Valuing Marine Data Assets: A Composite Multi-Period Valuation Framework Under the Blue Economy

Abstract

Marine data assets are increasingly recognized as important drivers of value creation in the blue economy, yet their valuation remains challenging due to difficulties in isolating data-related earnings in capital-intensive maritime enterprises. This study proposes a methodological valuation framework that integrates the multi-period excess earnings method with the Analytic Hierarchy Process (AHP) and the Fuzzy Comprehensive Evaluation (FCE) approach, incorporating both financial and non-financial dimensions. The framework follows a “total synergistic return–data contribution separation” logic to isolate data-related excess earnings and applies an AHP–FCE-based adjustment coefficient to account for data quality, application value, and risk. A representative container shipping enterprise is used as an illustrative application to demonstrate the implementation logic of the framework. The results indicate that marine data assets can constitute a non-negligible component of enterprise value under reasonable parameter settings, while sensitivity analysis highlights the influence of key parameters such as the data contribution coefficient and discount rate. The proposed framework provides a transparent methodological reference for marine data asset valuation and supports sustainability-oriented research and practice in the blue economy.

1. Introduction

1.1. Question Raised

In the era of big data, data is hailed as the “oil” of the 21st century. In April 2020, the “Opinions of the Central Committee of the Communist Party of China and the State Council on Building a More Perfect Market-Based Allocation Mechanism for Production Factors” explicitly recognized data as a production factor for the first time and emphasized its important role in the composition of productive forces. In December 2022, the Central Committee of the Communist Party of China and the State Council issued the “Opinions on Establishing a Data Governance System to Better Leverage the Role of Data as a Production Factor,” outlining strategies to fully unlock the potential of data as a production factor and accelerate corporate data accumulation. As countries increasingly focus on the blue economy, the marine economy is viewed as a key driver of future economic growth, with global competition in the blue economy intensifying. In 2023, China’s marine industry output reached 10.5 trillion yuan, accounting for approximately 9.5% of the national GDP. Within the marine industry, the potential of data is increasingly evident. Against this backdrop, marine data has become a critical strategic resource for enhancing industrial competitiveness. Additionally, driven by technologies such as big data and artificial intelligence, the potential of data is being unleashed to an unprecedented extent. The “14th Five-Year Plan for Marine Economic Development” explicitly states the need to deepen the application of big data in the marine sector and promote the flow of data elements to marine-related enterprises, clearly indicating that in the rapid rise of the marine economy, data assets will present enterprises with unprecedented opportunities. Despite these developments, a critical methodological question remains unresolved: how can the economic value of marine data assets be systematically identified and measured within capital-intensive maritime enterprises?

However, despite the immense potential of marine data assets, current challenges such as inadequate integration of data resources and incomplete data infrastructure hinder the full realization of marine data’s value, limiting its application scope and benefits to levels below their potential. Given this, developing a data asset valuation model tailored to the characteristics of the marine industry holds significant theoretical and practical significance for promoting the comprehensive development and application of marine data. Marine data assets play a critical role in promoting the sustainable development of the blue economy. These assets support decision-making in marine resource management, ecosystem protection, and climate change mitigation. By leveraging advanced data technologies, marine data enables industries to optimize operations while adhering to sustainable practices. As such, valuing marine data assets is not only essential for economic growth but also for fostering a resilient and environmentally sustainable blue economy. This study addresses the challenge of valuing marine data assets by proposing a composite valuation framework that integrates the Multi-period Excess Earnings Method with the Fuzzy Analytic Hierarchy Process. Methodologically, the study aims to construct and formalize a valuation framework rather than to conduct large-sample empirical hypothesis testing. Its primary theoretical contribution lies in overcoming the limitations of traditional income-based approaches, which often fail to capture the dynamic and multi-dimensional value characteristics of marine data assets. Specifically, this framework fills the gap in the literature by offering a comprehensive method that integrates both financial and non-financial factors, such as data quality, application value, and associated risks, tailored to the unique characteristics of marine data assets. Based on the financial and operational data of representative leading shipping enterprises in recent years, this study identifies the independent value contribution of data assets by constructing a dynamic earnings stripping model, combined with industry characteristic parameter optimization and expert correction coefficient calibration, with the aim of breaking through the limitations of the traditional assessment paradigm’s adaptability to unstructured data. The methodological contributions of this study are twofold. Rather than aiming at empirical generalization, this study focuses on methodological transparency, internal consistency, and operational feasibility of marine data asset valuation. First, it extends traditional income-based valuation by introducing a multi-period excess earnings framework that relaxes steady-state assumptions and is better suited to the dynamic characteristics of marine data assets. Second, it formalizes the synergistic relationship between marine data assets and physical assets through an explicit earnings separation mechanism, providing a structured analytical basis for data asset valuation in the blue economy. The research findings aim to provide a basis for enterprise data asset rights confirmation, financial institution pledge financing pricing, and regulatory authority classification standard design, driving the paradigm shift of marine data from “resources” to “capital.” By focusing on the valuation of marine data assets within the blue economy, this methodological framework directly contributes to Sustainability’s core concerns regarding sustainable resource utilization, digital transformation, and long-term value creation in marine industries.

Based on the above discussion, the objective of this study is to develop a systematic valuation framework for marine data assets that can be applied to capital-intensive maritime enterprises. Specifically, this study seeks to address the following research questions:

RQ1: How can the economic value of marine data assets be isolated from the returns generated by traditional tangible and intangible assets in maritime enterprises?

RQ2: How can a multi-period excess earnings framework better capture the dynamic and synergistic value creation mechanisms of marine data assets?

RQ3: How can the proposed valuation framework be operationalized through an illustrative application to demonstrate its practical implementation?

1.2. Literature Review

The literature related to the research topic of this paper can be broadly categorized into three types: the first type focuses on research into data assets, the second type addresses methods for assessing the value of data assets, and the third type pertains to applicable methods for evaluating the value of marine data. Currently, the inclusion of corporate data assets in financial statements in China is still in its experimental and initial stages, with most research focusing on theoretical aspects. Since the International Data Management Association proposed the concept of “data as a critical corporate asset” in 2009, domestic scholars have increasingly delved into data asset research, covering areas such as data definition, classification, characteristics, and value assessment (Ma et al., 2023) [1]. Existing research findings on data assets indicate that they can optimize enterprise resource allocation methods, enhance operational certainty (Cai et al., 2022) [2], and improve operational efficiency (Hu et al., 2022) [3]. However, these methods fall short when applied to marine data assets, which involve complex environmental, legal, and industry-specific factors that traditional models fail to account for. Therefore, the Multi-period Excess Earnings Method and the integration of Fuzzy Comprehensive Evaluation (FCE) are proposed as more suitable alternatives for valuing marine data assets, as they capture dynamic value fluctuations and the contribution of non-traditional assets. Additionally, the application capabilities of data assets play a significant role in enabling high-quality enterprise development (Shi, 2022) [4]. Regarding the valuation of data assets, some scholars have improved the traditional valuation model by constructing it from a cost perspective using the analytic hierarchy process, or by using the market approach as the core to construct a market-based data asset valuation model (Liu et al., 2016; Li et al., 2018) [5,6]. Another group has attempted to adopt non-traditional models from fields such as computer science and economics, such as using the B-S theory model as the core for evaluating the value of data assets in internet finance companies (Wang et al., 2019) [7]. Therefore, it can be observed that there is still a lack of unified measurement standards for data asset value assessment, further highlighting the theoretical gaps in the field of marine data asset assessment. Most domestic and international research has focused on data asset valuation in the internet and financial sectors, while systematic theoretical frameworks and practical guidelines for data asset valuation in the marine industry remain lacking (Feng, 2024) [8]. As a special type of production factor, marine data has significant differences in valuation methods compared to traditional fields such as the internet and finance. Therefore, existing valuation methods are not fully applicable to marine data assets, leading to challenges in pricing and valuation in practical applications, which further constrains the marketization and capitalization of marine data.

The unique characteristics of marine data assets—such as cross-border legal issues, data flow restrictions, and the sector-specific application in industries like marine transportation, fisheries, and environmental monitoring—distinguish them significantly from other data assets. These characteristics imply that the value realization of marine data assets is highly context-dependent, long-term, and synergistic with physical assets, making it difficult for static or single-period valuation approaches to adequately capture their economic contribution. Unlike conventional data assets, which primarily serve the financial or internet sectors, marine data assets have far-reaching implications for global marine resource management and environmental sustainability. These distinct characteristics necessitate the development of a new and specialized valuation framework for marine data assets, rather than relying on traditional income-based methods. In recent years, some scholars have proposed new valuation models based on traditional valuation methods and made innovations. For example, they have combined Grey Relation Analysis (GRA) and the Analytic Hierarchy Process (AHP) to establish a more appropriate data asset valuation method (Song et al.,2021) [9]. The data asset valuation framework based on the Fuzzy Comprehensive Evaluation Method (FCE) compensates for the shortcomings of traditional methods (Li Yonghong et al., 2018) [6] and provides new ideas for the valuation of marine data assets. However, while these studies have to some extent advanced the development of data asset evaluation methods, they lack in-depth exploration of the unique characteristics of marine data assets. In fields such as marine resource development, dynamic evaluation of data assets still faces significant challenges.

