Study on the comprehensive evaluation of regional financial risk

: Since 2011, the Chinese government began to pay attention to the prevention of regional financial risk in the financial field. The effective assessment of regional financial risk is an indispensable link in risk prevention, which is of great significance to maintain economic operation. This paper selects 5 first-level indicators and 15 second-level indicators to build a regional financial risk evaluation system. Based on the economic data of Jiangsu from 2015 to 2020, AHP and entropy weight method are used to calculate the corresponding weights of indicators at all levels. The multi - level fuzzy evaluation model is used to evaluate the financial risk in Jiangsu province. The results show that the membership degree of financial risk to the "general security" level is 0. 5050 at most, among which the proportion of macro environment is 0. 3932 at most, and the degree of external environment is 0. 0625 at least. It can be seen that the economic operation of Jiangsu is relatively stable and the financial risk is small.


Introduction
In July 2021, the General Office of the CPC Central Committee and The General Office of the State Council issued the Opinions on Strictly cracking Down on Illegal securities Activities in accordance with the Law, requiring a strict crackdown on illegal securities activities in accordance with the law to effectively prevent regional financial risks.Financial risks have strong linkage and self-reinforcing transmission characteristics [1], which are easy to accumulate, spread and then become regional financial risks.At present, the economic situation at home and abroad is changing with each passing day.In order to effectively prevent the occurrence of this situation and maintain the steady development of national economy, it is very important to carry out further comprehensive evaluation and research on regional financial risks.
Chinese scholars have carried out in-depth study on regional financial risk.Tan Zhongming et al. [2] found that Internet finance would impact the "steady state" of regional finance through the direct and indirect mechanisms of Internet finance's impact on regional financial risks, which would gradually accumulate regional financial risks.Ma Ruowei et al. [3] built a linear probability model to analyze the impact of Internet finance based on industrial structure adjustment on regional financial risks.They found that Internet finance has a positive impact on financial risks, and the adjustment of industrial structure has a positive moderating effect on this positive impact, and different industrial structures have different moderating effects.Chen Lei [4] et al. studied the relationship between the level of fintech development and regional financial risks through the spatial Dubin model and panel threshold model.After their empirical study, they found that the two showed an "inverted U-shaped" relationship, and digital development was helpful to reduce the occurrence of regional financial risks.Li Kaifeng et al. [5] studied the debt risk level of various regions in China through entropy weight TOPSIS method and fuzzy comprehensive evaluation method.Based on Gansu Province, they built a VAR model to evaluate their impact on regional financial risks.The results show that the level of local debt risk in Gansu Province is directly proportional to the possibility of regional financial risk.But this phenomenon has the characteristic of time delay.
At present, there are few studies on the assessment of financial risks in Jiangsu and lack of a complete and effective assessment system.In this paper, the analytic hierarchy process is used to find the weight of the first-level index, and the entropy weight method is used to find the weight of the second-level index, so as to improve the accuracy of the weight results.The five -level evaluation criteria are selected and the multilevel fuzzy comprehensive evaluation model is used to comprehensively evaluate the financial risk in Jiangsu Province.It provides strong reference value for the formulation of financial risk policies in Jiangsu Province [6].

Evaluation index analysis
The stability of macroeconomic operation is closely related to the risk state of the financial market.The stability of the macro environment indicates that the macro environment has strong resistance to financial risks.This paper selects the local GDP growth rate;Inflation rate;Growth rate of local fixed assets;Local fiscal revenue/expenditure.
The real estate market is of great significance to the study of financial risks.The unstable real estate market may cause social and economic turmoil.This article chooses the index to have the local real estate investment growth rate;Growth rate of commercial housing sales.
The soundness of currency and financial market is a micro factor affecting regional financial risk, and the soundness of its operation is directly related to the possibility of regional financial risk.This paper selects the capital adequacy ratio;Liquidity ratio;Non-performing loan ratio;Leverage ratio, M2/GDP.
A better external environment is a prerequisite for economic stability, and the impact of the external environment will inevitably lead to financial market turbulence.This paper selects the gross value of import and export /GDP;Growth rate of total exports.
The financial activities in the stock market have become an important cause of financial risks.Studying the activities in the stock market can effectively evaluate financial risks.This paper selects the average P/E ratio of Shanghai Stock Index and the total market value of A shares /GDP.

Construction of index system
Based on the above analysis and combined with the characteristics of regional financial risks, this paper selects 5 first-level indicators and 15 second-level indicators to construct the following regional financial risk comprehensive evaluation index system.As shown in Table 1 Table 1: Comprehensive evaluation index system of regional financial risk

Macro environment B1
Growth rate of gross regional product C1 (Last year's GDP-previous year's GDP)/previous year's GDP*100%

Rate of inflation C2
The degree to which the general level of prices has risen in a given period

Capital adequacy ratio C7
The ratio of a bank's total capital to its risky assets

Ratio of liquidity C8
Current ratio is the ratio of current assets to current liabilities Non-performing loan ratio of banks C9 Non-performing loans/total loans leverage C10 The ratio of total assets to equity capital in a balance sheet M2/gdp C11 In economic transactions, the proportion of exchange in the medium of money

