^{1}

^{1,2}

^{1}

^{2}

We introduce a new perspective to systematically investigate the cause-and-effect relationships among competition, innovation, risk-taking, and profitability in the Chinese banking industry. Our hypotheses are tested by the structural equation modeling (SEM), and the empirical results show that (i) risk-taking is positively related to profitability; (ii) innovation positively affects both risk-taking and profitability, and the effect of innovation on profitability works both directly and indirectly; (iii) competition negatively affects risk-taking but positively affects both innovation and profitability, and the effects of competition on risk-taking and profitability work both directly and indirectly; (iv) there is a cascading relationship among market competition and bank innovation, risk-taking, and profitability.

Managing a bank is a complex process that involves interactions of numerous factors. Among all of these factors, risk-taking and profitability are the most two important indicators of bank performance.

Innovation is regarded as an important factor which influences both risk-taking and profitability, because bank’s innovation activities improve the efficiency of the screening and monitoring of borrowers, which then reduces risk-taking and enhances profitability. Besides, financial innovation generates technology-based products which have several advantages, such as high returns and low risk.

Competition is widely accepted as an important factor that influences all of innovation, risk-taking, and profitability. Specifically, it is generally acknowledged that market competition changes a bank’s operating conduct, which subsequently influences bank risk-taking and profitability. For example, the market competition stimulates bank’s innovation activities and makes bank more efficient in screening and monitoring borrowers, which in turn affects bank risk-taking and profitability.

Though there are lots of empirical works about the cause-and-effect relationships in banking, most of them focus on one-on-one relationships between the factors. We contribute to the literature by clearly presenting a technical framework of structural equation modeling (SEM) applied in the banking industry and systematically investigating the relationships among competition, innovation, risk-taking, and profitability. Accordingly, our purpose is to reveal the interactional mechanism of these four factors through the application of SEM.

We select the Chinese banking industry as a sample, which is impacted by financial innovation in many aspects, such as organization, management, production, and business [

The structure-conduct-performance (SCP) hypothesis from traditional industrial organization theory states that a firm’s performance is determined by its business strategy which is influenced by industry structure [

A standard view of banking supervision is that competition is detrimental to bank stability. On the one hand, competition erodes a bank’s franchise value which is equivalent to the cost of bankruptcy and encourages bank to pursue risky policies, such as lowering capital levels and softening the terms of loans, which increase nonperforming loans and result in credit risk [

The relationship between market competition and innovation is a primary focus of industrial organization theory. Schumpeter first states that market competition discourages innovation by diminishing monopoly rents and large firms are able to afford more capital for innovation activities [

The efficiency hypothesis (EH) posits that the bank profitability depends on the bank’s degree of efficiency, whereas the bank’s degree of efficiency is affected by its financial innovation activities [

Chen states that a bank’s innovation activities improve the efficiency of the screening and monitoring borrowers and eventually reduce the quantity of nonperforming loans and the bank’s credit risk [

The capital asset pricing model (CAPM) provides the first coherent framework for interpreting how the risk of an investment affects its expected return and depicts that the expected return is calculated by adding the risk free interest rate to the product of the investment’s beta and the expected market risk premium [

SEM is a collection of procedures that tests hypothesized relationships among observed variables and is often related to causal modeling [

Thus, we use the structural equation modeling (SEM) to systematically test the cause-and-effect relationships among competition, innovation, risk-taking, and profitability.

The path model involves hypothesized cause-and-effect relationships among variables. These relationships are usually based on theoretical and empirical evidences from existing literature. Figure

A hypothetical path model.

Generally, an exogenous variable is always only a cause, but an endogenous variable can be a cause and/or an effect [

In this scenario, competition is hypothesized to have one direct effect and three indirect effects on profitability. The indirect effects are mediated by innovation and risk-taking, and the indirect paths are competition-innovation-profitability, competition-risk-taking-profitability, and competition-innovation-risk-taking-profitability, respectively. Innovation and risk-taking constitute the intermediate part of the indirect paths, indicating a cascading relationship among competition, innovation, risk-taking, and profitability.

