Summary of Key Points
Tencent reported impressive profits for the second quarter of 2026 (56 billion yuan in net profit attributable to the parent company, and 68.4 billion yuan using non-international accounting standards), but its free cash flow was negative by 13.8 billion yuan. The main reason for this is not that AI has “devoured” all the profits; rather, it prepaid approximately 51.4 billion yuan for purchasing computing power—this money has already been spent on future AI infrastructure projects such as the Hunyuan large model and WeChat AI. AI currently plays a dual role: it helps drive growth in advertising and cloud services while also incuring costs. While Tencent’s cash reserves are sufficient to cover these current investments, the company will need to monitor whether the prepaid payments can be converted into actual revenue and balance its cash expenditures on AI, share repurchases, and dividends.
1. Negative Free Cash Flow: Money Didn’t Disappear—It Was Used to Purchase Computing Power
Many people may be confused by the figures of “56 billion yuan in profit and -13.8 billion yuan in cash flow.” Profit and cash flow are two different things:
- Profit is the money earned according to accounting rules: For example, if you sell an ad for 100 yuan and the cost is 30 yuan, your profit is 70 yuan (regardless of whether the money has been received).
- Cash flow is the actual money coming in and going out: For instance, if you prepay a supplier 1000 yuan for servers, the money is paid out before the equipment is delivered, so it is not yet deducted from your profits.
Tencent’s negative cash flow in the second quarter was mainly due to:
- A significant increase in capital expenditures (59.3 billion yuan, up 176% year-over-year), mostly for purchasing AI servers.
- An additional prepaid payment of 51.4 billion yuan for computing power (the company stated that excluding this amount, the cash flow would be positive at 37.6 billion yuan).
The 51.4 billion yuan was not a waste; it was an investment in securing future computing resources (such as chips and data centers), but the money was indeed spent, though it has not yet been accounted for as an expense in the current period.
2. The Dual Nature of AI: Boosting Business Growth While Incurring Costs
AI is not only costly; it also begins to generate revenue for Tencent:
- Revenue-generating aspects: Advertising revenue increased by 22% (in part due to AI-driven precision advertising), and cloud business grew by 9% (driven by AI demand).
- Costly aspects: New AI products (such as Hunyuan and CodeBuddy) reduced operating profits by about 10.5 billion yuan (the company’s reported profit would be higher if these products were excluded). Additionally, the depreciation and operational costs of AI servers lowered the gross margin of some businesses.
In short, while AI has helped traditional businesses generate more revenue, the new AI products are not yet profitable and continue to draw on profits and cash resources.
3. Cash Reserves Are Still There, but Future Decisions Will Become More Challenging
Tencent does not currently lack funds, but its cash pressure is increasing:
- Current Financial Position: Total cash is 511.2 billion yuan, with total debt at 453 billion yuan, resulting in net cash of 58.2 billion yuan (887 billion yuan less than at the end of March).
- Expenditures: 56.3 billion yuan was spent on share repurchases and dividends in the second quarter (roughly equivalent to the prepaid amount for computing power), and further investments in AI are planned.
If Tencent continues to make large prepaid payments for computing power, it will need to balance three options:
- Reduce capital expenditures?
- Cut back on share repurchases or dividends?
- Borrow more money?
AI is no longer just a matter for the product departments; it has become a financial consideration that the finance department must manage.
4. The Future of Prepaid Payments: How Much Can Be Recouped?
The 51.4 billion yuan in prepaid payments will not be lost forever; it will gradually be converted:
- Conversion Paths: It may be used to purchase servers (with the cost being deducted over time through depreciation) or converted into revenue from AI services.
- Key Factors: Whether the equipment is delivered on time, whether the servers can be fully utilized (e.g., providing cloud services to customers), and whether new AI products can generate revenue.
If the equipment remains unused or customers do not purchase the services, the prepaid payments will be a costly investment with no immediate return. If they can be converted into revenue, they will represent an early investment that pays off in the future.
5. The Essence of AI Competition: Technology and Cash Resilience
The competition in AI is not just about which model is better; it’s also about who can sustain their investments until they start generating returns:
- Tencent’s Advantages: Its existing businesses (WeChat, advertising, gaming) continue to generate revenue, allowing it to make prepaid payments for computing power.
- Challenges for Smaller Companies: Without stable cash flow, early payments may lead to elimination from the market.
In the end, it will not be the least effective AI models that are eliminated in the industry; rather, those companies that cannot wait for their investments to pay off will be phased out. Tencent has the capability to wait, but the market will not wait indefinitely for results.
Conclusion
Tencent’s negative cash flow in the second quarter is not an alarm; it reflects its strategic investment in AI. However, the company needs to focus on several key aspects:
- Whether the prepaid payments are converted into usable computing power.
- Whether the revenue generated by AI can cover the associated costs.
- Whether free cash flow can return to a positive range.
These factors will determine whether this investment is a worthwhile expense or a waste of resources. The market will not reward Tencent solely based on its involvement in AI; real financial returns are what matter most.