The Analysis of Fintech Risks in China: Based on Fuzzy Models

Abstract

Fintech has not changed the nature and risk attributes of financial business. Its openness, interoperability, and other characteristics trigger the concealment, infectivity, universality and sudden characteristics of financial risk more obviously, and the potential systemic risk more complexly. This paper adopts fuzzy set analysis to conduct a comprehensive review of Fintech risk. It is found that the technology risk, moral hazard and legal risk, with a weight of up to 80%, are the dominant factors affecting the Fintech risk, while other prominent credit risk, market risk and operational risk in the traditional financial field account for a small proportion, but they still cannot be ignored. The research of this paper enriches the relevant research on the quantification of Fintech risk, and helps to strengthen risk prevention from the aspect of enhancing the safety of Fintech infrastructure, improving the legal system, sound the regulatory framework and strengthening industry self-discipline.

1. Introduction

In recent years, Fintech, coupled with finance and technology, has greatly changed the financial industry and continues to reshape the modern financial system [1,2]. According to the Financial Stability Board (FSB), Fintech refers to emerging business models, new technology applications, new products and services driven by emerging cutting-edge technologies such as big data, the blockchain, cloud computing and artificial intelligence, which have a significant impact on the financial market and the financial services sector. With its unique technical support, business model and value creation mode, Fintech influences the concept, business, structure and risk control mode of traditional financial institutions, and gradually becomes a force to be reckoned with in the whole financial ecosystem [3,4,5]. Due to its advantages of light assets, high innovation, large scale and easy compliance, Fintech is developing rapidly worldwide [6]. Fintech has higher requirements for the more thorough application and more profound and extensive impact on emerging technologies, including the Internet. It is a further expansion and deepening on the basis of Internet finance. It can be said that Fintech is evolving from a channel to facilitate financial development to a core part of financial development [7,8].

However, Fintech has not changed the nature and risk attributes of financial business. Its openness, interoperability, science and technology and other characteristics make the concealment, infectivity, universality and sudden characteristics of financial risks more obvious, and the potential systemic risks more complex [9,10,11]. Fintech has led to fundamental changes in the financial industry, from the institutional operation mode to capital raising mode and even the currency itself [12], making financial innovation more active. However, financial innovation may also blur the existing industry boundaries, subvert the industry pattern, accelerate financial disintermediation and induce new financial risks [13,14,15]. Figure 1 shows the main stages of China’s Fintech development and corresponding risk characteristics.

The rapid development of Fintech has had a great impact on traditional financial business. The process of the integration of finance and technology has not only changed the term conversion, credit conversion, income conversion and risk conversion of the traditional financial business [16,17,18], but also greatly reshaped the risk characteristics of the financial system itself [19]. Based on the existing research and the reality of Fintech development, Fintech may trigger traditional financial risks such as credit risk, liquidity risk and operational risk [20,21,22], and it contains risks triggered by non-financial factors such as underlying information technology and technology ethics due to its own characteristics [23,24,25,26]. Many innovations in Fintech have greatly promoted the openness, online and virtualization of modern financial services, remote customer transactions and real-time capital flow, which also increases the dependence of the financial market on the Internet and information technology [27,28,29,30]. Once there are any technical loopholes in the Internet business operation process, technology management and network maintenance, it can lead to information technology risks such as data theft, privacy infringement and website attack, which can bring serious economic losses to customers and enterprises [31,32,33,34]. Obviously, Fintech has not only changed the traditional form of financial risk, but also changed the distribution and weight of financial risk, making the technical risks caused by the technical loopholes of Fintech, such as incomplete technology, hidden dangers of data security, network security and so on increasingly severe [35,36,37].

Therefore, scholars have conducted a lot of valuable research: the first is to investigate the new risks brought by the introduction of Fintech into the traditional financial system [38,39,40,41,42]. Second, a study was undertaken to define the risk characteristics of Fintech and the construction of risk control mechanisms [43,44,45,46]. The third explored the risk prevention of Fintech platform operation [47,48,49,50,51,52].

