Study on the Role of Blockchain Technology in Supply Chain Carbon Emission Transparency Enhancement and Financial Regulation

Currently, the world is facing the serious challenge of climate change, and carbon emission reduction has become an urgent task. As a link between upstream and downstream enterprises, the supply chain has huge potential to reduce emissions. However, the supply chain also faces problems such as carbon leakage and an imbalance of the input-output ratio in the process of emission reduction. Blockchain technology provides a new idea for realizing efficient and collaborative emission reduction by all parties in the supply chain.

To achieve the goal of carbon peak and carbon neutrality, the key lies in the joint efforts of all sectors to reduce emissions, among which the supply chain, as an important link connecting upstream and downstream enterprises, aggregates a large number of production and trade activities, which is an important part of carbon emissions. The proposal to build a new model of “blockchain + supply chain + emission reduction” is of great significance to the early realization of the goal of “double carbon” by promoting emission reduction cooperation among all parties in the supply chain through leveraging the advantages of connectivity and trustworthiness of blockchain technology. As a new type of decentralized protocol, blockchain relies on multiple independent nodes to jointly maintain its distributed database system, which can securely store all kinds of data under the transaction of bitcoin, with shared and highly secure data, non-falsifiable and non-tamperable information, and automatic execution of smart contracts [1–3]. Because of the above characteristics, the use of blockchain can effectively improve the degree of trust between the two parties to the transaction and avoid the risk of inter-enterprise transactions, thereby reducing the complexity and cost of the transaction. At the same time, blockchain is very attractive in promoting inter-enterprise transactions and cooperation, providing an emerging solution for realizing product information transparency, traceability, and quality assurance [4–5]. In addition, the establishment of a supply chain finance regulatory system under blockchain technology and the formation of a long-term mechanism in the field of financial regulation are not intended to inhibit supply chain financial services and financial innovation but rather to provide a guarantee for them, so that the financial innovation services can walk more safely and soundly [6–7].

The use of blockchain technology for emission reduction cooperation at the supply chain level can leverage inter-firm synergies and achieve 1+1>2 synergies. Bai, C. et al. emphasized that the uncertainty of supply chain organizations and the need for sustainable transparency condition the selection of blockchain technology and provided supply chain managers with the ability to evaluate and select the most suitable by establishing blockchain technology performance indicators and proposing a hybrid swarm decision-making methodology for decision-making tool for blockchain technology [8]. El Hathat, Z. et al. investigated the storage of greenhouse gas emissions in the peanut supply chain and constructed a machine learning-based predictive model to achieve the sustainability goals, proposing a new approach of blockchain-enabled off-chain machine learning aimed at enhancing the data transparency and security of the process [9]. Chandan, A. et al. illustrated the advantages played by food systems designed with blockchain technology support for supply chain management. Blockchain technology not only enables traceability transparency and reduces the environmental impact in the food supply chain but also contributes to the achievement of sustainable development goals [10]. Guo, S. et al. investigated the information disclosure game in fashion supply chain environmental endeavors, and the use of blockchain technology can declare the environmental quality of fashion products to consumers and achieve transparent and open supply chain management, which will have an important impact on the sustainable practices of the entire fashion supply chain [11]. Wei, G. explores the implications and potential benefits of supply chain management based on blockchain technology, providing valuable insights into the advantages and challenges of blockchain technology by providing an in-depth discussion on traceability and transparency, carbon footprint reduction and emissions tracking [12].

