Will Off-Balance-Sheet Business Innovation Affect Bank Risk-Taking under the Background of Financial Technology?

Abstract

Given the rapid development of financial technology, the off-balance-sheet business innovations of banks may potentially impact bank risk-taking. This issue is of great importance to commercial banks and financial regulators. This paper analyzed the relationship between off-balance-sheet business innovation (OBI) and Bank Risk-Taking (BRT) in Chinese commercial banks, as well as the mediation role of the Bank Agency Cost (BAC), the impact of a bank’s Internal Control Quality (ICQ) on this relationship, and the moderating role of Bank Competition (BCMP) by analyzing panel data from a sample of 130 Chinese commercial banks from 2009 to 2019. The results of this empirical exercise showed that (1) OBI has a significant negative correlation with BRT, evidencing that off-balance-sheet business innovation can improve bank risk management processes and enhance the bank’s operating performance, thereby reducing their willingness to transfer risks, restraining the BRT level. Compared with state-owned and joint-stock banks, OBI has a more significant inhibitory effect on BRT in urban and rural commercial banks. (2) BAC showed a mediation role in the relationship between OBI and BRT levels. Bank OBI can inhibit BRT levels by BAC reduction, demonstrating an effective mediation channel. (3) The degree of BCMP displayed a positive moderation effect on the relationship between the explained and explanatory variables, which means that, at higher BCMP levels, the inhibitory effect of OBI on BRT levels becomes more significant. (4) Additionally, this exercise also found that a bank’s ICQ can enhance the impact of OBI on BRT. The research contributions of this paper constitute an important theoretical significance and reference value for researchers exploring mechanisms that can improve innovation in the commercial banking industry and give importance to financial supervision and financial system risk control.

1. Introduction

In recent years, the Chinese economy has been gradually entering a new stage of slower growth and structural adjustment, where real economy financial support and innovation vitality have become determinants of a successful supply-side reform [1]. The Chinese financial sector has traditionally been dominated by bank credit, and some structural problems have emerged around supporting the real economy. Fund loan and deposit term mismatch can quickly increase bank risk, and the original intention of providing financial support for the real economy becomes more difficult to achieve [2]. In order to face this challenge, the Chinese banking industry is giving play to off-balance-sheet business innovation in (1) improving the efficiency of banking operations, (2) promoting bank risk management reform, and (3) improving the profitability of the industry; thus, anti-risk capabilities within the banking industry have acquired higher relevance.

Off-balance-sheet business innovation (OBI) generally refers to applied innovation in financial products, processes, services, and organizational forms; it is an activity that produces and popularizes financial instruments, institutions, markets, and technologies [3,4]. During the global financial crisis of 2008, off-balance-sheet business innovation became one of the most important objectives for the national policies, legal systems, and regulatory frameworks that supported off-balance-sheet business innovation and financial liberalization [5]. However, the Asian financial sector received criticism because of its poor levels of innovation. The introduction of new policies, techniques, skills, and other new ideas brought through innovation is termed as “fresh blood”. Scholars have pointed to the absence of “fresh blood” in the development of the financial industry, economic repression, and underdeveloped infrastructure as factors that restricted healthy development in the financial sector. Nevertheless, since the global financial crisis, a less optimistic approach to innovation has argued that OBI might be precisely one of the main factors leading to this situation [6,7,8]. The ramifications of OBI have included unprecedented challenges in terms of its risks and its implementation in traditional financial models [9].

Commercial banking plays a crucial role in the financial sector and in the global economy; continuous off-balance-sheet business innovation and fresh blood are already advantages in these fiercely competitive environments. From this angle, the level of development presented by Chinese commercial banks has not been as competitive as those of its counterparts in developed countries. Some researchers maintain that OBI represents an important mechanism for enhancing Chinese banks’ competitiveness at the global level [10]. The rapid development of financial technology has promoted the transformation and upgrading of commercial banks, making the proportion of off-balance-sheet business conducted by banks higher. In addition, financial technology has also intensified market competition and increased financial risks. Therefore, it is urgent and necessary to consider the impact of OBI on the level of BRT against the backdrop of the vigorous development of financial technology. However, the key predicament lies in the relationship between bank risk-taking (BRT) and OBI. Does OBI have a significant effect on BRT? If yes, is this effect conducive to higher BRT, or does it have a suppressing effect? Moreover, some questions worthy of analysis emerge for this study: Is there heterogeneity in the impact of OBI on BRT? Through which channels does OBI affect BRT? Does the competitive environment in the Chinese banking industry in any way influence the relationship between OBI and BRT? Is the quality of a bank’s internal control (BIC) a factor with a significant effect on restraining BRT? This study’s significance lies in analyzing bank risk prevention in the Chinese banking industry for the ultimate purpose of contributing to its function of serving the real economy.

Throughout the existing research, there are many pieces of research on the financial innovation of banks, but there are still few studies in the literature focusing on OBI. Moreover, the existing research mainly considers the problem of banks’ risk-bearing from perspectives such as bank competition, interest rate marketization, and economic policy uncertainty. Few studies discuss the impact of OBI on the BRT level in financial technology. This paper attempts to analyze from this perspective to promote the research on the economic consequences of the OBI of banks. This empirical exercise contributes to actual research by exploring the following aspects of the effect of OBI on BRT: (1) With a focus on the relationship between OBI and BRT, and its subjacent mechanisms, panel data from 130 Chinese commercial banks in China between 2009 and 2019 were also analyzed to assess the impact of digital finance development on BRT; (2) Chinese commercial banks were classified into three categories, state-owned, join-stock, and urban–rural, allowing for a comparison of the nature of the OBI–BRT relationships among categories; (3) based on the recursive model for mediation proposed by [11], this study examined the intermediary role of the Agent Cost (AGC) in the relationship between OBI and BRT; (4) this study also examined the moderating effect of bank competition in the relationship between OBI and BRT; (5) by considering the importance of Banks’ Internal Control (BIC), this exercise explored the potential the differential impact of OBI on BRT for different quality levels of BIC.

The sustainability of the banking system is dependent on monitoring the risks, especially the off-balance-sheet risks, and there is an emerging need to address these issues to really make the banking system sustainable in the future. The results presented in this paper are of relevant theoretical significance and can be used as a reference for decision-making regarding off-balance-sheet business innovation-related policies toward systemic risk prevention and financial security in the Chinese banking industry. This paper presents its literature review and research hypothesis in Section 2 and its empirical research design in Section 3, while its empirical test results and analysis are presented in Section 4, and Section 5 presents the conclusions and recommendations of the study.

2. Literature Review and Research Hypothesis

2.1. Effects of Off-Balance-Sheet Business Innovation

Lee-Ying Tay et al. found that developing countries, mainly Asian countries, embrace and improve digital financial inclusion to help reduce poverty. However, the results indicate that, in developing countries, a persistent divide exists between different genders, the wealthy and the poor, and urban and rural areas regarding access to and usage of digital financial services [12]. Xiangrui Chao et al. reviewed the application of smart technology in financial stability regulation, analyzed the objects and results of the technology’s applications, and formed a clear context for its development, and they serve as the support and development foundation for financial stability research [13]. Current research on off-balance-sheet business innovation (OBI) is focused on its economic consequences, benefits, and associated risks for the traditional financial market. By emphasizing the benefits of OBI for the economy and approaching the issue from the perspective of taxation, Miller and Tufano proposed that OBI can provide financial institutions with a reasonable degree of tax avoidance and may play a role in regulatory arbitrage by reducing financial turbulence [14,15]. McConnell and Schwartz analyzed a case of liquid interest-bearing bonds and argued that OBI may provide an appropriate investment plan for commercial banks’ consumers and lower their transaction costs [16]. Mohd Javaid et al. focused on blockchain technology and its importance for financial services, and their results indicate that blockchain-based systems enable the faster, more cost-effective, and more customized issuance of digital securities [17]. Jameson, Dewan, et al. analyzed the income obtained from securitization in collateralized mortgage obligations, maintaining that their proposal may reduce mortgage rates and promote the conclusion of transactions between banks and customers [18]. Hasanul Banna et al. investigated whether a higher degree of fintech-based financial inclusion (FFI) intensifies banks’ risk-taking by analyzing data from 534 banks from 24 OIC countries, and the results indicate that a higher degree of FFI controls banks’ risk-taking behavior [19]. Hasanul Banna aims to examine the role of digital financial inclusion in promoting sustainable economic growth through banking stability in Bangladesh, and the results suggest that digital financial inclusion leads to economic growth and that digital financial inclusion by banks must be implemented carefully for the economic stability of the bank itself [20].

By studying equity-linked notes, Tufano and Sevick determined that the proposed innovative financial product may reduce bank costs by deferring the payment of capital income tax [21]. Guner considers that OBI may reduce the borrowing cost of loans in the sales of banks’ wealth management products [22]. Hirtle argued that the use derivatives by banks increases the credit supply to large companies and reduces the average loan spread between companies [23]. Weisbach argued that, during the securitization process, OBI could reduce the borrowing cost of loans [24].