This paper argues that the Multi-period Excess Earnings Method is particularly appropriate for valuing marine data assets, as it explicitly accommodates the multi-stage, cumulative, and uncertainty-intensive value realization processes inherent in marine data utilization. By contrast, conventional single-period or steady-state income approaches tend to either underestimate long-term data value or arbitrarily allocate residual earnings, especially in capital-intensive maritime enterprises. Unlike single-period models, which may fail to capture long-term value fluctuations, the multi-period framework accounts for dynamic changes in the marine economy, providing more accurate and flexible valuation results over extended periods. This method considers the different sources and paths of excess earnings, separates the total excess earnings of an enterprise among various assets, and calculates the contribution value of each asset separately. It is suitable for enterprises that continue to make profits and generate excess earnings (Chen et al., 2025) [10]. However, despite the widespread use of asset valuation methods in sectors like the internet and finance, the unique characteristics of marine data assets necessitate the development of a new valuation framework to address the challenges of assessing their long-term and multi-dimensional value. This method can better reflect the value fluctuations of data assets in different periods and market environments, and is particularly suitable for the marine economy, an industry with a high degree of uncertainty and dynamic changes. The Multi-period Excess Earnings Method not only provides enterprises with more accurate economic earnings expectations, but also fully reflects the long-term value of data assets in areas such as long-term resource development and environmental management (Chen et al., 2021) [11].

Combining the special characteristics of marine data assets, this paper proposes the applicability of the Multi-period Excess Earnings Method in the marine economy. The value of marine data is not only reflected in the direct economic benefits it brings to enterprises, but also includes its potential value in environmental protection and climate change response. By introducing the Multi-period Excess Earnings Method, this paper is able to more comprehensively measure the value of marine data assets, including their contribution to long-term strategies such as resource development and environmental management. In addition, this paper innovatively combines the Fuzzy Comprehensive Evaluation Method with the Analytic Hierarchy Process to propose a multidimensional evaluation model aimed at more effectively responding to the dynamic changes and complexity of data assets in the marine economy.

In summary, the valuation of marine data assets requires breaking free from the limitations of traditional valuation methods and adopting a more flexible and comprehensive framework. Through the innovative application of the excess return method and the construction of a multi-dimensional evaluation framework, this paper provides marine enterprises with more precise valuation tools and offers theoretical support and practical guidance for the digital transformation of the marine economy. As technology advances and market demands evolve, the value of marine data assets will become increasingly significant. Future research will continue to drive progress in this field, offering more insights and solutions to support the sustainable development of the global marine economy.

The structure of the subsequent sections is as follows: Section 2 presents the theoretical analysis and methodological choices, explaining the data, models, and their applicability selected for this study; Section 3 outlines the research design, constructing a model for assessing the value of marine data assets, including relevant validation tests; Section 4 presents an illustrative application based on a representative container shipping enterprise. Based on 2024 baseline data and projections for the 2025–2029 forecast period, it completes the valuation of data assets; Section 5 concludes with findings and policy implications.

Despite the growing body of literature on data assets and intangible asset valuation, several unresolved methodological gaps remain. First, existing studies predominantly focus on internet, platform-based, or financial enterprises, while the valuation of data assets in capital-intensive maritime enterprises has received limited attention. Second, many valuation approaches rely on static assumptions and fail to capture the multi-period and synergistic value creation mechanisms between data assets and traditional physical assets. Third, current studies rarely integrate financial valuation models with qualitative dimensions such as data quality, application value, and risk. These gaps motivate the development of a dynamic, multi-period valuation framework tailored to marine data assets. Addressing these gaps requires a methodological rather than purely empirical contribution, which frames the scope and design of this study. Taken together, these gaps highlight the need for a valuation framework that can jointly address multi-period earnings attribution, asset synergy, and qualitative heterogeneity in marine data assets.

This study is methodological in nature. Its primary purpose is to construct and formalize a valuation framework for marine data assets rather than to conduct large-sample statistical hypothesis testing. The empirical materials are employed in an illustrative manner to demonstrate the internal logic, parameter setting, and implementation procedure of the proposed model, rather than to provide statistical validation or cross-method comparison.

2. The Conceptual Definition of Data Asset Value and Its Evaluation Methods

2.1. The Conceptual Content and Scope of Data Assets

In the wave of global economic transformation driven by digital technologies, traditional industrial economic models are accelerating their transition toward digital economic forms, reshaping the operational paradigms of modern economic systems (Yuan et al., 2022) [12]. As a core carrier of new production factors, data assets have gradually transcended traditional resource boundaries and become important elements of corporate strategic decision-making and value creation, playing a critical role in resource allocation across production and operational processes (Li et al., 2024) [13]. According to the National Data Bureau’s definitions of common terms in the data field, data assets refer to data resources legally owned or controlled by a specific entity that can be monetarily measured and generate economic or social benefits. More broadly, data assets generally refer to data with specific application scenarios that are repeatedly or continuously used for more than one year in production activities (Xu et al., 2022) [14].

The core attribute of data assets lies in their potential to generate quantifiable value. However, data assets are often characterized by unstructured formats and incomplete information, which poses significant challenges for valuation and measurement (Sui et al., 2024) [15]. Consequently, data asset management differs fundamentally from the management of traditional physical assets or human resources and cannot rely on highly standardized and unified management approaches. Effective data asset management typically requires layered classification and detailed data analysis to reflect heterogeneous usage scenarios and value realization pathways. From a transaction perspective, prior classification of datasets can also facilitate information disclosure and reduce decision-making costs for potential buyers.

The classification of data assets provides an analytical foundation for understanding their value creation mechanisms and ownership allocation. However, marine data assets exhibit distinctive characteristics that differentiate them from data assets in other sectors. These include their reliance on environmental factors, satellite data collection methods, and their role in sustainable resource management, which require a new framework for valuation that fully accounts for their multi-dimensional and long-term value. Existing studies primarily classify data assets along dimensions such as usage purpose, data source, ownership entity, and acquisition method (Ding et al., 2024) [16]. While these classification frameworks exhibit theoretical completeness, they are often insufficient to capture the multidimensionality, dynamic evolution, and scenario-dependent characteristics of marine data assets. In particular, the economic value of marine data assets is closely intertwined with enterprise operations and physical assets, which may complicates direct value separation. Therefore, this study focuses specifically on marine data assets and defines them as data resources legally controlled by enterprises, monetarily measurable, and continuously serving maritime logistics and supply chain decision-making. This definition provides a clear analytical scope for subsequent value separation and model construction.

2.1.1. Marine Data

With continuous advances in marine observation technologies, including satellite remote sensing and sensor networks, the scale of marine data has expanded rapidly, with growth rates exceeding those observed in many other industries. The development of networked information systems has further accelerated the accumulation of marine social science data. As an important component of scientific big data, marine big data is increasingly characterized by the integration of natural and social science dimensions rather than being confined to traditional natural science domains (Guo et al., 2014) [17]. Based on data type, marine big data can be broadly divided into natural science data and marine social science data. While natural science data has been widely studied, the scope of marine social science data remains less clearly defined in both academic research and policy practice. Existing studies generally acknowledge that marine social science data encompasses information related to marine economic activities, governance, and socio-cultural processes. Among these categories, marine economic data is most closely linked to enterprise operations and value creation, and therefore constitutes the primary focus of this study.

Marine economic data mainly refers to information related to marine industries such as marine fisheries, sea salt industry, marine transportation industry, marine shipbuilding industry, marine oil and gas industry, coastal tourism industry, and marine service industry (Hou et al., 2017) [18]. This article focuses on the value of data assets corresponding to marine economic data.

2.1.2. Marine Data Assets

Although data possesses intrinsic value, not all data can be classified as data assets (Li et al., 2021) [19]. Generally, three conditions must be satisfied for data to be recognized as an asset: clearly defined ownership, the ability to generate economic benefits, and reliable cost measurement. At present, there is no unified academic definition of marine data assets. Considering both the technical attributes of data—such as storability, processability, shareability, and value appreciation—and the general conditions for asset recognition, marine data assets can be defined in this study as marine-related data resources recorded in physical or electronic form that are legally owned or controlled and capable of generating expected economic benefits for their owners.

2.2. Factors Affecting the Value of Marine Data Assets

With the advancement of digital transformation, marine data assets have become increasingly important in areas such as data management, marine resource development, and environmental monitoring. Marine data are mainly generated through observation systems, satellite remote sensing, shipborne sensors, and operational information systems, covering multiple dimensions related to ecology, resources, and climate. Existing studies generally suggest that the value of data assets can be analyzed from both benefit and risk perspectives, reflecting their dual attributes as economic resources and regulated information assets.