Following the existing research, the incremental work in this paper may lie in: (1) the existing research focuses on the identification and classification of Fintech risks and then lacks the necessary quantification. We attempt to comprehensively assess the traditional risks contained in Fintech and the new risks caused by factors such as information technology; (2) We also use the fuzzy set analysis method to comprehensively measure the financial science and technology risk, and find that the technical risk, ethical risk and legal risk are the dominant factors affecting the financial science and technology risk; while other prominent credit risk, market risk and operational risk in the traditional financial field accounts for a small proportion, but they still cannot be ignored.

2. Models Setting

2.1. Establish Fintech Risk Assessment System Based on Delphi Method

The definition of Fintech risk is a qualitative concept. Different personnel have different views and lack a unified standard [53,54,55]. Therefore, it is necessary to seek experts who have studied the relevant aspects to make the definition results more authoritative. Therefore, the Delphi method is preferred. In this paper, 30 experts in the field of Internet finance were invited to score and modify the first-level indicators in two rounds, and finally designed a two-level Fintech risk evaluation system, as shown in

Table 1. Fintech Risk Assessment System.

Evaluation Object	Primary Indicator	Secondary Indicator
Fintech risk: U	Technical risk: U1	Network system security risk: u11
		Technical support risk: u12
		Data transmission security risk: u13
	Ethical risks: U2	Social ethical risk: u21
		Liability ethical risk: u22
		Technical ethical risk: u23
	Management risk: U3	Consumer operational risk: u31
		Supplier operational risk: u32
		Mediator operational risk: u32
		Payment method innovation risk: u34
	Operational risk: U4	Internal management risk: u41
		Liquidity risk: u42
		Associated risk: u43
	Credit risk: U5	Internal fraud risk: u51
		External fraud risk: u52
		Credit risk: u53
		Credit information abuse risk: u54
	Market risk: U6	Interest rate risk: u61
		Exchange rate risk: u62
		price movement risk: u63
	Legal risk: U7	Laws and regulations absence risk: u71
		Regulatory vacancy risk: u72
		Subject qualification risk: u73
		Virtual currency risk: u74
		Online money laundering risk: u75

.

2.2. Determine the Weight of Fintech Risk Indicators According to Analytic Hierarchy Process

2.2.1. Construct a Fuzzy Judgment Matrix

Satty (American operations research scientist) proposed to use the numbers 1–9 as the scale to judge the relative importance of the two indicators, as shown in

Table 2. Satty Analytic Hierarchy Process Scale 1–9.

Scale aij	Meaning
1	i factor is as important as j factor
3	i factor is slightly more important than j factor
5	Compared with factor j, factor i is obviously more important than factor j
7	i factor is more important than j factor
9	i factor is absolutely more important than j factor
2,4,6,8	Intermediate value of the above adjacent judgment
reciprocal	Indicates that the former is less important than the latter

[56,57,58].

Using the results in Table 2, the relative importance of each risk was obtained by comparing the risk factors contained in each primary indicator and secondary indicator, and the primary and secondary judgment matrices were constructed.

2.2.2. Calculate Indicator Weight

(1) Multiply the element values of fuzzy judgment matrix A by rows to obtain the product P_i of each row [59,60].

$$P_i={\displaystyle \prod _{j=1}^n{U}_{ij}},i=1,2,\dots ,n$$

(2) Open P_i to the nth root and get Q_i.

$$Q_i=\sqrt[n]{P_i}$$

(3) Normalize Q_i to obtain W_i.

$$W_i=\frac{Q_i}{{\displaystyle \sum _{i=1}^n{Q}_i}}$$

The vector W = (W₁, W₂,..., W_n) is the feature vector, that is, the weight corresponding to the indicator.

(4) Calculate the maximum eigenvalue of the judgment matrix ${\lambda }_{\max }$.

$${\lambda }_{\max }\approx {\displaystyle \sum _{\mathrm{i}=1}^{\mathrm{n}}\frac{{(A\varpi )}_{\mathrm{i}}}{\mathrm{n}{\varpi }_{\mathrm{i}}}}$$

where, (AW)_i represents the ith component of vector AW.

