has been consistently overtaken by Meituan—revenue grew from RMB 44.9 billion in FY2022 to RMB 59.8 billion in FY2024, but EBITA losses have exceeded RMB 10 billion annually, earning it the market label of "bottomless pit." And new retail: Intime and Sun Art were once the showcase of the "new retail" vision, now sold off in bulk—an official admission that this vision wasn't realized. Each of these involved asking shareholders for money and then diluting existing shareholders. Now it's AI and instant retail's turn, following the same old playbook. Buffett's framework applies here: even the best farmer can't grow good crops in bad soil. Entertainment, local life services, and even new retail are not easy businesses. Ultimately, it's the shareholders who foot the bill.

Returning to the original question: why is Alibaba issuing these HK$80 billion in new shares? Because it's simultaneously fighting two battles it can't afford to lose—AI and instant retail; because the tough e-commerce business forces it to constantly seek new battlegrounds; because its love for grand narratives demands real capital for every vision. And every vision requires money, which can only come from shareholders, time and again. From the B2B listing in 2007, to the U.S. IPO in 2014, the Hong Kong listing in 2019, and now this HK$80 billion raise—Alibaba's shareholders have consistently paid for these grand visions at a steep cost. If you bought on the IPO day in 2014 and held until today, your annualized return including dividend reinvestment would be approximately 3.2%, with cumulative gains of roughly 46-50%. Over the same period, the S&P 500 has delivered annualized returns of 10-11%, with cumulative gains exceeding 258%. Alibaba has underperformed the S&P 500 by over 200 percentage points, with a maximum drawdown of 80%. Twelve years, 3.2% annualized, enduring an 80% drawdown—this return doesn't even beat U.S. Treasuries. Shareholders have absorbed all the volatility of "grand narratives" while receiving only "bond-level" returns.