Alibaba plans to raise HKD80bn to expand full-stack AI investment; Nomura sees manageable dilution and maintains Buy

Overview

This report assesses the impact of Alibaba's proposed HKD80bn equity placement on shareholder dilution, its balance sheet, AI competitiveness, and cash flow, while using its June-quarter results to explain developments in AI Cloud, T-Head chips, AI applications, and e-commerce. Nomura believes that although the financing method was unexpected, its economic rationale is understandable and it does not alter the firm's positive investment thesis.

Core views

Alibaba announced on August 23 that it plans to conduct an HKD80bn equity placement, equivalent to approximately USD10.2bn, with all net proceeds to be used for full-stack AI capabilities, including the expansion and upgrading of AI infrastructure. The transaction remains subject to market and other conditions. According to information cited in the report, the company plans to place approximately 710mn shares at HKD112.70 per share, an 8.4% discount to the previous trading day's Hong Kong closing price. The offering attracted substantial demand, including from sovereign wealth funds, and was upsized after being oversubscribed. Nomura estimates that the new shares will dilute existing shareholders' equity by approximately 3.7% and considers the dilution manageable. Upon completion of the transaction, total cash and cash equivalents are expected to increase from USD70bn at the end of the June quarter to USD80bn. The market had already widely expected Alibaba to raise new capital, given the intensive fundraising by U.S. hyperscale cloud providers during the year. Nomura believes this expectation may have contributed to the 8.6% plunge in BABA's U.S. ADRs on August 21, even though management had provided an upbeat AI Cloud outlook only one day earlier. The report judges that the formal announcement should help remove the financing overhang and refocus market attention on fundamentals. What genuinely surprised Nomura was the use of equity rather than debt financing. Given the company's share valuation at the time, it had previously expected the company to rely more on debt financing. However, the global AI capex boom has driven a surge in bond issuance by hyperscale cloud providers, widening credit spreads, increasing new-issue concessions, and reducing investor subscription coverage ratios. This has weakened the relative economics of debt financing, making the decision to use equity financing understandable. The proceeds will strengthen Alibaba's existing full-stack AI positioning. Nomura believes the company occupies a relatively favorable position in China's AI ecosystem, with capabilities spanning foundation models, MaaS, cloud infrastructure, and proprietary AI chips. External cloud revenue grew 45% y-y in the June quarter, accelerating for the ninth consecutive quarter. Management expects growth to exceed 50% in the September quarter and to accelerate further in the December quarter and the following March quarter. As of August, MaaS annualized recurring revenue had exceeded CNY16bn, and management said it could surpass CNY30bn by the end of the current fiscal year. AI Cloud's growth has not come at the expense of profitability. Its adjusted EBITA margin reached 11.6%, up from approximately 7% about a year earlier. Management expects continued improvement over the next few quarters and a gradual move toward its long-term target of approximately 20%. The report attributes the improvement to strong revenue growth, higher utilization and pricing, an increased contribution from high-margin MaaS revenue, and greater adoption of proprietary T-Head chips. T-Head chips could become an important competitive advantage for AI Cloud. Its products cover GPUs, CPUs, storage, and networking chips, enabling Alibaba to jointly optimize computing, storage, and networking at the system level rather than relying entirely on third-party hardware. Cumulative shipments of previous generations of T-Head AI chips have exceeded 500,000 units. The latest generation of domestically produced chips began deployment in supernode configurations on Alibaba Cloud's platform in August and has entered the large-scale commercialization stage. Management expects the next generation of domestically produced AI chips to begin tape-out in 2H26, targeting both large-scale model training and inference. Nomura believes that if the supply of advanced GPUs remains constrained and T-Head chips can support the training of large-parameter foundation models at commercial scale, their capabilities will significantly enhance the competitiveness of Alibaba Cloud and its foundation-model business. The newly established AI Labs and Applications (ALA) segment consolidates the Qwen foundation model and both traditional and emerging AI applications. The segment recorded an adjusted EBITA loss of CNY13.9bn in the June quarter, reflecting continued heavy investment in foundation models and the Qwen app. Management said this quarter may represent the peak loss for the current fiscal year. The September-quarter loss is expected to narrow to the low-teens range in billions of renminbi, then broadly stabilize or continue improving. Model training accounted for less than half of the segment's loss, with the remainder primarily arising from Qwen app inference, marketing, and operating costs. Nomura therefore believes that continued improvements in training efficiency can allow model capabilities to keep advancing without losses increasing in direct proportion to model scale. Qwen's commercialization remains at an early stage, with external monetization currently occurring mainly through Alibaba Cloud's MaaS services. Over the long term, management believes additional business models beyond MaaS can be developed. The mature e-commerce businesses are positioned as the group's profit and cash-flow foundation. Alibaba Ecommerce Group (AEG) includes China e-commerce, quick commerce, and international e-commerce, and its adjusted EBITA was broadly stable at CNY39.7bn in the June quarter. China e-commerce customer management revenue declined 7% y-y but grew 1% y-y on a comparable basis, broadly in line with market expectations. Management said China e-commerce performance in the September quarter to date had improved slightly from the June quarter. Quick commerce unit economics also continued to improve, with the September-quarter loss expected to remain flat or decline q-q and the full-year loss expected to narrow by approximately 50% y-y, while the FY29E profitability target remains unchanged. Nomura believes the company is placing greater emphasis on profit stability and cash generation in its mature e-commerce businesses rather than pursuing growth at the expense of profitability, thereby providing a more stable earnings and cash-flow foundation for AI investment. The key financial pressure remains AI capital expenditure and free cash flow. Capital expenditure rose from CNY26.9bn in the previous quarter to CNY67.7bn in the June quarter, equivalent to approximately USD10bn, while free cash flow recorded a net outflow of CNY44.7bn. Management cautioned against simply annualizing the quarter's capital expenditure because the timing of hardware procurement and delivery can create quarterly volatility. Meanwhile, management said incremental AI infrastructure investments can typically be recouped within three years. Improving AI product margins, increased use of T-Head chips, customer prepayments, and alternative computing arrangements could further enhance capital efficiency. Nomura therefore concludes that the company is willing to tolerate negative free cash flow in the near term to capture AI growth opportunities. On valuation, Nomura uses a sum-of-the-parts approach: it values China e-commerce at USD87bn based on 5x FY27F P/E, Alibaba Cloud at USD278bn based on 7x FY28F P/S, and assigns a net value of USD40bn to non-core assets, including international e-commerce, deriving a target price of USD178. The report maintains its Buy rating, believing that approximately 3.7% dilution and near-term cash-flow pressure do not alter its positive investment thesis.

