Data Finance: Valuation Assessment of the Potential Scale of Data Assetization

I. Overview of the development of the data assets market

1.1 Our countryData AssetizationStarting out on the road

At present, the construction of China's data factor market is gradually embarking on the fast track.As early as August 31, 2015, the State Council issued the Outline of Action for Promoting the Development of Big Data (Guo Fa [2015] No. 50, hereinafter referred to as the "Outline"), proposing to guide the cultivation of the big data trading market, and to carry out the pilot of the application-oriented data trading market. Since then, the Central Committee of the Communist Party of China, the State Council, and the NDRC have issued a number of documents to guide the development of the data factor market.2022 In December 2022, the Central Committee of the Communist Party of China and the State Council issued the "Opinions on Building a Data Base System to Better Play the Role of Data Factors" (hereinafter referred to as the "Twenty Articles on Data"), which determined the principles for the development of the data factor market and pointed out the direction of the development of the data factor market.2023 In March 2023, the CPC Central Committee and the State Council issued the Party and State Institutional Reform Program, which made it clear that a National Data Bureau would be formed to coordinate and promote the construction of the data foundation system, coordinate thedata resourceIntegrate, share and develop and utilize, and promote the planning and construction of digital China, digital economy and digital society in a coordinated manner.The establishment of the National Data Bureau means that the development of the data market will be transitioned from the previous principle guiding documents to the concrete implementation of the landing, China's data factor market on the fast track.

In terms of the process of moving data elements from resources to assets.The development of the data factor market can be divided into three stages, which are, in order, data resourcing, data assetization, and data capitalization (PwC, 2022; ICTA, 2023).data resourcingIt is the process of making disorganized and chaotic raw data into an organized and useful data resource.Data AssetizationIt is to take data resources as a starting point, process and develop them for established application scenarios and business purposes, and form data products that can be applied or traded by enterprises.Data capitalizationIt is the process of endowing data assets with more financial attributes on the basis of data resourcing and assetization, for example, they can be used as collateral for financial activities such as credit financing. At present, China is at the stage where data are moving from resources to assets, and the development of China's data factor market at this stage is characterized as follows:

One is that data can already bring significant economic benefits.According to the National Industrialinformation securityDevelopment Research Center and Peking University Guanghua School of Management in November 2022 published the "China Data Elements Market Development Report (2021-2022)" measured that data elements enable industrial enterprises to increase business growth by an average of 41.181 TP3T, increase productivity by an average of 42.81 TP3T, shorten the product development cycle by an average of 15.331 TP3T, and increase energy efficiency by an by an average of 10.19%.

Second, the data element trading market has taken shape.According to our report "National Data Trading Platforms and Commercial Banks' Business Opportunity Analysis"[1] released in April 2023, as of February 2023, according to incomplete information statistics, there are 40 data trading platforms across the country, involving more than 30 provinces, municipalities, or regions, and according to the goals and positioning of each data exchange, the initial formation of the international data trading hub, national data exchange, regional data trading centers, and industry-level data trading platforms is a hierarchical structure. According to the objectives and positioning of each data exchange, a hierarchical structure of international data trading hub, national-level data exchange, regional-level data trading center and industry-level data trading platform has been initially formed.

Thirdly, the data asset entry process is progressively under way.On December 9, 2022, the Ministry of Finance ("MOF") issued the Interim Provisions on Accounting Treatment for Enterprise Data Resources (Exposure Draft), exploring the adaptation of corresponding enterprise accounting standards to the accounting treatment of data resources in the table at this stage. At the same time, according to the way of realizing economic benefits in accounting, enterprise data assets are further subdivided into two categories: "data resources used internally by the enterprise" and "data resources traded externally by the enterprise". In recent years, a joint-stock bank [2], Guiyang Big Data Exchange [3] and others are also trying to promote the data assets into the table through data valuation.

According to our previous study [4], the asset properties of data as an important class of production factors have been recognized by various parties.With reference to the definition in the Accounting Standards for Business Enterprises, an asset is a resource that is formed by past transactions or events of an enterprise, owned or controlled by the enterprise, and expected to bring economic benefits to the enterprise. At present, there is no official and unified definition of data assets within China. From the perspective of overseas experience, the National Institute of Standards and Technology (NIST) defines data assets as any entity that is comprised of data, including systems, applications, databases, documents, web pages, and application-based data services [5].