The measurement model is employed to quantify linkages between latent variables and observed variables.

Latent variable is a conceptual variable that is not directly observed but inferred from other observed variables. In our model, competition, innovation, risk-taking, and profitability are all regarded as latent variables.

Observed variables represent a specific latent variable in the form of a linear combination. The selection of observed variables must consider validity and reliability. Validity is the accuracy of an observed variable for representing a latent variable, and reliability is consistency and stability of an observed variable for representing a latent variable.

The measurement model of the exogenous variable is defined as follows:

In our paper, competition is defined as

Similarly, the measurement model of the endogenous variable is defined as follows:

We define (i) innovation as

The structural regression (SR) model combines the path and measurement models and is the most general kind of core model widely applied in SEM to take measurement errors of observed variables into account. The general form of SR model is

Figure

A simple form of SEM model.

In our paper, (

We follow Lam and Maguire and choose four basic indices to assess model fit [

The chi-square test, which is based on the test statistic

The root mean square error of approximation (RMSEA): a value range from 0.05 to 0.08 which indicating a reasonable approximation and a value

The standardized root mean square residual (SRMR): the SRMR < 0.10 which is considered a good model fit

Joreskog-Sorbom goodness of fit index (GFI): The GFI which ranges from 0 to 1.0 with 1.0 indicating the best fit

We test the hypothesis with the financial statement data from 14 listed commercial banks in China, namely, Bank of China, Industrial and Commercial Bank of China, China Construction Bank, Agricultural Bank of China, Bank of Communications, China CITIC Bank, China Merchants Bank, China Minsheng Bank, Industrial Bank, China Everbright Bank, Hua Xia Bank, China Guangfa Bank, Ping An Bank, and Shanghai Pudong Development Bank. All the relevant data is obtained from the China Statistical Yearbook and the annual reports of the banks from 2004 to 2014. These 14 banks account for nearly 70% share in both the deposit and loan markets (see Table

Market shares of sample banks.

2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | |
---|---|---|---|---|---|---|---|---|---|---|---|

Deposit | 0.81 | 0.78 | 0.77 | 0.74 | 0.73 | 0.72 | 0.70 | 0.70 | 0.69 | 0.67 | 0.66 |

Loan | 0.76 | 0.73 | 0.71 | 0.70 | 0.68 | 0.69 | 0.68 | 0.68 | 0.67 | 0.66 | 0.65 |

This table reports the market share of sample banks in both deposit and loan markets.

We select return on assets (ROA), return on equity (ROE), and net interest margin (NIM) to measure bank profitability.

ROA is the ratio of net profit to average total asset, which reflects the ability of a bank to generate profits by using its asset.

ROE is the ratio of net profit to shareholder equity, which reflects the ability of a bank to generate profits by using investment.

NIM is the ratio of net interest income to the bank’s interest-earning assets, which reflects the profitability of the bank’s credit business.

Bank risk-taking is the risk resulting from banks’ policies. We uses nonperforming loan (NPL) ratio, risk-weighted asset ratio, and

Nonperforming loan (NPL) is the loan that is in default or close to being in default, and the nonperforming loan ratio is the ratio of NPL to total loan.

Risk-weighted asset ratio is the ratio of risk-weighted asset to total asset, which measures bank risk-taking through capital constraint. The value of the risk-weighted assets can be calculated by dividing the sum of core capital and supplementary capital by capital adequacy ratio.

The greater the

In order to measure bank innovation, we employ the cost efficiency (CE) and the technology gap ratio (TGR) and construct bank’s cost function to examine bank’s ability to minimize costs via innovation. Because the cost function describes the relationship between outputs and inputs, the set of estimated parameters reflects the state of technology and a parallel shift of the cost curve refers to the technical change triggered by innovation.

Figure

Cost efficiency and technology gap ratio.

The cost efficiency represents the ratio of the minimum cost on the annual cost frontier to the bank cost for a certain level of output, and the technology gap ratio represents the ratio of the minimum cost on the metafrontier to that on the cost frontier for a certain level of output. The cost efficiency equals 1 when bank operates on the annual cost frontier, which means there is no inefficiency. Because the metafrontier is invariably below the cost frontier, the value of the technology gap ratio is bounded between 0 and 1, where the latter is reached when bank operates on the metafrontier. Innovation generates technical improvement and consequently increases the value of cost efficiency and technology gap ratio.

