Research on Intelligent Methods and Strategies of Corporate Financial Risk Assessment in the Digital Economy Era

In the era of digital economy, financial risks do show new characteristics and forms, these risks not only from the traditional financial management field, but also by digital technology, big data, cloud computing, artificial intelligence and other advanced technologies. Enterprise financial risks exist in all aspects of production and operation, including financing, investment, capital withdrawal and profit distribution [1-4]. With the in-depth development of market economy and the increasing expansion of enterprise scale, the problem of enterprise financial risk has gradually come to the fore, becoming an important factor affecting the sound operation and sustainable development of the enterprise, so it is of great significance to assess the financial risk to the sustainable development of the enterprise [5-8].

The traditional financial risk assessment method mainly relies on manual judgment and experience, and this method has the problems of strong subjectivity and easy to be influenced by human factors. The financial risk assessment based on artificial intelligence realizes the automatic processing and analysis of massive information through the use of big data and machine learning algorithms, thus providing more accurate and comprehensive risk assessment results [9-12]. Intelligent financial risk assessment is a very important research field in modern enterprise management. With the continuous development of artificial intelligence technology, financial risk assessment through the use of artificial intelligence technology has become a trend. Intelligent financial risk assessment models, analyzes and predicts the financial data of an enterprise through the use of machine learning and natural language processing and other technologies, thus assessing and identifying financial risks and improving the effectiveness of financial decision-making [13-16].

Literature [17] proposes a fuzzy comprehensive evaluation algorithm in order to constitute the financial control evaluation index system within the enterprise, which is considered in terms of the control subject, content and other aspects. The case study shows that there is ambiguity in the internal financial control of the enterprise, it is an effective strategy to introduce fuzzy comprehensive evaluation method in the process of evaluating the efficiency of internal financial control[18] introduces the early warning method of enterprise financial risk based on DS-RF model. The results point out that the DS-RF model has a higher warning accuracy, which is conducive to the improvement of the efficiency of the enterprise financial risk early warning as well as the efficient and scientific decision-making of the enterprise managers. Literature [19] based on the theory of mobile payment and the principle of enterprise financial leverage, collected data through questionnaires, and introduced a multi-level evaluation and analysis method. The enterprise is analyzed to conclude that capital risk has the largest proportion of risk in financial management, and effective suggestions are put forward from multiple perspectives. Literature [20] mentions the risk assessment and risk control algorithm of small and medium-sized enterprises in the era of big data. And the GA-PSo algorithm is applied to the program design of enterprise’s risk control by constructing the risk assessment index system. The experimental results yielded that the risk control efficiency of this control algorithm is relatively high, and the accuracy of its risk assessment is even as high as more than 95%, which shows good performance. Literature [21] created data mining algorithms to improve the enterprise financial accounting data processing problems. The financial information cloud platform was designed by using Internet technology, and the financial risk indicator coefficients of enterprises were judged by association rules. Comparison experiments based on the st classification standard show that data mining technology effectively improves the processing efficiency of massive accounting data information.

This paper focuses on the intelligent assessment scheme of enterprise financial risk, comprehensively consider the solvency, profitability, operational efficiency, growth potential and cash flow status, and construct the financial risk assessment framework of enterprises. Aiming at the shortcomings of AHP in calculating the weights of indicators, this paper adopts the method of combining hierarchical analysis and decision laboratory analysis to assign weights to the first and second level indicators of the assessment system. Comprehensively consider the solvency, profitability, operational efficiency, growth potential and cash flow status, and construct the financial risk assessment framework of enterprises. The cloud model is introduced to integrate and analyze a large amount of fuzzy and random information in the enterprise financial risk, reflecting the uncertainty of financial information by calculating the cloud expectation, cloud entropy and cloud super entropy in the cloud model, and realizing the intelligent assessment of enterprise finance. Company G, a large-listed private science and technology enterprise, is selected as the case study object of this paper, and the overall overview and financial status of Company G are briefly described. A number of experts are invited to assess the financial risk of Company G according to the established enterprise financial risk assessment system, and the evaluation indexes are empowered using the obtained data combined with the AHP and DEMATEL methods. Calculate the cloud number characteristics of each evaluation index of Company G. According to the calculation results, the financial risk status of Company G is reasonably assessed and the corresponding financial risk control measures are provided.

