The influences of non-interest income on risks between
Chinese and US commercial banks
U
NIVERSITY VAN
A
MSTERDAM
Amsterdam Business School
Student Name: Liuzhu Meng
Student Number: 10916172
Thesis Supervisor: Dr. Dijkstra Mark
1
Statement of Originality
This document is written by Student Liuzhu Meng who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
2 content Abstract ... 3 1. Introduction ... 4 2. Literature review ... 6 2.1 Important knowledge ... 7 2.1.1 Non-interest income ... 7
2.1.2 Risk of commercial banks ... 7
2.2 Development of non-interest income business in the United States ... 8
2.3 Development of non-interest income business in China... 10
2.4 Influence of non-interest income in banks’ operating risk in America ... 14
2.5 Influence of non-interest income in banks’ operating risk in China ... 17
3. Theory analysis ... 18
3.1 The portfolio theory ... 18
3.2 Non-interest income business model ... 19
4. Empirical method and Data ... 21
4.1 Empirical method ... 21
4.2 Data ... 23
4.2.1 Data sources and sample selections ... 23
4.2.2 Explanatory variables ... 24
4.2.3 Descriptive statistics ... 26
5 Results ... 28
6. Robustness check ... 31
7 Policy Suggestion and Conclusion ... 35
7.1 Policy suggestion ... 35
7.2 Conclusions ... 37
3 Abstract
In the late 19th century and early 20th century, the connectivity of the world's economies and cultures grew very quickly, and globalization began to establish itself. Globalization can bring greater global competition, increased opportunities for success, and greater innovations. With the advance of globalization and accompanying financial liberalization, the traditional commercial bank income structure has been challenged. Traditional interest businesses have provided lower profits, and therefore the development of high profit non-interest income business has become a new option of choice. While profitability and security have been the two basic criteria of the bank prudent operation principle, this principle doesn’t account for unrealized profits, but only for the possible loss. Consequently, non-interest income business has high risk, so that the expansion of non-interest income business will have a significant impact on the risks facing with commercial banks.
This paper focuses on the impact of banks’ non-interest income on operating risks, using data on Chinese banks and US banks for the period between 2007-2014. It examines the relevant literature concerning the impact of non-interest income on banks income and risks, then categorizes the sources of non-interest income of commercial banks and analyzes the non-interest income characteristics and specific risks. After that, this paper will review the development of commercial banks’ non-interest income in China and the United States. Next, empirical models and results will be described in detail. This paper will try to test the interaction of country style and non-interest income ratio to explore the relationship. An expected important finding of this examination is that there is no great difference between Chinese commercial banks and US commercial banks in this regard; that is, the increase in non-interest income would increase risks.
4 1. Introduction
Since 1949 the Chinese economy has developed quickly. Especially after 2000, there have been numerous economic developments in China, including financial reforms such as the marketization of interest rates1, banking reforms, the internationalization of RMB2, and strict capital requirements. At the same time, financial disintermediation is the inevitable result of financial and economic development, and with the establishment of Shanghai and Shenzhen stock exchanges this financial disintermediation has been highlighted. Meanwhile, the profit of the traditional deposit and loan business has been gradually squeezed for commercial banks, with competition becoming increasingly fierce. For instance, during the early years of banking reform, interest income accounted for more than 85% of total bank income and remained consistently at this high level. However, commercial banks have begun to develop non-interest businesses and to pursue diversified strategies. In 1987, credit card transaction volume was only 20 billion RMB, however, 20 years later, this volume was more than 3,000,000 billion RMB. The proportion of non-interest income business increased hugely from 3%-5% in 2000 to around 20% in 2014. Through the exploration of business diversification, commercial banks in China have kept making efforts to go beyond the boundaries of the traditional services provided by the commercial banking business and, as a result, non-interest income businesses have developed widely. This paper tries to explore the influence of the increasing non-interest business in commercial banks.