Research evidence also shows that Chinese scholars have also researched the role of commercial banking OBI development in non-interest income growth, the reduction of the effect of macroeconomic fluctuations on banks’ operating income, and systemic risk reduction. Zhang and Mao, as well as Wu, pointed out that Internet finance, supported by the advantages of information technology, may reduce transaction costs and promote financial reform [25,26]. Quan and Wang classified OBI in project supervision cost reduction and ROI improvement [27]; this study showed that project supervision cost-oriented off-balance-sheet business innovation could improve banks’ capital utilization and reduce BRT. Liu also proposed that the swift development of Internet finance may have reduced banks’ bankruptcy risk and systemic financial risks while increasing the overall stability of the financial system [28]. The empirical research by Shen and An found that non-interest cost reduction may contribute to curbing BRT levels [29].

2.1.1. The Role of OBI Inhibiting BRT

Based on the references mentioned above, it is possible to infer that OBI in banking may inhibit BRT through (1) bank risk management reform and (2) banks’ operational efficiency as follows:

Marcelin et al. explore how information-sharing and financial inclusion influenced bank risk in 84 countries from 1996 to 2020, and the results show that greater information-sharing and financial inclusion lessen bank risk levels [30]. From the bank risk management reform angle, OBI entails using emerging technologies to break the traditional data structure within commercial banks, increasing the dimension of obtaining data, and improving data accuracy and customer screening ability. Secondly, using emerging technologies in OBI may improve the measurement and follow-up perspective of BRM-related data, improving risk identification and risk assessment accuracy; moreover, the role of technology in OBI, by making banking services more convenient, can also improve a bank’s work efficiency and reduce bank management costs. Overall, OBI can improve risk management and control-related processes and simplify the bank risk management process through better information platforms, improving its efficiency [28].

From the perspective of banks’ operational efficiency, through the use of these information technology advantages, OBI in banking can promote bank financial reforms, transaction costs reduction [26], and operating efficiency improvements; in the same way, through technological sophistication and new service concepts, OBI can encourage commercial banks to improve their business model, technology, and operational efficiency, thus having a positive effect on bank operating profits. In addition, commercial bank OBI can also safeguard bank financing effectiveness, maintain the financial system’s stability, and promote bank development and bank risk resistance.

2.1.2. Negative Effects of Off-Balance-Sheet Business Innovation

As for the negative effects of OBI on the economy, OBI may be a way for financial companies to evade financial supervision and control, consistent with the avoidance financial innovation theory [31]. Henderson and Pearson proposed that OBI may cause investors to have an incomplete or imprecise understanding of the capital market [7]. Financial professionals may use this weakness to design financial products that induce investors to improve financial product valuation and generate financial risks. Aghion, Comin, et al. consider that pursuing high-risk and high-profit products in OBI may configure innovator rent-seeking behavior, increasing the risk level of financial enterprises [6]. Henderson and Pearson believe that economic prosperity may hide risks brought on by financial product innovation [7]. Moreover, when commercial banks face higher credit-related or market-related risks, financial product innovations may increase the possibility of losses within commercial banks [32,33]. Evidence also showed that OBI might be seen as equivalent to financial industry resource reallocation, where the constant search for the optimal combination of factors may have a negative impact on financial security [34]. Chen suggested that financial product innovation may be conducive to concealed financial risks, which may be detrimental to financial supervision [35]. Another perspective suggests that bank OBI may provide investors with higher liquidity. However, the leverage created by this additional liquidity may bring with it subjacent systemic financial risks, so OBI not only can increase the asset leverage of commercial banks, but also will take funds away from supervision tasks, increasing the BRT level of commercial banks [36].

While studying the effect of financial product innovation on banking business operations, Feng and Huang proposed that financial product innovation may negatively affect banks; even its complexity and information opacity will increase banks’ risk-bearing levels [37]. Dai and Fang see OBI as a product of interest rate regulation, and its development may accelerate interest rate marketization, increasing bank financing costs and commercial banks’ risks [38]. Evidence also showed that OBI might significantly and immediately negatively impact BRT [39]. Another empirical study found that the relationship between internet finance and BRT displayed a U-shaped effect [40], with BRT reduction in the short term motivated by management cost reduction, followed by a further BRT increase in the longer term motivated by the subsequent rise in capital costs.

2.1.3. Off-Balance-Sheet Business Innovation and Bank Risk-Taking

Based on the above-described evidence, it is possible to infer that OBI may promote BRT from the perspective of risk contagion, exposing banks to risks. Specifically speaking, (1) in terms of credit risk, financial products produced by OBI usually are replicable, leading banks to compete in common target segments. This fiercer competition is conducive to reductions in bank profitability, excess loan granting, and over-investment in pursuit of higher profits, causing subsequent credit risks. (2) In terms of the liquidity risk, on-balance and off-balance assets work as a bank’s leverage while attracting deposits. However, if either a reversed economic situation or a financial crisis emerges, large-scale capital run and capital transfer may spread the risk to the commercial banking system. (3) In terms of operational risk, OBI implementation inevitably exposes the banking system to risks, with its subsequent effects on system security and stability. Given the banking system’s timeliness and convenience, implementing unmatured technology also risks information security. In other words, OBI is likely to increase operational risk.

Given the respective literature review and theoretical analysis, competing hypotheses H1a and H1b can be stated as shown in Table 1 as follows:

2.2. Moderating Effect of Banking Market Competition

Since the implementation of the financial reform in China in 1984 and the market liberalization of 2014, while interest rate marketization and competition have been continuously promoted and strengthened, regulatory authorities and academic circles have brought attention to the relationship between market competitiveness and financial system security. Given this context and the research evidence, it is possible to infer from a positive perspective that OBI may not only reduce BRT through risk change management and operational efficiency, but also by increasing market competitiveness. However, the opposite perspective suggests that OBI may increase BRT due to a decline in profitability and risk contagion as banking competition stimulates bank profit pressure and business model change, so it is necessary to pay attention to the mechanism whereby competition influences the relationship between OBI and BRT.

Research evidence proposed that, in a concentrated industry, bargaining advantages allow banks to obtain higher loan interest rates, further increasing the financing cost of enterprise loans [41]. In contrast, higher industry competition levels and continued reductions in loan interest margins may be conducive to conservative credit and investment strategies that allow bank managers to reduce the liquidity mismatch of short-term and long-term loans; in this way, by strengthening credit and investment supervision, banks may be able to reduce BRT.

In addition, a bank with strong market power will have an easier to acquire invisible bankruptcy protection, leading banks to invest in high-risk areas. The complexity of these banks’ business leads to higher supervision difficulty and, thus, to higher risk. On the other hand, due to limited market resources, a highly competitive bank would tend to improve its operating efficiency, as well as its early risk warning mechanism and its risk management level, all conducive to BRT reduction [42]. Consequently, when OBI presents an inhibitory effect on BRT, higher competition in the banking industry will be conducive to more conservative investment strategies, and regulators will also strengthen the banking business supervision. Since banking competition can constitute an external governance mechanism, it can further improve bank risk management reform and operating efficiency to reduce risk accumulation and promote the inhibitory effect of OBI on BRT.

Nevertheless, it is also important to consider the franchise value hypothesis, where increased bank competition may reduce deposit and loan spreads, weakening bank market power and franchise rights values. Banks may be forced to seek profits to face this profit-shrinking situation, increasing their BRT levels by investing in riskier projects [43]. Using empirical evidence from 60 developing countries, [38] supported this hypothesis: bank competition level had a significant positive correlation with BRT.

While OBI inhibits BRT levels, a higher bank competition level (BCMP) will motivate banks to prefer aggressive investment strategies. In this case, OBI may hinder the transmission path of risk management changes and operating efficiency. Thus bank competition may lead to further accumulation of banking risk, thereby reducing the inhibitory effect of OBI on BRT. Based on the above reasoning, this paper proposes its second competitive hypothesis, refer Table 2.

2.3. Agent Cost Mediating Effect

The information asymmetry between bank shareholders and creditors and between shareholders and managers causes a principal–agent problem that constitutes an important factor for BRT [44]. Bank governance is responsible for solving adverse selection and moral hazard issues; FIN extends the traditional financial service network chain, often reducing management costs through technology, work efficiency improvement, and service model innovation. In addition, OBI has also produced new wealth management tools that have influenced bank pricing mechanisms, affecting the supply and demand of capital traditionally dominated by commercial banks, and have promoted the marketization of financial factor prices. All of these effects contributed to the reduction of information asymmetry and, subsequently, to bank agency cost (BAC) reduction.

According to the principal-agent theory, a lower bank agency cost (BAC) can increase the bank’s profitability. In other words, reducing profitability pressure can subsequently motivate banks to reduce transfer risks to depositors. At the same time, with the support of emerging technologies, bank loan risk can be identified and monitored more efficiently, thus allowing banks to control BRT more effectively [45]. Given this logical reasoning, it is possible to define hypothesis H3, refer Table 3.

3. Research Design

3.1. Data Sources and Sample Selection

This empirical exercise analyzed data and macro data from Chinese commercial between 2009 and 2019, refining the sample by applying two basic filters: (1) policy banks and foreign banks were excluded from the sample; then (2) samples with missing data for three consecutive years were also excluded. The refined sample result showed unbalanced panel data for 130 banks, including 5 state-owned, 12 joint-stock commercial, and 113 urban and rural commercial banks, to obtain an overall sample of 1,188 annual observations. The corresponding microdata for this paper come from the Wind database [46], the Bankfocus database [47], and the EPS database [48].