External environment B4
Total value of imports and exports /GDP C12 Describe the proportion of imports and exports in national production Growth rate of total exports C13 Total exports this year/Total exports last year -1

Determination method of index weight
(1) AHP determines the first-level index weight 1) Construct the judgment matrix: The nine-degree scoring method is used to make pairwise comparison of the first-level indicators [7].By collecting the scores of several experts on the importance of the indicators, the average of the scores is taken as the final evaluation result, and the judgment matrix can be obtained: 2) Check the consistency of the judgment matrix, and the calculation steps are as follows: Step 1: Calculate the consistency index CI max 1 Step 2: Find the corresponding average random consistency index RI,As shown in Table 2.
If CR<0. 1, the consistency of the judgment matrix can be considered acceptable.Otherwise, the judgment matrix needs to be modified 3) Calculate the weight: If the consistency test of the judgment matrix can be accepted, then the eigenvalue method can be chosen to obtain the weight of the first-level index.The calculation procedure is as follows: Step 1: Find the maximum eigenvalue of judgment matrix A and its corresponding eigenvector.
Step 2: to normalize the feature vectors can get the desired weight.
(1) Entropy weight method is used to determine the weight of secondary index Let's assume that there are n years, m evaluation indicators constitute the original matrix X, and put X forward, the results are as follows: 1) The matrix Z is obtained by standardizing the forward matrix X and judging whether the input matrix has negative values.If so, it needs to be re-standardized to the non-negative interval.Where each element of Z: 2) Calculate the proportion of the ith data under the JTH index, and regard it as the probability used in the calculation of information entropy: calculate the probability matrix P, and each element in P: 3) The information entropy of each index and its corresponding information utility value, namely the difference coefficient, are calculated, and the entropy weight of each index is obtained by normalization.
For the JTH index, the calculation formula of information entropy is: A a a a    .
(1) Determine the fuzzy comprehensive judgment matrix: For each index, the membership degree of each comment is the fuzzy shadow set on V.In this paper, the membership degree of the index is determined by fuzzy statistical method.the evaluation of the first level index i U is denoted as 12 [ , ,..., ] , then the fuzzy comprehensive judgment matrix of each index is: If there is a fuzzy relation from U to V, then R can be used to get a fuzzy transformation, so as to get the final comprehensive evaluation result: The comprehensive evaluation result can be regarded as a fuzzy vector on V, denoted as:

Index weight determination
The 12 experts adopted the 9-degree scoring method to construct the judgment matrix respectively, and assigned 1/12 weight to each expert for weighted average.The final judgment matrix result was obtained by comprehensively considering the impact of each index on regional financial risk assessment, as shown in Table 3: The eigenvector corresponding to the largest eigenvalue is normalized to obtain the weight of the first-level index:0.3932, 0. 0987, 0. 2694, 0. 0625, 0. 1762.
The entropy weight method determines the weight of the second-level index, in which the data comes from CSMAR, and the relevant data is input into Matlab to obtain the entropy weight of the second-level index.
In summary, the index weight results are shown in Table 4.

Comprehensive evaluation result
(1) Determining factor set: The first level factor set:U={ B1,B2,B3,B4,B5}, set of second-level elements.Take B1 as an example: B1={C1,C2,C3}, and so on. ( Similarly, the fuzzy comprehensive evaluation matrix of real estate market, money and financial market, external environment and stock market can be obtained as follows: research object is 0. 5670, which belongs to the "general safety level", in which the proportion of the average price-earnings ratio is 0. 6695, and the proportion of the total stock market value /GDP is 0. 3305.

Conclusion
This paper establishes an evaluation system from five aspects of macro environment, monetary and financial market, stock market, real estate market and external environment.Through screening, 15 risk indicators are obtained and the relevant weights of the indicators are determined to determine a complete and efficient regional financial risk evaluation system.On this basis, the multi-level fuzzy comprehensive rating model is used to further evaluate the financial risk status in Jiangsu province, so as to help the people's government of Jiangsu Province effectively prevent risks and promote the stable and coordinated development of regional economy [9].

1 ,( 1 )( 2 )
information utility value is normalized to obtain the entropy weight of each indicator: Set of determined factors:U：First order factor set. Assuming that the number of first-level indicators is n, then: Set of second-order elements.If the number of second-level indicators corresponding to the ith first-level indicator is k, then: Determine the set of comments:Suppose that there are m levels, and thus the set of comments is denoted as:

) 3 )( 4 )
Determine the evaluation set: This paper adopts the 5-level 5-point system and divides the evaluation set into 5 levels, V={very safe, general safe, critical state, general dangerous, very dangerous} (Determine the weight: 0.3932,0.0987,0Determine the fuzzy comprehensive evaluation matrix: Take the macro-environment B1 as an example, determine the fuzzy comprehensive evaluation matrix according to the fuzzy statistical method 0.4 0.3 0.2 0 0.2 0.6 0.15 0.05 0 0.2 0.5 0.25 0.05 0

Table 2 :
Average random consistency index

Table 4 :
Index weight table