Blockchain technology has a wide range of application scenarios in the financial sector, and blockchain supply chain finance is one of the more mature application areas at present. Min, H. argues that in an era of increased risk and uncertainty, utilizing the properties that blockchain technology possesses to mitigate the risks associated with intermediary interventions in financial services is of great importance in enhancing supply chain resilience [13]. Sun, W. pointed out that supply chain finance faces challenges such as credit identification while providing flexible financial products and services for related enterprises, and the innovative application of blockchain technology in supply chain finance is conducive to solving the information asymmetry and credit crisis in the chain of financial transactions and provides strong vitality for digital financial development such as [14]. Liu, Z. showed that supply chain finance is a financial service established in the environment of linking multiple parties and relatively closed, while blockchain technology has the characteristics of decentralization, immutability, etc., combining the two, providing controllable and trustworthy data as a supporting factor for supply chain finance, and providing new solutions for the problems existing in its development [15]. Chen, Y. H. et al. designed a financial management mechanism for supply chains that can check corporate credit as well as manage the loan book, which is studied from three aspects: credit information preservation and supervision, post-loan management and financial supervision chain [16]. Rijanto, A. explored the application of blockchain technology in supply chain finance based on a technology acceptance model combined with a multi-case approach, and experiments showed that blockchain technology, with its smart contract, transparent and secure distributed ledger data characteristics, provides solutions to automation problems faced by supply chain finance [17]. Zhao, H. et al. investigated the strategic choices of supply chain finance in an uncertain environment by constructing a game model; in a non-blockchain environment, the supply chain dominated by core firms will increase the price of the products due to the fraud problem, whereas the strategic choices of the firms in blockchain environment are divided into two scenarios in which third-party service model or platform model is the equilibrium strategy [18].

The study first proposes that blockchain technology can play a role in reducing the credit risk of enterprise supply chain finance, improving transparency of supply chain carbon emissions and financial regulation. Accordingly, this paper uses factor analysis and logistic regression analysis to establish an initial model for assessing supply chain financial credit risk. 105 sample enterprises’ relevant data is used to construct the final assessment model. Z enterprise is used as an example to verify the effectiveness of blockchain technology by comparing the calculated risk-bearing capacity value with the standard value of the risk-bearing capacity of enterprises without the application of blockchain technology.

2

Application of blockchain technology in supply chain management

2.1

Application of Blockchain Technology in Supply Chain Carbon Emissions

In the supply chain, there are many types of relationships, such as games, cooperation, and competition between different subjects, which leads to a problem of trust. The trust problem exists not only between the government and enterprises, but also between enterprises, among enterprises, and between consumers and enterprises. The authenticity and validity of carbon emissions is the main indicator of the trust issue between government and enterprises. The government needs to ensure the compliance and honesty of enterprises’ carbon emission behaviors through regulation and incentive mechanisms, while enterprises need to prove their carbon emission results and contributions through information disclosure and third-party verification. The trust problem between enterprises is mainly reflected in transaction benefits and risk sharing. Enterprises need to maximize their interests while considering the interests and risks of their supply chain partners and establish a fair, reasonable, and transparent transaction relationship to avoid problems such as moral hazard and adverse selection. The green level and environmental attributes of products are the main indicators of the trust issue between consumers and enterprises. Consumers need to judge whether the products of enterprises meet their green needs and preferences through product labeling, certification, evaluation, etc., and enterprises need to win the trust and recognition of consumers by improving the green quality and performance of their products, as well as providing reliable environmental information. Therefore, trust is an important factor that affects transparency in supply chain carbon emission enhancement, which needs to be established and maintained through a variety of ways and means. [19]

The decentralized, tamper-proof, transparent, and trustworthy features of blockchain technology can effectively facilitate information sharing among subjects, solve the trust crisis among subjects, and have an important impact on supply chain carbon emissions. The application of blockchain technology to the monitoring, accounting and trading of carbon emissions, the production, transmission, consumers and certification of green power, and the affirmation, circulation and traceability of carbon sinks can provide reliable data support and incentives for supply chain carbon emissions. It further improves the credibility and transparency of supply chain carbon emission data, thus enhancing market competitiveness, improving brand image, and promoting green financial innovation. Therefore, combining blockchain technology with supply chain carbon emissions can effectively enhance corporate supply chain financial credit.

2.2

Application of Blockchain Technology in Supply Chain Finance Regulation

Blockchain is a distributed ledger that combines blocks of data in chronological order sequentially linked into a chained data structure that is cryptographically guaranteed to be untamperable and unforgerable. A blockchain is a technical solution that allows multiple participating nodes to maintain a reliable database through decentralization. All data stored in the blockchain will be time-stamped with a time stamp that accurately records the time of occurrence of each transaction activity, which can effectively reflect the sequence between each transaction activity, thus enabling all transaction activities to be traceable. Based on the traceability of blockchain, it can provide an effective means of supervision for supply chain finance regulators. The blockchain stores all the electronic accounts receivable note financing and payment transaction information, as well as transaction time. The regulatory authorities can use big data technology to analyze the massive transaction information as well as mining to realize the risk of pre-warning, reducing the enterprise supply chain finance credit risk, but also can provide strong evidence support for the risk of post-disposal [20].