3.2. Research Model

3.2.1. Benchmark Model

To assess the impact of FIN on BRT, this exercise constructed the following Benchmark Model:

(1)${\mathrm{BRT}}_{\mathrm{i},\mathrm{t}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1{\mathrm{OBI}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_2{\mathrm{X}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_3{\mathrm{M}}_{\mathrm{t}}+{\mathsf{\alpha }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{i},\mathrm{t}}$

where ${\mathrm{BRT}}_{\mathrm{it}}$ represents the explained variable (bank risk-taking) for each ith bank for the jth year; the proportion of bank weighted risk assets and the Z score are defined as proxy variables. The core explanatory variable is represented as ${\mathrm{OBI}}_{\mathrm{it}}$ (Off-balance-sheet business innovation), and the proportion of bank non-interest income was defined as the alternative variable. Its coefficient represents the impact of OBI on BRT, the most relevant in this study.

${\mathrm{X}}_{\mathrm{it}}$ represents the control variables that change over time at the bank level; ${\mathrm{M}}_{\mathrm{t}}$ the control variables at the macro level; ${\mathsf{\alpha }}_{\mathrm{i}}$ the corresponding fixed effect used to control the individual static characteristics; and ${\mathsf{\mu }}_{\mathrm{it}}$ the random error term for this model.

3.2.2. Heterogeneity Test

The empirical exercise for this paper divided the sample according to three different classifications for commercial banks, (1) state-owned banks, (2) joint-stock banks, and (3) urban–rural banks, to assess the heterogeneity level of the impact of OBI on BRT.

3.2.3. Moderating Effect Test

To further research the relationship between OBI and BRT, this paper sets the following model to assess the moderating effect of BCMP on this relationship.

(2)${\mathrm{BRT}}_{\mathrm{i},\mathrm{t}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1{\mathrm{OBI}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_2{\mathrm{BCMP}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_3{\mathrm{OBI}}_{\mathrm{i},\mathrm{t}}·{\mathrm{BCMP}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_3{\mathrm{X}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_4{\mathrm{M}}_{\mathrm{t}}+{\mathsf{\alpha }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{i},\mathrm{t}}$

where $\mathrm{BCMP}$ (Bank competition level) represents the moderating variable, and higher values represent a stronger competitive position. Given a significant negative correlation between OBI and BRT, if BCMP presents a significant positive moderation effect in the proposed relationship, then the interaction coefficient ${\mathsf{\alpha }}_2$ will be negative and significant, indicating a degree of change if OBI impacts BRT for each unit of BCMP. Otherwise, if ${\mathsf{\alpha }}_2$ is significant and positive, it will indicate that BCMP has a negative moderation effect on the relationship between OBI and BRT.

3.2.4. Mediation Effect Test

The tests mentioned above only assess the effects of OBI on BRT; however, they do not explore the transmission channel of this effect. To further analyze this aspect of the model, this empirical exercise proposed Bank Agency Cost (BAC) as the mediation variable and constructed the progressive model described in Equations (3) and (4).

(3)${\mathrm{BAC}}_{\mathrm{it}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{OBI}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }}_2{\mathrm{X}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\beta }}_3{\mathrm{M}}_{\mathrm{t}}+{\mathsf{\alpha }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{i},\mathrm{t}}$

(4)${\mathrm{BRT}}_{\mathrm{it}}={\mathsf{\delta }}_0+{\mathsf{\delta }}_1{\mathrm{OBI}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\delta }}_2{\mathrm{BAC}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\delta }}_3{\mathrm{X}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\delta }}_4{\mathrm{M}}_{\mathrm{t}}+{\mathsf{\alpha }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{i},\mathrm{t}}$

This testing process is composed of three steps. The first step will assess the significance of the relationship between OBI and BRT through the model described above in Equation (1); then, the use of the model described in Equation (3) assesses the significance of OBI as an explanatory variable for the mediator BAC. Finally, the explained variable BRT, the explanatory variable OBI, as well as the mediator BAC are included in the regression model as described in Equation (4).

3.3. Variable Definition

3.3.1. Explained Variable: Bank Risk-Taking

To determine the proxy variable that explains BRT, the authors explored relevant existent literature on risk level in commercial banks and BRT [49,50], selecting the weighted risk assets RWA [40] and the Z value [51] as proxy variables for BRT. The proportion for RWA is determined by the ratio of bank net-weighted assets to the total of assets; a higher value means a higher risk taken by a determined bank in a determined time. The Z score indicates the bank’s bankruptcy probability, specifically referring to the standard deviation of the bank’s return on total assets divided by the sum of the return on total assets and the equity ratio. A higher value for Z means a higher risk.

This paper also analyzed the bank loan loss reserve adequacy ratio as a substitute ratio for the robustness test applied in this empirical exercise. This value is determined by the ratio of the amount of the loan loss reserves to the amount of the required loan loss provision. A larger value will mean a lower level for BRT.

3.3.2. Core Explanatory Variables

To define OBI, this paper used the proportion of bank non-interest income to total operating income as a proxy variable for off-balance-sheet business innovation [52]. This perspective argues that the development of an intermediary business constitutes an important indicator of measure of the bank’s off-balance-sheet business innovation capability, as the income generated by the intermediary business, investment, and consulting is linked to innovation and constitutes most of the non-interest bank income.

3.3.3. Control Variables

Based on previous studies [48,49,50], this paper identified a set of control variables at the bank and macroeconomic levels, detailed as follows. On the bank level, this exercise included the growth rate of bank loans; bank leverage, in terms of equity-to-asset ratio; operational efficiency in terms of cost-to-income ratio; bank deposit-to-loan ratio; return on assets (ROA), expressed as the ratio of net income to total assets; and capital adequacy ratio (CAR), as the ratio of bank capital to risk-weighted assets.

Regarding the macroeconomic level, this study considered the consumer price index percentage (CPI); the Shanghai composite Index (Stkdex), expressed as its year-end value natural logarithm; and gross domestic product growth (GDPg).

3.3.4. Moderator Variable: Bank Competition Level

Based on [53,54,55], this empirical exercise used the Lerner Index as a proxy for the bank competition level (BCMP); it was calculated following the procedure described.

(1) The first step determines bank marginal cost as proposed by [55], where a transcendental logarithmic cost function determines the cost function, and its first derivative defines the marginal cost for each bank. The estimation of the bank cost function is defined by a panel data stochastic model and described as follows.

(5)$\ln {\mathrm{TC}}_{\mathrm{it}}={\mathsf{\lambda }}_0+{\mathsf{\lambda }}_1\ln {\mathrm{TA}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\lambda }}_2\ln {\mathrm{TA}}_{\mathrm{it}}{}^2+{\displaystyle \sum }_{\mathrm{j}=1}^3{\mathsf{\eta }}_{\mathrm{j}}\ln {\mathrm{w}}_{\mathrm{j}\mathrm{i},\mathrm{t}}{}^2+{\displaystyle \sum }_{\mathrm{j}=1}^3{\displaystyle \sum }_{\mathrm{k}=1}^3{\mathsf{\delta }}_{\mathrm{jk}}\ln {\mathrm{w}}_{\mathrm{j}\mathrm{i},\mathrm{t}}\ln {\mathrm{w}}_{\mathrm{k}\mathrm{i},\mathrm{t}}+{\displaystyle \sum }_{\mathrm{j}=1}^3{\mathsf{\mu }}_{\mathrm{j}}\ln {\mathrm{TA}}_{\mathrm{i},\mathrm{t}}\ln {\mathrm{j}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}}$

where ${\mathrm{TC}}_{\mathrm{it}}$ represents the total cost for bank i for the t year, including bank interest expenditure and non-interest expenditure. ${\mathrm{TA}}_{\mathrm{i},\mathrm{t}}$ represents bank total output, expressed in terms of total assets; ${\mathrm{w}}_{\mathrm{j}\mathrm{i},\mathrm{t}}$ represents the capital cost and labor of bank i for year t, where j-values from 1 to 3 determine the type of cost, specifically; ${\mathrm{w}}_1$ represents capital cost, expressed as the ratio of interest expense to total deposits; ${\mathrm{w}}_2$ is associated with labor cost, expressed as the ratio of management expenses to fixed assets; and ${\mathrm{w}}_3$ represents asset cost, expressed as the ratio of non-interest expenses to fixed assets.