From the perspective of economic benefits, marine data assets are fundamentally different from data assets in other sectors, such as internet or finance, due to their unique characteristics. These include their reliance on environmental factors, the use of satellite remote sensing for data collection, and the significant role they play in sustainable marine resource management. Unlike internet or financial data, marine data is often subject to environmental variability, making it essential to adopt a new, dynamic valuation framework that accounts for the long-term, multi-dimensional value of these assets in the marine industry. The application value of data assets is closely related to data quality attributes, including accuracy, timeliness, completeness, and consistency, which directly affect their usability and reliability [15,16]. High-quality data are more likely to generate stable economic returns by supporting operational optimization and strategic decision-making.

In addition to application benefits, the value of marine data assets is influenced by multiple intrinsic factors, such as data authenticity, integrity, accuracy, and related acquisition and storage costs. Data authenticity and integrity determine the credibility and application potential of data assets, while accuracy affects the reliability of analytical results and decision outcomes. At the same time, data collection, processing, and storage costs impose practical constraints on the market transaction value of data assets and influence their valuation boundaries.

Furthermore, the value realization of marine data assets is shaped by application-related characteristics, including scarcity, timeliness, multidimensionality, and scenario dependence. Scarce data resources with limited accessibility tend to exhibit higher economic value, while timeliness determines the effectiveness of data in dynamic decision-making contexts. Multidimensional data structures and specific application scenarios further enhance the economic attributes of data assets by enabling cross-domain value creation and synergistic effects.

Finally, marine data assets are subject to various risk factors that constrain their circulation and utilization. Legal and regulatory requirements, data security concerns, confidentiality obligations, and ethical standards all play important roles in determining the feasible scope of data asset exploitation. In particular, marine data often involve sensitive issues related to ecological protection, national security, and international cooperation, which necessitate strict compliance with legal and institutional frameworks. These risk factors highlight the need to incorporate both qualitative and quantitative considerations into the valuation of marine data assets.

2.3. Analysis of the Value Assessment of Marine Enterprise Data Assets

Marine enterprises operate within a highly complex economic and institutional environment characterized by industrial diversification, technological dependence, policy sensitivity, and internationalization. Their business activities span multiple sectors, such as shipping, fisheries, offshore energy, and marine resource development, and are often accompanied by high capital intensity and long investment cycles. As marine economic activities expand into deep-sea and polar regions, enterprises increasingly rely on data-driven technologies to enhance operational efficiency while addressing environmental and regulatory constraints. These characteristics differentiate marine enterprises from typical data-intensive firms in internet or platform-based industries.

Within this context, the value of marine enterprise data assets is jointly shaped by data quality, cost structures, and value realization potential. High-quality data—characterized by accuracy, timeliness, and completeness—directly affects the reliability of decision-making and the effectiveness of operational optimization. At the same time, data acquisition, processing, storage, and management costs impose practical constraints on the economic returns that data assets can generate. In addition, the value-added potential of marine data assets is often realized over extended time horizons through repeated use, sharing, and integration across business processes, rather than through one-time transactions.

From a valuation perspective, these features pose significant challenges for traditional asset valuation approaches. Marine data assets exhibit strong dependence on enterprise operations, long-term value realization paths, and pronounced synergy with physical assets and organizational capabilities. Moreover, their economic contributions are influenced not only by financial indicators but also by non-financial dimensions, such as environmental monitoring functions, regulatory compliance, and risk exposure. As a result, conventional static valuation methods and purely financial approaches may have difficulty fully capturing the dynamic and multidimensional value of marine data assets.

In the digital economy, data has become a key production factor driving socio-economic development, and the accurate valuation of data assets is increasingly important for enterprise management and policy design (Chen et al., 2025) [10]. Recent studies emphasize the growing importance of incorporating sustainability factors into asset valuation, particularly for industries such as marine resource management. According to Manioudis & Meramveliotakis (2022) [20], a comprehensive theory of sustainable development in asset evaluation must take into account not only traditional economic metrics but also long-term environmental and social impacts. Similarly, Klarin (2018) [21] argues that sustainable development should be an integral part of value assessment frameworks, which aligns with the idea that marine data assets play a pivotal role in addressing ecological and climate-related challenges. These insights highlight the necessity of a new valuation framework that incorporates the broader, long-term value of marine data assets in supporting sustainable development within the blue economy. However, the valuation of marine data assets remains in its infancy, facing challenges such as limited comparability across enterprises, data confidentiality, and methodological adaptability. These challenges emphasize the need for a valuation framework that can accommodate multi-period value realization, asset synergy, and both quantitative and qualitative factors, providing a more suitable basis for assessing the value of marine enterprise data assets.

2.4. Marine Enterprise Data Asset Evaluation Method

Existing approaches for valuing data assets can be broadly classified into cost-based, market-based, and income-based methods. The cost approach evaluates data assets based on their historical acquisition and development costs, but it often fails to reflect the full economic value of data assets and does not adequately account for future value realization. The market approach determines asset value by referencing transaction prices of comparable assets; however, due to the immaturity of marine data trading markets and the limited availability of comparable transaction cases, its practical applicability in the marine sector remains constrained. By contrast, the income approach estimates the future economic benefits generated by data assets and discounts them to present value, making it more suitable for capturing the long-term contribution of data assets, although its accuracy depends on forecasting assumptions and parameter uncertainty.

In addition to these classical approaches, several extended valuation methods have been applied in the assessment of marine data assets, including real option analysis, excess earnings methods, neural network models, and scenario analysis. The real option approach is effective in addressing uncertainty and managerial flexibility, particularly in resource development and environmental monitoring contexts. Neural network models can capture complex nonlinear relationships in data valuation but require large volumes of high-quality training data and involve high model complexity. Scenario analysis evaluates potential asset value under different future conditions and is well suited for highly uncertain environments. However, these methods are often applied in isolation and may have limitations in systematically capturing the long-term, multi-stage risk heterogeneity and dynamic synergistic effects between marine data assets and other enterprise assets.

To address these challenges, this study constructs a multi-period excess earnings valuation framework based on the income approach, following a “total synergistic earnings–data contribution separation” logic. The multi-period approach is particularly effective for valuing marine data assets, as it allows for adjustments based on varying market conditions and incorporates the long-term benefits and risks associated with environmental and regulatory changes, which are central to the marine industry. The framework first estimates the total synergistic returns generated by enterprise intangible assets and then isolates the excess earnings attributable to data assets by introducing a contribution coefficient (α). These excess earnings are subsequently discounted to present value using stage-specific discount rates. Compared with single-period or static valuation methods, the proposed framework allows for dynamic risk adjustment across different stages of value realization and reduces reliance on individual forecasting parameters through multi-period earnings decomposition. Moreover, as an income-based approach, the framework is compatible with existing accounting principles for data asset recognition, facilitating its practical application in marine enterprises.

3. Construction of a Data Asset Valuation Model Based on the Multi-Period Excess Earnings Method

3.1. Model Construction for Marine Data Asset Valuation: Multi-Period Excess Earnings Method

Building on the theoretical discussion presented in the previous section, this study develops a valuation framework for marine data assets based on the multi-period excess earnings method. The core rationale of the framework is that the economic value of marine data assets is not fully captured by traditional static methods or single-period income measures. Instead, the value emerges over time, as data assets contribute to enterprise earnings through multiple periods. This dynamic approach allows for more accurate valuation by considering the long-term and multi-dimensional nature of data assets, which are influenced by varying market conditions, environmental changes, and evolving regulatory landscapes.

In this study, “marine data assets” refer to data resources generated from marine-related operations (e.g., vessel navigation, port operations, cargo tracking, and marine environmental monitoring) that are (i) lawfully owned, controlled, or licensed by an enterprise, (ii) governed by sector-specific compliance constraints (e.g., cross-border data transfer rules, data security, and maritime regulatory requirements), and (iii) capable of generating identifiable economic benefits through internal use or permitted external utilization (e.g., service enhancement, operational optimization, risk management, or authorized data-sharing arrangements). This definition distinguishes marine data assets from general enterprise data by explicitly considering ownership/control rights, permissible use and transferability, and legal–regulatory constraints that affect value realization.

Accordingly, the multi-period excess earnings method captures the dynamic value realization process of marine data assets by estimating the excess earnings attributable to intangible assets over multiple periods. A key feature of this approach is the separation of data asset contributions from the overall synergistic returns generated by enterprise operations. In Step 1, we calculate total synergistic earnings by subtracting the normal returns of traditional assets, and in Step 2, we isolate the excess earnings attributable specifically to data assets using the data contribution coefficient α. This process ensures that the value of data assets is not conflated with the broader contributions of other intangible assets (e.g., brand value, organizational capital, or proprietary know-how), allowing for a clearer understanding of the specific economic impact of data assets on the enterprise. To account for heterogeneity in data quality, application value, and risk characteristics, the framework incorporates the Analytic Hierarchy Process (AHP) and the Fuzzy Comprehensive Evaluation (FCE) method as adjustment mechanisms. AHP is used to structure and weight key influencing factors, while FCE is applied to derive a correction coefficient that reflects qualitative attributes not directly captured by financial indicators.