2.2.3. Checking the Consistency of Fuzzy Judgment Matrix

After the Calculated λ_max and the corresponding eigenvector were calculated, the consistency of the judgment matrix needs to be checked. If the consistency test is passed, the eigenvector can be regarded as the relative weight of the next level indicator to the upper level indicator. Otherwise, the values of each element in the judgment matrix should be readjusted until it meets the consistency test. The results were checked by calculating the consistency ratio CR. The calculation formula of CR is:

$$CI=\frac{CR}{RI}$$

Among them, CI is the consistency indicator and RI is the random consistency indicator. The calculation formula of CI is:

$$CI=\frac{{\lambda }_{\max }-n}{n-1}$$

The value of RI can be found in

Table 3. Values of random consistency indicator RI.

n	1	2	3	4	5	6	7	8	9	10	11
RI	0	0	0.58	0.90	1.12	1.24	1.32	41	1.45	1.49	1.51

. If the consistency ratio CR < 0.1, it passes the consistency test.

2.3. The Value of Fintech Risk Assessment Is Determined Based on the Fuzzy Comprehensive Evaluation Method

2.3.1. Determine Evaluation Factor Set

Assuming that the factor set of Fintech risk is u, according to the evaluation system in Table 1:

U = {technical risk u₁, ethical risk u₂, management risk u₃, operational risk u₄, credit risk u₅, market risk u₆, legal risk u₇} each factor can be further decomposed into:

U₁ = {network system security risk u₁₁, technical support risk u₁₂, technical ethical risk u₁₃}

U₂ = {Social ethical risk u₂₁, liability ethical risk u₂₂, technical ethical risk u₂₃}

U₃ = {consumer operational risk u₃₁, supplier operational risk u₃₂, mediator operational risk u₃₃, Payment method innovation risk u₃₄}

U₄ = {internal management risk u₄₁, liquidity risk u₄₂, associated risk u₄₃}

U₅ = {internal fraud risk u₅₁, external fraud risk u₅₂, credit risk u₅₃, credit information abuse risk u₅₄}

U₆ = {interest rate risk u₆₁, exchange rate risk u₆₂, price movement risk u₆₃}

U₇ = {laws and regulations absence risk u₇₁, regulatory vacancy risk u₇₂, subject qualification risk u₇₃, virtual currency risk u₇₄, online money laundering risk u₇₅}

2.3.2. Determine Evaluation Set

The evaluation set refers to the set of fuzzy evaluation levels of each Fintech risk, and the set evaluation set is:

V = {V₁, V₂, V₃, V₄, V₅} = {very high, high, medium, low, very low},

The scores corresponding to V₁, V₂, V₃, V₄ and V₅ are 9, 7, 5, 3 and 1, respectively.

2.3.3. The Single Factor Fuzzy Evaluation Matrix Is Established

The fuzzy relationship matrix R_i between u and V is established, which is expressed as follows:

$$R_i=\left[\begin{array}{ccccc}{r}_{11}& r_{12}& r_{13}& \dots & r_{1n}\\ r_{21}& r_{22}& r_{23}& \dots & r_{2n}\\ r_{31}& r_{32}& r_{33}& \dots & r_{3n}\\ \dots & \dots & \dots & \dots & \dots \\ r_{m1}& r_{m2}& r_{m3}& r_{m4}& r_{mn}\end{array}\right] \left(0<r_{ij}<1,i=1,2,\dots m\right)$$

The membership degree, denoted by r_ij, reflects the membership relationship between the evaluation factor set (various risks of Internet finance) and the evaluation set (risk fuzzy evaluation level). The membership degree R_ij of the i-th risk in the factor set to the evaluation set {V₁, V₂, V₃, V₄, V₅}, that is, the probability of belonging to the jth risk level for the i-th risk.