(2) The bank’s marginal cost ${\mathrm{MC}}_{\mathrm{it}}$, a constituent of the Lerner index, is calculated from the combination of panel data, the regression fixed effect model, and its respective model parameters, as expressed in the following equation:

(6)${\mathrm{MC}}_{\mathrm{i},\mathrm{t}}=\frac{{\mathrm{TC}}_{\mathrm{i},\mathrm{t}}\left({\mathsf{\lambda }}_1\ln {\mathrm{TA}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\lambda }}_2\ln {\mathrm{TA}}_{\mathrm{it}}{}^2+{{\displaystyle \sum }}_{\mathrm{j}=1}^3{\mathsf{\mu }}_{\mathrm{j}}\ln {\mathrm{w}}_{\mathrm{j}}\right)}{{\mathrm{TA}}_{\mathrm{i},\mathrm{t}}}$

(3) Then, the Lerner index is calculated according to Equations (7) and (8) as follows:

(7)${\mathrm{P}}_{\mathrm{i},\mathrm{t}}=\frac{{\mathrm{IntInc}}_{\mathrm{i},\mathrm{t}}+{\mathrm{NIntInc}}_{\mathrm{i},\mathrm{t}}}{{\mathrm{TAsset}}_{\mathrm{i},\mathrm{t}}}$

(8)${\mathrm{Lerner}}_{\mathrm{i},\mathrm{t}}=\frac{{\mathrm{P}}_{\mathrm{i},\mathrm{t}}{\mathrm{MC}}_{\mathrm{i},\mathrm{t}}}{{\mathrm{P}}_{\mathrm{i},\mathrm{t}}}$

where ${\mathrm{P}}_{\mathrm{i},\mathrm{t}}$ represents the average price of the bank’s output in year t, calculated from the interest income ${\mathrm{IntInc}}_{\mathrm{i},\mathrm{t}}$, non-interest income ${\mathrm{NIntInc}}_{\mathrm{i},\mathrm{t}}$, and the total of assets ${\mathrm{TAsset}}_{\mathrm{i},\mathrm{t}}$.The values from the Lerner index vary from 0 to 1, where 0 represents a perfectly competitive market, and 1 means a monopolistic market; in other words, higher values mean a weaker bank competition level n. Nevertheless, for the empirical exercise described in this paper, the variable BCMP has been defined as $\mathrm{BCMP}=1-{\mathrm{Lerner}}_{\mathrm{i},\mathrm{t}}$, meaning that higher BCMP values will mean a higher bank competition level.

3.3.5. Mediation Variable: Agency Cost

Bank agency cost (BAC) can be an indicator of the absence of corporate governance; based on [56], and this empirical exercise used the management expense ratio (the proportion of management expenses to operating income) as a proxy for BAC.

Overall, all variable definitions are summarized in Table 4 as follows.

3.4. Descriptive Statistics

Descriptive statistics for the entire set of variables within the selected sample are given in Table 5. The results for the explained variable showed a notable variation in RWA and Z-score within the sample. The explanatory variable OBI displayed a mean value of 0.1995 and a SD of 0.01673 with a median of 0.1559, also reflecting a certain degree of difference in OBI among samples. In general terms, these results are consistent with the existing literature.

3.5. Unit Root Test

In addition, the unit root test is performed to ensure the reliability of the empirical results, and the results of the unit root test are given as Table 6, which shows that, at the 10% significance level, the statistical values of the series are less than the critical value, and the p values are less than 0.05, so there is no unit root and the series are stationary time series.

4. Empirical Test and Result Analysis

4.1. Impact of OBI on BRT

The F-test and Hausman test results indicated that a fixed effect model was appropriate for this sample for analyzing OBI’s impact on BRT empirically. According to the results in Table 7, the coefficient for OBI on RWA column (1) was −0.032, significant at a 1% level. For each additional unit of OBI, the proportion of RWA will decline by 0.032%, supporting a significant negative correlation between the bank OBI and the proportion of bank RWA. The results in column (2), with a coefficient value of −0.760, significant at 5%, showing that, for each additional unit of OBI, the bank’s Z-score will drop by 0.760%, also supported a significant negative correlation between OBI and BRT, in this case represented by the Z-score. Based on these results, it is possible to support H1 for the whole sample. As Chinese commercial banks are actively involved in risk management reforms and operating efficiency improvements, these activities can make them able to compensate for the adverse shocks caused by OBI, reducing the BRT level. Consequently, it is reasonable to infer that bank OBI may effectively restrain the risk behavior in commercial banks and promote the financial system’s stability.

The regression results for RWA discriminated by type, displayed in columns (3) to (8), assessed the role of the heterogeneity of the bank’s nature on the impact of bank OBI on BRT. The coefficient of OBI in columns (3), (4), and (6) is not significant at 10%, but the coefficient in columns (5), (7), and (8) is significant; this shows that, compared with state-owned banks and joint-stock banks, the inhibition effect of OBI on BRT is more significant in the sample of urban–rural commercial banks. For each unit of OBI, the proportion of urban–rural commercial banks’ RWA will decline by −0.025%, and their Z-score will decline by 0.639%. This can be explained by the fact that Chinese state-owned and joint-stock banks have the characteristics of systemic relevance, such as fuzzy ownership, stricter supervision, larger scale, and broader business scope. Therefore, on one side, these banks tend to have a greater degree of OBI. Nevertheless, their response to OBI measures may be slower, facing a delay in the risk response strategy adjustment and operating efficiency improvement necessary to manage the impacts of OBI.

On the other side, regional banks’ smaller scale, narrower business scope, and flexible management allow them to respond flexibly to the impact of OBI. These banks use their geographic and business expertise to accumulate information and cost-related advantages. Therefore, their risk identification ability is less disturbed, resulting in lower levels of BRT.

4.2. Extended Model: The Moderation Effect of Competition Level

The intensity of bank competition may influence a bank’s risk management, business model, and operating efficiency, which may be reflected in a significant moderation effect on the relationship between OBI and BRT. As explained previously, this empirical exercise assessed BCMP through the Lerner Index, and the result is displayed in Table 8.

In regard to the combined effect of OBI and BCMP, column (2) showed a multiplier coefficient of −0.221 significant at 10%, indicating that the negative impact of bank OBI on the proportion of RWA increases by 0.221 percentage points per each unit of increase on BCMP. In other words, BCMP can positively moderate the inhibitory effect of off-balance-sheet business innovation on a bank’s risk level. While assessing the Z-score, the results in column (4) showed a multiplication coefficient of −12.516, significant at the level of 1%, indicating that the negative effect of bank off-balance-sheet business innovation on a bank’s Z value increases by 12.516 percentage points for each point of increases in BCMP.

Based on these results, an increase in BCMP can significantly promote the restraining effect of OBI on the BRT level, meaning that BCMP can effectively increase the bank risk resistance level while facing an environment that promotes OBI; it is possible to support H2a for the whole sample. Two reasons may explain this: (1) Increased banking competition can diminish bank interest margins, reducing bank management expectations for future cash flows and promoting a proclivity to adopt strategies in favor of credit and investment strategies, thus resulting in a restraining effect of OBI on bank BRT. On the other hand, (2) given higher industry competition and limited market resources, banks will tend to improve their operating efficiency and risk management levels, in the end reducing BRT and thereby moderating the effect of OBI on BRT.

4.3. Expansion Model: Identifying the Mediation Mechanism

The results supported that bank OBI may reduce BRT level by showing a significant effect caused by both measured proxy variables, the RWA and Z-score; however, to deepen the analysis of the transmission channel of this relationship, this paper hypothesized that the bank agency cost might constitute a suitable mediator for this model.

This study uses a recursive model to test the intermediary effect of OBI on BRT, as displayed in Table 5; the significance of the direct effect between OBI and BRT has already been explained in the previous section and in Table 8, as well as in columns (1) and (2) in Table 9. As shown in column (3), the effect of OBI on BAC showed a coefficient of −0.001 significant at the 5% level, which means that OBI can significantly reduce the BAC, thus reducing the bank’s management expense rate. Furthermore, columns (4) and (5) both show a significant and negative coefficient for OBI as explanatory for BRT (−0.026 significant at 1% level for RWA and −0.631 significant at 5% level for Z-Score), but also a significant positive effect when testing BAC as explanatory for BRT (6.367 for RWA and 210.309 for Z-Score, all significant at 1% level).

This indicates that OBI tend to encourage banks to improve their technical levels and work efficiency and urges them to inspect and optimize risk control procedures and processes, reduce shareholding-management agency costs, reduce bank information asymmetry, and reduce their willingness to take risks. In addition, this reduction in agency costs generated by OBI can also help banks to improve their operating efficiency through demonstration effects, reducing banks’ willingness to pass on risks, or their risk behavior restriction, and subsequently reducing BRT. Therefore, the mechanism of “OBI-BAC-BRT” is established, and Hypothesis H3 is verified.

4.4. Model Robustness Test

To examine the robustness of the restraining effect of OBI on BRT, this exercise applied multiple methods, described as follows:

• Firstly, the explanatory variable BRT was substituted by the loan loss reserve adequacy ratio (LLR) to re-examine the relationship proposed in this model, refer Table 10. LLR is the ratio of the current amount of the bank loan loss reserves to the amount of the loan loss reserves that should be withdrawn, so higher values for LLR are associated with lower risk exposure; the corresponding results are displayed in Table 11. Column (1) indicates that OBI can significantly increase LLR (2.001 significant at a 5% level), also indicating a reduction of BRT, demonstrating the robustness of the core findings of this empirical exercise.

• Secondly, to alleviate possible endogenous problems in the relationship between OBI and the BRT level, this paper added a one-period lagging term of the explained variable to the model. As shown in columns (2) to (5), OBI showed significance at 1% and 5% levels for RWA and Z-Score with negative coefficients (−0.0253, −0.800, −0.026, and −0.0651), consistent with the inhibitory role of OBI on BRT.

• Thirdly, supplementary control variables such as the Banking Industry Prosperity Index (BPI), the Banker’s Macroeconomic Confidence Index (BCI), and the Lending ratio (PLR) were added to the analysis to re-examine the relationship between OBI and BRT. The respective empirical results showed that, regardless of the proxy variable used for the explained variable BRT (RWA or Z-Score), the OBI coefficients were all significant at a 5% level, indicating once more a negative correlation between OBI and BRT, as is consistent with previous findings.

