Data Assetization: Institutional Deepening and Market Realities Behind a Frenzy

Data, as a critical element in the digital economy, is experiencing unprecedented attention regarding its place on corporate balance sheets. The core question is: Are the methods of data asset recognition, valuation, and liquidity undergoing definitional changes? We distinguish between 'deepening' and 'revolution': the former refers to optimization and adjustments within existing asset definition frameworks, such as adding new accounting categories; the latter implies fundamental changes to core definitions like 'what constitutes an asset, who owns it, how it is valued, and how it circulates.' For precise judgment, we establish four criteria: Q1 Does the asset definition accommodate non-exclusive, reproducible, and combinable data? Q2 Is the valuation basis shifting towards verifiable future earnings? Q3 Is control/exclusivity detached from legal ownership? Q4 Does a separable data market exist, independent of entity credit? If three or more of these criteria undergo definitional changes, it can be termed a 'revolution'.

Under traditional accounting paradigms, data is typically treated as an operational byproduct, with its costs directly expensed and not booked on the balance sheet. International Accounting Standard (IAS 38) explicitly defines an asset as a resource controlled by an entity as a result of past events and from which future economic benefits are expected to flow to the entity, with control typically stemming from legal rights. Internally generated intangible assets, such as self-generated goodwill and customer relationships, are generally expensed, and can only be capitalized during the development phase if strict conditions are met. The cornerstone of this model lies in the assumptions that 'control depends on legal ownership' and 'data can be infinitely replicated, is difficult to ascertain rights for, and thus doesn't flow.' These old paradigmatic assumptions form the historical starting point and intellectual shackles for data assetization, and they are precisely the bottlenecks that current new mechanisms seek to break through.

China's data assetization process exhibits a notable policy-driven characteristic at the institutional level. The 'Opinions on Building Basic Systems for Data and Better发挥 the Role of Data Elements' (i.e., the 'Data Twenty Articles'), issued in December 2022, creatively proposed a 'three-tiered rights system' framework: 'data resource holding rights, data processing and usage rights, and data product operating rights.' This aims to circumvent legal disputes over traditional 'data ownership' and provide an institutional basis for the market-oriented flow of data elements.【已查证·人民网】Subsequently, the National Data Administration was established in October 2023, and the Ministry of Finance also issued the 'Interim Provisions on Accounting Treatment for Enterprise Data Resources' in August of the same year, which officially took effect on January 1, 2024. These measures have constructed the institutional backbone of data assetization from a top-down design perspective.

However, crucially, the 'Interim Provisions' explicitly state that they 'apply to current enterprise accounting standards and do not alter the accounting recognition and measurement requirements of existing standards,' merely achieving presentation by adding a 'data resources' item under 'inventory/intangible assets/development expenditures.' This indicates that changes at the accounting level are more supplementary adjustments within the existing framework rather than a fundamental breakthrough in asset definition.【已查证·财政部/安永解读】The China Appraisal Society also issued the 'Guidance Opinions on Data Asset Valuation' (中评协〔2023〕17号), effective October 1, 2023, attempting to provide guidance for data valuation. In practice, data asset valuation predominantly relies on the cost method, with the income and market methods having limited applicability due to a lack of comparable transaction data.【已查证·证券时报引业内访谈】

Despite strong policy promotion, the actual scale and pervasiveness of data being booked on balance sheets are still in their early stages. According to data from the National Data Administration, the national data market transaction volume is expected to exceed 160 billion CNY in 2024, representing a year-on-year increase of over 30%.【已查证·国家数据局】However, the total amount of data resources disclosed by A-share listed companies on their balance sheets in 2024 was only approximately 2.2 billion CNY. Among these, 38 central and state-owned enterprises accounted for 1.471 billion CNY, while the three major telecommunication operators (China Mobile, China Unicom, China Telecom) alone booked 1.6 billion CNY, making up 60.7% of the total 2.64 billion CNY booked in their 2025 semi-annual reports. For non-listed entities, local state-owned enterprises accounted for 82.67% of the 375 disclosing entities.【已查证·证券时报；高金报告/中证网】These figures indicate that current data assetization practices are primarily concentrated among large state-owned enterprises and leading companies with vast data resource advantages, suggesting significant room for growth in overall market scale and penetration.

During the advancement of data assetization, multiple 'braking' mechanisms have limited its progression towards a paradigm shift. Firstly, accounting changes are not revolutionary; the positioning of the 'Interim Provisions' determines that data booking is essentially a 'deepening that applies old asset definitions to new objects.' For instance, data resources under the inventory account decreased from approximately 200 million CNY in Q3 2024 to 0 for the full year, reflecting the extremely low market liquidity of 'data as saleable inventory.'【已查证·证券时报】Secondly, the 'labeling data and effectively conducting debt financing' in the liquidity layer previously led to excessive speculation. For example, in 2025, 11 data asset ABS issuances totaled 4.59 billion CNY, with the first four months of 2026 already exceeding the full-year total. However, regulatory window guidance subsequently tightened, aiming to rectify the chaotic situation where some underlying assets were still traditional accounts receivable or credit, backed only by the entity's credit or local government guarantees.【已查证·上海证券报/新华财经】Data from the China Academy of Information and Communications Technology (CAICT) showing that exchange-traded transactions account for less than 5% also corroborates the weakness of the data trading market.【单一来源·待核对】

