Alibaba's Share Placement Reveals a Longer AI Buildout Cycle, but the Market-Implied Value Loss Far Exceeds the Transaction's Actual Cost

Overview

Using Alibaba's August 2026 equity financing as a starting point, the report analyzes how China's three leading internet cloud service providers are funding the AI infrastructure cycle. Its core conclusion is that the financing suggests elevated capital expenditure may persist for longer, but the dilution and intrinsic value loss caused by Alibaba's transaction are far smaller than the market reaction. What truly warrants monitoring is the use of proceeds, compute utilization, cloud service pricing and project returns, rather than depreciation or short-term liquidity alone.

Core views

On August 24, 2026, Alibaba issued 710 million shares at HK$112.70 per share, raising HK$80 billion (US$10.3 billion). The issue price represented an 8.4% discount to the prior Friday's closing price, and the new shares were equivalent to 3.7% of existing shares. On August 25, Alibaba's Hong Kong shares fell 8.5%, while Tencent and the Hang Seng TECH Index declined 2.2% and 2.1%, respectively. The report estimates that the discounted issuance itself explains only about 0.3 percentage points of the share-price move: existing shareholders lost approximately US$0.9 billion due to the discount, peer market movements explain about US$6.7 billion, and a further approximately US$18 billion to US$19 billion represented abnormal repricing. Even after including the US$10.3 billion of cash in post-transaction value, the company's pro forma market capitalization remained close to US$286 billion, meaning the value abnormally erased by the market was about twice the amount raised. If the entire abnormal decline were attributed to the proceeds, it would imply approximately RMB185 of value destruction for every RMB100 raised; even a total loss of the proceeds would explain only about US$10 billion. The report therefore believes the market was primarily reassessing the scale and duration of the entire AI capital expenditure program rather than merely calculating the placement discount. What JPMorgan has changed is its view of the duration of the AI investment cycle, not its FY27 capital expenditure forecast or assessment of compute returns. The report continues to forecast RMB200 billion of FY27E capital expenditure for Alibaba and treats the June quarter's RMB270 billion annualized run rate as a stress scenario. At RMB200 billion of capital expenditure, Alibaba's annual uses exceed operating cash flow by approximately RMB100 billion; under the stress scenario, the shortfall is about RMB170 billion. This financing can cover approximately 8.4 months of the base-case shortfall or about five months of the stress-case shortfall. Alibaba also has approximately RMB150 billion of haircut-adjusted available liquidity and debt capacity costing around 2% to 3%, so the report does not consider liquidity or solvency the main issue. Net bank borrowings increased by RMB11 billion in the June quarter, while current and non-current prepayments rose by approximately RMB69 billion in aggregate. The report interprets these as including prepayments for chips and capacity, indicating that cash commitments are already running ahead of reported capital expenditure. Over the next 12 months, the report expects Alibaba to add debt and potentially seek cloud-business-level financing; another parent-company share placement would be viewed as a negative signal that returns are falling short of expectations. The three companies have adopted different financing models. Following the share-price rerating, Alibaba is using parent-company equity supplemented by debt, resulting in direct dilution but with a cost that is transparent and easy to measure. Tencent relies primarily on operating cash flow, bonds and its investment portfolio: first-half operating cash flow was RMB154 billion, or approximately RMB310 billion annualized, versus the report's RMB200 billion capital expenditure forecast. It has also paid RMB41.6 billion in dividends, while its buyback run rate is close to RMB45 billion. Tencent's quarterly net cash declined from RMB146.9 billion to RMB58.2 billion, primarily because dividends and RMB51 billion of compute prepayments occurred concurrently. The company also issued US$2.45 billion and RMB15 billion of long-term notes in June, while short-term borrowings increased to RMB86 billion and accounts payable rose from RMB121 billion at the end of last year to RMB176 billion. Supporting this is RMB487 billion of listed investments and RMB388 billion of unlisted investments, totaling RMB875 billion. The report therefore considers Tencent's parent-company equity financing risk close to zero. However, selling investments would forgo potential returns and incur tax leakage; whether such disposals are worthwhile depends on the spread between an approximately 20% capital expenditure return and the investment portfolio's high-single-digit return. Baidu has the weakest internal financing capacity. Capital expenditure in the June quarter was RMB11.4 billion, while operating cash flow was only RMB3.4 billion, with the shortfall supported by RMB17.4 billion of financing inflows. Short-term loans increased from RMB7.6 billion to RMB26.3 billion, long-term loans rose from RMB3.4 billion to RMB20.8 billion, and right-of-use lease assets doubled from the end of last year to RMB18.6 billion. Even after applying a haircut to long-term deposits and excluding iQIYI, the report estimates that Baidu still has approximately RMB100 billion of liquidity, sufficient to cover the base-case financing shortfall for many years. The constraint is not cash but Baidu's low parent-company valuation and management's reluctance to dilute at the parent-company level. Its equity financing channel is Kunlunxin, which can reduce parent-company dilution, but Baidu shareholders would also surrender part