The valuation framework follows a “total synergistic earnings–data contribution separation” logic and consists of three main steps.

Step 1: Calculation of total synergistic earnings:

$$R_{syn,t}={FCF}_t-\left(V_{wc,t}+V_{fa,t}+V_{ia,t}+V_{labor,t}\right)$$

(1)

where $R_{syn,t}$ denotes the total synergistic earnings in period $t$, representing the excess returns generated jointly by enterprise intangible assets after deducting the normal returns attributable to traditional production factors. ${FCF}_t$ is the free cash flow of the enterprise in period $t$, while $V_{wc,t}$, $V_{fa,t}$, $V_{ia,t}$, $V_{labor,t}$ represent the values of working capital, fixed assets, amortizable intangible assets, and labor input, respectively. Operationally, $V_{labor,t}$ is measured using total employee compensation and converted into a value-contribution term using a cost-plus multiplier calibrated to reflect labor-enabled value creation in the firm’s operating context.

Step 2: Identification of data asset excess earnings:

$$\Delta R_{data,t}=R_{syn,t}\times \alpha$$

(2)

where $\Delta R_{data,t}$ represents the excess earnings attributable to data assets in period $t$. The parameter α denotes the data-asset contribution coefficient (0 < α < 1), representing the marginal share of total synergistic earnings that is attributable to marine data assets. To reduce arbitrariness, α is determined using a two-stage triangulation procedure that combines structured expert elicitation with industry benchmarking. First, a panel of domain experts (the same expert group used in the AHP–FCE procedure) provides independent judgments on the relative contribution of marine data assets to operational synergies under a standardized rubric (covering data-driven efficiency gains, service improvement, and risk mitigation). The individual assessments are aggregated using a consensus-based approach (e.g., median/trimmed-mean aggregation after iterative feedback) to obtain an initial interval for α. Second, this interval is cross-checked against benchmarking evidence from comparable maritime enterprises with disclosed digital transformation practices, ensuring that the selected baseline α is consistent with industry-level plausibility. The baseline scenario adopts α = 0.25, and its uncertainty is explicitly examined through sensitivity analysis (α ∈ [0.20, 0.30]) to assess the robustness of valuation conclusions.

To operationalize the expert elicitation, experts were provided with a short briefing note that defined marine data assets and the scope of “synergistic earnings” in this study, together with a standardized scoring rubric. Each expert first proposed an interval estimate for α (lower bound–most likely–upper bound) based on three dimensions: (i) operational efficiency gains enabled by data-driven optimization, (ii) service and revenue enhancement enabled by data-enabled offerings, and (iii) risk mitigation effects (e.g., safety, compliance, and disruption resilience). To reduce anchoring and dominance effects, ratings were collected independently and anonymously in the first round, all ratings were collected independently and anonymously, and the final baseline α was determined using robust aggregation (median and interquartile range) rather than relying on a single point estimate. The final baseline α was set using the median of the “most likely” values, while the sensitivity interval [0.20, 0.30] was aligned with the interquartile range and cross-validated against industry benchmarking evidence from comparable maritime enterprises.

Step 3: Valuation of marine data assets:

$$V_{data}=\sum _{t=1}^T {\displaystyle \frac{\Delta R_{data,t}}{(1+i_t{)}^t}}\times K$$

(3)

where $V_{data}$ denotes the market value of marine data assets, $i_t$ denotes the discount rate applicable to data asset returns in period $t$, reflecting period-specific risk and market conditions, and $K$ is a correction coefficient derived using the fuzzy comprehensive evaluation method. The correction coefficient $K$ is derived through the fuzzy comprehensive evaluation (FCE) method, where expert judgment is combined with quantitative analysis. This coefficient adjusts the initial valuation results by incorporating qualitative factors that are difficult to capture using traditional financial metrics. These factors include data quality attributes, such as accuracy and timeliness, as well as external risks, such as environmental changes, market volatility, and regulatory compliance. The coefficient K is crucial for providing a more accurate and comprehensive valuation that reflects both financial and non-financial elements.

Through this three-step structure, the proposed framework enables a refined separation of data asset-specific earnings from overall synergistic returns, thereby mitigating the risk of attributing total intangible value to a single asset category. This approach provides a systematic basis for capturing the dynamic and multi-period value characteristics of marine data assets.

3.2. Return on Investment in Marine Data Assets

In the valuation of marine data assets, the discount rate is a critical parameter, as excess earnings are realized over future periods and must be adjusted to reflect market risk and capital costs. Directly applying an enterprise-level discount rate to data assets may lead to biased valuation results, given the heterogeneous risk characteristics of different asset categories. Therefore, an optimized estimation process is required to derive a discount rate that better reflects the risk profile of data assets.

This study first estimates the cost of equity capital using the Capital Asset Pricing Model (CAPM):

$$R_e=R_f+\beta \left(R_m-R_f\right)$$

(4)

where $R_f$ denotes the risk-free interest rate, proxied by government bond yields; $R_m$ represents the average market return, calculated based on the long-term performance of the Shanghai and Shenzhen stock indices; and $\beta$ is the equity beta coefficient obtained from the Choice Financial Terminal.

Based on the cost of equity and the cost of debt, the weighted average cost of capital ($WACC$) is calculated as:

$$WACC=R_e\times {\displaystyle \frac{E}{D+E}}+R_d\times {\displaystyle \frac{D}{D+E}}\times \left(1-T\right)$$

(5)

where $R_d$ is the cost of debt capital, proxied by the benchmark loan interest rate published by the People’s Bank of China; $E$ and $D$ denote the market values of equity and debt, respectively; and $T$ represents the corporate income tax rate.

To mitigate potential bias arising from firm-specific characteristics, this study adopts an industry benchmarking approach. Specifically, the return on data assets is inferred by reverse-engineering the return on intangible assets of representative industry benchmark firms, such as Maersk and MSC. This approach allows the discount rate to better reflect industry-level risk characteristics and improves the robustness of the valuation framework.

Furthermore, the return on investment (ROI) of data assets is estimated using the Weighted Average Return on Assets ($WARA$) framework. Given the absence of a direct and observable return rate for data assets, the $WARA$ method is employed to infer the return on intangible assets:

$$WARA=W_C\times i_c+W_f\times i_f+W_j\times i_j$$

(6)

When the weighted average return rate equals the weighted average cost of capital, the return on intangible assets can be derived as:

$$i_j={\displaystyle \frac{WACC-\left(W_f\times i_f\right)-\left(W_c\times i_c\right)}{W_j}}$$

(7)

where $W_f$, $W_c$, and $W_j$ denote the proportions of fixed assets, current assets, and intangible assets in total assets, respectively, and $i_f$, $i_c$, and $i_j$ represent the corresponding rates of return. The derived $i_j$ is subsequently used as a reference return rate for data assets in the valuation model.

3.3. Determination of Weights and Parameter Settings in the Valuation Model

To enhance the practical applicability of the valuation framework, the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation (FCE) methods are employed to derive the correction coefficient ($K$). Based on the factor analysis presented in Section 2, the influencing factors of marine data asset value are structured into three criteria layers: data quality, application value, and risk. Expert judgment is used to quantify the relative importance of these factors.

Specifically, ten experts are invited to conduct pairwise comparisons of the evaluation indicators to construct the judgment matrix. The expert panel was constituted to balance valuation expertise and domain knowledge of marine digitalization. Experts were selected based on predefined criteria (e.g., relevant professional experience in maritime operations/data management or intangible asset valuation, familiarity with data governance/compliance, and the ability to provide independent judgments). To mitigate subjectivity, the elicitation process followed a standardized scoring protocol and consistency checks ($CR$ < 0.1) were enforced for the AHP judgment matrices. The initial weights are calculated using the geometric mean method:

$$W_i={\displaystyle \frac{\sqrt[n]{{\prod }_{j=1}^n a_{ij}}}{{\sum }_{i=1}^n \sqrt[n]{{\prod }_{j=1}^n a_{ij}}}}$$

(8)

After the consistency test is satisfied ($CR<0.1$), the fuzzy comprehensive evaluation is performed by combining the weight vector $W$ with the expert scoring membership matrix $R$. The correction coefficient is obtained as: $K=W^T\times R$. The coefficient $K$ is used to adjust the initial valuation results in order to incorporate qualitative factors that are not directly captured by financial indicators, such as data quality fluctuations, application conditions, and policy-related risks.

Experts were selected to ensure balanced coverage of (a) intangible asset valuation and financial modeling and (b) maritime digitalization and data governance. Inclusion criteria required (i) at least five years of relevant professional experience, (ii) familiarity with data-enabled maritime operations (e.g., shipping/port/logistics digital systems) or data valuation practices, and (iii) the ability to provide independent judgments without conflicts of interest related to the illustrative firm. The AHP pairwise-comparison task was conducted using a standardized 1–9 Saaty scale, and each expert’s judgment matrix was screened for internal consistency. Matrices not meeting the CR < 0.1 criterion were returned for revision with guidance, and the final weights were obtained by aggregating consistent matrices using the geometric mean method.

