Computational Path and Risk Analysis of Performance Changes of Chinese Listed Commercial Banks under Diversification of Interest Rate Marketization

The positive impact is mainly reflected in three areas:

First, it promotes the development of deposit business and the improvement of loan quality. Under the marketization of deposit interest rates, commercial banks can appropriately increase the percentage of deposit interest rates to absorb deposit funds and promote the development of deposit business. Under the marketization of loan interest rates, commercial banks can lower the loan interest rate and lend to SOEs, listed companies, and large enterprises with good credit, thus continuously improving the level of loan quality.

Second, increase profitability and improve risk prevention. Banks in order to increase loan interest income, increase bank profits, can to a certain extent floating loan interest rates, increase profitability. Financial institutions to take differentiated interest rate strategy to fight for quality customers, less good benefits of small and medium-sized enterprises have loan demand, the bank can put forward higher loan conditions, that is, a higher level of lending interest rates, while those good benefits, good prospects for the development of small and medium-sized enterprises to give strong support, so that it will reduce the rate of non-performing loans, and thus high-risk prevention ability.

Thirdly, it is to encourage banks to improve their service systems. Under the marketization of interest rates, the price of funds has been transparent, customers will compare the level of interest rates of banks, commercial banks have to compete for customer resources by enriching the product features, improve service quality, provide a full range of financial services and other measures to strengthen market competitiveness, and promote the upgrading of the bank [27].

The negative impact is mainly reflected in two aspects:

The first step is to increase the intensity of vicious competition. In the context of the marketization of interest rates, the development of bank deposits will face enormous pressure and challenges, commercial banks continue to raise deposit rates, “high amount of reserves” way to fight for deposits, at the same time, in order to optimize the allocation of credit funds, through the lowering of lending rates to fight for high-quality credit customers to issue loans, deposit and loan spreads Getting smaller and smaller, commercial banks in the development of more and more emphasis on the price war, which is the main cause of increasing competition in the financial market.

Secondly, the risk of operation increases. Under the marketization of interest rates, banks are facing a very complex business development environment, faced with more uncertainty factors and the risk of increased burden. Bank interest rates change more frequently, which leads to a more frequent flow of funds within the market, which increases liquidity risk. Uncertainty in interest rates also causes certain interest rate losses to commercial banks, resulting in a sharp increase in credit risk.

In this paper, a total of 25 commercial banks listed on the A-share market from 2012 to 2023 are selected as the research sample, including 5 state-owned banks, 11 joint-stock banks, 8 city commercial banks, and 1 agricultural and commercial bank, totaling 322 research samples. The data sources of this paper include the wind database and the Cathay Pacific database, and the missing data is collected from the banks’ annual reports of the past years.

1)

Diversity metrics. In the existing literature studies, there are two forms of diversity metrics.

First, the Herfindahl index, expressed in the form: (1)H=∑i=1nPi2

Among them, I indicates the business type of the enterprise, P_i indicates the proportion of the income of the i type of business to the total business income, the larger the value of H, indicating that the business type of the enterprise is more homogeneous, and if the value of H is 1, indicating that the enterprise has only one type of business, the degree of diversification of the enterprise is zero.

Second, entropy index, entropy index is expressed as: (2)E=∑i=1nPiln(1pi)=−∑i=1nPilnPi

n is the number of businesses operated by the bank, and P_i is the share of operating income of business item i. Usually in practical analysis, n = 2, P₁ is the proportion of net interest income to operating income, and P₂ indicates the proportion of non-interest income to operating income. The degree of diversification increases with the increase of entropy index, and when the entropy index is 0, it indicates that the company has not diversified.

The Herfindahl index is the sum of the squares of the revenue share of different businesses, so it will make the gap between the businesses be amplified and the phenomenon of polarization, called the Matthew effect [28]. China’s diversification time is not yet long, interest income occupies most of the operating income, so then the use of Herfindahl index will amplify the gap between interest income and non-interest income, so that the results are biased. Therefore, this paper chooses entropy index to measure diversification. 2)

Performance Indicators

Performance indicators are a reflection of a bank’s profitability. In the existing literature, the most commonly used performance indicators are ROA and ROE. ROA indicates the ability of an enterprise to utilize its existing assets to obtain profits, and the larger the ROA is, the higher the bank’s performance level is [29]. ROE is also called return on equity, which means that the bank obtains profits based on its net assets, and the larger the ROE is, the stronger the enterprise’s profitability is. For commercial banks, the liquidity of their daily operations usually comes from loans, so using ROE as an indicator of performance measurement is not accurate enough, so this paper chooses ROA as an indicator of bank performance.The formula for ROA is: (3)ROA=Net profitTotal opening assets+Ending total assets / 2