2.3.4. Fuzzy Comprehensive Evaluation

First, the evaluation set t is obtained

$$T_i=W_i\cdot R_i$$

where, W_i is the weight set of each secondary and R_i is the membership matrix. The values of T₁, T₂, T₃, …, T_m are obtained respectively, and then the primary indicator evaluation matrix T can be obtained by sequentially arranging T_i (i = 1, 2, 3,..., m) as the row vector.

$$T=\left[\begin{array}{c}{T}_1\\ T_2\\ T_3\\ \dots \\ T_5\end{array}\right]=\left[\begin{array}{ccccc}{t}_{11}& t_{12}& t_{13}& \dots & t_{1n}\\ t_{21}& t_{22}& t_{23}& \dots & t_{2n}\\ t_{31}& t_{32}& t_{33}& \dots & t_{3n}\\ \dots & \dots & \dots & \dots & \dots \\ t_{m1}& t_{m2}& t_{m3}& t_{m4}& t_{mn}\end{array}\right]$$

Then, the fuzzy comprehensive evaluation matrix B can be obtained from the primary weight W.

B = W·T = (b₁, b₂, b₃, …, b_n)

b_i is the probability that the evaluation object U (Fintech risk) makes the i-th evaluation on the evaluation set {V₁, V₂, V₃, V₄, V₅}, after considering all indicators. According to the maximum membership principle, take b = max (b₁, b₂, b₃, …, b_n), then the risk level corresponding to b is the risk level of Internet finance as a whole.

3. Empirical Analysis of Fintech Risk Based on Fuzzy Analytic Hierarchy Process

3.1. Determine Indicator Weight

3.1.1. Primary Indicator Weight

Based on satty1–9 scoring standards, the judgment matrix of the first-level indicators was obtained in this paper by considering the opinions of thirty experts, as shown in

Table 4. Fintech risk judgment matrix.

U	U2	U3	U4	U5	U6	U7
U1	5	6	7	6	8	5
U2	1	4	5	4	6	2
U3	1/4	1	2	2	3	1/4
U4	1/5	1/2	1	1/2	3	1/5
U5	1/4	1/2	2	1	3	1/5
U6	1/6	1/3	1/3	1/3	1	1/5
U7	1/2	4	5	5	6	1

.

That is, the judgment matrix A of Fintech risk is:

$$A=\left[\begin{array}{ccccc}1& 5& 3& 3& 6\\ 1/5& 1& 1/3& 1/4& 2\\ 1/3& 3& 1& 1/2& 2\\ 1/3& 4& 2& 1& 2\\ 1/6& 1/2& 1/2& 1/2& 1\end{array}\right]$$

Calculated by MATLAB software, the maximum eigenvalue, CI and CR of matrix A are respectively: λ_max = 7.57, CI = 0.09, CR = 0.07 < 0.1.

Therefore, the matrix passes the consistency test. The corresponding eigenvector of λ_max, that is, the weight distribution of the primary indicator, is: W = (0.4397, 0.1984, 0.0696, 0.0430, 0.0553, 0.0260, 0.1680).

3.1.2. Secondary Indicator Weight

(1) Technical risk. In the same way as the weight calculation method of primary indicators, the judgment matrix and indicators of technical risk are:

$$A_1=\left[\begin{array}{ccc}1& 1/3& 1/5\\ 3& 1& 1/2\\ 5& 2& 1\end{array}\right]$$

λ_max = 3, CI = 0, CR = 0 < 0.1.

Therefore, it passes the consistency test. The weight distribution of secondary indicators under technical risk is: W₁ = (0.1095, 0.3090, 0.5816).

(2) Ethical risk. The judgment matrix and indicators of operational risk are:

$$A_2=\left[\begin{array}{ccc}1& 3& 1/3\\ 1/3& 1& 1/6\\ 3& 6& 1\end{array}\right]$$

λ_max = 3.02, CI = 0.01, CR = 0.02 < 0.1.

Therefore, it passes the consistency test. The weight distribution of secondary indicators under operational risk is: W₂ = (0.2499, 0.0953, 0.6548).