3

Credit Risk Assessment Model for Supply Chain Finance

The analysis above shows that the use of blockchain technology for supply chain carbon emission transparency improvement and financial regulation can effectively reduce enterprise supply chain financial credit risk. The method of constructing a supply chain financial credit risk assessment model is introduced in this chapter and the initial model is constructed.

3.1

Factor analysis

Multivariate statistical analysis deals with multivariate problems due to the presence of a large number of variables, leading to an increase in the complexity of the analysis problem. Generally, in practical use, there is a certain correlation between the variables. It’s possible for multiple variables of information to overlap. In order to reduce the overlap of data, it is necessary to put forward the main factors affecting the variables to achieve the least number of variables to replace the original variables, which is a kind of “dimensionality reduction” process [21]. Factor analysis is a statistical method that reduces the dimensionality and simplifies the data.

3.1.1

Factor analysis model

The common factors in factor analysis are influences that are difficult to observe and access directly, and each variable can be represented as a linear function and a special factor after the common factor, i.e: 1 $X_{i} = A_{i 1} F_{1} + a_{i 2} F_{2} + \dots A_{i n} F m + ε_{i} (i = 1, 2, \dots p)$ Where: F₁,F₂⋯F_m is called the common factor. ε_i is called the special factor of X_i and the model is expressed as: 2 $X = A F + ε$ 3 $X = [\begin{matrix} X_{1} \\ X_{2} \\ ⋮ \\ X_{p} \end{matrix}], A = [\begin{matrix} a_{11} & a_{12} & \dots & a_{1 m} \\ a_{21} & a_{22} & \dots & a_{2 m} \\ \dots & \dots & \dots & \dots \\ a_{p 1} & a_{p 2} & \dots & a_{p 3} \end{matrix}], F = [\begin{matrix} X_{1} \\ X_{2} \\ ⋮ \\ X_{m} \end{matrix}], ε = [\begin{matrix} ε_{1} \\ ε_{2} \\ ⋮ \\ ε_{p} \end{matrix}]$

The following conditions also need to be met: 1)

m ≤ p

2)

Cov(F,ε) = 0 , i.e., there is no correlation between the male and special factors.

3)

$D F = D (F) = [\begin{array}{l} 1 & 0 \\ 1 \\ ⋱ \\ 0 & 1 \end{array}] = I m$ , the variables are uncorrelated and satisfy variance equal to 1.

4)

$D ε = D (ε) = [\begin{array}{l} σ_{1}^{2} & 0 \\ σ_{2}^{2} \\ ⋱ \\ 0 & σ^{2} \end{array}]$ , the special factors are not correlated with each other and the variances need not be equal.

The matrix A in the factor analysis model is called the “factor loadings” matrix, and a_ij is called the factor “loadings”, which are the loadings of the i th variable on the j th factor.

3.1.2

Solving for factor analysis

There are many ways to solve the factor matrix, and only the most commonly used method of principal component analysis is introduced here. The steps for solving using principal component analysis are as follows: 1)

Calculate the covariance matrix Σ of the original data.

2)

Calculate the eigenroot of the covariance array Σ by substituting it into λ₁ ≥ ⋯ ≥ λ_p ≥ 0, and the corresponding eigenvector becomes T1, T2, ⋯ T_p.