2

Enterprise financial risk assessment model

2.1

Design of the assessment system and selection of indicators

This paper conducts financial risk assessment of enterprises from five dimensions: profitability, solvency, operating ability, growth ability, and cash ability [22], and the overall evaluation index system is shown in Table 1. In evaluating the profitability of enterprises, we mainly use two key indicators: gross profit margin on sales and net profit margin on assets. For solvency considerations, we use the equity ratio, current ratio, quick ratio, and guaranteed interest rate ratio to make a comprehensive assessment. In terms of measuring the operating efficiency of enterprises, we choose the total assets turnover, current assets turnover, and business cycle as the core indicators. In order to evaluate the growth potential of an enterprise, we refer to the growth rate of total assets and the growth rate of main business revenue. As for the assessment of cash capacity, we rely on indicators such as the proportion of cash from sales, the proportion of cash from main business income, the proportion of cash from net profit and the proportion of cash dividend protection. It should be noted that, in addition to the equity ratio and business cycle are regarded as positive indicators, the rest of the indicators are regarded as negative indicators.

Table 1.

Financial risk assessment indicator system

Primary indicator	Secondary indicator
Profitability(A1)	Gross margin(A11)
Profitability(A1)	Net profit rate(A12)
Solvency(A2)	Equity ratio(A21)
	Mobility ratio(A22)
	Speed ratio(A23)
	Overspeed ratio(A24)
	Interest coverage ratio(A25)
Operational ability(A3)	Total asset turnover ratio(A31)
	Current asset turnover ratio(A32)
	Business cycle(A33)
Growth ability(A4)	Total asset growth ratio(A41)
Growth ability(A4)	Main business revenue growth ratio(A42)
Cash ability(A5)	Sales cash ratio(A51)
	Main business income cash content(A52)
	Net profit cash content(A53)
	Cash dividend guarantee multiple(A54)

2.2

Composite indicator weights

Analytic hierarchy Process (AHP) is a decision-making method for difficult quantification problems [23], the principle of operation is to disassemble a complex system with multiple goals and guiding principles in a hierarchical manner until it is refined to a single indicator. Then, by comparing the importance of the peer indicators, we determine their respective weights. However, it should be noted that when AHP is used for weight calculation, it is assumed that there is no correlation among the influencing factors, which ignores the possible mutual influence in practice. In contrast, decision lab analysis allows for more comprehensive consideration of the interaction between these factors[24]. This paper combines AHP and DEMATEL analysis methods. First, the AHP method is used to clarify the basic weight ranking of each index. Then, the DEMATEL method is used to analyze the interaction between various factors, and the comprehensive influence weight of each index is adjusted and calculated accordingly, which greatly improves the accuracy, rationality and fairness of weight distribution.

2.2.1

Initial weight calculation

At the same time, according to the established evaluation criteria, the relative importance of the indicators at the same level is compared in pairwise, and then the judgment matrix is constructed. We assume that the resulting judgment matrix is as follows A = (a_ij)_n×n: (1) $A = {(\begin{matrix} a_{i j} \end{matrix})}_{n \times n} = (\begin{matrix} a_{11} & a_{12} & \dots & a_{1 n} \\ a_{21} & a_{22} & \dots & a_{2 n} \\ \dots & \dots & \dots & \dots \\ a_{n 1} & a_{n 2} & \dots & a_{n n} \end{matrix})$

The judgment matrix A is regularized, see equation (1), to obtain the regularized judgment matrix B = (b_ij)_n×n: (2) $b_{i j} = \frac{a_{i j}}{\sum_{i = 1}^{n} a_{i j}} (i, j = 1, 2, \dots, n)$

In order to ensure the logical rigor of indicators at all levels, the reliability of the constructed judgment matrix needs to be evaluated closely, and then the consistency ratio is obtained to verify its accuracy: (3) $C R = \frac{C I}{R I}$

Among them is the consistency indicator: (4) $C I = \frac{λ_{\max} - n}{n - 1}$

where λ_max is the maximum eigenvalue of the judgment matrix; R₁ is the stochastic consistency index, which is a constant and varies with n .

