An Empirical Research Of The Impact Of Non-interest Income On Systemic Risk In China

From 2007 to 2009 of the international financial crisis shows that the risk is no longer just limited to the micro individuals, but shows the characteristics of the systematic cluster. Widespread collapse of financial institutions because of the market risk, liquidity risk, closely related with financial institutions excessive innovation and infection. Commercial bank as an important part of the financial industry, have a high value in the financial regulatory policy and regulatory practice, "too big to fail" and "too associated to fail" become a hot topic in the theory and practice to discuss. The main business of commercial banks is the net interest and non-interest business. Though commercial bank non-interest business started late in our country, the development is very rapid. The Notice About Pushing The Reform of Interest Rate Marketization Forward Further, published in July 2013, to break the" loan interest rate limits lower, deposit interest rate limits upper ", further interest rate marketization, commercial bank spreads narrow space, the traditional commercial banking face great challenge. At the same time,13th Five Year Plan clearly pointed out that our country should build a multi-level, broad coverage, discrepant bank organization system, improve the proportion of direct financing. Under this background, the commercial Banks in China must seize the opportunity, actively adjust the business model, enlarge the proportion of non-interest income, so as to better position in market competition.Will the commercial banks expand non-interest income affect the financial system to systemic risk? Cause what kind of impact? International and domestic scholars carry on the extensive discussion, but has yet been reached. Some scholars think the non-interest income helps reduce systemic risk, as Baele et al. (2007) using extreme value theory method to measure systemic risk, and use it to study of non-interest income and systemic risk. Zhang Xiaomei and Mao Yaqi (2014) using LRMES and the data of listed commercial banks in China draw similar conclusions. Other scholars think the non-interest income will increase systemic risk, such as Ivashina and Lerner (2010) think most non-interest income has the pro-cyclical, such as investment banking, thus enhancing systemic risk. As Brunnermerier and Oehmke (2012) use ?CoVaR method to measure of systemic risk, also come to the conclusion that non-interest income will increase systemic risk, Wang Haiyou (2011) using data of listed commercial banks in China concluded similarly with Brunnermerier and Oehmke (2012).Commercial banks are adjusting the income structure at present stage, to discuss the relationship between commercial banks'non-interest income and the systemic risk has theoretical significance and urgency of the regulatory practice level, but the domestic relevant empirical research is relatively scarce. Jonghe (2010) microscopic analysis on the stability of the banking system, found the relationship between commercial banks'non-interest income and systemic risk exists nonlinear effect, the nonlinear effect may caused by bank scale or institutional factors. Smaller banks corporate governance structure is imperfect, has poor quality of information disclosure, lack of risk management techniques and experienced risk management, are more likely to engage in risky activities, lead to systemic risk increase (Flannery et al.,2004; Milbourn et al.,1999). At the same time, the smaller bank in business has less external supervision, is more likely to carry out the risky business (Freixas et al.,2007). Starting from this thinking, this paper tries to from the perspective of scale heterogeneous, to explore further the influence commercial bank's non-interest income on systemic risk in our country.This paper builds the systemic risk analysis framework, using the TARCH, DCC-GRACH calculate four kinds of commonly used as a single measure ?Co VaR, MES, SRISK, CES, through principal component analysis method from the four kinds of single measure to extract the information construction of comprehensive index, to avoid a single index measure respectively. On this basis, based on the perspective of the size of the heterogeneous, threshold panel model of non-interest income and the non-linear model of systemic risk, through the self-help method sample build statistics method of threshold effect testing, using the lattice point search method obtained the corresponding threshold value, and fixed effect panel regression. After decomposition of non-interest income structure, this paper discusses nonlinear relationship of causes and non-interest income of different part of the difference influence to the systemic risk.By using TARCH, DCC-GRACH to compute four kinds of common used single measurement index(?CoVaR, MES, SRISK, CES),at first, this article constructs analysis framwork of systemic risk. Through principal component analysis method, it extracts information to constrcut comprehensive indicator, avoiding bringing the one-sideness which comes from single indicator of measurement. Based on this fondation and the prespective of heterogeneous scale, through self-help method sampling to constrcut statistic of method to measure the threshold effect and by using point search method obtaining the corresponding threshold value, it builds a nonlinear relationship modelling of threshold panel model of non-interest income and systemic risk and carry the fixed effect panel regression. After that, through the decomposition of non-interest income structure, it discuss the causs of nonlinear relationship and systemic risk difference effect due to different parts of nonlinear income.With empirical analysis of the capital market of Listed commercial Banks in China from 2008 to 2015 in the third quarter, the balance sheet data and special data regarding non-interest income effects on systemic risk, this paper reveals three main points. (1) Principal component extraction method to extract of existing commonly used indicators could construct a better measure of systemic risk indicators. From the view of measurement results, in our country size effect still needs to be focused on systemic risk. But at the same time, the difference of systemic risks between big banks and small banks is narrowing. (2)There is nonlinear characteristics under the scale heterogeneousness in non-interest income and systemic risk. Bigger banks can reduce systemic risk in non-interest business because of the strong profitability, risk management level and a higher quality of information disclosure. On the other side, Smaller Banks are more likely to engage in risky activities, and they will increase systemic risk instead due to carrying out non-interest income business. (3) No matter the bank size, other business net income are positively related to the systemic risk. The main source of non-interest income and nonlinear effect of systemic risk is handling charge and commission charge. Meanwhile,there is no threshold effect in other business net income, and other business net income business will increase systemic risk.This article is divided into five parts for empirical research on influence