1.2 Challenges to data assetization

Data assetization is typically characterized by the fact that data can not only generate value within the data-generating subject, but can also be used multiple times through a wider range of subjects after the data has been circulated, thus giving full play to its asset value.Therefore, the orderly circulation of data is an important pivot to enhance the energy level of data assetization. And from the perspective of data circulation and use, data assetization currently faces many challenges:

First, the top-level design of the data system is not yet perfect, and the existence of unclear data property rights and an imperfect classification and grading system for data is the biggest challenge facing the road to data assetization.On December 2, 2022, the Central Committee of the Communist Party of China (CPC) and the State Council issued Twenty Articles on Data, stating that data-related legal rights include the right to hold data resources, the right to use data processing, and the right to operate data products. Although the Twenty Articles on Data has not yet given detailed rules for the determination of data property rights and the division of ownership, it encourages prioritizing the promotion of data assets to explore deep and extensive use in economic production.

Secondly, data flow rights, data application andData SecurityTechnology still needs to be further upgraded so that the construction and implementation of data-related systems can be guaranteed by means of mature and abundant technology.The lack of data confirmation technology hinders the protection of the relevant interests of all kinds of rights holders in the flow of data, the lack of data application technology is not conducive to the full release of the value of data, and the lack of data security technology increases the risks and obstacles to the flow of data and reduces the willingness of data subjects to provide data.

Thirdly, and the issue that is the focus of this paper, is how the value of data is determined.The biggest difference between an asset and a resource is that an asset can bring economic benefits to an organization. And how to measure the economic benefits brought by data, i.e.Valuation of data assetsand how to reach a consensual transaction price (i.e., facilitate both parties to agree on the valuation of the data) during the transaction flow is a key topic that needs to be addressed on how the data can be further empowered externally.It should be noted that fair value is the price that a market participant would receive to sell an asset or pay to transfer a liability in an orderly transaction occurring at the measurement date, and that the valuation of data assets below refers to assessing their fair value.

The data-related system needs to be clarified by the national regulatory authorities through legislation, the issuance of normative documents, etc.; the development of data-related technology also needs time; and the difficulty in determining the value of data assets leads to the difficulty in forming the price of data, which will directly result in the blockage of the circulation of data transactions.

There are many difficulties in the current valuation of data assets:

On the one hand, data value is difficult to measure directly. First, it is not possible to directly determine the value of data in terms of the data itself.Neither the dimensionality nor the volume size of the data can directly determine the contribution of the data in data analysis, and the number of data entries has some direct role in measuring customer-level data, but it cannot determine the quality of the data. For example, if an asymmetric data set has an uneven distribution of data, this type of data has less value in data analysis. In addition, issues such as sparse data can lead to large volumes and numbers of data, but not much value in practical applications.Second, from the perspective of data use, the statistical framework lacks a methodology for quantitative comparison of data.In the classical statistical framework, the model evaluation system is more complete, and there are numerous model evaluation indicators. In contrast, due to the lack of data quality evaluation indicators, some of the existing indicators can not play a role in measuring the quality of data, such as Variance Inflation Factor (VIF) is only for the specific data problem of multicollinearity; Fisher's information matrix contains the amount of information that the data can be used for parameter estimation, but it is only for the specific parameter estimation and cannot measure the value of the whole data; deep learning and reinforcement learning frameworks use methods such as data augmentation [6] to increase the amount of data and thus improve the model performance, and are even more lacking in data evaluation metrics.Third, there is a lack of a reference system for mapping data asset values to currencies.Even if we have a metric for data value, how it maps to being measured in monetary terms still lacks the backing of a reference system, i.e., there is a lack of connectivity between the data value measure and the monetary measure.

On the other hand, it is difficult to form a consensus among parties when indirectly valuing data assets through various valuation techniques. First, in terms of data supply and demand, theAt present, most of the domestic data transactions are based on bilateral "peer-to-peer" aggregation, and for the data provider, based on the starting point of its own data assets in the table, there may be a tendency to overestimate the value of data assets. For data users, the value of data is reflected in the enhancement of their business capabilities, and the enhancement effect of data on business before data procurement is highly uncertain, so they have a tendency to underestimate the value of data assets.Second, in terms of the industries in which the data are usedDue to the differences in the needs and usage methods of different industries, there will be differences in the added value recognized for the same data. For example, the consumption data of individual customers has high application value in the Internet and financial industries, while it has limited application value in the traditional manufacturing industry.Third, in terms of the industries in which data assets are generated, theThe cost of data asset generation is different. As represented by the e-commerce industry, a large amount of user data can be obtained through business operations at a low cost, while geological exploration and governmental data require a large cost in the process of collection and organization. In addition, classical valuation methods, such as the cost method, face certain limitations when valuing data assets.