3)

The factor loading array is calculated using the eigenroots and eigenvectors of the covariance array: 4 $A = (\sqrt{λ_{1}} T_{1}, \sqrt{λ_{2}} T_{2}, \dots, \sqrt{λ_{p}} T_{p})$

The process of factor analysis is to reduce the number of original variables, and in general the number of factors m should be smaller than the number of original variables p. Therefore, only the first m eigenroots and eigenvectors are extracted to represent the factor loading array in the final data analysis: 5 $A = (\sqrt{λ_{1}} T_{1}, \sqrt{λ_{2}} T_{2},\dots, \sqrt{λ_{m} T_{m}})$

3.1.3

Statistical significance of factor analysis

1)

Factor loadings

For the factor model X = A_i1F_i + a₂F₂ + ⋯ A_inF_m + ε_i (i = 1,2,…p) we have that the covariance of X_i with F_j is: 6 $C o v (X i, F j) = C o v (\sum_{i = 1}^{i} a_{i k} F k + ε_{i}, F j) = a_{i j}$

2)

Factor naming and factor rotation

Factor models that do not explicitly explain the actual meaning of the factors usually require factor rotation, which results in the original variables having a common factor with a large loading. Each common factor (i.e., each column of the loading matrix) has a larger loading on some of the original variables but has a smaller loading on other variables, which is expressed as a bifurcation on each column of the loading matrix, either close to 1 or close to 0. This bifurcation shows the connection between each common factor and the original variables, and in this way, the actual meaning of the factors can be reasonably interpreted.

Generally, there are two types of factor rotation: orthogonal and skewed rotations, and this paper focuses on orthogonal rotation. The orthogonal rotation of a factor is the orthogonal transformation of the loading matrix A, and the right-hand side is multiplied by the orthogonal transformation matrix to obtain the rotated factor loading matrix.

Orthogonal rotation is generally used in practice as a maximum variance rotation method, where the variance of each column resulting from a maximized rotation of the factor matrix is summed so that the loadings on the same columns are close to 1 or close to 0. It should be noted here that the rotated factor series changes the variance contribution g₂ of each factor compared to the loadings matrix before rotation and is no longer equal to the j th largest eigenroot of the sample covariance matrix, but rather all m factors are extracted.

Factor scores are the definitive reflection of the outcomes of factor analysis. Factor scores are specific values that can be calculated for the original variables on each sample after determining the array of factor loadings. After obtaining the factor scores, principal component analysis and other methods can be used instead of the original variables to reduce dimensionality.

3.2

Logistic regression analysis

The logistic model assumes that the probability of company compliance obeys a logistic distribution, and a series of indicators (Xk, k = n) in the supply chain finance risk assessment system empowered by blockchain technology is used as the independent variable to establish a logistic regression model, which predicts the company’s risk-taking ability as P. According to the standard P* (60%) set by the risk-taking ability of enterprises in the supply chain [22], P is compared with P* for comparison to determine whether blockchain technology can reduce corporate supply chain finance credit risk.

The logistic model is represented as follows:

The logistic model is expressed as follows: let there exist K factors affecting the value of Y, denoted X₁X₂X₃ ⋯ X_k, respectively: 7 $L N \frac{P}{1 - P} = β_{0} + β_{1} X_{1} + \dots β_{k} X_{k}$ Where: β₀, β₁ ⋯ β_k is the unknown parameter to be estimated, which can be obtained by organizing the above equation: 8 $\frac{P}{1 - P} = e^{β_{0} + β_{1} X_{1} + \dots β_{k} X_{k}}$

The above equation can be organized to obtain the formula for calculating the risk-taking capacity P of the enterprise as: 9 $P = \frac{1}{1 + e^{- (β_{0} + β_{1} X_{1} + \dots β_{k} X_{k})}}, i = 1, 2, 3 \dots k$

A logistic function is an increasing function that takes values in the range of (0, 1). The closer the calculated P-value is to 1, the better the company’s credit is, and vice versa, the worse the credit is.

4

Empirical analysis

4.1

Sample Selection and Data Sources

In terms of time selection, since the rise of the blockchain concept, the degree of its application has shown a slow increase in the first few years. However, since the end of 2015, the degree of its application has shown an exponential growth trend and reached its peak at the end of 2017. This paper therefore selects the timeframe of data as 2017-2023.

In terms of industry selection, due to the obvious organizational structure characteristics of the manufacturing industry, the clear division of labor and mature development of upstream and downstream enterprises in the supply chain, and the high level of carbon emission and financial regulation in the supply chain, this paper selects small and medium-sized enterprises (SMEs) in the manufacturing industry as the object of study.