(3) Management risk. The judgment matrix and indicators of operational risk are:

$$A_3=\left[\begin{array}{cccc}1& 1/4& 1/3& 1/6\\ 4& 1& 2& 1/3\\ 3& 1/2& 1& 1/5\\ 6& 3& 5& 1\end{array}\right]$$

λ_max = 4.10, CI = 0.03, CR = 0.04 < 0.1.

Therefore, it passes the consistency test. The weight distribution of secondary indicators under operational risk is: W₃ = (0.0631, 0.2348, 0.1360, 0.5661).

(4) Operational risk. The judgment matrix and indicators of credit risk are:

$$A_4=\left[\begin{array}{ccc}1& 1/3& 1/2\\ 3& 1& 3\\ 2& 1/3& 1\end{array}\right]$$

λ_max = 3.05, CI = 0.03, CR = 0.05 < 0.1.

Therefore, it passes the consistency test. The weight distribution of secondary indicators under credit risk is: W₄ = (0.1571, 0.5936, 0.2493).

(5) Credit risk. The judgment matrix and indicators of credit risk are:

$$A_5=\left[\begin{array}{cccc}1& 1/2& 1/4& 1/7\\ 2& 1& 1/3& 1/6\\ 4& 3& 1& 1/5\\ 7& 6& 5& 1\end{array}\right]$$

λ_max = 4.15, CI = 0.05, CR = 0.06 < 0.1.

Therefore, it passes the consistency test. The weight distribution of secondary indicators under credit risk is: W₅ = (0.0610, 0.0963, 0.2076, 0.6351).

(6) Market risk. The judgment matrix and indicators of credit risk are:

$$A_6=\left[\begin{array}{ccc}1& 3& 1/5\\ 1/3& 1& 1/6\\ 5& 6& 1\end{array}\right]$$

λ_max = 3.09, CI = 0.05, CR = 0.08 < 0.1.

Therefore, it passes the consistency test. The weight distribution of secondary indicators under credit risk is: W₆ = (0.1947, 0.0881, 0.7172).

(7) Legal risk. The judgment matrix and indicators of legal risk are:

$$A_7=\left[\begin{array}{ccccc}1& 2& 1/2& 1/5& 1/7\\ 1/2& 1& 1/5& 1/6& 1/8\\ 2& 5& 1& 1/2& 1/5\\ 5& 6& 2& 1& 1/4\\ 7& 8& 5& 4& 1\end{array}\right]$$

λ_max =5.28, CI = 0.07, CR = 0.06 < 0.1.

Therefore, it passes the consistency test. The weight distribution of secondary indicators under legal risk is: W₇ = (0.0649, 0.0384, 0.1320, 0.2270, 0.5377).

3.2. Fuzzy Comprehensive Evaluation

In this paper, thirty experts in the field of Fintech risk were invited to score the membership of each secondary indicator. The results are shown in

Table 5. Single factor fuzzy evaluation of Fintech risk.