When CR < 0.1 , it passes the reliability analysis. The judgment matrix A is regularized to obtain the regularized judgment matrix B , and the initial weights of each index can be obtained by assuming that B = (b_ij)_n×n is normalized by the sum method on the regularized matrix B : (5) $X_{i} = \frac{{\bar{X}}_{i}}{\sum_{j = 1}^{n} X_{i}} (i, j = 1, 2, \dots, n)$

In the formula, ${\bar{X}}_{i} = \sum_{j = 1}^{n} b_{i j} (i, j = 1, 2, \dots, n)$

2.2.2

Calculation of the degree of impact

Suppose that the relationship between the factors is divided into five levels: none, weak, average, strong, and extremely strong, and the corresponding scores are 0, 1, 2, 3, and 4 points respectively. By evaluating and scoring the influence among different indicators, experts construct the direct influence matrix and the hypothetical judgment matrix N = (n_ij)_n×n by evaluating and scoring the influence relationship that exists between the various indicators: (6) $N = {(n_{i j})}_{n \times n} = (\begin{matrix} 0 & n_{12} & \dots & n_{1 n} \\ n_{21} & 0 & \dots & n_{2 n} \\ \dots & \dots & \dots & \dots \\ n_{n 1} & n_{n 2} & \dots & 0 \end{matrix})$

The judgment matrix N is normalized to obtain its canonical direct impact matrix O , which is further processed to obtain the combined impact matrix T = (t_ij)_n×n : (7) $T = O \cdot {(I - O)}^{- 1}$

Where O is the canonical direct impact matrix of matrix N ; I is the unit matrix.

According to the comprehensive impact matrix T calculates the value of each influencing factor and normalizes it to obtain the weight of each indicator’s impact degree: (8) $Y_{i} = \frac{U_{i}}{\sum_{i = 1}^{n} U_{i}}$

where centrality U = D+C , D is the degree of influence, $D = \sum_{j} = 1^{n} t_{i j}$ , C is the degree of being influenced, and $C = \sum_{i = 1}^{n} t_{i j}$ .

2.2.3

Calculation of composite indicators

Using one method alone to determine weights may have limitations. AHP ignores the interaction between indicators when calculating weights, and DEMATEL’s accuracy in weight calculation needs to be improved. Therefore, in order to evaluate the weight more comprehensively, we combined AHP and DEMATEL methods to carry out comprehensive calculation. The initial weight X_i of each level of indicators calculated by AHP and the center degree Y_i calculated by DEMATEL are combined to get the comprehensive influence weight of each indicator: (9) $Z_{i} = \frac{X_{i} Y_{i}}{\sum_{i = 1}^{n} X_{i} Y_{i}}$

2.3

Cloud Modeling

A large amount of information needs to be aggregated into one place for analysis, and this information usually has non-deterministic characteristics, including expressions of ambiguity, randomness, and complexity. In order to deal with the problem of ambiguity and randomness, the academic circle extracts the concept of "cloud" from probability theory and fuzzy mathematics, and deduces its analytic formula. The concept innovatively merges randomness and fuzziness into a single mathematical model through a normal distribution, the so-called cloud model. Since then, the model has flourished in the fields of artificial intelligence and data mining [25], in recent years, its application scope has been further expanded to various assessment scenarios. In recent years it has been extended to be applied to various types of assessments.

For a quantitative argument U , C is a qualitative concept on U . The quantitative value x is subordinate to U , x the degree of subordination of qualitative concepts C is expressed by the affiliation function μ_x ∈ [0,1], then x multiple random occurrences on the domain U become cloud, each x occurrences are counted as such, regarded as a cloud droplet, the degree of deviation from the ideal state of the cloud droplets just reveals the fuzzy characteristics of the cloud. Therefore, the cloud model, a mathematical model, contains both randomness and fuzziness.