4.5. Subsample Testing: OBI, Internal Control Quality, and BRT

OBI’s impact on commercial banks can be expressed in two dimensions. From one angle, OBI can be associated with banks’ technical improvements, efficiency increases, and management cost reduction, reducing banks’ risk-taking willingness. From another angle, the emergence of innovations can also be associated with bank deposit diversion and price competition. This phenomenon poses a significant challenge to the banking and financial supervision systems.

Commercial banks rely on internal control as an important managerial tool associated with bank risk management and control to achieve their goals by following the formulation and implementation of administrative systems and methods. An effective internal control procedure can mitigate the risks caused by information asymmetry, contributing to bank asset safety and stability maintenance in bank operations and management [57]. Based on the definition from Basel Bank and the COSO Committee (Committee of Sponsoring Organizations), and considering the characteristics of the Chinese environment, the People’s Bank of China issued the “Guidelines for Internal Control of Commercial Banks” to contribute to the standardization and improvement of internal control requirements in Chinese commercial banks. While considering internal control, can this factor influence the restraining effect of OBI on BRT?

To address this question, this paper explored the influence of the Internal Control Quality (ICQ) on the inhibitory effect of OBI on BRT levels. To define the variable for this analysis, this exercise selected the “DIB Corporate Internal Control Index” [58] as a proxy for ICQ. Nevertheless, since this index only assessed a sample of the listed companies, only that sample of the listed banks was included in this analysis that included a crossover term, including OBI and ICQ. The corresponding model for this assessment is described in Equation (9).

(9)${\mathrm{BRT}}_{\mathrm{i},\mathrm{t}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1{\mathrm{OBI}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_2{\mathrm{ICQ}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_3{\mathrm{FIN}}_{\mathrm{i},\mathrm{t}}·{\mathrm{ICQ}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_3{\mathrm{X}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\alpha }}_4{\mathrm{M}}_{\mathrm{t}}+{\mathsf{\alpha }}_{\mathrm{i}}+{\mathsf{\mu }}_{\mathrm{i},\mathrm{t}}$

If the mentioned crossover term ($\mathrm{OBI}·\mathrm{ICQ}$) is positive and significant, it means that ICQ can weaken the inhibitory effect of OBI on BRT level; otherwise, a significant and positive crossover term will mean that ICQ can further promote the inhibitory effect of OBI on BRT. For this section, the explained variable BRT has been measured through RWA, and the corresponding results are displayed in Table 12 as follows.

The results in column (1) showed a negative and significant coefficient for the corresponding crossover term (−0.584 significant at 10% level), indicating that higher values for ICQ can strengthen the inhibitory effect of bank OBI within the proposed model. Columns (2) and (3) differentiated subsamples with higher and lower ICQ levels using the median as a reference. The results showed that a higher ICQ showed a negative significant value (−0.418 significant at 1% level), while the corresponding coefficient for the lower ICQ group was significant and positive (0.132 significant at 1% level). This further verifies mentioned analysis; ICQ can enhance the inhibitory effect of bank OBI on BRT.

5. Conclusions and Recommendations

Commercial banks represent an important influence in the Chinese financial structure system. In consequence, the effects of bank OBI are not just limited to the boundaries of banks. It also has an important relationship with the Chinese financial system. It constitutes a source of supply-side reforms and is related to the lifeblood of Chinese economic development; however, financial development not only brings opportunities to banks, but may also represent certain risks to banks. By using panel data for banks from 2009 to 2019, this paper empirically analyzed the relationship between off-balance-sheet business innovation (OBI) and bank risk-taking level (BRT) and obtained the following conclusions and recommendations:

• Off-balance-sheet business innovation (OBI) can improve bank risk management methods, enhance risk management processes, and, at the same time, bring new technologies and improve operating efficiency, thus reducing banks’ willingness to take risks, and subsequently inhibiting their risk-taking level. Therefore, commercial banks should attach great importance to the analysis and application of financial technology, promote the digital transformation of the front desk, middle office, and back office of banks, use emerging technologies to create digital risk control models, and promote financial technology to enable business management and risk management of commercial banks. At the same time, they should further strengthen the standard formulation of financial technology policies and industry norms to guide the development of financial technology and combine the difficult problems of commercial banks’ off-balance-sheet business innovation to launch a guide for financial technology to guide banks’ off-balance-sheet business innovation. In addition, banks should improve the innovation level of off-balance-sheet business in credit systems, credit products, and credit management technology, and improve the risk management level and operational efficiency of banks through the improvement of the innovation level to reduce the risk-bearing level of banks.

• From the perspective of banks’ natures, this paper analyzed the heterogeneity of the impact of bank off-balance-sheet business innovation on bank risk-taking level. The empirical results of this study showed that, compared with state-owned and joint-stock banks, the inhibitory effect of off-balance-sheet business innovation on bank risk-taking appeared more evident in the subsample of urban and rural commercial banks. In this regard, for state-owned and joint-stock banks, we should strengthen business risk prevention capabilities, build a dynamic risk evaluation system based on data mining, machine learning, and other technologies, timely intervene and adjust high-risk businesses, and improve risk compensation programs. Moreover, OBI further strengthens the information acquisition capabilities, risk assessment, and management levels of state-owned and joint-stock banks.

• Off-balance-sheet business innovation can help firms reduce management expenses and agency costs by inhibiting banks’ willingness to transfer credit risk to customers, reducing bank risk-taking, and developing a three-element transmission path (Off-balance-sheet Business Innovation > Agent Cost > Bank Risk-Taking). In this regard, financial supervision departments can establish a bank agency cost (BAC) monitoring mechanism to determine the regulatory threshold of the management expense rate to prevent agency-related problems and risk accumulation effects that may be caused by high bank agency costs, contributing, in so doing, to the control of bank risk-taking levels.

• From the perspective of bank competition, this article examined whether a bank’s competition level can play a moderation role in the relationship between off-balance-sheet business innovation (OBI) and bank risk-taking (BRT). The empirical results in this paper show that higher bank competition (BCMP) levels can enhance OBI’s inhibitory effect on BRT levels. This implies that banking reforms need to consider the relationship between banking competition and financial security, and that a gradual reduction in banking industry concentration levels, control of monopolistic practices, appropriate control of entry barriers, as well as measures that guide private banks to compete with state-owned banks fully can be conducive to control banking risk.

• This paper used a sample of listed banks to investigate whether a bank’s Internal Control Quality (ICQ) affects the risk suppression effect of off-balance-sheet business innovation, finding that, in the sample for this research, a higher ICQ strengthens the inhibitory effect of off-balance-sheet business innovation (OBI) on the bank’s risk exposure. This finding explains the importance of BIC within the presented financial innovation model. From this perspective, banks may improve their risk management systems’ effectiveness and reduce their risk exposure levels by improving their internal control mechanisms and operating efficiency. Simultaneously, regulatory agencies could strengthen internal control supervision, encouraging banks to improve their risk management and control capabilities toward a secure and stable financial system.

In addition, this study still has some limitations. On the one hand, this paper uses the proportion of non-interest income to measure OBI. With the continuous development of artificial intelligence and data mining tools, advanced science and technology can be used to gradually improve the indicators of OBI, and a scientific and comprehensive measurement method can be developed. On the other hand, this paper selects bank competition as the moderating variable. However, the impact of OBI on the level of BRT may be affected by multiple situational factors. Future research can explore whether the moderating effect of other situational variables exists. The discussion of the above issues will help to promote further understanding of the influence scope of OBI.

Author Contributions

Conceptualization, S.G. and H.G.; methodology, S.G. and G.A.B.; software, S.G. and H.H.; validation, S.G. and H.G.; formal analysis, S.G. and H.G.; investigation, S.G. and H.G.; resources, S.G.; data curation, S.G. and G.A.B.; writing—original draft preparation, S.G.; writing—review and editing, G.A.B. and H.H.; visualization, S.G. and H.H.; supervision, H.G.; project administration, H.G.; funding acquisition, H.G. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The corresponding author can provide the data on request.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Tables

Table 1

Hypothesis H1a–H1b.

Code	Hypothesis	Involved Variables
H1a	There is a significant negative relationship between Off-balance-sheet Business Innovation and Bank Risk-Taking Level.	OBI-BRT
H1b	There is a significant positive relationship between Off-balance-sheet Business Innovation and Bank Risk-Taking Level.	OBI-BRT

Table 2

Hypothesis H2a–H2b.

Code	Hypothesis	Involved Variables
H2a	Bank competition has a significant positive moderating effect on the relationship between off-balance-sheet business innovation and bank risk-taking level.	OBI-BCMP-BRT
H2b	Bank competition has a significant positive moderating effect on the relationship between off-balance-sheet business innovation and bank risk-taking level.	OBI-BCMP-BRT

Table 3

Hypothesis H3.

Code	Hypothesis	Involved Variables
H3	Bank agency cost plays a mediating effect in the relationship between off-balance-sheet business innovation and bank risk-taking level.	OBI-BAC-BRT

Table 4

Variable Definition.

