Sinolink Securities released a research report stating that there is no significant bubble in current AI development. Supply chain financing for major AI industry leaders remains in its early stages and primarily involves equity investments, reducing risks compared to the debt-driven supply chain financing seen during the dot-com bubble.
The B2B AI business model is gradually forming a closed loop, with cloud providers accelerating growth due to AI adoption. Industrial leaders such as TSMC, UPS, and Walmart have already begun using AI to enhance production efficiency. High capital expenditures (CAPEX) by cloud service providers (CSPs) are sustainable, and the AI business model is maturing. Strong financing enthusiasm for AI indicates that current stock price growth is primarily driven by earnings per share (EPS) expansion. Sinolink Securities maintains a positive outlook on core AI hardware stocks.
Key points from the report include: 1. **Sustainable High CAPEX by Major CSPs**: The four leading North American CSPs—Google, Meta, Microsoft, and Amazon—are expected to sustain and even increase capital expenditures. Their combined CAPEX in Q3 2025 reached $97.3 billion, up 65% year-over-year, with further increases anticipated in 2026. 2. **Revenue and Profit Growth**: Combined Q3 2025 revenue for the four CSPs was $396.6 billion (+17% YoY), with net profits at $86.6 billion (+6% YoY). Excluding Meta’s one-time non-cash loss of $15.9 billion due to tax policy changes, net profits would have been $91.9 billion (+27% YoY). Gross margins for Microsoft, Google, and Meta remained stable, while Amazon continued to improve. 3. **Healthy Financial Metrics**: The CSPs’ CAPEX/EBITDA ratio stood at 62% in Q3 2025, below historical peaks. Amazon’s ratio has previously exceeded 100% in multiple quarters. Debt ratios for most CSPs (excluding Google) have declined since 2023, indicating no significant financial deterioration. 4. **AI-Driven Efficiency Gains**: Industrial firms are adopting AI for equipment maintenance, improving employee productivity by 5–20% and reducing equipment downtime by 15%. As more enterprises follow suit, AI adoption is expected to boost cloud service demand, reinforcing the B2B business model. 5. **Rapid Growth in AI Model Firms**: OpenAI’s revenue surged from $2.8 million in 2022 to an estimated $13 billion in 2025, with projections of 230%, 200%, and 160% YoY growth in 2026–2028. Anthropic aims for $70 billion revenue and $17 billion cash flow by 2028. 6. **Strong Investor Sentiment**: AI firms continue to attract significant funding, with Anthropic’s valuation rising from $61.5 billion in February 2025 to $183 billion by September 2025. 7. **EPS-Driven Stock Performance**: Unlike the dot-com bubble, current AI stock gains are supported by EPS growth. Nvidia’s stock price and EPS have risen in tandem since Q1 2023, with forward P/E ratios for major AI stocks and CSPs below 2024 highs.
**Risks**: AI model stagnation, tightening U.S. financing conditions, macroeconomic pressures, slower-than-expected AI hardware development, and deteriorating U.S.-China tech policies.
Comments