of their ownership in a core asset underpinning the AI investment thesis. The report evaluates Alibaba's share placement using four independent tests. First, whether the compute project itself creates economic value. The model assumes that, at full utilization, each RMB1 of capital expenditure generates RMB1 of annual revenue, with a 45% incremental EBITDA margin, a 20% tax rate, a five-year economic life and straight-line depreciation, while utilization ramps through 60%, 90% and 100%. For every RMB100 invested, the corresponding after-tax cash flows are -100, 25.6, 36.4, 40.0, 40.0 and 40.0, producing a 22% project internal rate of return, a 2.9-year payback period and RMB36 of net present value at a 10% discount rate. These results align with management's stated payback period of less than three years and aggregate return on invested capital of over 13%. Second, whether the transaction is accretive to earnings per share depends on the use of proceeds. Alibaba stated that the proceeds would fund full-stack AI capabilities, including infrastructure, but model training, application inference and customer acquisition expenses are recorded under AI Labs and Applications. This business currently incurs quarterly losses of RMB13.9 billion and has no auditable returns. The report therefore assumes that only 60% of the proceeds are allocated to infrastructure, with the remainder generating no quantifiable return. When infrastructure accounts for 80% or more, the placement becomes accretive to earnings per share from the second year onward. Under the 60% base case, FY28E earnings per share are diluted by 0.5% and are broadly neutral thereafter. Expressed in terms of return on invested capital, once most of the proceeds have been deployed, a mature-stage ROIC of approximately 15% is required for third-year earnings-per-share accretion. Idle proceeds dilute earnings in all scenarios, making actual capital expenditure and deployment progress in the September quarter particularly important. Third, whether equity is the lowest-cost funding source—the report's answer is no. Because RMB bond coupons are only 2% to 3%, while the equity yield is approximately 12%, the debt counterfactual produces earnings per share that are approximately 3 to 3.5 percentage points better in each year. The rationale for choosing equity comes from factors beyond earnings per share, including obtaining offshore currency for overseas procurement, preserving ratings and covenant headroom while consolidated free cash flow is negative, and securing permanent capital that does not need to be refinanced during the capital expenditure cycle. These benefits are real, but they do not mean the financing is costless. Fourth, whether the transaction increases intrinsic value per share for existing shareholders. Based on a pre-transaction fair value of HK$205, if all proceeds are invested in infrastructure and its net present value is included, post-transaction value per share is HK$203.1, down 0.9%. At a 60% infrastructure allocation, value per share is HK$202.6, down 1.2%. The transaction becomes accretive to value per share only if pre-transaction fair value is below approximately HK$153 and 100% of proceeds are allocated to infrastructure, or below approximately HK$137 under the 60% base-case portfolio allocation. Thus, whether compute projects have positive net present value, whether the transaction is accretive to earnings per share, whether equity is preferable to debt, and whether intrinsic value per share rises for existing shareholders are four separate questions. The report believes the transaction is approximately earnings-per-share neutral under the base case but transfers about 1% of intrinsic value; this cost remains far below the market's approximately US$19 billion of abnormal repricing. Expanding capital expenditure does not guarantee proportionate growth in returns. The report's sensitivity analysis shows that project returns remain above the 10% cost of capital under most utilization and pricing combinations. However, if utilization is only 80% and prices decline by 10%, the project's internal rate of return falls to approximately 11%; if utilization is 70% and prices decline by 20%, the return falls to only about 3%. Therefore, RMB270 billion of capital expenditure is treated as a risk scenario rather than an upside scenario: greater scale creates more value only if unit economics remain stable, while declines in utilization or pricing caused by supply expansion are the key threats. The income statement must absorb the buildout cycle, but the report does not consider depreciation itself the core risk. Under the same cohort model, total depreciation equivalent to one year of base-case capital expenditure represents 28% of Alibaba's FY27E net profit, 13% of Tencent's and 23% of Baidu's. However, incremental EBITDA exceeds depreciation from the first full year onward. Therefore, EBITDA, equipment utilization and compute pricing—not the financing structure or depreciation expense itself—are what truly determine returns. In relative positioning, Tencent's higher valuation reflects a largely internally funded buildout plan and its RMB875 billion investment portfolio. Alibaba has removed part of the financing uncertainty, and the report believes the market has over-discounted the negative signal conveyed by the share placement. With the stock currently trading at 12 times FY27E earnings and eight times FY28E earnings, the report recommends gradually increasing positions in Alibaba ahead of the September-quarter results. Baidu's parent-company dilution risk remains low because its equity financing is expected to occur at the Kunlunxin level. The report maintains Overweight ratings on all three companies but emphasizes that its conclusions are conditional; capital expenditure, cloud growth, AI business losses and financing actions over the next two quarters will test those conditions.