Type	Variable	Definition	Construction
Explained	BRT	Risk-Weighted Assets Ratio	$\mathrm{Net}\mathrm{RWA}/\mathrm{Total}\mathrm{Assets}$
		Z-score	$\frac{\mathrm{SD}\mathrm{of}\mathrm{Return}\mathrm{on}\mathrm{Total}\mathrm{Assets}}{\mathrm{Return}\mathrm{on}\mathrm{Total}\mathrm{Assets}+\mathrm{Equity}\mathrm{Ratio}}$
Explanatory	OBI	Non-Interest Income NIIR	Proportion of Bank Non-interest Income
Control(Bank Level)	Growth	Bank Growth Rate	Growth Rate of Bank Loan
	LEV	Bank Leverage	Equity-to-Asset Ratio
	OpEff	Operational Efficiency	Cost-to-Income Ratio
	LDR	Loan-to-Deposit Ratio	Loan-to-Deposit Ratio
	ROA	Return on Assets	$\mathrm{Bank}\mathrm{Net}\mathrm{Profit}/\mathrm{Total}\mathrm{Assets}$
	CAR	Capital Adequacy Ratio	$\mathrm{Bank}\mathrm{Capital}/\mathrm{RWA}$
Control(Macro Level)	CPI	Consumer Price Index	$\mathrm{Consumer}\mathrm{Price}\mathrm{Index}/100$
	Stkdex	Shanghai Composite Index	ln (year-end Shanghai Composite Index)
	GDPg	Gross Domestic Product Growth	Nominal GDP Growth Rate
Moderator	BCMP	Bank Competition Level	Lerner Index
Mediator	BAC	Bank Agency Cost	$\mathrm{Management}\mathrm{Expenses}/\mathrm{Operating}\mathrm{Income}$

Table 5

Descriptive Statistics.

Variable	Mean	p50	sd	Min	Max
BRT (RWA)	0.4579	0.4689	0.0964	0.2047	0.6433
BRT (Z-score)	1.7681	1.3563	1.5353	0.0783	10.2883
OBI	0.1995	0.1559	0.1673	−0.0105	0.8460
Growth	0.0111	0.0100	0.0077	−0.0057	0.0473
LEV	0.0727	0.0704	0.0188	0.0375	0.1664
OpEff	0.3365	0.3296	0.0713	0.1888	0.5616
LDR	0.6518	0.6675	0.1150	0.3314	0.9505
ROA	0.0106	0.0101	0.0044	0.0004	0.0239
CAR	0.1336	0.1297	0.0232	0.0939	0.2587
CPI	1.0224	1.0210	0.0139	0.9930	1.0540
Stkdex	7.9414	8.0229	0.1761	7.6573	8.1717
GDPg	0.0779	0.0730	0.0138	0.0610	0.1060
BCMP	0.5682	0.5706	0.0597	0.4084	0.7484
BAC	0.0095	0.0091	0.0031	0.0041	0.0217

Table 6

Unit Root Test.

Variables	Fisher Test		Conclusion
BRT (RWA)	Statistics	−4.4684	Stable
	p value	0.0000
BRT (Z-score)	Statistics	−11.3335	Stable
	p value	0.0000
OBI	Statistics	−3.1828	Stable
	p value	0.0007
Growth	Statistics	−10.7110	Stable
	p value	0.0000
LEV	Statistics	−7.3557	Stable
	p value	0.0000
OpEff	Statistics	−4.3759	Stable
	p value	0.0000
LDR	Statistics	−1.4299	Stable
	p value	0.0000
ROA	Statistics	−9.0968	Stable
	p value	0.0000
CAR	Statistics	−8.9236	Stable
	p value	0.0000

Table 7

Assessment of the effect of OBI on BRT.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	RWA	Z-Score	RWA	RWA	RWA	Z-Score	Z-Score	Z-Score
	Full Sample	Full Sample	State-Owned	Joint Stock	Urban–Rural	State-Owned	Joint Stock	Urban–Rural
OBI	−0.032 ***	−0.760 **	−0.125	0.011	−0.025 **	−1.513	−4.872 **	−0.639 *
	(−3.33)	(−2.39)	(−1.12)	(0.15)	(−2.51)	(−0.60)	(−2.01)	(−1.93)
Growth	1.894 ***	−13.679 **	4.503 *	−0.153	1.995 ***	49.130	−42.751	−11.865 *
	(9.44)	(−2.09)	(1.96)	(−0.17)	(9.63)	(0.95)	(−1.40)	(−1.72)
LEV	0.518 ***	−9.183 *	−0.118	2.324 ***	0.464 ***	−18.592	−43.819 *	−8.335 *
	(3.53)	(−1.92)	(−0.11)	(3.25)	(3.08)	(−0.78)	(−1.85)	(−1.66)
OpEff	0.096 ***	1.466	−0.500 *	0.381 ***	0.071 **	−0.309	7.578 **	0.646
	(3.38)	(1.57)	(−1.81)	(3.37)	(2.35)	(−0.05)	(2.03)	(0.64)
LDR	0.545 ***	−0.440	0.118	0.315 ***	0.572 ***	−3.808 *	3.190	−0.588
	(31.58)	(−0.78)	(1.30)	(4.73)	(30.68)	(−1.87)	(1.45)	(−0.95)
CPI	−0.529 ***	−9.964 **	−0.582	−1.253 ***	−0.512 ***	19.024 **	17.442	−10.838 **
	(−4.34)	(−2.50)	(−1.55)	(−3.19)	(−3.85)	(2.24)	(1.35)	(−2.44)
Stkdex	−0.036 ***	0.976 ***	−0.028	−0.005	−0.041 ***	1.835 ***	2.348 **	0.900 ***
	(−3.88)	(3.25)	(−1.39)	(−0.18)	(−4.09)	(4.09)	(2.36)	(2.70)
GDPg	1.281 ***	30.812 ***	0.531	2.851 ***	1.266 ***	−11.450	−2.077	30.772 ***
	(9.48)	(6.99)	(0.90)	(4.62)	(8.78)	(−0.86)	(−0.10)	(6.41)
ROA	1.242 **	−17.957	0.726	5.143	1.227 **	−108.209	−90.794	−18.123
	(2.48)	(−1.10)	(0.24)	(1.42)	(2.40)	(−1.62)	(−0.76)	(−1.06)
CAR	−0.438 ***	−2.521	0.798	0.735	−0.456 ***	−9.048	−19.964	−2.816
	(−4.64)	(−0.82)	(1.64)	(1.39)	(−4.68)	(−0.82)	(−1.14)	(−0.87)
_cons	0.789 ***	3.141	1.264 **	0.946	0.809 ***	−25.687 **	−32.394 *	5.054
	(4.42)	(0.54)	(2.67)	(1.62)	(4.17)	(−2.41)	(−1.68)	(0.78)
Individual fixed affected	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	1188	1188	55	118	1015	55	118	1015
R2	0.526	0.023	0.452	0.544	0.553	0.454	0.325	0.112

Note: ***, **, * respectively indicate significant at the level of 1%, 5%, and 10%. T value is in parentheses.

Table 8

Assessment of the effect of OBI on BRT.

	(1)	(2)	(3)	(4)
	RWA	RWA	Z-Score	Z-Score
OBI	−0.032 ***	0.098	−0.760 **	6.566 **
	(−3.33)	(1.33)	(−2.39)	(2.48)
BCMP		0.025		3.715 **
		(0.51)		(2.08)
OBI × BCMP		−0.221 *		−12.516 ***
		(−1.77)		(−2.80)
Growth	1.894 ***	1.903 ***	−13.679 **	−11.891 *
	(9.44)	(10.30)	(−2.09)	(−1.80)
LEV	0.518 ***	0.514 ***	−9.183 *	−11.396 **
	(3.53)	(3.80)	(−1.92)	(−2.35)
OpEff	0.096 ***	0.113 ***	1.466	1.343
	(3.38)	(3.82)	(1.57)	(1.27)
LDR	0.545 ***	0.581 ***	−0.440	−0.103
	(31.58)	(35.23)	(−0.78)	(−0.17)
CPI	−0.529 ***	−0.623 ***	−9.964 **	−8.266 **
	(−4.34)	(−5.51)	(−2.50)	(−2.04)
Stkdex	−0.036 ***	−0.038 ***	0.976 ***	1.029 ***
	(−3.88)	(−4.50)	(3.25)	(3.39)
GDPg	1.281 ***	1.417 ***	30.812 ***	31.127 ***
	(9.48)	(11.28)	(6.99)	(6.93)
ROA	1.242 **	1.522 ***	−17.957	−2.154
	(2.48)	(2.62)	(−1.10)	(−0.10)
CAR	−0.438 ***	−0.410 ***	−2.521	−1.872
	(−4.64)	(−4.64)	(−0.82)	(−0.59)
_cons	0.789 ***	0.841 ***	3.141	−1.472
	(4.42)	(4.97)	(0.54)	(−0.24)
Individual fixed affected	Yes	Yes	Yes	Yes
N	1188	1165	1188	1165
R2	0.526	0.593	0.023	0.141

Note: ***, **, * respectively indicate significant at the level of 1%, 5%, and 10%. T value is in parentheses.

Table 9

Mediation effect of BAC.