Primary Indicator (Weight)	Secondary Indicator (Weight)	Membership Degree
		Very High	High	Medium	Low	Very Low
Technical risk U1 (0.4397)	Network system security risk u11 (0.1095)	0.60	0.30	0.05	0.05	0.00
	Technical support risk u12 (0.3090)	0.30	0.50	0.15	0.05	0.00
	Technical ethical risk u13 (0.5816)	0.50	0.30	0.10	0.10	0.00
Ethical risk U2 (0.1984)	Social ethical risk u21 (0.2499)	0.10	0.20	0.50	0.15	0.05
	Liability ethical risk u22 (0.0953)	0.05	0.40	0.50	0.05	0.00
	Technical ethical risk u23 (0.6548)	0.05	0.60	0.20	0.15	0.00
Management risk U3 (0.0696)	Consumer operational risk u31 (0.0631)	0.00	0.10	0.20	0.60	0.10
	Supplier operational risk u32 (0.2348)	0.00	0.05	0.25	0.60	0.10
	Mediator operational risk u33 (0.1360)	0.00	0.05	0.50	0.40	0.05
	Payment method innovation risk u34 (0.5661)	0.05	0.40	0.40	0.10	0.05
Operational risk U4 (0.0430)	Internal management risk u41 (0.1571)	0.00	0.05	0.50	0.30	0.15
	Liquidity risk u42 (0.5936)	0.00	0.05	0.45	0.35	0.15
	Associated risk u43 (0.2493)	0.00	0.05	0.55	0.30	0.10
Credit risk U5 (0.0553)	Internal fraud risk u51 (0.0610)	0.00	0.05	0.40	0.40	0.15
	External fraud risk u52 (0.0963)	0.00	0.05	0.55	0.30	0.10
	Virtual currency risk u53 (0.2076)	0.00	0.20	0.50	0.20	0.10
	Credit information abuse risk u54 (0.6351)	0.10	0.70	0.15	0.05	0.00
Market risk U6 (0.0260)	Interest rate risk u61 (0.1947)	0.00	0.05	0.55	0.35	0.05
	Exchange rate risk u62 (0.0881)	0.00	0.05	0.15	0.50	0.30
	Price movement risk u63 (0.7172)	0.00	0.05	0.60	0.20	0.15
Legal risk U7 (0.1680)	Laws and regulations absence risk u71 (0.0649)	0.00	0.05	0.60	0.25	0.10
	Regulatory vacancy risk u72 (0.0384)	0.00	0.05	0.65	0.15	0.15
	Subject qualification risk u73 (0.1320)	0.05	0.30	0.50	0.10	0.05
	Virtual currency risk u74 (0.2270)	0.00	0.65	0.20	0.10	0.05
	Online money laundering risk u75 (0.5377)	0.25	0.50	0.15	0.10	0.00

.

According to the fuzzy comprehensive evaluation formula ${\mathrm{T}}_{\mathrm{i}}={\mathrm{W}}_{\mathrm{i}}\cdot {\mathrm{R}}_{\mathrm{i}}$, the evaluation results of each secondary indicator are as follows:

T₁ = W₁·R = (0.4491, 0.3618, 0.1100, 0.0791, 0.0000)

T₂ = W₂·R = (0.0625, 0.4810, 0.3036, 0.1405, 0.0125)

T₃ = W₃·R = (0.0283, 0.2513, 0.3658, 0.2898, 0.0649)

T₄ = W₄·R = (0.0000, 0.0500, 0.4828, 0.3297, 0.1375)

T₅ = W₅·R = (0.0635, 0.4939, 0.2764, 0.1266, 0.0395)

T₆ = W₆·R = (0.0000, 0.0500 0.5506, 0.2556, 0.1437)

T₇ = W₇·R = (0.1410, 0.4612, 0.2560, 0.1116, 0.0302)

The evaluation matrix of secondary indicators can be obtained as follows:

$$T=\left[\begin{array}{ccccc}0.4491& 0.3618& 0.1100& 0.0791& 0.0000\\ 0.0625& 0.4810& 0.3036& 0.1405& 0.0125\\ 0.0283& 0.2513& 0.3658& 0.2898& 0.0649\\ 0.0000& 0.0500& 0.4828& 0.3297& 0.1375\\ 0.0635& 0.4939& 0.2764& 0.1266& 0.0395\\ 0.0000& 0.0500& 0.5506& 0.2556& 0.1437\\ 0.1410& 0.4612& 0.2560& 0.1116& 0.0302\end{array}\right]$$

Since the weight distribution of criteria layer indicators is W = (0.4397, 0.1984, 0.0696, 0.0430, 0.0553, 0.0260, 0.1680), according to the formula B = W·T:

B = (0.2391, 0.3802, 0.2274, 0.1294, 0.0239)

According to the principle of maximum subordination, Fintech risk b = max (b₁, b₂, b₃,...,b₄) = 0.3802, corresponding to the first risk level of “high”, so the overall evaluation level of Fintech risk is “high”.