Analysis framework

The report first quantifies the placement size, discount, equity dilution, and post-transaction cash level, then compares the relative costs of equity and debt financing to explain the change in financing method. It subsequently uses June-quarter operating data to test the AI investment thesis, sequentially analyzing cloud revenue and margins, T-Head chips, AI application losses and commercialization, e-commerce cash-generation capacity, and the capital-expenditure payback period, before linking its operating assessment to the USD178 target price through a sum-of-the-parts valuation.

Methodology notes

• Valuation methodologySOTP valuation

  Sum-of-the-parts valuation

  The report separately values China e-commerce, Alibaba Cloud, and non-core assets, then aggregates the value of each component to derive Alibaba's target price.

• Valuation methodologyPE/PEG valuation

  FY27F P/E valuation

  The report values China e-commerce at USD87bn based on 5x FY27F P/E.

• Valuation methodologyPS valuation

  FY28F P/S valuation

  The report values Alibaba Cloud at USD278bn based on 7x FY28F P/S.

• Company fundamentals and financial framework

  AI infrastructure investment payback-period analysis

  The report uses the typically less-than-three-year payback period for incremental AI infrastructure, together with product margins, proprietary chips, customer prepayments, and alternative computing arrangements, to assess the economic returns and capital efficiency of high capital expenditure.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Alibaba Group Holding Ltd. (BABA.US)

  The equity placement will directly dilute existing shareholders, but all net proceeds will be invested in AI infrastructure and full-stack capabilities. Nomura believes the support for long-term competitiveness outweighs the impact of manageable dilution.

  Strengths

  Possesses full-stack AI capabilities spanning foundation models, MaaS, cloud infrastructure, and proprietary chips; accelerating cloud revenue growth and improving margins; mature e-commerce businesses can provide a relatively stable profit and cash-flow foundation.