In addition, the determination of the revenue period and cash flow structure is a key component of the valuation process, as the economic value of data assets is realized through long-term operations. Free cash flow is calculated using a standard financial formulation:

$${FCF}_t={EBIT}_t\times \left(1-T\right)+{Dep}_t+{Amort}_t-{CAPEX}_t-\Delta {WC}_t$$

(9)

where ${EBIT}_t$ denotes earnings before interest and taxes in period $t$, ${Dep}_t$ and ${Amort}_t$ represent depreciation and amortization, respectively, ${CAPEX}_t$ is capital expenditure, and $\Delta {WC}_t$ denotes the change in working capital.

By integrating the revenue period setting with the multi-period cash flow structure, the proposed framework provides a consistent basis for capturing the long-term economic contribution of marine data assets.

4. Illustrative Application of the Proposed Valuation Framework (Company C)

4.1. Basic Information About the Case

When selecting case study companies, the following principles were primarily followed: First, the companies must be representative of their industries. China’s marine economy is developing rapidly, encompassing multiple sectors such as fisheries, transportation, energy, and equipment manufacturing. To ensure the broad applicability of the research, the case study companies must be industry leaders in terms of market share, exhibit strong competitive advantages, and have demonstrated substantial progress in digital transformation, evidenced by their operational platforms and data-driven initiatives. Second, the selected companies must serve as benchmarks for digital transformation. As the digitalization of the marine industry progresses, the importance of data assets in corporate value assessment has become increasingly prominent. Although many companies have not yet fully achieved digital transformation, leading companies have already established operational platforms based on marine big data to conduct data collection, storage, analysis, and application. Finally, case study companies should have sound operational performance. Given that this paper employs a multi-period excess return model for evaluation, which primarily relies on financial indicators, the selected companies must possess stable profitability. In summary, during the case selection process, three principles must be adhered to: industry representativeness, digital transformation benchmarking, and operational stability.

This study selects Company C, the world’s third-largest container shipping company, accounting for about 11.5% of the global market share. Company C is an anonymous code name referring to COSCO Shipping Holdings Co., Ltd. (601919.SH). Data sources are detailed in its 2018–2023 annual reports (Shanghai Stock Exchange). The company was established on 3 March 2005, and successfully listed on the Main Board of the Hong Kong Stock Exchange on 30 June 2005. Its operations cover 556 ports across 137 countries and regions worldwide, with accumulated vessel sensor data exceeding 500 TB. Its electronic booking platform processes over 20 million orders annually, reflecting the significant scale of its data assets in the global shipping industry. After years of development, the company has demonstrated strong business resilience through industry cycle fluctuations, bolstered by its ongoing investment in digital technologies such as smart shipping systems, automated port operations, and data management platforms. Its business spans multiple sectors including shipping, logistics, ports, finance, and energy, and it has established a robust global logistics and shipping network. Therefore, Company C effectively meets the three key criteria for case selection.

Based on the “multi-period excess return-AHP-FCE” composite model constructed in this paper, the excess return of data assets from 2025 to 2029 is predicted by introducing the contribution coefficient ($\alpha$) and finely stripping out the contribution value of traditional assets such as current assets and fixed assets; The discount rate $i_j$, derived through the CAPM-WACC-WARA composite framework, is 15.0%. The adjustment factor K was quantified using the AHP-FCE method, yielding a composite adjustment factor $K_{\mathrm{adjustment}}=0.824$. Through quantitative model testing, the appraised value of Company C’s data assets as of 31 December 2024 was finally determined to be $3390 million, and the result was illustrated to be robust through sensitivity analysis. It is important to note that this valuation outcome is not intended to serve as a definitive estimate, but rather to illustrate the internal logic, parameter interactions, and operational feasibility of the proposed valuation framework. The valuation outcome derived from the illustrative case serves as a methodological reference for demonstrating how the proposed framework can be operationalized in capital-intensive shipping enterprises, rather than as an industry-wide benchmark. In the future, it is necessary to further explore the dynamic pricing mechanism and cross-border circulation rules of data assets, so as to promote their transformation from “hidden resources” to “visible capital” thereby supporting sustainable value creation and data-driven governance in the blue economy. The selection of Company C does not aim to provide industry-wide empirical inference, but to ensure that the illustrative application is grounded in a representative, data-intensive, and digitally advanced shipping enterprise where the proposed valuation logic can be meaningfully demonstrated in practice.

4.2. Business Structure and Operating Conditions of Case Companies

First, Analysis of the Case Company’s Core Business Operations: As shown in

Table 1. Composition of Company C’s Main Business Operations for 2024 and 2023.

Year	2024		2023
	Amount (RMB billion yuan)	Specific Gravity	Amount (RMB billion yuan)	Specific Gravity
Total Operating Revenue	2338.59	100%	1754.48	100%
By Industry
Container shipping business	2259.71	96.63%	1681.26	95.83%
Terminal operations	108.10	4.62%	103.96	5.93%

, Company C’s main business operations in 2023 and 2024 are primarily concentrated in container shipping, with terminal operations constituting a smaller but stable share of total revenue.

As shown in in

Table 2. Detailed Revenue Composition Breakdown for Company C in 2024 and 2023.

Branch	2024 Revenue (RMB Billion Yuan)	2023 Revenue (RMB Billion Yuan)
Container shipping	2259.71	1681.26
Terminal operations	108.10	103.96
Other business	30.78	29.26

, Company C’s main business structure has been highly focused and stable in recent years. The container shipping business is the dominant segment, contributing 96.63% of revenue in 2024. This business generates extensive data on ship trajectories, cargo status, and port operations, which are integral to optimizing global logistics and supporting data-driven decision-making. As an important synergistic segment, terminal business accounts for about 4.6% of the revenue, and its production operation and logistics service data effectively complement and synergize with the main business of shipping. Together, these segments form an integrated marine logistics data ecosystem.

In terms of operating conditions, after experiencing the industry’s high freight rate cycle in 2021–2022, Company C will see a rational retracement of container freight rates from 2023 to 2024 as the global supply chain pressure eases and market supply and demand gradually return to the norm, with the Company’s operating revenue and profit growth rate subsequently stabilizing, but it still maintains a solid profitability and strong cash flow level. In this process, the Company has continued to increase its investment in digitalization and intelligence, including the development and construction of intelligent ship systems, digital freight platforms and automated ports, aiming to improve operational efficiency, control costs and enhance service resilience through technological means. This shows that amid cyclical industry fluctuations, data assets play an essential role in stabilizing operations and enhancing efficiency, particularly through refined management, capacity optimization, and predictive analytics, which are vital in maintaining profitability and mitigating operational risks. This business scenario provides a realistic basis for the transformation of marine data assets from “operational by-products” to “strategic capital”, and also provides a clear business basis for the application of subsequent valuation models.

4.3. Assessment of the Value of Company C’s Marine Data Assets

Company C’s shipping and logistics data accumulated during its digital transformation has gradually formed significant data assets. With the development of intelligent shipping platforms and big data technologies, the market value of data assets is becoming more apparent. The present value approach is used to estimate the economic value by discounting future earnings contributions of the data assets. Using the Multi-period Excess Earnings Method, the valuation focuses on estimating the data assets’ specific earnings contribution, with adjustments made to account for non-data asset components. As data asset transaction types diversify, data assets are expected to play an increasingly important role in supporting Company C’s future value creation. Therefore, this paper defines the valuation subject as the data assets of Company C, with the valuation date set as 31 December 2024, and the value type determined as market value.

4.3.1. Profit Period

Regarding the determination of the profit period, Company C formed a large shipping enterprise through its merger with China Shipping Group in 2016, thereby optimizing its equity structure. In response to shifts in the global shipping market and economic landscape, the company progressively adjusted its business structure, optimized resource allocation, and enhanced competitiveness. Since 2020, the Company has accelerated the construction of information technology and digital platforms, significantly enhancing its capabilities in intelligent shipping, data management and analysis. According to its 2023–2024 annual report, the company has continued to invest in the areas of “intelligent shipping”, “digital supply chain” and “green and low-carbon operation”, and its digital infrastructure and application scenarios have been deepening. The company has continued to invest in areas such as “smart shipping”, “digital supply chain” and “green and low-carbon operation”, and has deepened its digital infrastructure and application scenarios. Given that the release of the benefits of digital transformation usually has a lag of 3–5 years, and taking into account the medium-term planning cycle of the shipping industry, this paper expects that the period from 2025 to 2029 will be a key stage for the realization of the value and growth of its data assets. Therefore, the benefit period is projected from 2025 to 2029, aligning with the anticipated timeline for the realization of value from digital transformation initiatives such as intelligent shipping and digital supply chain management.