Following Bos and Schmiedel and Bos et al., we initially employ the Stochastic Frontier Analysis (SFA) to estimate the cost frontiers in each year and then use a linear program to obtain the metafrontier [

We express total cost as

Because_{it}. The cost efficiency is obtained as follows:

We model bank production by employing the intermediation approach because of banks’ financial intermediary role. The intermediation approach regards banks’ fixed asset, labor, and funds as inputs (

Definitions of variables.

Variable | Definition |
---|---|

Total cost | |

Capital cost | Interest expense |

Operating cost | Business and management cost, loss of impairment of asset |

Labor cost | Employee salary and welfare |

Input | |

Labor price | The ratio of cash payments for salaries and staff expenses to amount of labor |

Fixed assets price | The ratio of depreciation to net fixed asset |

Funds price | The ratio of interest expense to the sum of customer deposit and borrowed fund |

Output | |

Investment | Financial assets at fair value through profit or loss, available-for-sale financial assets, held-to-maturity investment, receivables-bond investment |

Loan | Personal loan, corporate loan, other loans |

This table reports the definitions and measurements of the variables in (

We obtain the metafrontier by enveloping the cost frontiers and utilize the parameter estimates for the cost frontiers

The constraint condition in (

We employ concentration ratio (CR), Herfindahl-Hirschman Index (HHI), and Boone index to measure competition.

Concentration ratio (CR) is a reverse indicator of competition and measures the total market share occupied by a specified number of the largest banks in the industry. We select

Herfindahl-Hirschman Index (HHI) is a commonly accepted measure of market concentration. The value of the HHI is calculated by squaring the market share of all banks in the industry and then summing the result.

Boone index is an indicator of competition proposed by Boone and associates performance with the difference in efficiency [

With respect to the banking industry, the profit (

To allow for time variation and capture

Although a comprehensive reform is placed to facilitate the liberalization, the monetary authority still plays a leading role in Chinese banking industry. The market structure mainly resulted from the policies of the authority, rather than market discipline. Thus, we obtain several hypotheses for the empirical work.

(i) The SCP hypothesis assumes that bank profitability is influenced by market structure or competition. A concentrated market indicates collusion and monopoly, which make a bank earn high profits. In other words, a bank in a concentrated market has a high profitability. Thus, we hypothesize that competition is detrimental to bank profitability.

Competition has a negative relationship with bank profitability in the Chinese banking industry.

(ii) Competition has both positive and negative effects on bank risk-taking at the same time. The intensities of these two effects vary with competition; thus the relationship between competition and bank risk-taking is unclear. We hypothesize that competition is negatively related to risk-taking, considering that the Chinese banking industry is highly concentrated and bank’s market interest rate primarily results from administration policies rather than market discipline.

Competition has a negative relationship with bank risk-taking in the Chinese banking industry.

(iii) Banks in a concentrated market can afford more capital for innovation activities due to the monopoly rents, but a competitive market provides more incentive factors for bank innovation than in a concentrated market. In Chinese banking industry, the domination of banks mainly resulted from the policies of the authority, rather than market discipline. Therefore, we hypothesize that competition is positively related to banks’ innovation activities.

Competition has a positive relationship with bank innovation in the Chinese banking industry.

(iv) The efficiency hypothesis (EH) states that bank profitability depends on its efficiency, which is affected by bank’s innovation activities. In addition, new technology-based products generated by innovation also enhance profitability. Therefore, we hypothesize that innovation activities are beneficial to bank profitability.

Innovation has a positive relationship with bank profitability in the Chinese banking industry.

(v) Though innovation can trigger technical progress and efficiency improvement, whether innovation is beneficial or not depends on why and how it is used by banks. Based on the empirical work about Chinese commercial banks, we hypothesize that innovation activities increase bank risk-taking behaviors.