1.3 Principles of accounting for data assets

2022 On December 9, 2012, the Ministry of Finance ("MOF") issued the Interim Provisions on Accounting Treatment Related to Enterprise Data Resources (Exposure Draft) ("Interim Provisions").Its scope of application includes the accounting treatment related to data resources that are recognized by an enterprise as an asset category such as intangible assets or inventories in accordance with the relevant provisions of enterprise accounting standards, as well as the accounting treatment related to data resources that are legally owned or controlled by an enterprise and are expected to bring economic benefits to the enterprise but are not recognized as an asset due to the fact that they do not satisfy the conditions for recognition of assets under the relevant provisions of enterprise accounting standards.The Interim Provisions clarify that at this stage the accounting treatment of data resources should be implemented in accordance with enterprise accounting standards and further subdivided into two categories, namely "data resources used internally by the enterprise" and "data resources traded externally by the enterprise", in accordance with the way of realizing the economic benefits for accounting purposes. (b) Data resources used internally by an enterprise and data resources traded externally by an enterprise.

Data resources used within the enterpriseIntangible assets need to be recognized as intangible assets in accordance with Accounting Standard for Business Enterprises No. 6 - Intangible Assets (Caixin [2006] No. 3, hereinafter referred to as "Intangible Assets Standard"). According to the method of acquisition, they can be classified as intangible assets developed by the enterprise itself, intangible assets acquired through outsourcing, and intangible assets acquired in other ways, including non-monetary asset exchanges, debt restructuring, government grants, and intangible assets acquired through business combinations. Intangible assets are defined in the Intangible Assets Standard [7] as identifiable non-monetary assets owned or controlled by an enterprise that have no physical form. Recognition of intangible assets first needs to meet the "identifiability criteria"[8], and determine "it is probable that the economic benefits associated with the intangible asset will flow to the enterprise" and "the cost of the intangible asset can be measured reliably The cost of the intangible asset can be measured reliably."

Data resources for corporate external transactionsInventories are recognized as inventories in accordance with Accounting Standard for Business Enterprises No. 1 - Inventories (Caijing [2006] No. 3, hereinafter referred to as "Inventory Standard"), and can be classified as self-processed inventories, purchased inventories and inventories acquired by other means according to the method of acquisition. Inventories are recognized as inventories according to the method of acquisition. Inventories are defined in the Inventory Standards[9] as finished goods or merchandise held for sale in the ordinary course of an enterprise's activities, work in progress in the production process, and materials and supplies consumed in the production process or in the provision of services. Recognition of inventories requires a determination that "it is probable that the economic benefits associated with the inventory will flow to the enterprise" and "the cost of the inventory can be measured reliably".

It should be noted that the Interim Provisions specify the accounting standards applicable to data assets, but the accounting treatment of data assets is only one of the many application scenarios for the valuation of data assets. In addition to the accounting treatment necessary in the process of entering assets into the table, assessing the fair value of data assets will be of greater significance for the effective and smooth flow and use of data, and is a necessary condition for the establishment of an open and active trading market. Therefore, the following section focuses on the application of various valuation methods for data assets in the process of recognizing the value of data assets.

II. Analysis of data asset valuation methods

Accounting Standard for Business Enterprises No. 39, Fair Value Measurement (2014) ("AS 39") states that the valuation techniques used by an enterprise to measure the underlying assets or liabilities at fair value primarily include the market approach, the income approach and the cost approach. Enterprises should measure fair value using an approach that is consistent with one or more of these valuation techniques [10].The China Asset Appraisal Association (2020) states that the valuation methods of data asset value include three basic methods, namely the cost method, the income method and the market method, and their derivatives.We will analyze in detail the advantages and disadvantages as well as the practical applications of the three basic methods for data asset valuation, and propose the development direction of the technical approach to data asset valuation based on the basic methods.