Furthermore, the inverse cost-benefit ratio of data assetization is also a prevalent issue. The cost of booking a single data resource item can be as high as 200,000 to 500,000 CNY. Coupled with difficulties in defining ownership and inconsistent valuation methods, corporate enthusiasm is suppressed.【已查证·证券时报引业内人士】Legally, the 'three-tiered rights system,' as a policy-driven arrangement of rights, has not replaced the traditional concept of ownership; issues such as the legality of data sources, personal information protection, and cross-regional recognition of ownership rights remain significant hard constraints.【已查证·业内访谈；推断】These challenges collectively constitute a formidable barrier for data assetization to transition from 'vision' to 'reality'.

For businesses, data assetization signifies a shift from data as an 'IT cost center' to a 'financeable operating asset,' but this applies only to data capable of generating independent, attributable, and verifiable cash flows. The dividing line is not the sheer volume of data, but whether data can be monetized in the market, who pays for it, and whether its cash flows can withstand rigorous scrutiny. Internationally, the EU Data Act (Regulation (EU) 2023/2854), effective September 12, 2025, mandates data holders to provide 'access by design' to users and third parties, aiming to promote data flow by reducing exclusivity. This differs significantly in practical approach from China's 'three-tiered rights system' and offers another perspective for examining data value realization.【已查证·Clifford Chance/Withers】

At the organizational level, data assetization drives enterprises to establish new functions such as rights confirmation, registration, valuation, auditing, operations, and monetization, thereby promoting the iteration of organizational structures and talent capability models. Notably, local state-owned enterprises account for as high as 82.67% of non-listed entities booking data assets, indicating that current data assetization is primarily the revitalization and reallocation of resources within the state-owned system rather than a widespread market behavior.【已查证＋推断】For individuals, personal data itself is unlikely to become a core item on their balance sheets. However, the emphasis on 'legality of data sources' and the explicit 'authorization-revenue' relationship will grant individuals greater data control and bargaining power, especially in data interactions with platforms, where compliance becomes a prerequisite for the realization of personal data value.

As a one-person company (OPC) and a research institution collaborating with multi-tool AI, the vast majority of our data assets (client lists, corpora, prompt libraries, evaluation datasets, workflow logs, etc.) are internally generated. According to IAS 38 and the 'Interim Provisions,' such expenditures are generally expensed, and the conditions for capitalization are extremely stringent. This warns us not to view 'booking data assets' as a financing narrative or a shortcut to massive profits for an OPC. Given that OPCs lack corporate credit and guarantees, and the current data ABS market is tightening, the path of financing through 'booking' is neither realistic nor free from potential regulatory risks. A pragmatic approach is to eschew accounting complexities and focus on the commercial value and productization capabilities of data.

Therefore, an OPC should transform data into 'deliverable data products' rather than merely 'book assets.' A reusable lesson from the first data asset ABS cases is to establish an ownership chain comprising 'registration certificates + legal opinions + source documents.' An OPC should build its own data ledger and ownership chain, which serves as a low-cost, auditable moat. Furthermore, 'independent cash flow' should be the sole rigorous standard for filtering data value. Only data that can create value for customers and generate verifiable payments (such as industry insight reports, anonymized behavioral data services, evaluation benchmarks) can be truly termed an asset; all others are merely operational byproducts. Treating 'verification and accountability' as preconditions for data assetization ensures that data can undergo thorough scrutiny and be held accountable, aligning perfectly with our emphasis on 'verifiability and attributability for results-based payment.'

The road to data assetization remains long and fraught with variables. The following key observation points will help us continuously assess its progress: Firstly, pay attention to the scale of A-share data resources booked in the 2025 annual reports and 2026 semi-annual reports, as well as the performance of inventory accounts, and whether the proportion of the three major operators in the total booked amount declines, to ascertain concentration trends. Secondly, closely monitor the full-year issuance volume of ABS after the 2026 window guidance and be vigilant for potential **regulatory penalty cases involving data asset recognition or valuation fraud**. Thirdly, the progress of the national unified data property rights registration system, data asset registration platforms, and cross-regional mutual recognition rules will be critical infrastructure for building a data element market. Fourthly, whether data asset valuation can shift from being cost-method-dominated to incorporating income or market approaches, and whether unified valuation and impairment standards are introduced, will directly impact the authenticity of its asset attributes.

Furthermore, attention should be paid to the alignment between 3-5 year straight-line amortization and the actual data lifecycle, as well as impairment provisions, to identify potential 'under-amortization' issues. Lastly, the actual compliance costs and data sharing traffic during the first year of the EU Data Act's implementation will serve as a control experiment for the 'mandated flow' pathway, offering valuable experience and lessons for global data governance. These observation points will collectively outline the evolutionary trajectory of data assetization from institutional exploration to market maturity.