of country commercial bank's non-interest income on systemic risk. The first chapter is introduction. Introduction part mainly introduces the background paper topics, research significance, shows that in the post-financial crisis era, with strengthening prudential regulation and commercial banks facing transition in China, non-interest income ratio rising gradually, studying on the significance of the impact of commercial bank's non-interest income on systemic risk is foundation of my analyses, meanwhile, expounding the research content, research methods and innovation points in this paper. The second chapter is literature review. This part based on defining the systemic risk and the non-interest income, respectively, sum up the related articles about researches of systemic risk, analysis of the importance of system, relations between non-interest income and business risk, the relationship between non-interest income and systemic risk, then points out the shortage.The third chapter is the research design. This chapter first describes the analytical framework about systemic risk.Systemic risk referred to herein includes consideration of the bank's own risk and the infection risk caused by association relationships with others. And including the latter is an important feature of this paper that is different from extensive literature merely measuring individual banks risks.This chapter describes each of the four single indicators used to measure systemic risks, including ?CoVaR, MES, SRISK, CES and describes the characteristics,the evolution and the calculation methods of these indicators.Considering that the above four indicators are able to portray different aspects of the systemic risk, this chapter applies the principal component analysis method to construct the index, and some details of PCA method have been described, including the standard of extraction of principal components, processing principal component analysis for panel data in the application.After elaborates systemic risk analysis framework,this chapter describes the empirical model of systemic risk related to non-interest income and variable selection.In order to examine possible non-linear threshold effect between non-interest income and systemic risk, and to avoid artificial division threshold by subjectivity, this paper adopts the panel threshold model proposed by Hansen (1999) to study the relations between systematic risk and non-interest income, and builds models separately to investigate different parts of non-interest income how to affect the systemic risk including fee,commission income,and other net income.This chapter also describes threshold effect test method, solving method and confidence intervals in detail, and explains the core variable, dependent variable, the control variable definitions and calculation methods. On this basis, the sample selection and data sources used herein is also described.The fourth chapter is empirical analysis. Firstly, it calculates the index value of the four indicators of systemic risk used the actual data and extracts information from the four indice by principal component analysis, the results show that the composite indicator by principal component analysis method can be able to explain "too big to fail" and "too associated to fail" based on which the time-varying characteristics of systemic risk in the banking system is analyzed.Secondly, the chapter describes descriptive statistical characteristics of each variable and analyses the evolution trend and current situation of non-interest income of China's commercial banks.in the view of the total size,the non-interest business have been expanding at the same time it may also gather a lot of risk;in the view of the scale of various commercial banks, large state-owned commercial banks and joint-stock commercial banks have much difference in the growth rate and business structure; from the perspective of non-interest income constitution, the fee and commission business are main components of non-interest business.Thirdly, the threshold effect of empirical model is tested, according to the threshold effect test results it builds threshold model and calculates the threshold value and confidence intervals.the results indicate the presence of non-linear threshold effect between non-interest income and systemic risk, the threshold value, that is 10.10 trillion yuan, quite match with actual situations of China's systemically important banks.Fourthly, according fixed effects panel regression model, regression results show that for large-scale banks due to strong profitability, risk management and higher quality of information disclosure, to carry out non-interest business can reduce systemic risk.but to carry out non-interest income business will increase systemic risk for the small-scale because of easier to engage in risky activities.Finally,it explores the causes of non-linear relationship and different impact on systemic risk generated by different components of non-interest income by non-interest income structural decomposition.The results show that the main source of non-linear effect is fee and commission income, and there is no threshold effect between systemic risk and other net income.to carry out other net income operation business will result in systemic risk to rise. The fifth chapter is the conclusion. The fifth chapter summarizes the results of this study, thereby makes relevant policy recommendations, points out some defect and the direction for future study.The innovation of this study is mainly reflected in the following aspects:(1)It applies a variety of methods for the measurement of systemic risk and conducts principal component analysis to extract useful information, to better describe individual financial institutions' contribution to systemic risk.(2)In the view of the size, it explores the non-linear relationship between systemic risk and non-interest income by the panel threshold regression model,which shows that impact on systemic risk of non-interest income is different for various-sized commercial banks,to provide new ideas for related research.(3)the influence on systemic risk of parts of non-interest income is discussed in detail. It is found that fee and commission income is the cause of non-linear relationship and to carry out other net income business will result in increasing systemic risk.Although this paper makes some innovations on the basis of the relevant literature and draws useful conclusions, here are still many shortcomings, specifically summarized as follows. (1) The measurement of systemic risk may not be entirely accurate. Although the paper chooses four commonly used methods, while extracting information by constructing a composite index, the method does not contain other typical methods (extreme value theory etc.). and there is no systemic risk measurement method widely accepted. (2)The samples have certain limitations. The paper only selects China's listed commercial banks as samples,thus ignoring large number of city commercial banks, credit unions and other deposit-taking financial institutions, which may have a greater impact on systemic risk. Hence the relationship between systemic risk and non-interest income in China may not be fully characterized.