Analysis for informational purposes only. Capital at risk.
The AI Funding Test: Alibaba and Tencent are both ramping AI investment, but their ability to fund the transition differs materially. Tencent’s capex‑to‑core‑profit ratio stays comfortably below 50%. Alibaba’s ratio has spiked above 100%, temporarily driving free cash flow negative in mid‑2025.
Contrasting Core Engines: The divergence traces back to their core businesses. Tencent’s cash flow is underpinned by a resilient, countercyclical online gaming franchise that grew 21% in 4Q25. Alibaba is still rebuilding e‑commerce margins after recent price wars, leaving it more exposed to capex stress.
Agentic AI and Capital Trade‑offs: Both firms are embedding agentic AI (eg. OpenClaw‑style agents) into WeChat and Taobao to accelerate monetisation and improve AI ROI. To fund the AI cycle, Tencent is rebalancing capital away from share repurchases toward capex, and Alibaba is facing a similar capital‑allocation dilemma.
The Central Question on AI Investment
Investors often view Tencent and Alibaba as two sides of the same coin—both dominant and pivoting to AI, and with the same challenges: rising capex and margin pressure.
Yet their latest financial results reveal a different situation.
They are both ramping up AI investment.
The key question is whether they can fund the cycle from operating cash flow or will need to draw on reserves and raise leverage.
Their ability to fund the AI cycle from operations differs materially.
The gap sits in their core businesses, gaming versus e-commerce, which determines how they are paying for this transition.
Distinct AI Strategies
Tencent and Alibaba adopt two distinct AI strategies.
Tencent was historically prudent on AI, prioritising using AI to enhance existing products. That delivered margin improvement but raised questions about long-term competitiveness. It is now shifting to a more proactive strategy, focusing on foundation models, the Yuanbao AI chatbot, Weixin AI, and productivity agents.
Alibaba has been more aggressive. It is open-sourcing Qwen to drive developer adoption and cloud usage and embedding agentic experiences across its payment and commerce ecosystem.
That strategic difference flows directly into their capex profiles and cash flow burden.
The Capex Test
Tencent’s 2025 total capex was RMB 79bn, of which RMB 18bn was AI. AI capex could rise toward RMB 40bn in 2026, pushing total capex above RMB 100bn (capex/revenue ~13–14%).
Alibaba’s 2025 capex was ~RMB 120bn, and its 2026 capex could exceed RMB 150bn (capex/revenue ~13–14%).
The capex/revenue ratios are converging. The cash flow behind them is not.
The Cash Flow Burden
Alibaba’s free cash flow has been consistently lower than Tencent’s and turned negative in 2Q25–3Q25, suggesting difficulty covering rising capex from core operations.
Furthermore, Alibaba’s capex-to-core profit ratio has exceeded 100% in recent quarters. Tencent’s has stayed below 50%.
Tencent, on the other hand, benefits from materially higher profit margins, converting revenue to operating cash more effectively.
Balance‑sheet Cushion
On a positive note, both companies retain sizable balance sheets to support AI investment. Their net cash and equity investments cover roughly 3–4x projected annual capex (or 1–2x if counting only immediately deployable cash).
Core Engines: Gaming vs. E-Commerce
These distinct cash flow realities stem from their core businesses.
Online gaming—resilient
Online gaming remains a bright spot and a defensive segment in the China internet market.
Tencent’s online gaming reported a 21% revenue growth in 4Q25. Domestic revenue experienced 15% YoY growth, significantly outperforming NetEase’s 3.4% gaming revenue growth, implying ongoing market share gains. International gaming revenue reported strong 32% growth and accounted for 36% of total gaming revenue.
Why gaming cash flows are well protected
- Market concentration: Tencent and NetEase dominate; the rest of the market is fragmented among smaller developers.
- Entry barriers: Regulatory approvals and long development cycles limit new competition.
- Social network effects: WeChat’s social graph keeps players engaged because their social network is embedded in the platform.
- Countercyclical demand: Gaming often substitutes for costlier leisure activities during downturns, supporting spend even in weak consumer environments.
E-commerce—margin still under pressure
Alibaba’s e-commerce grew 6.0% in 4Q25 amid soft consumer sentiment. Quick commerce was the standout—56.0% revenue growth, now 13% of total e-commerce revenue.
Government pressure is easing the competition, and margins are recovering sequentially. Taobao/Tmall’s operating margin recovered to 22% in Q4 2025, up from 8% the prior quarter but still well below the 40%+ level before the price war.
The AI Agentic Upside
AI cloud is the strategic focus for both companies, and the near-term monetisation argument is the same: agentic AI applications.
Alibaba Cloud grew 36.0% in 4Q25, with AI-related products delivering triple-digit growth for the tenth consecutive quarter. The company targets to achieve USD100bn Cloud and AI revenue by 2030, roughly 5x of current levels.
Tencent was a deliberate late mover — it deployed AI internally first before selling externally — and is now accelerating. External cloud revenue should rise from 2026.
Agentic AI, such as OpenClaw, could be a near-term “killer app”.
Compared with costly model training, integrating agentic applications into existing ecosystems (WeChat, Taobao) should offer faster monetisation via token consumption and platform-level services.
Agentic sessions consume tokens at 10–50x the rate of basic chat, turning users into recurring cloud revenue. Tencent has 1.4 billion WeChat users as distribution.
Alibaba has Qwen connected to Alipay, Taobao, and Ele.me — a complete commerce loop. A user can ask Qwen to find, pay for, and arrange delivery without leaving the app. Meituan has no equivalent closed loop and could face disintermediation risk.
Capital Allocation Trade‑offs
Chinese internet companies increased buybacks over the past two years, driven by depressed valuations and limited post-COVID investment opportunities.
AI is changing that priority.
Tencent has confirmed it will scale back repurchases to fund the capex cycle while maintaining its dividend. Other incumbents generally face a similar dilemma and might follow suit.
Cutting buybacks shifts the shareholder base. Yield-focused investors might exit due to lower shareholders’ return. Growth-oriented investors might enter at different valuation levels. That transition creates near-term volatility separate from the fundamentals.
This article is a “periodical publication” for information only and is not investment advice or a solicitation to buy or sell securities. This article does not constitute a “personal recommendation” or “investment advice” under UK FCA regulations. Investing in equities involves significant risk. The author holds NO position in the securities mentioned. There is no warranty as to completeness or correctness. Please do your own due diligence or consult a licensed financial adviser. Please read the Full Disclaimer before acting on any information. Images created with the assistance of Gemini AI.