In terms of sample selection, this paper chooses the manufacturing enterprises of GEM as the object of SME research. For the following two considerations, on the one hand, the SME board is the predecessor of GEM. On the other hand, for the consideration of data availability. In summary, this paper selects 2017-2023 as the initial research object for the SICI industry classification of manufacturing enterprises on the Shenzhen Exchange Gem Board. Financial data is derived from the wind database and word frequency data from the company’s annual report are analyzed through text analysis. After screening, a total of 105 sample enterprises were obtained.

4.2

Blockchain-based Credit Risk Indicator System for Supply Chain Finance

As an emerging technology, blockchain technology, with its decentralization, distributed ledger, smart contract, anti-tampering, traceability and other characteristics applied to supply chain carbon emission and financial regulation, has a very good effect on supply chain financial risk control. Therefore, when choosing the qualitative indicators of the credit risk evaluation system of supply chain finance empowered by blockchain technology, it will also be based on these characteristics of blockchain and the opinions of experts in the industry. The constructed credit risk evaluation index system of supply chain finance empowered by blockchain technology is shown in Table 1. The first-level indicators are set as supply chain carbon emission transparency, supply chain finance regulation level, and blockchain technology utility. The total number of indicators is 9 second-level indicators and 23 third-level indicators.

Table 1.

Supply chain financial credit risk evaluation system based on block-chain

Primary index	Secondary index	Tertiary index
Supply chain carbon transparency	Carbon data accuracy	Data collection frequency (X1)
	Carbon data accuracy	Data verification mechanism validity (X2)
	The degree of disclosure of carbon emissions	The timeliness of disclosure (X3)
		The comprehensiveness of the disclosure (X4)
		Standardization of disclosure formats (X5)
	The setting and achievement of carbon targets	Clarity of target reduction (X6)
	The setting and achievement of carbon targets	Effectiveness of mitigation measures (X7)
Supply chain financial supervision level	The perfection of the regulatory framework	Coverage of laws and regulations (X8)
		Authority of regulators (X9)
		The timeliness of policy updates (X10)
	The stricture of regulatory execution	Frequency of compliance checks (X11)
	The stricture of regulatory execution	The severity of the punishment (X12)
	The degree of application of regulatory technology	Data analysis capability (X13)
	The degree of application of regulatory technology	Information sharing mechanism (X14)
Block chain technology utility	The node enterprise is decentralized	Point to point trading (X15)
		Distributed ledger (X16)
		Information sharing degree (X17)
	Business function	Intelligent payment (X18)
		Prudential survey (X19)
		Traceability (X20)
	Business mechanism security	Enterprise information completeness (X21)
		Trade consensus mechanism (X22)
		Encryption security (X23)

4.3

Factor analysis

The weighted average results of the evaluation scores of 23 indicators for 105 sample enterprises from 2017 to 2023 are compiled using the constructed credit risk evaluation index system. The summarized data are subjected to the KMO and Bartlett sphericity test, and the test results are shown in Table 2. The significance result of the Bartlett sphericity test is much lower than 0.05, indicating that these variables have passed the pre-procedure of factor analysis. The result of the KMO sampling suitability measure is 0.659, which is greater than 0.6, suggesting that the correlation between variables is strong.

Table 2.

KMO and Bartlett spherical test results

KMO sampling availability number		0.659
Bartlett ball test	Approximate Chi-square	1126.524
	df	314
	Sig.	0.000

In this paper, principal component analysis was used to extract the common factors, and the results of the principal component analysis are shown in Table 3. The cumulative contribution rate of the seven principal component variables extracted with initial eigenvalues greater than 1 is 75.679%, which is greater than 60%. The above results indicate that all seven principal component variables extracted by principal component analysis have a small loss rate and meet the requirements of Logistic regression analysis.

Table 3.