2.3.1

Numerical Characterization of Cloud Models

The cloud model characteristic parameters are calculated as in equation (10). Where Ex, En, He denotes cloud expectation, cloud entropy and cloud super entropy respectively, and information uncertainty is reflected by the three parameters together: (10) ${\begin{matrix} E x = \frac{1}{n} \sum_{i = 1}^{n} x_{i} \\ E n = \frac{1}{n} \sqrt{\frac{π}{2}} \sum_{i = 1}^{n} | x_{i} - \bar{x} | \\ H e = \sqrt{\frac{1}{n - 1} \sum_{i = 1}^{n} {(x_{i} - \bar{x})}^{2} - E n^{2}} \end{matrix}$

where x_i and n denote the sample observations and the number of samples, respectively. Expectation Ex is the desired numerical quantity for the qualitative problem, It represents the core of the distribution of cloud droplets in the argument space. En entropy, as a bridge between qualitative and quantitative, reveals the degree of deviation between the actual value and the ideal expectation, and this deviation can reflect the fuzzy characteristics of information. Hyperentropy He is reflected in the cloud diagram as the degree of cohesive looseness of the cloud droplets. For the number of intervals for which there is an interval threshold [L_min,L_max], the characteristic parameters can be expressed as in equation (11): (11) ${\begin{array}{l} E x = (L_{\min} + L_{\max}) / 2 \\ E n = (L_{\max} - L_{\min}) / 6 \\ H e = k \end{array}$

k is a constant, selected according to the fuzzy threshold of the variable, often take 0.01, 0.02, 1, the fuzzy threshold of the indicators in this paper this paper takes k = 0.01 . In order to realize qualitative and quantitative transformation of cloud modeling, it is necessary to use forward and reverse cloud algorithm. From qualitative to quantitative, it is often achieved through the forward cloud algorithm, that is, the transformation of abstract concepts into concrete numerical points, a typical normal cloud diagram and its digital features are shown in Figure 1.

2.3.2

Weighted Integration of Cloud Models

The process of weighted integration of n cloud into one cloud is called weighted integration of clouds, let E_i(Ex_i,En_i,He_i)(i = 1,2,⋯,n) be n different clouds and μ_i(i = 1,2,⋯,n) be the corresponding weights of each cloud respectively E₀(Ex₀,En₀,He₀) be the integrated composite cloud after integration, then the weighted integration equation is: (12) ${\begin{matrix} E x_{0} = \sum_{i = 1}^{n} μ_{i} E x_{i} \\ E n_{0} = \sqrt{\sum_{i = 1}^{n} {(μ_{i} E n_{i})}^{2}} \\ E e_{0} = \sqrt{\sum_{i = 1}^{n} {(μ_{i} H e_{i})}^{2}} \end{matrix}$

3

Empirical case studies

3.1

Overview of cases

Company G is a large listed private science and technology enterprise, founded in September 2005. The high-tech company invests heavily in product research and development and market expansion, accounting for 20%-30% of the annual sales profit, and attaches great importance to the protection of current assets, which accounts for the proportion of total assets increased year by year. By the end of 2023, Company G had 380 active employees, with R&D personnel accounting for about 34%, management personnel accounting for 12%, and the rest being technicians and ordinary skilled workers. Figure 2 shows the financial status of the stock price in 2022 in an open-high-low-close curve for Company G. Company G successfully went public in 2021, and from September-November 2022, the company’s stock price experienced unprecedented turbulence with fluctuating ups and downs. The graph shows that in September-November 2022, Company G’s stock price had a high opening price of about $18 and a low closing price of about $16.40. Overall the company’s stock price started a downward trend after rising until late October. Such a stock price trend had a significant impact on the finances of Company G. The management of the company decided that it was necessary to assess the company’s financial situation and financial risk in order to clarify the direction of the enterprise’s capital management and to promote the further development of the enterprise.

3.2

Determination of indicator weights

For the financial situation of Company G, according to the established enterprise financial risk assessment system, a number of experts are invited to assess the financial risk of Company G. By combining AHP and DEMATEL method, the comprehensive weights and specific index weights among factors can be obtained. Using AHP-DEMATEL grouping method, the comprehensive weights of each index can be calculated accurately, and Table 2 shows the results of weight calculation. Table 2 shows the results of the weight calculation. From the table, we can see that among the comprehensive weights of the first-level indicators, the weight of the growth capacity (A4) indicator is the highest at 0.33, while the weight of the operating capacity (A3) is the lowest at 0.07. It shows that when the enterprise develops to a certain extent, the most important factor affecting the enterprise’s financial risk is whether the enterprise has enough growth capacity or not.