	(1)	(2)	(3)	(4)	(5)
	RWA	Z-Score	BAC	RWA	Z-Score
OBI	−0.032 ***	−0.760 **	−0.001 **	−0.026 ***	−0.631 **
	(−3.33)	(−2.39)	(−2.17)	(−2.87)	(−2.01)
BAC				6.367 ***	210.309 ***
				(6.94)	(6.74)
Growth	1.894 ***	−13.679 **	0.024 ***	1.797 ***	−17.106 ***
	(9.44)	(−2.09)	(3.67)	(9.41)	(−2.64)
LEV	0.518 ***	−9.183 *	0.034 ***	0.311 **	−17.600 ***
	(3.53)	(−1.92)	(7.33)	(2.18)	(−3.64)
OpEff	0.096 ***	1.466	0.016 ***	−0.005	−2.052 *
	(3.38)	(1.57)	(17.81)	(−0.15)	(−1.96)
LDR	0.545 ***	−0.440	0.001 **	0.561 ***	−0.697
	(31.58)	(−0.78)	(2.13)	(33.94)	(−1.24)
CPI	−0.529 ***	−9.964 **	0.007 *	−0.579 ***	−11.191 ***
	(−4.34)	(−2.50)	(1.81)	(−5.02)	(−2.86)
Stkdex	−0.036 ***	0.976 ***	−0.001 *	−0.034 ***	1.085 ***
	(−3.88)	(3.25)	(−1.73)	(−3.86)	(3.66)
GDPg	1.281 ***	30.812 ***	0.007	1.327 ***	29.062 ***
	(9.48)	(6.99)	(1.52)	(10.36)	(6.67)
ROA	1.242 **	−17.957	0.276 ***	−0.331	−76.940 ***
	(2.48)	(−1.10)	(17.27)	(−0.62)	(−4.22)
CAR	−0.438 ***	−2.521	−0.026 ***	−0.237 **	3.329
	(−4.64)	(−0.82)	(−8.58)	(−2.55)	(1.05)
_cons	0.789 ***	3.141	−0.002	0.785 ***	3.460
	(4.42)	(0.54)	(−0.42)	(4.66)	(0.60)
Individual fixed affected	Yes	Yes	Yes	Yes	Yes
N	1188	1188	1174	1174	1174
R2	0.526	0.141	0.413	0.581	0.170

Note: ***, **, * respectively indicate significant at the level of 1%, 5%, and 10%. T value is in parentheses.

Table 10

Robustness Test Variable Definition.

Type	Variable	Definition
Explained *	LLR	Loan Loss Reserve Adequacy Ratio
Control	BPI	Banking Industry Prosperity Index
	BCI	Banker’s Macroeconomic Confidence Index
	PLR	Lending Ratio

* As a substitute for BRT defined proxies, RWA, and Z-score.

Table 11

Model Robustness Test.

	(1)	(2)	(3)	(4)	(5)
	LLR	RWA	Z-Score	RWA (Lag)	Z-Score (Lag)
OBI	2.001 **	−0.023 ***	−0.800 ***	−0.026 ***	−0.651 **
	(2.21)	(−2.88)	(−2.72)	(−2.65)	(−2.03)
L.RWA2		0.350 ***
		(16.33)
L.Z1			0.368 ***
			(12.82)
Growth	105.807 ***	0.788 ***	−1.164	1.549 ***	−19.382 ***
	(5.65)	(4.14)	(−0.17)	(7.33)	(−2.78)
LEV	−19.319	0.722 ***	−8.769 *	0.613 ***	−7.940 *
	(−1.40)	(5.88)	(−1.92)	(4.23)	(−1.66)
OpEff	−5.120 *	0.012	−0.182	0.095 ***	1.763 *
	(−1.83)	(0.50)	(−0.20)	(3.34)	(1.87)
LDR	−10.974 ***	0.415 ***	−0.145	0.519 ***	−0.104
	(−6.80)	(25.43)	(−0.27)	(29.23)	(−0.18)
CPI	32.415 ***	0.129	−17.092 ***	−0.497 ***	−10.145 **
	(2.85)	(0.78)	(−2.78)	(−4.06)	(−2.51)
Stkdex	−4.389 ***	−0.031 ***	0.556 **	0.013	0.947 **
	(−5.24)	(−4.17)	(1.98)	(1.11)	(2.54)
GDPg	−47.367 ***	0.494 ***	18.258 ***	0.002	20.504 ***
	(−3.73)	(3.56)	(3.53)	(0.01)	(2.66)
ROA	134.815 ***	−0.013	2.946	0.875	−27.666
	(2.90)	(−0.03)	(0.18)	(1.64)	(−1.57)
CAR	32.514 ***	−0.627 ***	−5.070 *	−0.509 ***	−2.371
	(3.69)	(−8.03)	(−1.74)	(−5.39)	(−0.76)
BPI				0.325 ***	3.530 **
				(6.90)	(2.27)
BCI				−0.005	−2.215 ***
				(−0.31)	(−4.63)
PLR				0.135	21.926 ***
				(0.76)	(3.76)
_cons	15.669	0.124	14.373 *	0.255	2.092
	(0.95)	(0.61)	(1.90)	(1.33)	(0.33)
Individual fixed affected	Yes	Yes	Yes	Yes	Yes
N	1068	1101	1101	1166	1166
R2 _a	0.156	0.669	0.241	0.547	0.159

Note: ***, **, * respectively indicate significant at the level of 1%, 5%, and 10%. T value is in parentheses.

Table 12

Subsample Test for OBI, internal control quality, and BRT.

	(1)	(2)	(3)
	Listed Banks	High ICQ Group	Low ICQ Group
OBI	0.188	−0.418 ***	0.132 ***
	(0.83)	(−5.42)	(2.87)
ICQ	0.190 *
	(1.91)
OBI × ICQ	−0.584 *
	(−1.71)
LEV	2.672 ***	2.770 ***	5.049 ***
	(4.87)	(3.76)	(9.50)
CAR	−0.191	−0.655	−1.016 **
	(−0.51)	(−1.54)	(−2.39)
ROA	−3.004	−2.369	3.642
	(−1.14)	(−0.64)	(1.42)
CPI	−0.626 **	−0.154	0.595
	(−2.24)	(−0.55)	(0.90)
Stkdex	−0.019	0.084 ***	−0.044 *
	(−0.84)	(2.67)	(−1.95)
_cons	1.067 **	0.016	−0.087
	(2.51)	(0.03)	(−0.11)
Individual fixed affected	Yes	Yes	Yes
N	201	101	100
R2	0.052	0.222	0.633

Note: ***, **, * respectively indicate significant at the level of 1%, 5%, and 10%. T value is in parentheses.

References

1. Zheng, L. Deepening the supply-side structural reform of China’s financial system: A financial function perspective. Financ. Trade Econ.; 2019; 40, pp. 66-80.

2. Ruan, S.; Zhen, X. Research on the measurement and evaluation of innovation efficiency of urban commercial banks in China. Jiang-Huai Trib.; 2015; 2, pp. 39-45.

3. Tufano, P. Chapter 6: Financial Innovation. Handbook of the Economics of Finance; Elsevier: Amsterdam, The Netherlands, 2003; pp. 307-335.

4. Frame, W.S.; White, L.J. Empirical studies of financial innovation: Lots of talk, little action. J. Econ. Lit.; 2004; 42, pp. 116-144. [DOI: https://dx.doi.org/10.1257/002205104773558065]

5. Xia, W.; Wang, L. The trend, experience and enlightenment of the reform of international financial supervision system before and after the financial crisis. Huabei Financ.; 2017; 1, pp. 31-35.

6. Aghion, P.; Comin, D.; Business, H.; Howitt, S.P.; Tecu, I.; Acemoglu, D.; Tsyrennikov, V. When Does Domestic Saving Matter for Economic Growth?. Harvard Business School Working Papers; Harvard Business School: Cambridge, MA, USA, 2009.

7. Henderson, B.J.; Pearson, N.D. The dark side of financial innovation: A case study of the pricing of a retail financial product. J. Financ. Econ.; 2011; 100, pp. 227-247. [DOI: https://dx.doi.org/10.1016/j.jfineco.2010.12.006]

8. Wang, Y.; Zhang, T. Financial innovation, Audit quality and bank’s risk taken: Empirical evidence from Chinese commercial bank. Account. Res.; 2014; 4, 81–87+96

9. Ferguson, R.W. Alternative Approaches to Financial Supervision and Regulation. J. Financ. Serv. Res.; 1999; 17, pp. 297-303. [DOI: https://dx.doi.org/10.1023/A:1008156828977]

10. Hu, W.; Zhang, L.; Li, X.X.; Wang, Z.J. Financial innovation, risk tolerance and profitability of commercial banks. Financ. Forum; 2019; 24, pp. 31-47.

11. Wen, Z.; Ye, B. Analysis of Mediating Effects: The Development of Methods and Models. Adv. Psychol. Sci.; 2014; 22, pp. 731-745. [DOI: https://dx.doi.org/10.3724/SP.J.1042.2014.00731]

12. Tay, L.-Y.; Tai, H.-T.; Tan, G.-S. Digital financial inclusion: A gateway to sustainable development. Heliyon; 2022; 8, e09766. [DOI: https://dx.doi.org/10.1016/j.heliyon.2022.e09766]

13. Chao, X.; Ran, Q.; Chen, J.; Li, T.; Qian, Q.; Ergu, D. Regulatory technology (Reg-Tech) in financial stability supervision: Taxonomy, key methods, applications and future directions. Int. Rev. Financ. Anal.; 2022; 80, 102023. [DOI: https://dx.doi.org/10.1016/j.irfa.2022.102023]

14. Miller, R.M. Deformation and Shock Deformation in Rocks from the Roter Kamm Crater SWA/Namibia. Meteoritics; 1986; 21, pp. 456-458.