4.3.2. Corporate Free Cash Flow Forecast

For operating revenue forecast, this paper is based on the ten-year operating revenue data of Company C (stock code 601919) from 2015–2024, while taking into account the strong cyclicality of the shipping industry, and adopting the dual validation of trend extrapolation and industry consensus for forecasting. First, a linear trend equation $y=890.95+207.35x$ is fitted using the ten-year data to obtain the trend prediction (317.185, 3379.20, 358.655, 379.389, and 400.124 billion yuan for 2025–2029, respectively). However, the R² of this linear trend is low (0.283) and the forecasts are not in line with the general expectation of a cyclical pullback in the industry. Therefore, this article refers to Drewry and a number of brokerage firms on the global container shipping market will enter a low-speed growth phase in the future, the forecast will be adjusted to 2024 as the base, the average annual growth rate of 4% of the steady growth rate. The operating revenue forecast is based on Company C’s historical data (2015–2024), adjusted for the cyclical nature of the container shipping industry, and calibrated using a conservative industry growth forecast. A linear trend is first estimated to provide a mechanical baseline; however, the model fit is limited (R² = 0.283) and the implied path is inconsistent with the industry expectation of a post-cycle normalization. Therefore, the baseline forecast adopts a prudential “industry-consensus” adjustment, taking 2024 as the base year and applying a 4% steady annual growth rate, consistent with publicly available industry outlooks indicating a normalization phase after the recent cycle. The cost and expense forecasts are based on historical averages (2020–2024) to reflect steady operational trends while smoothing out short-term fluctuations. Among them, the ratio of operating costs refers to the average value of 79.85% in 2023–2024, reflecting the optimization of cost structure brought about by the Company’s digital operation; the ratio of financial expenses continues the characteristics of net interest income in recent years and takes a negative value.

Depreciation and amortization reflect the lagging relationship between assets and revenue, so its forecast adopts the “proportion of operating income method”, the ratio of which is taken as the average value in the past five years (2019–2023), and according to the structure of assets in 2024, it is split into depreciation of fixed assets, amortization of intangibles, and amortization of long-term amortization expenses. The alternative depreciation approaches and their applicability are summarized in

Table 3. Comparison and Selection of Depreciation Methods.

Method	Advantages	Disadvantages	Applicable Scenarios
Percentage of Operating Revenue Method	Simple and quick, relying on the correlation between historical revenue and depreciation	Ignoring changes in asset structure, assume that income is linearly correlated with depreciation	Income has grown steadily, with asset investments fluctuating in tandem with income
Depreciation Forecast Based on Original Asset Value	Accurately reflect asset status, taking into account depreciation policies and asset additions and disposals	High data requirements (original asset value, depreciation policy, useful life, etc.)	Complex asset structure or plans for major investments/disposals

.

The forecast of capital expenditure and increase in working capital also adopts the “proportion of operating income method”, and the proportion is taken as the average of the last five years (2020–2024) and the last four years (2021–2024) respectively, in order to reflect the company’s normal rhythm of investment and the efficiency of working capital management. The resulting forecast of depreciation and amortization for each period is reported in

Table 4. Forecast Breakdown of Depreciation and Amortization (Unit: RMB 100 million).

Annual	2023	2024	2025	2026	2027
Depreciation of Fixed Assets	79.55	92.33	105.11	117.89	130.67
Amortization of Intangible Assets	4.45	5.16	5.88	6.59	7.31
Amortization of Deferred Expenses	1.58	1.83	2.09	2.34	2.59
Total	85.58	99.32	113.05	126.79	140.53

.

In order to ensure that the forecast parameters are representative of the long-term and reflect the recent operating characteristics, the following principles are followed in this paper for the selection of the historical data window for each indicator: (1) indicators revealing the long-term trend and cycle (such as operating income) are based on ten-year data; (2) indicators characterizing the current operating structure (such as cost and expense ratio and capital expenditure ratio) are based on the recent five-year data; (3) indicators that need to avoid the interference of extreme values (such as changes in working capital) are selected from more representative recent subintervals (e.g., four years). All parameters are determined with the primary consideration of matching the economic meaning of the indicators and enhancing the robustness of the forecast. Based on historical data and following the principles of prudence and sustainability, this paper determines the following core forecast parameters: (1) depreciation and amortization as a percentage of operating income is 4.58%, which is based on the historical average for the period of 2019–2023; (2) capital expenditure is 6.00%, which is set with reference to the company’s longer-term historical average level of capital expenditure and industry norms, compared with the longer-term historical average level of capital expenditure and industry norms in the last three years, which is attributable to the concentrated investment in green ships and digitization. Compared to the higher expenditures incurred in the recent three years due to the concentrated investment in green ships and digitalization (10–20% range), this ratio better reflects its long-term sustainable investment pace; (3) The increase in working capital as a percentage of operating income is set at 2.00%, based on the Company’s working capital trends over 2021–2024, excluding outliers caused by extreme freight-rate fluctuations (e.g., a sharp increase in 2022 and a significant decrease in 2023), which better reflects normal working capital management efficiency; (4) Taxes and surcharges are set at 0.54% of operating income, based on the average level observed over the past five years (2020–2024).

Finally, free cash flow is calculated for each of the years 2025–2029 based on the free cash flow formula (Free Cash Flow = EBITDA × (1 − Income Tax Rate) + Depreciation and Amortization − Capital Expenditures − Working Capital Additions) (see

Table 5. Free Cash Flow Forecast (Unit: RMB billion).

Annual	2025	2026	2027	2028	2029
Operating Revenue	3008.10	3128.42	3253.56	3383.70	3519.05
Less: Operating costs (79.85%)	2402.97	2498.54	2597.97	2701.38	2809.46
Less: Taxes and surcharges (0.54%)	16.24	16.89	17.57	18.27	19
Less: Selling expenses (0.46%)	13.84	14.39	14.97	15.56	16.19
Less: Administrative expenses (2.46%)	74	76.96	80.04	83.24	86.57
Less: R&D expenses (0.42%)	12.63	13.14	13.66	14.21	14.78
Earnings before interest and taxes (EBIT)	488.42	508.5	529.35	551.04	573.05
Less: Income tax expense (25%)	122.1	127.12	132.34	137.76	143.26
Profit after tax	366.31	381.37	397.01	413.28	429.79
Add: Depreciation and amortization (4.58%)	137.77	143.28	149.01	154.97	161.17
Less: Capital expenditures (6.00%)	180.49	187.71	195.21	203.02	211.14
Less: Increase in working capital (2.00%)	60.16	62.57	65.07	67.67	70.38
Corporate Free Cash Flow (FCF)	263.43	274.37	285.74	297.56	309.44

).

When forecasting capital expenditures, as shown in the table below, the ratio of capital expenditures to revenue exhibits significant fluctuations. The 2018 capital expenditure figure represents a notable anomaly that deviates from the trend observed in subsequent years. Therefore, it is necessary to exclude this outlier year when calculating a reasonable average. This also reflects the current conservative nature of the investment strategy. Should the company plan to increase investments in the future, the ratio may be incrementally adjusted upward.

When forecasting free cash flow: By substituting the calculated values into the formula for a company’s free cash flow, the projected free cash flow for Company C from 2023 to 2027 can be determined.

4.3.3. Discount Rate

The discount rate calculation involves two steps. First, calculate the cost of equity capital $Re$, where the beta coefficient is obtained from Choice Financial Terminal data. The risk-free rate $R_f$ is selected as the yield on China’s five-year government bonds at 3.20% (2023 data), reflecting the long-term risk-free benchmark. The market rate of return $R_m$ is based on the geometric average return of the CSI 300 Index over the ten-year period from 2015 to 2024. $R_m$ is proxied by the long-run average market return based on the CSI 300 index over 2015–2024, and the parameter choice follows a prudential mid-range assumption consistent with public market performance statistics. Subsequently, the weighted average cost of capital ($WACC$) is calculated using a capital structure weighting approach. Then, the weighted average return on investment ($WARA$) is used to derive the intangible asset return rate $i_j$ for each enterprise through a backward calculation method.

In order to ensure the homogeneity of the $WARA$ model parameters, the benchmark companies selected in this paper need to meet the following conditions: (1) their main business is container shipping; (2) they have publicized digital transformation strategies and practices; and (3) their financial data are complete and transparent. Based on this, COSCO (A-share), OOCL (HK-share) and SITC (HK-share) are finally selected as benchmark companies.

The calculated theoretical value of $i_j$ for the three benchmarking firms ranges between 21.70% and 39.40%, with a mean value of 33.30% (see

Table 6. Analysis of Equity Capital Costs for Peer Companies.