Analysis framework

The report first separates the placement discount, peer market movements and the abnormal share-price reaction to determine whether the market is repricing the transaction itself or a longer-term capital expenditure cycle. It then calculates each company's internal funding shortfall, available liquidity, financing capacity and funding coverage period. For Alibaba's transaction, the report sequentially tests project net present value, the impact on earnings per share, the cost of equity versus debt and intrinsic value per share for existing shareholders. It also uses utilization and pricing sensitivity analyses to examine return risks as scale expands, before mapping the results to relative company valuations and subsequent verification events.

Methodology notes

• Valuation Methodology

  Project Net Present Value and Internal Rate of Return Model

  The report forecasts compute project cash flows based on the revenue-to-capital-expenditure ratio, incremental EBITDA margin, tax rate, utilization ramp, economic life and discount rate, using IRR, payback period and NPV to determine whether infrastructure investment creates economic value.

• Corporate Fundamentals and Financial FrameworkROIC–WACC spread

  Comparison of Project Returns and Cost of Capital

  The report compares the compute project's approximately 22% IRR and mature-stage ROIC with an approximately 10% cost of capital, and uses an approximately 15% mature-stage ROIC as an important threshold for third-year earnings-per-share accretion.

• Corporate Fundamentals and Financial FrameworkFree cash flow analysis

  Analysis of Funding Sources, Uses and Coverage Periods

  The report distinguishes annual internal funding surpluses or shortfalls, immediately available liquidity, incremental debt or equity capacity and the period for which each funding pool can cover the shortfall, avoiding confusion between a cash shortage and long-term solvency risk.

• Valuation MethodologyPE/PEG valuation

  Relative Positioning Based on Forward P/E

  The report uses Alibaba's 12 times FY27E P/E and eight times FY28E P/E to illustrate that the market has already priced in a strongly negative signal, combining this valuation with its assessment of financing risk.

• Event Strategies and Behavioral FinanceEvent-driven analysis

  Share Placement Event and Subsequent Earnings Validation

  The report establishes event-based conditions for validating capital expenditure returns and financing paths around specific milestones, including placement settlement, September-quarter results, Tencent's Hunyuan 4 and Kunlunxin's listing.

Asset mapping & comparison

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

• Alibaba Group (09988.HK, BABA.US)

  Funds AI buildout through a parent-company share placement and debt, making its financing risk the most visible, although the report believes the market has over-discounted the placement's negative signal.

  Strengths

  Approximately RMB150 billion of available liquidity and substantial debt capacity; the compute infrastructure model indicates a 22% project IRR and positive net present value.

