Summary of Key Points
Recently, the stock price of Xunce Technology (HKEX: 03317.HK) has risen due to its first profit in the first half of the year (revenue increased by 389% to RMB 967 million, net profit of RMB 72.51 million). The underlying logic is as follows: General large-scale model companies (such as DeepSeek and Zhipu) have shifted from price wars to raising prices to pursue profits, while Xunce has already made money through its “scenario-based tokens” model. It processes private domain data into “high-value data blocks” that can solve specific industry problems, charging ten times more than general tokens. However, Xunce currently mainly generates revenue from its 1.0 Data Governance business (helping companies organize data and charging for services), with the 2.0 Token business (charged per usage) accounting for only 10% of its total revenue, and is still in the “from scratch” phase. Of the current market value of RMB 40.6 billion, most of it reflects expectations for the 2.0 business, which may be overestimated. It will be necessary to wait for three key indicators to confirm whether the investment is worthwhile before making a decision.
1. The Transformation of the Large-Scale Model Industry: From “Burn Money to Capture Market” to “Earn Real Money”
Previously, large-scale model companies acted like “price cutters,” competing for users with low prices (for example, general tokens were very cheap). Now, DeepSeek has announced a price increase for its API, and Zhipu’s revenue increased by 400% after raising prices by 83%—this indicates that the logic behind companies’ willingness to pay for AI services has changed: they no longer focus on settlement costs but on whether AI can solve real problems. It’s similar to how cheaper electricity and water usage leads to greater demand for specialized tools (scenario-based tokens) that can help businesses make money or avoid pitfalls. The shift towards profit-making by leading model companies marks the industry’s transition from a phase of “expanding territory” to one of “realizing value.”
2. Why Was Xunce Able to Profit First?
Xunce’s revenue model consists of two parts:
- 1.0 Business (Foundation): Helping Companies Organize Data and Charging for Services
Xunce started in the financial assets management sector, where there is a high demand for data timeliness and accuracy. It has since applied this capability to 11 industries, including telecommunications and energy. For example, it helps companies organize complex private domain data (customer information, production data, etc.) into standard formats and charges for projects or subscriptions. This business is already profitable: net profit in the second half of 2025 adjusted to RMB 50 million, and it increased to RMB 67 million in the first half of 2026. The customer retention rate is over 90%, and ARPU (average revenue per customer) has doubled from RMB 2.72 million to RMB 5.59 million, indicating that companies recognize its value. Even if the 2.0 business fails, Xunce will still have cash flow.
- 2.0 Business (Options): Selling “High-Value Data Blocks” and Charging per Usage
Since February this year, Xunce has turned the organized data into scenario-based tokens (e.g., risk warning data for the financial industry, equipment failure prediction data for the manufacturing industry), charging based on usage (RMB 10–100 per million tokens, ten times more than general tokens). In May, it launched the TokenOS system to standardize these token services. Currently, one strategic agreement is signed each week, but this only accounts for 10% of total revenue—it’s like drawing a pie; the 2.0 business has not yet become the main source of income.
3. The 2.0 Token Business: Great Potential, But Also Significant Risks
What are the potential advantages?
- High pricing: Scenario-based tokens are industry-specific solutions with scarce data, real-time functionality (in milliseconds), and the ability to solve practical problems (e.g., helping financial institutions avoid risks), so companies are willing to pay a higher price.
- Growing demand: The cheaper general tokens make AI more accessible, increasing the demand for high-quality data that can solve problems—similar to how cheaper electricity and water usage leads to greater demand for better household appliances.
What are the risks?
- Slow conversion: The time from signing a contract to receiving token revenue is variable depending on the industry, indicating that the standardized replication process is not yet mature.
- Customer loyalty needs verification: The token model has only been in operation for a few months; it’s unclear whether customers will continue using it over the long term.
- Low market share: Currently at 10%, and it is expected to rise to 20–30% by the end of 2026, with uncertainties along the way.
4. Valuation Analysis: How Much of the Market Value is Real, How Much is Imaginary?
Xunce’s current total market value is RMB 40.6 billion, which can be broken down into two parts:
- The “real” part from the 1.0 business: Up to RMB 11.2 billion
Based on a net profit of RMB 150 million for the entire year of 2026 and using a P/E ratio of 20–30 times for similar companies (such as China Software International), the 1.0 business is valued at RMB 2.7–4 billion. Even if compared to Palantir (a similar data platform in the U.S.) with a P/E ratio of 83.5 times, it would still be valued at RMB 11.2 billion.
- The “imaginary” part from the 2.0 business: RMB 29.3 billion
This reflects market expectations for the 2.0 business. Using a P/E ratio of 10–25 times for future revenue of data cloud companies, this portion implies that the 2.0 business could generate RMB 1.1–1.5 billion in revenue. However, by the end of 2026, the 2.0 business is expected to generate only RMB 600 million (based on annual revenue of RMB 2 billion and a 30% market share), indicating that the current stock price may already overestimate future growth.
5. Should You Invest? Wait for These Three Signals
Buying Xunce now is equivalent to investing in both the stable cash flow from its 1.0 business and the potential of its 2.0 business. However, whether the 2.0 business will be successful depends on three key indicators:
1. Mid-year report (end of August): Will token revenue be listed separately?
If the mid-year report does not provide detailed information on token revenue, gross margin, or customer numbers, it indicates that the business is not yet mature, which is a downside.
2. Order conversion: Can signed contracts turn into actual revenue?
Customers will only continue using the service if they believe the high price is worthwhile. If the agreements in the second half of the year are merely intentions and do not convert to actual revenue, expectations may decline.
3. TokenRouters platform: Can it facilitate cross-company transactions?
Xunce plans to launch a global token exchange platform in the second half of the year to enable the trading of tokens between different companies. If the platform is successful and there are real transactions (e.g., listing on a data exchange or third-party institution custody), the company’s valuation could shift from that of a product provider to a platform provider, representing a significant opportunity.
Conclusion: Xunce’s 1.0 business is stable and worth long-term tracking, but the 2.0 business has not yet been fully implemented. The current stock price largely reflects the potential of the 2.0 business. It is recommended to wait for these three indicators to be confirmed before making a decision, avoiding blind speculation.
Disclaimer: This article does not constitute investment advice; investments should be made with caution.