Total variance explanation

Constituent	Initial eigenvalue			Extracting the load of the load			Rotational load squared
Constituent	Total	Var%	Cum%	Total	Var%	Cum%	Total	Var%	Cum%
1	4.07	26.913	26.913	4.07	26.913	26.913	4.058	25.096	25.096
2	3.935	15.173	42.086	3.935	15.173	42.086	3.947	14.258	39.354
3	2.74	10.548	52.634	2.74	10.548	52.634	2.722	10.879	50.233
4	2.028	7.808	60.442	2.028	7.808	60.442	2.029	8.517	58.75
5	1.596	6.109	66.551	1.596	6.109	66.551	1.6	6.458	65.208
6	1.336	5.155	71.706	1.336	5.155	71.706	1.328	5.982	71.19
7	1.036	3.973	75.679	1.036	3.973	75.679	1.028	4.489	75.679
8	0.856	3.481	79.16
9	0.819	3.165	82.325
10	0.677	2.59	84.915
11	0.629	2.436	87.351
12	0.541	2.03	89.381
13	0.477	1.819	91.2
14	0.386	1.512	92.712
15	0.364	1.409	94.121
16	0.338	1.335	95.456
17	0.263	1.089	96.545
18	0.223	0.937	97.482
19	0.185	0.696	98.178
20	0.162	0.626	98.804
21	0.151	0.523	99.327
22	0.101	0.398	99.725
23	0.087	0.275	100

The matrix of component score coefficients is shown in Table 4. Each of the seven principal component variables extracted with eigenvalues greater than 1 is named F1, F2, F3, F4, F5, F6, and F7. The specific loading values of each indicator on the seven principal component variables can be seen in Table 4. For example, the data collection frequency (X1) has loading values of -0.484, 0.721, 0.207, -0.135, 0.056, -0.062, and 0.024 on each of the 7 principal component variables.

Table 4.

Component score coefficient matrix

Index	Constituent
Index	1	2	3	4	5	6	7
X1	-0.484	0.721	0.207	-0.135	0.056	-0.062	0.024
X2	-0.454	0.53	0.46	-0.211	-0.301	0.181	-0.062
X3	-0.305	0.72	0.09	-0.015	-0.067	0.232	0.042
X4	0.188	0.4	-0.314	0.015	-0.38	0.446	0.233
X5	-0.06	-0.498	0.694	0.03	-0.188	0.109	0.088
X6	0.209	0.499	-0.028	0.167	0.459	0.378	0.209
X7	-0.333	0.656	0.079	-0.107	0.106	-0.24	0.232
X8	-0.403	0.389	0.29	0.721	0.035	-0.123	0.009
X9	-0.086	0.186	0.15	0.917	0.082	-0.035	-0.077
X10	-0.456	0.506	0.466	-0.193	-0.254	0.238	-0.098
X11	0.191	-0.048	0.39	-0.148	0.528	0.261	-0.114
X12	-0.262	0.077	0.211	-0.237	0.742	0.091	-0.198
X13	0.007	-0.492	0.69	-0.132	0.117	0.073	0.274
X14	-0.246	-0.066	-0.136	0.049	0.221	-0.296	0.752
X15	-0.011	-0.601	0.364	0.298	-0.192	0.339	0.127
X16	0.216	-0.021	0.638	-0.192	-0.069	-0.323	0.182
X17	-0.032	0.289	0.33	-0.077	-0.141	-0.556	-0.142
X18	0.753	0.175	-0.058	-0.312	-0.009	-0.007	0.044
X19	0.77	0.328	-0.056	0.027	-0.025	-0.005	0.11
X20	0.817	0.172	-0.072	0.042	-0.134	0.074	-0.032
X21	0.779	0.061	0.161	-0.319	-0.082	0.083	0.183
X22	0.811	0.185	0.216	0.198	0.02	0.115	0.138
X23	0.819	0.187	0.121	-0.01	-0.001	-0.152	-0.168

When using IBM SPSS 28.0 software for principal component analysis, the rotation method was set to the maximum variance method, and the maximum number of convergence iterations was set to 22, and then the orthogonal rotation was performed on the component score coefficient matrix in Table 4, and the rotated component matrix that reached the convergence condition through 7 iterations is shown in Table 5. F1~F7 are named as blockchain application and innovation, data security and transparency, credit risk management, financial regulation intensity, supply chain operation efficiency, supply chain finance credit level, and market and industry environment adaptability, respectively. Table 5 clearly distinguishes the loading values of each factor for each indicator. For example, the loading values of the blockchain application and innovation factor on each indicator are 0.889, 0.868, 0.83, 0.826, 0.814, 0.785, 0.784, 0.781, and 0.746, respectively.