  Weaknesses

  At RMB200 billion of capital expenditure, there is an annual internal funding shortfall of approximately RMB100 billion; the placement directly dilutes existing shareholders, while non-infrastructure AI spending does not yet generate auditable returns.

  Comparison

  Its funding shortfall is larger than Tencent's; compared with Baidu, it is more willing to use parent-company equity, and its financing cost is also easiest for the market to observe and price directly.

  Risks

  A low share of proceeds allocated to infrastructure, prolonged idle cash, declining cloud service prices, insufficient utilization, or another parent-company share placement within 12 months.

• Tencent (00700.HK)

  Primarily funds AI capital expenditure through operating cash flow, bonds and its investment portfolio; the report considers parent-company equity financing risk close to zero.

  Strengths

  First-half operating cash flow annualizes to approximately RMB310 billion, and the company holds an RMB875 billion investment portfolio, allowing the capital expenditure plan to be largely internally funded.

  Weaknesses

  Selling investments would forgo potential asset returns and incur tax leakage; simultaneous dividends, buybacks and compute prepayments can reduce net cash.

  Comparison

  It has the most robust funding structure, with its higher valuation reflecting internal financing capacity and a large investment portfolio; depreciation represents a smaller proportion of FY27E net profit than at Alibaba and Baidu.

  Risks

  If combined capital expenditure, dividends and buybacks continue to exceed operating cash flow without corresponding asset disposals, the financing buffer will weaken.

• Baidu (09888.HK, BIDU.US)

  Capital expenditure primarily depends on borrowings, with equity financing planned at the Kunlunxin level, resulting in relatively low direct dilution risk for the parent company.

  Strengths

  Approximately RMB100 billion of adjusted liquidity can cover the base-case funding shortfall for many years, and equity financing can be conducted at the AI subsidiary level.

  Weaknesses

  It has the weakest internal operating cash flow support; June-quarter capital expenditure significantly exceeded operating cash flow, while borrowings and lease liabilities rose rapidly.

  Comparison

  It is more dependent on external financing than Tencent; unlike Alibaba, it is unwilling to issue parent-company equity at a low valuation, but subsidiary-level financing will dilute ownership in AI assets.

  Risks

  If Kunlunxin's listing is delayed beyond 2026 or the funds raised fall materially short of supporting the planned capacity expansion, the scale of AI buildout will be constrained.

Key data

• Alibaba Share Placement Size710 million shares, HK$80 billion (US$10.3 billion)The issue price was HK$112.70, an 8.4% discount to the previous closing price, with the new shares equivalent to 3.7% of existing shares.
• Alibaba Share-Price Reaction After Placement-8.5%Performance on August 25, 2026; Tencent fell 2.2% and the Hang Seng TECH Index declined 2.1% over the same period.
• Abnormal Value RepricingApproximately US$18 billion to US$19 billionApproximately twice the cash raised; the loss to existing shareholders caused by the discount itself was about US$0.9 billion.
• Alibaba FY27E Capital ExpenditureRMB200 billionThe forecast remains unchanged; RMB270 billion is treated as a stress scenario.
• Alibaba Annual Internal Funding ShortfallBase case approximately RMB100 billion; stress scenario approximately RMB170 billionThis share placement can cover approximately 8.4 months and five months, respectively.
• Alibaba Available LiquidityApproximately RMB150 billionThis represents haircut-adjusted available liquidity, with additional debt capacity costing approximately 2% to 3%.
• Compute Project Internal Rate of Return22%Based on a five-year economic life, a 60%/90%/100% utilization ramp and a 10% discount rate.
• Compute Project Payback Period2.9 yearsConsistent with management's stated payback period of less than three years.
• Compute Project Net Present ValueRMB36 of NPV for every RMB100 of capital expenditureApplicable to the infrastructure component; under a 60% infrastructure portfolio, aggregate NPV is approximately RMB22.
• EPS Impact of Alibaba Share Placement0.5% dilution to FY28EBased on 60% of proceeds being used for infrastructure; broadly neutral thereafter and accretive from the second year onward if infrastructure accounts for 80% or more.
• EPS Advantage of Debt FinancingApproximately 3 to 3.5 percentage pointsBecause bond coupons of approximately 2% to 3% are lower than the approximately 12% equity yield.
• Impact on Alibaba Intrinsic Value Per Share-0.9% to -1.2%Based on a pre-transaction fair value of HK$205, corresponding respectively to all proceeds being invested in infrastructure and the 60% base-case portfolio.
• Tencent Investment PortfolioRMB875 billionIncludes RMB487 billion of listed investments and RMB388 billion of unlisted investments.
• Tencent First-Half Operating Cash FlowRMB154 billionApproximately RMB310 billion annualized, versus the report's RMB200 billion capital expenditure forecast.
• Baidu June-Quarter Funding PositionCapital expenditure of RMB11.4 billion, operating cash flow of RMB3.4 billion and financing inflows of RMB17.4 billionIndicates that its capital expenditure is more dependent on borrowings and external financing.
• Stress-Scenario Project IRRApproximately 11% or 3%Approximately 11% at 80% utilization with a 10% price decline; approximately 3% at 70% utilization with a 20% price decline.
• Depreciation Corresponding to Base-Case Capital Expenditure as a Percentage of FY27E Net ProfitAlibaba 28%, Tencent 13%, Baidu 23%The report believes incremental EBITDA can exceed depreciation from the first full year onward.