Innovation has a positive relationship with bank risk-taking in the Chinese banking industry.

(vi) Based on the capital asset pricing model (CAPM), the expected return is positively related to the expected market risk premium. Commercial banks can be perceived as investor, because they must manage assets through investments. Therefore, the CAPM can be applied to investigate the relationship between bank risk-taking and profitability. We hypothesize that there is a positive relationship between bank risk-taking and profitability.

Risk-taking has a positive relationship with bank profitability in the Chinese banking industry.

(vii) We select competition, innovation, risk-taking, and profitability as variables and investigate the cause-and-effect relationships among them. The one-on-one relationship between any two factors is supported by both empirical and theoretical works. For example, market competition directly impacts bank innovation which has an effect on risk-taking; furthermore, the risk-taking is hypothesized to be related to profitability which is affected by competition. Thus, we deduce that innovation and risk-taking play mediating role in the relationship between competition and profitability. That is to say, there is a cascading relationship among these four factors.

There is a cascading relationship among market competition and bank innovation, risk-taking, profitability.

We verify the hypotheses by applying the Analysis of Moment Structure (AMOS) software, version 21. Owing to the difference of dimensions of the observed variables, the standardized processing of original data is required. The estimated path coefficients depict the effects among variables (Figure

The path coefficients of variables. The exogenous latent variable, competition, is measured by the observed variables (CR_{4}, HHI, and Boone index). The endogenous latent variables are innovation (measured by CE and TGR), risk-taking (measured by

The indices of model fit illustrated in Table

Indices of the model fit.

| RMSEA | GFI | SRMR |
---|---|---|---|

7.57 | 0.053 | 0.92 | 0.09 |

This table reports the value of indices of model fit.

Figure

Risk-taking-profitability: this path represents a direct effect of risk-taking on profitability (0.890). Specifically, one percent increase in risk-taking directly increases profitability by 0.890 percent. This result supports Hypothesis

Innovation-risk-taking-profitability: innovation has a direct effect on risk-taking (0.471), which supports Hypothesis

Competition-innovation-risk-taking-profitability: competition affects innovation directly (0.408), which supports Hypothesis

The competition-innovation-risk-taking-profitability path of the indirect effect of competition on profitability shows that innovation and risk-taking constitute the intermediate part of this effect. This result supports Hypothesis

We introduce a new perspective to systematically investigate the cause-and-effect relationships among competition, innovation, risk-taking, and profitability in the Chinese banking industry. Several conclusions are obtained based on our empirical work.

A positive relationship between risk-taking and profitability has been demonstrated, which indicates that the risk appetite of bank managers has significant influence on bank profitability, and the rising appetite for risk may increase the profitability of bank asset. Therefore, a balance between profitability and risk-taking is vital to banks.

The net effects of innovation on both risk-taking and profitability are positive. One percent increase in innovation amplifies risk-taking by 0.471 percent and profitability by 0.625 percent. Thus, it is concluded that innovation more intensely affects profitability than it does risk-taking, suggesting that financial innovation should be actively supported and promoted.

Competition affects all of innovation, risk-taking, and profitability. One percent increase in competition promotes innovation by 0.408 percent and profitability by 0.241 percent, and it reduces risk-taking by 0.542 percent. This outcome shows that competition can encourage innovation, enhance profitability, and reduce risk. This result indicates that competition in the Chinese banking industry is not sufficiently intense enough. Although comprehensive reform facilitated liberalization, the monetary authority still plays a leading role.

Innovation and risk-taking constitute the intermediate part of the competition-innovation-risk-taking-profitability path, suggesting a cascading relationship among competition, innovation, risk-taking, and profitability.

The authors confirm that they all have checked the manuscript and have agreed to the submission.

There is no conflict of interests regarding the publication of this paper.

This work was supported by the National Natural Science Foundation of China under Grants nos. 71373072 and 71501066; China Scholarship Council under Grant no. 201506135022; Specialized Research Fund for the Doctoral Program of Higher Education under Grant no. 20130161110031; and Foundation for Innovative Research Groups of the National Natural Science Foundation of China under Grant no. 71521061.