3.3. Empirical Conclusion Analysis

3.3.1. Primary Indicator Analysis

Firstly, according to the weight analysis W = (0.4397, 0.1984, 0.0696, 0.0430, 0.0553, 0.0260, 0.1680), the weight value of technical risk is 0.4397, which accounts for the largest proportion and is the most important risk. The second is ethical risk, which accounts for 0.1984. The third is legal risk, which accounts for 0.1680. The proportions of management risk, operational risk, credit risk and market risk are relatively small, which are 0.0696, 0.0430, 0.0553 and 0.0260, respectively. Their importance is relatively small, but we should also pay attention to prevention.

In terms of technical risk, according to the evaluation results T₁ = (0.4491, 0.3618, 0.1100, 0.0791, 0.0000) and the maximum membership principle, the risk evaluation value is 0.4491 and the corresponding risk evaluation level is “very high”. Furthermore, in the case of ethical risk, credit risk and legal risk, according to the maximum subordination principle, the risk evaluation values are 0.4810, 0.4939 and 0.4612, respectively, which correspond to the “high” level risk evaluation. In terms of management risk, operational risk and market risk, the risk assessment value are 0.3685,0.4828 and 0.5506 respectively, and the corresponding risk assessment level is “medium”. The final order is: technical risk > credit risk > ethical risk > legal risk > market risk > operational risk > management risk, which is also consistent with the conclusion of weight analysis. The higher risk evaluation level of each level of risk, the more it helps to improve overall risk of Fintech risk.

3.3.2. Secondary Indicator Analysis

(1)

Technical risk

Firstly, from W₁ = (0.1095, 0.3090, 0.5816), it can be seen that the technical ethical risk accounts for the largest proportion, which is 0.5816. The most important factor inducing investors’ capital loss is the network system security problems such as hacker attacks, phishing and trojan horses on the Fintech websites.

The second is technical support risk, accounting for 0.3090. Common Internet viruses include the Trojan horse virus, hacker virus, worm virus, etc. These viruses steal users personal information and cause property losses to investors by harming the Internet online trading system.

Finally, the network system security risk accounts for 0.1095. The data transmission security risk mainly refers to the risk that hackers and other criminals use system vulnerabilities to intercept and steal data in the transmission process, and even steal the database.

(2)

Ethical risk

First, according to the weight distribution W₂ = (0.2499, 0.0953, 0.6548), the risk of technical ethical is the most significant risk, accounting for 0.6548. The update of various types of Fintech devices leads to repeated changes in payment methods, and then are prone to cause operational risks in the process of change. In daily life, the most used third party payment methods used are Alipay, Fu Tong Tong, Yi Bao, Remittance World, etc.

The second is the social ethical risk, accounting for 0.2499. Due to information asymmetry, consumers easily fall into the trap of illegal malicious programs in the changing network environment, resulting in consumer operational risk. At present, there is a lack of unified operation standards among various Internet finance suppliers in China, and their operation processes are quite different, and the degree of supervision is also different. If the awareness of standardization is not strong, it may lead to the operation risk of suppliers.

Finally, the liability ethical risk accounts for 0.0953.

(3)

Management risk

Firstly, according to the weight distribution W₃ = (0.0631, 0.2348, 0.1360, 0.5661), the most important risk is the payment method innovation risk, which accounts for as high as 0.5661. Liquidity refers to the circulation and operation capacity of funds in a short time. If the platform is illiquid, it will inevitably produce a chain reaction, leading to a series of debt crises.

The second is supplier operational risk, accounting for 0.2348. Internal management mainly refers to the daily operation and maintenance, system updates, etc. Within the enterprise and e-commerce platform, some internal employees may be negligent and this may lead to the incorrect transfer of user funds. Correlation risks include business transactions between enterprises and other cooperative enterprises and investors.

Finally, there are intermediary operational and consumer operational risks, accounting for 0.1360 and 0.0631, respectively.