ROA denotes the return on total assets of the bank, σ_ROA then represents the standard deviation of ROA on total assets, the higher the value, the more volatile the returns and the higher the risk, E/A denotes the bank’s equity ratio. The value of Z represents the bank’s default distance, which is used to measure the bank’s bankruptcy risk, and it is the inverse of the default distance, and the higher the value of Z, the farther the bank is from the default distance, and the less likely it is to go bankrupt, so the lower the risk, the more stable the bank will be. 4)

Control Variables

Referring to related literature, the study selected bank size, shareholders’ equity ratio, and level of economic development as control variables to study the impact of diversification on performance and risk [30]. The level of macroeconomic development is represented by the growth rate of gross domestic product GDP. For the measurement of the scale of bank operations, this paper chooses the logarithmic value of bank assets as to measure the size of the bank. Non-performing loan ratio is an important indicator to measure the operation of the bank, and a high non-performing loan ratio indicates that the more loans the bank can not recover, the higher the operational risk of the bank. At the same time, this paper adds the year effect dummy variable to control the year effect.

In this paper, a panel model is chosen for regression analysis. The regression model is developed for the explanatory variables ROA and Z as follows: (5)ROAit=α0+α1DIVit+α2controlit+εit (6)Zit=β0+β1DIVit+β2controlit+εit

where i = 1, 2, ..., 25, denotes the 25 commercial banks in the sample, t = 1, 2, ..., 12, denotes the 12-year time series from 2012-2023, and α and β are the variable coefficients, respectively.ROA and Z are the explanatory variables, with ROA measuring the bank’s level of performance, and Z measuring the bank’s risk.DIV is the explanatory variable denoting the level of diversification of the commercial banks.CONTROL is a set of control variables, including bank size, cost-income ratio, NPL ratio, capital adequacy ratio, equity ratio, and economic growth rate. ε is a random disturbance term.

In this paper, the variables of 25 commercial banks from 2012-2023 were analyzed by descriptive statistics, and the minimum, maximum, mean, and standard deviation of each variable were calculated, and the results are shown in Table 1. The explanatory variables are return on total assets (ROA) and bankruptcy risk (Z).The standard deviation of Z is 23.425, indicating that the overall risk level of the banks has varied considerably in recent years, and the maximum and minimum values of Z are 139.508 and 11.798, respectively, with a large gap between them. The maximum and minimum values of the return on total assets are 2.28 and 0.133 respectively.The extreme values of the variables measuring the performance and risk of the banks in recent years have large differences, indicating that the performance and risk of different banks are obviously different, and the indicators selected in this paper are representative. The explanatory variable is the entropy index. The average value of entropy index of commercial banks in recent years is 0.447. From the maximum and minimum values, diversification is also significantly different among different banks. In terms of non-performing loan (NPL) ratio, the level of NPL ratio in the entire sample of banks is 1.365, which is lower than the warning line for NPL ratio of commercial banks at 4%. In terms of cost-to-income (CIR) ratio, the average cost-to-income ratio is 32.865%, which is in line with the cost-to-income ratio of no more than 45% as agreed in the regulatory standard, indicating that the profitability of Chinese commercial banks as a whole is relatively good. In terms of Capital Adequacy Ratio (CAR), the average value of banks’ CAR was 12.364%, higher than the regulatory target of 8%, which indicates that commercial banks are more capable of coping with risks.

The panel model used in this paper requires the data to be smooth, so before regression analysis, the smoothness of the data needs to be tested, this paper chooses the LLC unit root test method, the test results are shown in Table 2. The test results show that the P-value of all variables is 0.000, and all of them are smooth and can be analyzed in regression analysis.

In this paper, Kao method is used for cointegration test, and the results of Kao cointegration test are shown in Table 3. At the level of P=0.05, the Kao method passes the test, i.e., by the cointegration test, it can be seen that there is a cointegration relationship between the variables, which can be analyzed by panel regression.

In the regression process, this paper uses the F test and Hausman test to determine whether the specific selection of a fixed effect model, random effect model, or mixed effect model is correct. The regression results of equations (5) and (6) are shown in Tables 4 and 5. ***, **, and * indicate significant at the 1%, 5%, and 10% levels, respectively. Values in parentheses are p-values corresponding to the estimated coefficients.

The results of the model estimation of diversified operations on performance in the context of interest rate marketization are shown in Table 4. The empirical results for the overall sample group show that the diversification variable (DIV) is significant at the 5% level and the coefficient is positive, thus indicating that the conduct of diversified business has a significant performance enhancing effect and hypothesis 1 is valid. The asset size variable (SIZE) and capital adequacy variable (CAR) are both significant at 1% level with coefficients of 0.07236 and 0.03652 respectively, indicating that the expansion of their asset size and abundant capital are also important in enhancing performance. The non-performing loan ratio (NPL) variable NPL is significant at 1% level and has a coefficient of -0.05326, which shows that there is a negative correlation between the bank’s performance and the NPL ratio, and that too high NPL ratio is detrimental to the bank’s development.