Table 2.

Weight of each indicator

Primary indicator	AHP	DEMATEL	Synthesize	Secondary indicator	AHP	DEMATEL	Synthesize
A1	0.14	0.13	0.11	A11	0.060	0.036	0.084
A1	0.14	0.13	0.11	A12	0.080	0.054	0.034
A2	0.22	0.23	0.23	A21	0.044	0.093	0.030
				A22	0.047	0.046	0.014
				A23	0.043	0.068	0.013
				A24	0.038	0.077	0.026
				A25	0.048	0.067	0.073
A3	0.11	0.05	0.07	A31	0.060	0.083	0.026
				A32	0.020	0.049	0.042
				A33	0.030	0.081	0.094
A4	0.27	0.30	0.33	A41	0.160	0.040	0.092
A4	0.27	0.30	0.33	A42	0.110	0.093	0.095
A5	0.26	0.29	0.26	A51	0.080	0.073	0.094
				A52	0.070	0.035	0.101
				A53	0.100	0.018	0.089
				A54	0.010	0.087	0.093

3.3

Cloud Model Parameter Calculation

Five financial risk assessment-related domain experts are invited to assign scores on the 16 indicators affecting the financial risk of Company G. The cloud model parameters of each indicator are determined according to the formula, and the cloud digital features of Company G’s financial risk assessment are determined by combining the weight values of each indicator in Table 2 as shown in Table 3, where CDC denotes the cloud digital features and ICDC denotes the cloud digital features of A1-A5.

Table 3.

Cloud digital characteristics of financial risk assessment

Indicator	Expert score					CDC	ICDC
Indicator	P1	P2	P3	P4	P5	CDC	ICDC
A11	6.3	6.5	6.6	6.3	7.1	(6.8,0.5,0.2)	(7.3,0.6,0.2)
A12	7.2	7.1	6.1	6.8	6.1	(6.9,0.4,0.1)	(7.3,0.6,0.2)
A21	6.6	6.8	6.9	6.9	6.3	(7.1,0.6,0.2)	(7.2,0.6,0.1)
A22	7.3	6.6	6.2	6.6	6.2	(6.4,0.6,0.1)
A23	6.1	6.4	6.6	6.8	6.4	(6.2,0.4,0.2)
A24	6.5	7.2	6.9	6.4	6.6	(6.8,0.5,0.1)
A25	7.4	7.5	6.8	7.2	6.5	(7.1,0.6,0.2)
A31	8.1	8.2	7.9	8.3	7.8	(7.8,0.4,0.2)	(7.8,0.5,0.3)
A32	6.2	7.9	7.7	7.8	7.4	(7.3,0.3,0.1)
A33	6.1	7.7	8.5	7.6	8.3	(7.4,0.6,0.1)
A41	6.4	7.3	6.9	6.7	7.1	(6.9,0.5,0.2)	(7.1,0.5,0.1)
A42	8.2	6.8	7.1	6.8	6.8	(6.9,0.4,0.1)	(7.1,0.5,0.1)
A51	7.6	6.5	6.8	6.2	6.2	(6.3,0.3,0.1)	(7.4,0.6,0.2)
A52	6.1	6.9	6.4	6.4	7.1	(6.2,0.6,0.2)
A53	6.3	7.3	6.5	6.5	6.9	(7.1,0.5,0.3)
A54	6.2	7.6	6.7	6.2	6.4	(6.9,0.6,0.2)