(4)

Operational risk

First, according to the weight value W₄ = (0.1571, 0.5936, 0.2493), liquidity risk is the most important, accounting for 0.5936. Credit investigation refers to two-way credit investigations between enterprises and investors. Enterprises give priority to investors with strong guarantee ability and good credit records, and investors also give priority to large enterprises that have good reputations. At present, China has not established an authoritative and unified national credit investigation database, resulting in uneven credit investigation ratings among enterprises and lack of reference.

The second is the internal management risk and associated risk, accounting for 0.1571 and 0.2493, respectively. Due to the lack of effective supervision, it is easy to leak investors information. Some criminals steal or illegally purchase investors personal information and defraud investors’ money by telephone, e-mail fraud and other means.

(5)

Credit risk

First, according to the weight value W₅ = (0.0610, 0.0963, 0.2076, 0.6351), the credit information abuse risk is the most prominent credit risk, accounting for 0.6351. Although China has issued some guidelines related to Internet finance, they are still in the exploratory period. For example, the detailed rules on Fintech supervision issued in August 2016 is an interim measure to regulate P2P management.

The second is the risk of virtual currency, accounting for 0.2076. Virtual currency refers to the electronic currency provided by some trading platforms of Internet finance, which, to some extent, play a role in replacing cash. Although virtual currency improves the efficiency and liquidity of a transaction, it does not establish a strict correlation with real money and will bring huge property losses to investors in case of risk. At present, China lacks a relevant legal basis, and virtual currency has not been recognized by government departments. Network money laundering means that criminals taking advantage of the ability to conceal their identities on the Internet to hide the source of funds with the help of the Fintech transaction shell, and thereby transfer funds and whitewash the illegal income.

Finally, there are the risk of internal fraud and regulations and the risk of external fraud, accounting for 0.0610 and 0.0963, respectively. Due to the rapid development of Internet finance, the current legislative situation of China has a certain lag. Some emerging Internet finance fields have not issued special policies to regulate their development yet. They can only refer to other relevant laws and regulatory measures, resulting in a mismatch between legal supervision and the cases.

(6)

Market risk

First, according to the weight distribution W₆ = (0.1947, 0.0881, 0.7172), the most important risk is price movement risk, accounting for as high as 0.7172. The second is interest rate risk, which accounts for 0.1947. Finally, the exchange rate risk accounts for 0.0881.

(7)

Legal risk

It can be seen from W₇ = (0.0649, 0.0384, 0.1320, 0.2270, 0.5377) that the online money laundering risk accounts for the largest proportion, which is 0.5377. The second is the subject qualification risk and virtual currency risk, accounting for 0.1320 and 0.2270, respectively. Finally, the laws and regulations absence risk accounts for 0.0649. Meanwhile, the regulatory vacancy risk accounts for 0.0384.

The significance of modeling in this paper lies in the level of each risk level obtained according to the empirical analysis, so as to carry out risk prevention that is targeted and focused. The higher the risk evaluation value, the greater the prevention strength should be.

4. Conclusions

With the emergence of Fintech, the disruptive changes in the identification, acquisition and assessment of credit will inevitably lead to fundamental changes in the way financial transactions are undertaken, the logic of financial product design and the operation mechanism of the financial market. The complexity of Fintech’s business model has led to the continuous renovation of the manifestations and connotation of financial risk, which increases the difficulty of risk identification and the speed of risk transmission. Technology itself is not only the driving force in the development of Fintech, but also the risk point of Fintech.

Looking ahead, we should firstly clarify the basic characteristics of the of Fintech risks such as complexity, endogeneity, non-equilibrium and variability. Secondly, we should further identify and quantify the new risks brought by financial technology, especially the new risks involving technological change. Finally, we should use regulatory technologies (reg Tech) such as big data, artificial intelligence and cloud computing to enrich regulatory means and strengthen the construction of financial technology infrastructure from the construction of financial technology platforms, compliance technology applications, financial security and anti-fraud technology development, and prevent financial technology risks and their impact on the existing risk system.