The research results of the sample group of large banks in Table 4 show that the diversification variable (DIV) is significant at the 1% standard, but the coefficient is negative, which indicates that diversification of business has not led to the improvement of the profitability of large banks, but rather caused a bad effect to some extent, which is at variance with the empirical results of the overall sample group. The reason for the above may be due to the fact that large banks in China are much larger than small and medium-sized ones in terms of business scale, manpower scale, etc., and diversification may lead to a decrease in their profitability due to the increase in overhead costs. The non-performing loan ratio variable (NPL) is significant at the 1% level and has a coefficient of -0.00423, which indicates that there is a negative correlation between bank performance and non-performing loan ratio, and that too high a non-performing loan ratio is detrimental to the development of banks. The asset size variable (SIZE), cost income variable (CIR), and capital adequacy variable (CAR) are all significant at the 1% level and the coefficients are 0.71235, 0.71256, and 0.21489 respectively, which indicates that the diversification of business has not led to improved performance for large commercial banks, and that a greater portion of the improvement in performance relies on the traditional This indicates that for large commercial banks, diversification of business has not led to improved performance, and performance improvement depends more on the traditional credit business and more on the economies of scale generated by the expansion of its scale.

The research results of the sample group of small and medium-sized banks in Table 4 show that the Diversification Variable (DIV) is significant at the 5% criterion and the coefficient is positive, which leads to the conclusion that small and medium-sized banks in China are able to improve a certain level of performance through the diversification of their business, which is in line with the empirical results of the overall sample group. Since small and medium-sized banks have far less financial strength and business level than large banks, thus it is difficult to form the market power effect, so the above reason may be due to the economy of scope effect. The asset size variable (SIZE) and capital adequacy variable (CAR) are both significant at the 1% level, and the coefficients are 0.07452 and 0.07412, respectively, which suggests that for small and medium-sized banks, the performance improvement is not only due to diversification of business, but also that the expanded scale and abundant capital play an important role.

The results of the model estimation of diversified operations on risk in the context of interest rate marketization are shown in Table 5. The empirical results of the overall sample group in Table 5 show that the diversification variable (DIV) is insignificant at the 1%, 5%, and 10% levels, which indicates that diversification does not have a significant effect on the diversification of its risk and Hypothesis 2 is not valid. The asset size variable (SIZE) is significant at the 1% level and has a coefficient of -0.03654, which indicates that asset size is negatively related to risk, i.e., it is able to reduce risk by expanding the level of size. The Cost Income Variable (CIR) is significant at the 5% level and has a coefficient of 0.31236, which indicates that the larger the loan share, the higher the risk. Non-Performing Loan Ratio variable (NPL) is significant at 10% level and the coefficient is 0.00364, which indicates that the two show a positive effect, too high NPL ratio will lead to the bank’s risk to become bigger.

The empirical results for the sample group of large banks in Table 5 show that the Diversification Variable (DIV) is also insignificant at 1%, 5%, and 10% criteria, which suggests that diversification also does not have a significant diversifying effect on the risk of large banks. The Cost Income Variable (CIR) and Non-Performing Loan Ratio Variable (NPL) are both significant on the 1% criterion and have coefficients of 0.59632 and 0.02635, respectively, indicating that the larger the loan share, the higher the risk. Capital Adequacy Ratio variable (CAR) is significant on the 1% standard and has a coefficient of -0.03695, which indicates a negative relationship, and if the bank wants to make the bank less risky, it has to maintain abundant capital.

The findings of the sample group of small and medium-sized banks in Table 5 show that the diversification variable (DIV) is also insignificant at 1%, 5% and 10% criteria, thus indicating that diversification is also not significantly effective in diversifying the risks of small and medium-sized banks. The asset size variable (SIZE) is significant at the 1% criterion and has a coefficient of -0.03698, which indicates that asset size is negatively correlated with risk, i.e., it is able to reduce risk by expanding the level of size. The Cost Income Variable (CIR) is significant at 1% level and has a coefficient of 0.71256, which indicates that the larger the loan share, the higher the risk. The Non-Performing Loan Ratio variable (NPL) is significant at the 5% level and has a coefficient of 0.03652, which illustrates a positive correlation between the two, and an excessively high NPL ratio will lead to a greater risk for the bank, and if it wants to minimize the bank’s risk, its NPL ratio has to be controlled within a certain level.