  Weaknesses

  AI capital expenditure has increased significantly, with free cash flow recording a net outflow of CNY44.7bn in the June quarter. ALA still recorded an adjusted EBITA loss of CNY13.9bn, while foundation-model and application commercialization remains at an early stage.

  Comparison

  The report notes that amid intensive fundraising by global hyperscale cloud providers and potentially constrained supplies of advanced GPUs, Alibaba's proprietary chips and system-level optimization capabilities could create a differentiated advantage.

  Risks

  Investment expansion may depress margins; regulatory risks related to payments and internet finance could affect the core business and the value of its equity interest in Ant Group.

Key data

• Proposed equity financing sizeHKD80bn (approximately USD10.2bn)100% of the net proceeds are intended to be invested in full-stack AI capabilities
• Proposed new-share placementapproximately 710mn sharesPlacement price of HKD112.70 per share
• Placement discount8.4%Relative to the previous trading day's Hong Kong closing price
• Estimated equity dilution3.7%Nomura's estimate of dilution to existing shareholders
• Total cash and cash equivalents after the transactionUSD80bnNomura's estimate, up from the reported USD70bn at the end of the June quarter
• BABA U.S. ADR one-day performance-8.6%August 21, 2026; the report believes financing expectations may have been a drag
• External cloud revenue growth45% y-yJune quarter, accelerating for the ninth consecutive quarter
• September-quarter external cloud revenue growth guidanceabove 50% y-yManagement also expects further acceleration in the following two quarters
• MaaS annualized recurring revenueabove CNY16bn as of AugustManagement expects it to exceed CNY30bn by the end of the current fiscal year
• AI Cloud adjusted EBITA margin11.6%Approximately 7% about a year earlier, with a long-term target of approximately 20%
• Cumulative shipments of T-Head AI chipsmore than 500,000 unitsCumulative shipments across previous product generations
• ALA adjusted EBITA lossCNY13.9bnJune quarter; management believes this may be the peak loss for the current fiscal year
• AEG adjusted EBITACNY39.7bnBroadly stable in the June quarter
• China e-commerce CMR growth-7% y-y1% y-y growth on a comparable basis
• June-quarter capital expenditureCNY67.7bnCNY26.9bn in the previous quarter, equivalent to approximately USD10bn
• June-quarter free cash flowCNY44.7bn net outflowThe principal financial pressure arising from high AI capital expenditure
• Incremental AI infrastructure payback periodless than three yearsThe typical payback period stated by management
• Target priceUSD178Maintained unchanged; valuation based on a sum-of-the-parts approach

Impact & implications

Nomura believes the placement will exchange manageable dilution of approximately 3.7% for about USD10.2bn in AI funding and is expected to increase total cash to USD80bn. The financing announcement may remove market uncertainty regarding the method and scale of fundraising. If the cloud business continues to accelerate, margins improve, and proprietary chips gain wider adoption, the additional capital could consolidate Alibaba's full-stack AI position. The corresponding costs are elevated near-term capital expenditure, negative free cash flow, and AI applications that remain loss-making and at an early stage of monetization.

Risks

• Accelerated investment may cause margins to decline.
• Regulatory risks in the payments and internet-finance industries could harm Alibaba's core business and the value of its equity interest in Ant Group.
• High AI capital expenditure may continue to weigh on near-term free cash flow.

What to watch

• Whether the placement can be completed after market and other conditions are satisfied, as well as the final offering size and degree of dilution.
• Whether external cloud revenue can achieve more than 50% y-y growth in the September quarter and continue accelerating over the following two quarters.
• Whether AI Cloud's margin can continue improving from 11.6% toward its long-term target of approximately 20%.
• Progress in the large-scale commercialization of the latest generation of T-Head chips and whether the next generation of domestically produced AI chips can begin tape-out in 2H26.
• Whether ALA's loss can narrow from its June-quarter peak and whether Qwen can expand monetization methods beyond MaaS.
• Whether the full-year quick-commerce loss can narrow by approximately 50% y-y while maintaining the FY29E profitability target.
• Whether changes in AI capital expenditure, hardware delivery schedules, and free cash flow align with the expected investment payback period of less than three years.