Impact & implications

The report believes the more important implication of Alibaba's share placement is that the AI capital expenditure cycle may last longer, rather than that compute investment has lost its economic viability. All three companies may continue expanding AI investment, but financing costs take different forms: Alibaba bears direct equity dilution, Tencent forgoes part of its investment portfolio returns and incurs tax leakage, while Baidu may dilute ownership in AI assets such as Kunlunxin. Whether valuations recover will depend on whether proceeds are rapidly deployed into return-generating infrastructure, whether external cloud growth can keep pace with capital expenditure, and whether utilization and pricing can sustain project returns.

Risks

• If the mature-stage ROIC on Alibaba's deployed proceeds falls below approximately 15%, the report's base-case investment thesis will weaken.
• If infrastructure capable of generating auditable returns accounts for only approximately 40% of AI spending, returns on the overall funding portfolio may be insufficient.
• If Alibaba's three-year capital expenditure commitment materially exceeds RMB380 billion without external cloud growth guidance rising above 50%, it would constitute a negative signal.
• If the proceeds remain primarily in treasury balances for more than 12 months rather than flowing into fixed assets and prepayments, earnings dilution will persist.
• If losses at AI Labs and Applications fail to decline sequentially in the September quarter, it may indicate that infrastructure spending accounts for a lower-than-expected share.
• If supply expansion causes external cloud prices to decline, project IRR could fall to approximately 11% at 80% utilization and a 10% price decline.
• If Alibaba conducts another parent-company-level equity financing within 12 months, it would be a negative signal for expected project returns.
• Tencent's self-funding capacity will be tested if quarterly capital expenditure, dividends and buybacks together exceed operating cash flow without corresponding investment disposals.
• If Kunlunxin's listing is delayed beyond 2026 or the amount raised is materially insufficient, Baidu's ability to execute its planned capacity expansion will be constrained.

What to watch

• Monitor whether Alibaba's external cloud revenue growth exceeds 50% in the September quarter.
• Monitor whether losses at Alibaba's AI Labs and Applications decline sequentially in the September quarter.
• Monitor whether Alibaba's capital expenditure, fixed assets and prepayments indicate that the proceeds have been used as planned for advance procurement and infrastructure buildout.
• Monitor the share of proceeds allocated to infrastructure, mature-stage ROIC, utilization and compute pricing disclosed over the next two quarters.
• Monitor Tencent's Hunyuan 4 as a key evidence point for the effectiveness of its AI investment.
• Monitor Kunlunxin's listing progress and fundraising size, as this event will determine Baidu's financing capacity to expand AI production capacity.
• Monitor whether Alibaba needs additional debt, cloud-business-level financing or another parent-company share placement over the next 12 months.