4

Enterprise financial risk assessment

4.1

Financial risk judgment and early warning

According to the calculated cloud numerical characteristics can be obtained the evaluation cloud diagram of the first-level indicators, and the results are shown in Fig. 3, where (a)-(e) are the evaluation cloud diagrams for each first-level indicator from A1 to A5, respectively. From the figure, it can be seen that each level of indicators is in the medium range,the comprehensive score of financial risk cloud numerical characteristics calculation is above 7.0,which means that there is a low risk in the finances of G companyand overall good financial condition. Horizontally, the risk of solvency and growth capacity of Company G is relatively high, with a composite score value of 7.2 and 7.1, respectively. This is because Company G as a whole is in the expansion stage, and its solvency is more sensitive to financial risk. At the same time, the growth capacity of Company G in the expansion stage is still to be explored, and some of the development resolutions in the growth process will have a significant impact on the company’s finance. Therefore, Company G should pay great attention to matters related to solvency and growth capacity in the process of subsequent development, so as to reduce the financial risks faced by the enterprise and realize the healthy development of the enterprise.

4.2

Measures to reduce the financial risk of enterprises

In view of the low comprehensive scores of G Company’s "solvency" and "growth". It is recommended that the company strengthen credit assessment and management, and fully understand the credit status of customers and trading partners before cooperating with them, including payment history, financial status, etc. Avoid cooperating with enterprises or individuals with bad credit, improve the customer credit evaluation system, control the risk of bad debt, and reduce the uncertainty of accounts receivable. We will keep the scale and structure of debt under reasonable control. Company G shall, according to its financial strength, operating conditions and profitability, reasonably determine the amount of debt, optimize the debt structure, balance the proportion of long-term and short-term debt, and avoid the debt repayment pressure caused by excessive short-term debt. We will strengthen the monitoring and management of the liquidity of funds and improve the efficiency of the use of funds. Strengthen contract management and receivables collection by carefully reviewing the contract terms before signing, especially key terms such as payment method and time limit. Set up early warning indicators for accounts receivable risk, such as ageing and bad debt ratio. Conduct regular risk assessments of accounts receivable to identify potential risks in time. Establish a accounts receivable monitoring system to monitor the dynamic changes of accounts receivable in real time.

Actively adopting new technologies and tools can optimize the production process by introducing artificial intelligence technology, and achieve the combination of personalized and large-scale production. Pay close attention to market demand, continuously carry out product innovation, and introduce new products to meet user needs. Product development can be guided by customer feedback and market research to ensure that the product is competitive in the market. Market expansion and brand building can be achieved by entering new markets or expanding existing market share. You can consider expanding geographic markets or related industry markets through partnerships, acquisitions or investments. Focus on staff training and development to improve staff skills and capabilities to adapt to the rapidly changing market environment. The talent pool can be enriched by bringing in external talent or partnering with universities. Optimize organizational structure, improve operational efficiency, and improve management level. Ensure that the company can quickly respond to market changes and take advantage of growth opportunities. Establish and enhance the financial management system to guarantee the company’s stable financial situation. Through reasonable budget planning and cost control, improve the efficiency of capital utilization. Actively explore avenues of capital operations to raise additional funds for the company to support its growth and development.

5

Conclusion

This paper is based on the combined assignment method of AHP and DEMATEL, as well as the cloud model, to achieve intelligent assessment of enterprise financial risk. Company G is selected for the study and the method described in this paper is used to evaluate the financial risk of the enterprise, and the risk control strategy is given. at the same time, the risk control measures are put forward. This paper explores a set of intelligent methods to evaluate the financial risk of enterprises in the era of digital economy, which is feasible and replicable. In addition, the research results also have a certain reference significance for reducing the financial risk of enterprises.

Idioma:: Inglés

Calendario de la edición:: 1 veces al año
Temas de la revista:: Ciencias de la vida, Ciencias de la vida, otros, Matemáticas, Matemáticas aplicadas, Matemáticas generales, Física, Física, otros

RSS Feed de revista

Research on Intelligent Methods and Strategies of Corporate Financial Risk Assessment in the Digital Economy Era

Zhuochen Shi

Publicado en línea: 21 mar 2025

Recibido: 19 oct 2024

Aceptado: 04 feb 2025

DOI: https://doi.org/10.2478/amns-2025-0636

Palabras claveAHP, DEMATEL, Cloud model, Cloud digital features, Financial risk assessment

© 2025 Zhuochen Shi, published by Sciendo

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

Palabras clave
AHP, DEMATEL, Cloud model, Cloud digital features, Financial risk assessment