Company Name	${\mathit{R}}_{\mathit{f}}$	${\mathit{R}}_{\mathit{m}}$	$\mathit{\beta }$	${\mathit{R}}_{\mathit{e}}$	${\mathit{R}}_{\mathit{d}}$	D/E	Tax Rate	$\mathit{W}\mathit{A}\mathit{C}\mathit{C}$	${\mathit{i}}_{\mathit{j}}$
Company C	2.56%	8.06%	1.30	9.11%	3.20%	0.74	17.20%	6.37%	21.70%
SITC International Holdings Company Limited	4.25%	9.25%	1.29	10.03%	4.20%	0.32	1.91%	8.08%	38.80%
OrientOverseas (International)Limited	4.25%	9.25%	1.07	9.80%	2.02%	0.34	22%	8.14%	39.40%
Industry average (theoretical)									33.30%
Values (benchmarks) are used in this paper	-	-	-	-				-	15.00%

). This value is more sensitive to the weighting assumptions in the $WARA$ model. In order to maintain the prudence and robustness of the valuation conclusions, and to avoid over-discounting the value due to parametric assumptions, this paper finally determines the baseline scenario for the discount rate of the data assets as $i_t=15.0\%$.Given the high sensitivity of the WARA-implied return to weighting assumptions and to avoid overstating valuation volatility, the study adopts a conservative baseline discount rate for data assets ($i_t$ = 15.0%) and further examines its impact through sensitivity analysis. In the subsequent valuation calculations and sensitivity analyses, the discussion will be centered around this benchmark value.

4.3.4. Projected Values of Contributions from Current, Fixed, Amortizable Intangible Assets, and Labor Resources

In this paper, the simplified forecasting method based on the fixed income ratio is discarded in favor of the framework of “Dynamic forecast of asset size × economic rate of return” in order to enhance the scientificity and transparency of the contribution value separation process. The forecast of each asset is based on the company’s historical data from 2020–2024, and the key parameters are set as follows: (1) Current assets: Their size is forecasted through the current asset turnover ratio. The ratio has averaged 1.86 times in the last five years, but shows fluctuations related to industry cycles. Considering the potential for improvement of the company’s operational efficiency under digitalization, its turnover ratio is forecasted to gradually increase from 1.40 times in 2025 to 1.60 times in 2029. The return on current assets is set at 1.0% with reference to the company’s historical ratio of interest income to cash-based assets. (2) Fixed assets and intangible assets: The size of both is projected through a rolling model of net asset value (closing net value = opening net value + capital expenditure − depreciation/amortization). Its rate of return is set at 8% for fixed assets and intangible assets with reference to the long-term average profitability of the global shipping industry under the non-boom cycle and reflecting the prudent principle of valuation. (3) Labor resources: Total employee compensation is used as the cost basis, and its percentage of operating revenue has stabilized at 2.31% in the past five years. Labor contribution draws on human capital theory and adopts the cost-plus method, setting its value creation multiplier at 1.3. Based on the above assumptions, the projected contribution value of each non-data asset from 2025 to 2029 is shown in

Table 7. Forecasted Contribution Value of Non-Data Assets (Unit: RMB billion).

Year	Current Assets	Fixed Assets	Intangible Assets	Labor Force	Total
2025	21.49	101.45	5.98	90.34	219.26
2026	22.28	105.11	6.16	93.95	227.50
2027	23.02	108.89	6.35	97.71	235.97
2028	23.17	112.78	6.55	101.62	244.66
2029	24.37	116.78	6.75	105.68	253.58

.

4.3.5. Correction Factor

After detailing the concept and characteristics of data assets, this paper summarizes factors influencing the value of corporate data assets, such as data quality factors, utility value factors, and risk factors. Therefore, this project proposes to assess the importance of each factor to the assets through 10 domain experts, develop an appropriate analytic hierarchy process to quantify these factors, and revise the multi-period revaluation method to enhance the accuracy of the assessment results. In this paper, five experts in asset valuation and five experts in database management will be invited to evaluate each factor. All experts provided ratings independently following a unified evaluation guideline. For the FCE step, a discrete linguistic scale (e.g., very low/low/medium/high/very high) was mapped to numerical scores, and the corresponding membership degrees were constructed using predefined membership functions (e.g., triangular or trapezoidal functions) to transform expert scores into the membership matrix R. The aggregation procedure uses the weight vector from AHP and the membership matrix from FCE ($K=W^T\times R$), thereby incorporating qualitative dimensions in a transparent and replicable manner. Potential individual bias was reduced by anonymized scoring and by using aggregated statistics across experts rather than relying on any single judgment.

Quantify the impact of data asset quality, application value, and risk factors using AHP and FCE. Step 1: Conduct AHP weight calculations: Invite 10 experts to perform pairwise comparisons of indicators, construct a judgment matrix, and compute weights. Calculate weights for primary indicators of advancement:

The expert scoring judgment matrix is constructed as follows: The corresponding judgment matrix for the primary indicators is presented in

Table 8. Scoring Judgment Matrix.

Indicator	Quality Factors	Applied Value	Risk Factors
Quality Factors	1	3	5
Applied Value	1/3	1	2
Risk Factors	1/5	1/2	1

.

Based on expert ratings, the geometric mean method is first used to calculate the initial weights for quality factors, application value, and risk factors. These weights are then normalized. A consistency test is conducted, where $CR<0.1$ indicates passing the test. The resulting weights for primary indicators are obtained, followed by the calculation of secondary indicator weights. Taking quality factors as an example, the expert ratings are used to construct a judgment matrix: The resulting judgment matrix for quality factors is shown in

Table 9. Quality Factor Judgment Matrix.

Indicator	Real-Time Capability	Integrity	Accuracy
Real-time Capability	1	2	4
Integrity	1/2	1	3
Accuracy	1/4	1/3	1

.

Next, apply the geometric mean method to calculate the initial weights and perform normalization. The same approach applies to the application value and risk factors, ultimately yielding the summary table of secondary indicator weights: The weighting results of primary indicators and the corresponding decomposition of secondary indicators are summarized in

Table 10. Weighting of Primary Indicators and Decomposition of Secondary Indicators.

Primary Indicator	Weight	Secondary Indicators	Weight
Quality Factors	0.6483	Data Timeliness	0.5584
		Data Integrity	0.3196
		Data Accuracy	0.1220
Applied Value	0.2297	Value of Route Optimization	0.6000
		Supply Chain Collaboration Efficiency	0.3000
		Customer Service Enhances Value	0.1000
Risk Factors	0.1220	International Regulatory Compliance	0.6500
		Data Network Security	0.2500
		Geopolitical Risks	0.1000

.

The second step involves calculating correction coefficients using the Fuzzy Comprehensive Evaluation (FCE) method. First, ten experts independently evaluated Company C’s data asset performance using a five-level linguistic scale (very low, low, medium, high, and very high) to derive the evaluation matrices for data quality, application value, and risk factors. These linguistic ratings were mapped to numerical anchors (e.g., 1–5) and converted into membership degrees through predefined membership functions. Specifically, triangular (or trapezoidal) membership functions were adopted to translate expert scores into the membership matrix R, where each row corresponds to an evaluation indicator and each column corresponds to a linguistic grade. The membership functions were specified ex ante to ensure consistent interpretation across experts and to avoid post hoc tuning.

Subsequently, a two-level fuzzy comprehensive evaluation was conducted, with similar procedures applied to quality, value, and risk dimensions. Expert-provided membership degrees were aggregated indicator-wise using the mean, with robustness checks based on the median to mitigate the influence of extreme judgments. The comprehensive correction coefficient was then computed as $K$ = W^T × R and normalized to the interval [0, 1]. The FCE procedure yields a composite score of 8.236, which corresponds to a normalized correction coefficient of $K$ = 0.8236 ≈ 0.824.

Based on the integrated AHP–FCE procedure and expert scoring, the normalized correction coefficient for Company C is $K$ = 0.824. The expert evaluations indicate that data quality and application-value dimensions receive relatively higher scores, while the risk dimension is assessed as manageable under the assumed regulatory and operational conditions. The coefficient K is therefore used as a qualitative adjustment factor to incorporate non-financial attributes into the valuation results in a transparent and replicable manner.

4.3.6. Calculation of Data Asset Value

The purpose of this section is to measure the value of Company C’s marine data assets by applying the “Total Synergy Gain-Data Contribution Separation” composite model constructed in Chapter 3 of this paper. The valuation process strictly follows the three-step process set out in the model, and the key parameters and forecast data that have been measured in the preceding sections of Chapter 4 are used as inputs.

Step 1: Calculate total synergy gains ($R_{\mathrm{syn},t}$)

First, from the predicted free cash flow (FCF) of the firm, the normal return required by each non-data asset (current assets, fixed assets, intangible assets, and labor) is deducted to obtain the total excess return, i.e., the total synergistic return, that is, the total synergistic return, that is synergistically created by all intangible assets. Where $FC{F}_t$ is derived from the projections in Table 5, and the sum of the contribution values of each non-data asset $V_{\mathrm{wc},t},V_{\mathrm{fa},t},V_{\mathrm{ia},t},V_{\mathrm{labor},t}$ is derived from Table 7. The total synergy return $R_{\mathrm{syn},t}$ for each of the years from 2025 to 2029 is computed as shown in

Table 11. Synergy Earnings Calculator (2025–2029) (Unit: RMB billion).