15. Tufano, P.; Poetzscher, C.; Santangelo, R. Financial Engineering and Tax Risk: The Case of Times Mirror PEPS. Harvard Business Review; 20 December 1996; 9.

16. McConnell, J.J.; Schwartz, E.S. The Origin of Lyons: A Case Study in Financial Innovation. J. Appl. Corp. Financ.; 1992; 4, pp. 40-47. [DOI: https://dx.doi.org/10.1111/j.1745-6622.1992.tb00216.x]

17. Javaid, M.; Haleem, A.; Singh, R.P.; Suman, R.; Khan, S. A review of Blockchain Technology applications for financial services. BenchCouncil Trans. Benchmarks Stand. Eval.; 2022; 2, 100073. [DOI: https://dx.doi.org/10.1016/j.tbench.2022.100073]

18. Jameson, M.; Dewan, S.; Sirmans, C.F. Measuring welfare effects of “unbundling” financial innovations: The case of collateralized mortgage obligations. J. Urban Econ.; 1992; 31, pp. 1-13. [DOI: https://dx.doi.org/10.1016/0094-1190(92)90029-K]

19. Banna, H.; Kabir Hassan, M.; Rashid, M. Fintech-based financial inclusion and bank risk-taking: Evidence from OIC countries. J. Int. Financ. Mark. Inst. Money; 2021; 75, 101447. [DOI: https://dx.doi.org/10.1016/j.intfin.2021.101447]

20. Banna, H. The role of digital financial inclusion on promoting sustainable economic growth through banking stability: Evidence from Bangladesh. Dev. Rev.; 2020; 29, pp. 19-36.

21. Tufano, P.; Sevick, M. Board structure and fee-setting in the US mutual fund industry. J. Financ. Econ.; 1997; 46, pp. 321-355. [DOI: https://dx.doi.org/10.1016/S0304-405X(97)00033-0]

22. Guner, A.B. Loan sales and the cost of corporate borrowing. Rev. Financ. Stud.; 2006; 19, pp. 687-716. [DOI: https://dx.doi.org/10.1093/rfs/hhj017]

23. Hirtle, B. Macroprudential Supervision of Financial Institutions: Lessons from the SCAP; Working Paper EconStor: Leibnitz, Germany, 2009.

24. Weisbach, N. Did securitization affect the cost of corporate debt?. J. Financ. Econ.; 2012; 105, pp. 332-352.

25. Zhang, X.; Mao, Y. Research on the relationship between the systemic risks and noninterests income of commercial banks in China. Stud. Int. Financ.; 2014; 11, pp. 23-35.

26. Wu, X. Internet finance: The logic of growth. Financ. Trade Econ.; 2015; 2, pp. 5-15.

27. Quan, F.; Wang, X. The effect of financial innovation on commercial bank risk-taking—Based on the different types of financial innovation. Collect. Essays Financ. Econ.; 2016; 9, pp. 35-45.

28. Liu, Z. Research on the influence of internet finance on commercial banks’ risk-taking. Financ. Trade Econ.; 2016; 4, 71–85+115

29. Shen, Y.; An, L. Will the saving of non-interest cost help to reduce bank risks. Financ. Econ. Res.; 2017; 32, pp. 95-106.

30. Marcelin, I.; Sun, W.; Teclezion, M.; Junarsin, E. Financial inclusion and bank risk-taking: The effect of information sharing. Finance Res. Lett.; 2022; 50, 103182. [DOI: https://dx.doi.org/10.1016/j.frl.2022.103182]

31. Kane, E.J. Regulatory structure in futures markets: Jurisdictional competition between the sec, the cftc, and other agencies. J. Futures Mark.; 1984; 4, pp. 367-384. [DOI: https://dx.doi.org/10.1002/fut.3990040306]

32. Ashraf, D.; Rizwan, M.S.; L’Huillier, B. A net stable funding ratio for Islamic banks and its impact on financial stability: An international investigation. J. Financ. Stab.; 2016; 25, pp. 47-57. [DOI: https://dx.doi.org/10.1016/j.jfs.2016.06.010]

33. Khan, M.S.; Scheule, H.; Wu, E. Funding Liquidity and Bank Risk Taking. J. Bank. Financ.; 2016; 82, pp. 203-216. [DOI: https://dx.doi.org/10.1016/j.jbankfin.2016.09.005]

34. He, D.; Zhang, X. Financial services innovation in marketing: A literature survey. Econ. Res. J.; 2009; 44, pp. 138-154.

35. Chen, W. The enlightnment on government audit of monetary institutes from the subprime mortgage crisis—Thinking baesd on the audit essence of the immune system. Audit. Res.; 2009; 2, pp. 22-25.

36. Dam, K.W. The Subprime Crisis and Financial Regulation: International and Comparative Perspectives. Chic. J. Int. Law; 2010; 10, 581. [DOI: https://dx.doi.org/10.2139/ssrn.1579048]

37. Feng, Q.; Huang, X. Financial innovation, product intervention and conduct risk management of financial institutions. Financ. Forum; 2016; 21, pp. 21-32.

38. Dai, G.; Fang, P. The liberalization of interest rate and risks of banks—A study from the perspective os shadow banking and internet finance. Financ. Forum; 2014; 19, 13–19+74

39. Gu, H.; Zhang, Y. Financial innovation, loan environment and bank risk acceptance—Evidence from 2006–2016 Chinese industry. Stud. Int. Financ.; 2018; 9, pp. 66-75.

40. Guo, P.; Shen, Y. The impact of internet finance on risk taking of commercial banks: Theoretical interpretation and empirical test. Financ. Trade Econ.; 2015; 10, pp. 102-116.

41. Boyd, J.H.; De Nicolo, G. The theory of bank risk taking and competition revisited. J. Financ.; 2005; 60, pp. 1329-1343. [DOI: https://dx.doi.org/10.1111/j.1540-6261.2005.00763.x]

42. Amidu, M.; Wolfe, S. Does bank competition and diversification lead to greater stability? Evidence from emerging markets. Rev. Dev. Financ.; 2013; 3, pp. 152-166. [DOI: https://dx.doi.org/10.1016/j.rdf.2013.08.002]

43. Marcus, A.J. Deregulation and bank financial policy. J. Bank. Financ.; 1984; 8, pp. 557-565. [DOI: https://dx.doi.org/10.1016/S0378-4266(84)80046-1]

44. Adrian, T.; Shin, H.S. Liquidity and leverage. J. Financ. Intermediat.; 2010; 19, pp. 418-437. [DOI: https://dx.doi.org/10.1016/j.jfi.2008.12.002]

45. Repullo, R. Capital requirements, market power, and risk-taking in banking. J. Financ. Intermediat.; 2004; 13, pp. 156-182. [DOI: https://dx.doi.org/10.1016/j.jfi.2003.08.005]

46. Wind, Wind Financial Database. 2020; Available online: https://www.wind.com.cn/ (accessed on 25 December 2022).

47. Moody’s Analytics Inc, Bankfocus Database. 2008; Available online: https://bankfocus.bvdinfo.com/ (accessed on 25 December 2022).

48. EPS China Data Inc, EPS China Data. 2018; Available online: https://www.epsnet.com.cn/ (accessed on 25 December 2022).

49. Delis, M.D.; Kouretas, G.P. Interest rates and bank risk-taking. J. Bank. Financ.; 2011; 35, pp. 840-855. [DOI: https://dx.doi.org/10.1016/j.jbankfin.2010.09.032]

50. Boyd, J.H.; Graham, S.L.; Hewitt, R.S. Bank holding company mergers with nonbank financial firms: Effects on the risk of failure. J. Bank. Financ.; 1993; 17, pp. 43-63. [DOI: https://dx.doi.org/10.1016/0378-4266(93)90079-S]

51. Xu., M.; Chen., X. Monetary environment capital adequacy ratio and risk taking of commercial banks. J. Financ. Res.; 2012; 7, 489+50–62

52. Ma, Y. Research on institution environment in financial innovation. Contemp. Econ. Manag.; 2011; 33, pp. 87-91.

53. Berger, A.N.; Klapper, L.F.; Turk-Ariss, R. Bank Competition and Financial Stability. J. Financ. Serv. Res.; 2008; 35, pp. 99-118. [DOI: https://dx.doi.org/10.1007/s10693-008-0050-7]

54. Bikker, J.A.; Shaffer, S.; Spierdijk, L. Assessing Competition with the Panzar-Rosse Model: The Role of Scale, Costs, and Equilibrium. Rev. Econ. Stat.; 2012; 94, pp. 1025-1044. [DOI: https://dx.doi.org/10.1162/REST_a_00210]

55. Li, S.; Tian, G. Bank competition and bank risk-taking channel of monetary policy: Theoretical and empirical analysis. Manag. World; 2020; 36, pp. 149-158.

56. Zhen, H.; Zhang, X.; Chi, G. The effect of the institutional environment and ultimate control on corporate performance: Based on the test of mediator effect of agency costs. J. Financ. Res.; 2015; 12, pp. 162-177.

57. Yang, Z.; Yang, G. Internal control quality and banks’ risk-taking—Evidence from Chinese listed banks. Audit. Res.; 2017; 6, pp. 105-112.

58. DIB Corporate Risk Management Technology Ltd., DIB Corporate Internal Control Index. 2021; Available online: https://data.dibdata.cn/ (accessed on 25 December 2022).