Table 5.

Rotated component matrix

Index	Constituent
Index	1	2	3	4	5	6	7
X18	0.889	-0.03	0.143	0.126	0.017	-0.134	0.051
X15	0.868	-0.166	0.06	0.133	0.007	0.065	-0.022
X23	0.83	-0.138	-0.054	-0.028	0.09	0.088	-0.119
X19	0.826	-0.112	-0.086	-0.037	0.035	0.185	-0.186
X16	0.814	-0.033	-0.167	-0.039	-0.111	-0.081	0.06
X17	0.785	-0.12	-0.096	-0.086	-0.159	-0.095	-0.098
X21	0.784	-0.131	0.08	0.15	0.049	0.097	-0.108
X20	0.781	-0.024	0.137	-0.378	-0.001	-0.061	0.036
X22	0.746	-0.056	-0.145	-0.37	0.004	-0.015	0.018
X5	-0.173	0.869	0.13	0.006	-0.016	0.087	-0.177
X7	-0.208	0.862	0.176	0.03	0.011	0.056	-0.192
X4	-0.193	0.812	-0.258	0.122	0.14	0.161	0.145
X6	-0.046	0.751	-0.246	0.182	0.018	-0.135	0.057
X14	-0.097	0.607	-0.325	0.1	0.09	0.217	0.382
X10	-0.062	-0.019	0.871	0.036	0.029	0.157	-0.129
X9	-0.013	-0.052	0.82	-0.128	0.287	0.151	0.143
X12	-0.103	-0.266	0.768	0.169	-0.109	-0.226	-0.101
X11	0.053	-0.004	0.026	0.961	-0.049	-0.03	-0.031
X13	-0.152	0.375	-0.003	0.853	-0.01	0.098	0.086
X2	-0.214	0.134	-0.059	-0.037	0.858	0.063	-0.004
X1	0.207	0.012	0.242	-0.065	0.684	-0.077	-0.075
X3	0.125	0.306	-0.056	0.108	-0.072	0.682	0.004
X8	0.251	0.319	-0.181	-0.094	-0.415	-0.539	0.251

4.4

Logistic regression analysis

In this study, logistic regression was performed using SPSS software, and the regression results are shown in Table 6, with ***, **, and * denoting significance at the 1%, 5%, and 10% levels, respectively. The empirical results show that blockchain application and innovation (F1) has a negative correlation on the probability of default of enterprise credit risk, and in the logistic regression model, the β coefficient is -5.594, i.e., the larger the blockchain application and innovation factor is, the smaller the probability of default of the enterprise is, which corroborates the previous inference of blockchain technology can reduce the credit risk of SMEs’ supply chain finance.

Table 6.

Logistic regression

	β	Standard error	Sig.	Exp(β)
F1	-5.594	1.226	0.001***	0.011
F2	1.114	0.345	0.003***	0.318
F3	0.162	0.421	0.635	1.163
F4	2.368	0.518	0.003***	0.116
F5	-0.137	0.288	0.001*	0.987
F6	-1.839	0.528	0.005*	0.158
F7	2.603	0.46	0.000	13.553
Constants	5.366	0.261	0.005***	1.441

The results of the logistic regression show that only F1, F2, and F4 of the seven principal component variables are significant <0.05 with a constant of 5.366, so the regression equation of the logistic model can be expressed as follows: 10 $L N \frac{P}{1 - P} = 5.366 - 5.594 F_{1} + 1.114 F_{2} + 3.368 F_{4}$

Organized: 11 $P = 1 - \frac{1}{1 + e^{5.366 - 5.594 F_{1} + 1.114 F_{2} + 3.368 F_{4}}}$

In order to further ensure the reliability of the model results, this paper conducts the Omnibus test for the established model. The Omnibus test for the table model coefficients is shown in Table 7. It can be seen that all three p-values are less than the critical value of 0.05. Therefore, the results of the model parameters in this paper are reliable, and the model is statistically significant.

Table 7.