Year	$\mathbf{Enterprise} \mathbf{Free} \mathbf{Cash} \mathbf{Flow} \mathit{F}\mathit{C}{\mathit{F}}_{\mathit{t}}$	Total Value of Non-Data Asset Contributions	$\mathbf{Total} \mathbf{Synergy} \mathbf{Gains} {\mathit{R}}_{\mathbf{syn}\mathsf{,}\mathit{t}}$
2025	263.43	219.26	44.17
2026	274.37	227.50	46.87
2027	285.74	235.97	49.77
2028	297.56	244.66	52.90
2029	309.44	253.58	55.86

.

Step 2: Separate the Excess Returns from Data Assets ($\Delta R_{data,t}$)

Data assets, as part of intangible assets, have a value contribution that needs to be separated from total synergistic returns. This paper introduces the data asset contribution coefficient $\alpha$ to quantify this marginal contribution. Based on expert research and industry benchmarking, this paper sets the contribution coefficient $\alpha$ = 0.25 in the baseline scenario, and substitutes the total synergies in Table 11 into the above formula to get the annual excess returns attributable to data assets, and the calculation results are shown in

Table 12. Calculation of Excess Returns on Data Assets (2025–2029) (Unit: RMB billion).

Year	Total Synergy Gain (${\mathit{R}}_{\mathbf{syn}\mathsf{,}\mathit{t}}$)	Data Asset Contribution Factor (α)	Excess Return on Data Assets ($\mathbf{\Delta }{\mathit{R}}_{\mathbf{data}\mathsf{,}\mathit{t}}$)
2025	44.17	0.25	11.04
2026	46.87	0.25	11.72
2027	49.77	0.25	12.44
2028	52.90	0.25	13.22
2029	55.86	0.25	13.97

.

Step 3: Discounting and comprehensive correction to arrive at the final value

The appraised value of the data assets is obtained by discounting the excess earnings generated by the data assets in each forecast period to the appraisal base date (31 December 2024) and incorporating the comprehensive correction of non-financial factors. Where $i_t$ is the discount rate of data assets in period t (baseline: 15% for t = 1, …, 5), $K$ is the AHP–FCE correction coefficient (0.824), and T = 5 denotes the revenue period (2025–2029).

The sequence of $\Delta R_{\mathrm{data},t}$ in Table 12 is discounted and summed up with a discount rate of 15%, and then multiplied by the modification factor $K$. The complete calculation process and final results are summarized in

Table 13. Assessment of final value of data assets (Unit: RMB billion).

Annual	2025	2026	2027	2028	2029
Excess return on data assets ($\mathsf{\Delta }{R}_{\mathrm{data},t}$)	11.04	11.72	12.44	13.22	13.97
Discount rate ($i_t$ = 15%)	15%	15%	15%	15%	15%
discount factor	0.8696	0.7561	0.6575	0.5718	0.4972
present value of excess earnings	9.60	8.86	8.18	7.56	6.94
Total present value	41.14
Correction factor (K = 0.824)	0.824
Value of data assets (Vdata)	33.90

.

In summary, through the above three-step calculation, this paper concludes that the value of the marine data assets of Company C (COSCO Overseas Holdings) based on the valuation reference date (31 December 2024) is RMB 3.390 billion. The result is a quantitative valuation derived under prudent financial forecasts, asset return assumptions and value separation logic, and provides a core basis for subsequent analysis and discussion.

4.4. Analysis of Evaluation Results

Based on the proposed framework and the case inputs, the estimated market value of Company C’s data assets at the valuation date (31 December 2024) is RMB 3.390 billion. The case demonstrates that, in a capital-intensive shipping model enhanced by digital transformation, data assets represent a significant part of enterprise value, particularly when their contribution is isolated and adjusted through multi-period modeling. In recent years, Company C has continued to promote the application of intelligent shipping platform and logistics big data, and the value transformation path of its data assets has become increasingly clear. The valuation model constructed in this paper aims to provide an objective valuation reference for the company to identify, quantify and manage this strategic asset, and the methodology and conclusions adopted can also serve as a reference for the assessment of the value of data assets of enterprises in the same industry. In order to further test the reliability of the conclusions and explore the influence mechanism of key parameters, a systematic sensitivity analysis will be conducted in the following.

4.5. Sensitivity Analysis

In order to test the robustness of the valuation results, this paper conducts a one-way sensitivity analysis of the three key parameters in the model. The range of parameter variations is set based on reasonable grounds: the data asset contribution coefficient (α) fluctuates within the range of 0.20 to 0.30, which is derived from the feedback of expert research; the discount rate ($i_t$) is adjusted between 12% and 18% to reflect the possible changes in the market risk premium; and the rate of return on fixed assets is taken within the range of the industry’s historical fluctuation of 7% to 9%. The results show that when the contribution coefficient decreases to the lower bound of 0.20, the estimated value of data assets declines to RMB 2.712 billion; when it increases to the upper bound of 0.30, the value correspondingly rises to RMB 4.068 billion. Variations in the discount rate exhibit a negative relationship with valuation outcomes: when the discount rate is set at 12%, the estimated value reaches RMB 3.815 billion, whereas it decreases to RMB 3.011 billion at an 18% discount rate. Changes in the return on fixed assets have a relatively moderate impact, with the valuation ranging from RMB 3.650 billion to RMB 3.130 billion as the return rate varies between 7% and 9%. The data asset values remain positive under all scenarios tested and the range of fluctuations suggests good robustness of the benchmark result. Sensitivity analysis further reveals that the valuation is most sensitive to changes in the contribution factor (α), highlighting the importance of accurately assessing the business contribution margin of data assets. Overall, the valuation remains positive across all tested scenarios, and the fluctuation range (RMB 2.712–4.068 billion) suggests that the baseline estimate is reasonably robust to parameter uncertainty. The results also indicate that the valuation is most sensitive to α, highlighting the importance of carefully assessing the marginal contribution of data assets in operational synergies.

5. Conclusions and Insights

This study examines the valuation of marine data assets in the context of the blue economy and proposes a composite valuation framework integrating the multi-period excess earnings method with the AHP–FCE approach. Using Company C as an illustrative application, the study demonstrates how the proposed framework can be operationalized to separate and quantify the economic contribution of marine data assets within a capital-intensive shipping enterprise. The estimated valuation result (RMB 3.390 billion) does not aim to provide a definitive benchmark, but rather serves as a quantitative illustration of the valuation logic, demonstrating how data assets may represent a non-negligible component of enterprise value when their contribution is explicitly isolated and adjusted for risk and quality dimensions. Consistent with the methodological orientation of this study, the illustrative application is intended to demonstrate the internal logic and operational feasibility of the proposed framework, rather than to provide empirical generalization or hypothesis testing.

From a theoretical perspective, this study contributes to the literature on data asset valuation by addressing the long-standing attribution problem associated with synergistic returns in data-intensive enterprises. By introducing the logic of “total synergistic return–data contribution separation” within a multi-period framework, the proposed model extends existing income-based approaches that typically rely on static assumptions or residual allocations. The framework highlights the role of marine data assets as value-enabling capital that interacts nonlinearly with physical assets and organizational resources over time, thereby offering a structured analytical perspective for understanding data-driven value creation under uncertainty in the marine economy. Methodologically, the study integrates dynamic discounting, a contribution coefficient (α), and expert-based correction factors to reconcile financial valuation with qualitative attributes such as data quality, application value, and risk. Rather than claiming universal applicability, the framework is designed to be adaptable to marine enterprises with comparable operational structures and digitalization levels, providing a transparent and modular valuation logic that can be recalibrated as data governance regimes, technologies, and market conditions evolve.

Several limitations of this study should be acknowledged. First, the empirical application relies on a single illustrative firm, which constrains the generalizability of the numerical results across heterogeneous marine industries. Second, key parameters, including the contribution coefficient (α) and the discount rate, are partially informed by expert judgment and industry benchmarking, which may introduce a degree of subjectivity, although structured elicitation procedures and sensitivity analyses are employed to mitigate this concern. Third, the framework does not explicitly model extreme exogenous shocks, such as geopolitical disruptions or sudden regulatory changes, which may affect short-term data value realization.

Despite these limitations, the study offers meaningful implications for research and practice. Future research may extend the framework by incorporating real-time data sources (e.g., satellite remote sensing and Internet of Things data) and data-driven learning techniques to support dynamic parameter updating. In particular, by explicitly incorporating multi-period uncertainty, risk adjustment, and qualitative dimensions, the framework aligns data asset valuation with the long-term sustainability objectives of the blue economy, where environmental, operational, and regulatory considerations jointly shape value creation. In this sense, the framework supports a transition from viewing marine data as passive operational by-products toward recognizing them as strategic and value-relevant assets within the broader agenda of sustainable blue economic development.