The Omnibus test of the table model coefficient

		Chi-square	df
Step 1	Step	20.455	16
	Block	20.455	16
	Model	20.455	16

4.5

Example Validation Based on Logistic Modeling

The study assesses the credit risk of supply chain finance for Enterprise Z, which uses blockchain technology to manage carbon emissions and regulate financial regulation in the supply chain.

According to the collected data and the expert scoring data obtained through the questionnaire survey, the various indicators of Enterprise Z are shown in Table 8. The expert scoring adopts a “10-point system”, in which the frequency of compliance inspection, the strictness of violation penalties, the ability of data analysis, smart payment, and the completeness of enterprise information are all scored as 10 points.

Table 8.

Enterprise indicators

Primary index	Secondary index	Tertiary index	Index data
Supply chain carbon transparency	Carbon data accuracy	Data collection frequency (X1)	7
	Carbon data accuracy	Data verification mechanism validity (X2)	9
	The degree of disclosure of carbon emissions	The timeliness of disclosure (X3)	9
		The comprehensiveness of the disclosure (X4)	8
		Standardization of disclosure formats (X5)	6
	The setting and achievement of carbon targets	Clarity of target reduction (X6)	7
	The setting and achievement of carbon targets	Effectiveness of mitigation measures (X7)	8
Supply chain financial supervision level	The perfection of the regulatory framework	Coverage of laws and regulations (X8)	9
		Authority of regulators (X9)	9
		The timeliness of policy updates (X10)	7
	The stricture of regulatory execution	Frequency of compliance checks (X11)	10
	The stricture of regulatory execution	The severity of the punishment (X12)	10
	The degree of application of regulatory technology	Data analysis capability (X13)	10
	The degree of application of regulatory technology	Information sharing mechanism (X14)	7
Block chain technology utility	The node enterprise is decentralized	Point to point trading (X15)	7
		Distributed ledger (X16)	9
		Information sharing degree (X17)	9
	Business function	Intelligent payment (X18)	10
		Prudential survey (X19)	7
		Traceability (X20)	9
	Business mechanism security	Enterprise information completeness (X21)	10
		Trade consensus mechanism (X22)	7
		Encryption security (X23)	8

Based on the model constructed in the previous section, the specific values of the company’s three principal component variables, F1, F2, and F4, were calculated, and the results were substituted into the regression equation: 12 $P = 1 - \frac{1}{1 + e^{5.366 - 5.594 F_{1} + 1.14 F_{2} + 3.368 F_{4}}} = 92.05 %$

The calculated risk-bearing capacity is 92.05%, compared with the standard P* (60%) set by the risk-bearing capacity of enterprises in the supply chain, indicating that the application of blockchain technology in supply chain carbon emission and financial regulation can play a role in reducing the credit risk of enterprise supply chain finance.

5

Conclusion

The application of blockchain technology in supply chain carbon emission and financial regulation can effectively improve the financial credit of enterprises in their supply chains. The study designed a supply chain finance credit risk assessment model to validate this conclusion. The blockchain application and innovation factor in the logistic regression model has a β coefficient of -5.594.

Indicating that blockchain technology can reduce the financial credit risk of SMEs in the supply chain. The risk-taking capacity of Enterprise Z, which applies blockchain technology to manage carbon emissions and financial regulation in the supply chain, is 92.05%, which is much higher than the standard value of the risk-taking capacity of enterprises without blockchain technology augmentation (60%). This paper’s inference is further validated by it.

Langue:: Anglais

Périodicité:: 1 fois par an
Sujets de la revue:: Sciences de la vie, Sciences de la vie, autres, Mathématiques, Mathématiques appliquées, Mathématiques générales, Physique, Physique, autres

RSS Feed de la revue

Study on the Role of Blockchain Technology in Supply Chain Carbon Emission Transparency Enhancement and Financial Regulation

Xingyao Zhou

Fujun Wan

Qiaohui Wang

Wenting Song

Publié en ligne: 03 févr. 2025

Reçu: 02 sept. 2024

Accepté: 25 déc. 2024

DOI: https://doi.org/10.2478/amns-2025-0013

Mots clésFactor analysis, Logistic regression analysis, Blockchain technology, Supply chain

© 2025 Xingyao Zhou et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Mots clés
Factor analysis, Logistic regression analysis, Blockchain technology, Supply chain