Non-interest income is a mixture of heterogeneous components. These include trading and securitization income, investment banking and advisory fees, income from brokerage commissions, venture capital, and fiduciary income. In these activities,
1 What is called interest rate marketization refers to the transition procedure that both sides of supply
and demand in financial market decide the interest rate level mechanism on their own. Particularly speaking, market-oriented interest rate refers that the central bank controls the benchmark interest rate based on monetary policy, every commercial bank adjusts the interest rate of deposit and loan on their own.
2
According to the Society for Worldwide Interbank Financial Telecommunication (SWIFT), the path of RMB internationalisation can be divided into three phases—first as usage for trade finance, then for investment, and in the longer term, as reserve currency
5
banks are competing with other capital market intermediaries such as hedge funds, mutual funds, investment banks, insurance companies and private equity funds.
For commercial banks, raising the proportion of non-interest income could bring a new source of profit. The US banking industry provides a good example. It has already steadily shifted away from traditional sources of revenue to non-traditional sources of revenue. For instance, the industry increasingly earns fee income, service charges, trading revenue, and other types of non-interest income instead of loan income. Consequently, the non-interest income ratio in the US rose from merely 25% of net operating revenue in 1984 to 43% of net operating revenue in 2001. However, the banking industry in China began the transformation to this kind of business later. In 2011, the total non-interest income of Chinese commercial banks was 514.9 billion RMB, rising to 902.2 billion RMB by 2014. That is, the proportion of non-interest income business in China increased hugely from 3%-5% in 2000 to around 20% in 2014. The rate of increase in of non-interest income is considerably higher than the comparative rate (around 0%) of net interest income increase during the same period.
The increased share of non-interest income also increases the operating risks of commercial banks, as non-interest income business has a higher volatility compared with interest income businesses. Therefore, researchers such as Lown (2000), Stiron (2004), and Lepetit (2008) have begun to pay close attention to this field of operations, notably investigating how to ensure the rapid development of the non-interest income business without increasing the operating risks of banks. Thus, this paper studies the link between non-interest income ratio and banks’ operating risk using data on Chinese banks and US banks for the period between 2007 and 2014.
The outline of this article is as follows. Section 1 is the introduction. Section 2 systematically examines the related literature of bank income diversifications, bank performance, and bank risks. It summarizes the development of non-interest income business in both countries and provides several different opinions about the influence of non-interest business, covering positive relationships, negative relationships and no significant relationships. This part also explains some key works and links in the field. Section 3 uses theory to explain the influence of non-interest income business in
6
banks’ operating risks. Section 4 gives the empirical methodologies. In this empirical method, it creates a dummy variable to test the differences of both countries before providing the 10-year data sources and a summary of statistics. After that, Section 5 contains and main results. In order to obtain more convincing results, Section 6 performs further robustness checks, including replacing the dependent variable, choosing different time periods and narrowing the sample size. Section 7 provides the conclusions of this paper.
2. Literature review
The development of non-interest income business in the US differs from China. In the US, development has been ongoing for more than 100 years and has already reached a high level (around 35% of total commercial banks income). However, China only started to pay attention to this kind of business within that last 50 years, and the proportion of non-interest business has reached only 20% in recent years. Besides the differences in the processing of non-interest income business in both countries, the influence of non-interest income in the assessment of risk for commercial banks in the US is also different from that in China. The main opinions are as follows: for American commercial banks, research found that the increase in non-interest income would reduce the risk of commercial banks. In China, the studies of relationship between non-interest income and the levels of risk broadly generated two conclusions. One is that the increase in non-interest income of commercial banks would have no significant impact on the business risk of the bank. The other is that the increase in non-interest income of commercial banks would reduce the business risk of banks. (similar to the studies of the US banks).
7 2.1 Important knowledge
2.1.1 Non-interest income
Non-interest income is a form of income that is generated from the application of fees, rather than from interest that is applied to the outstanding balance of a financial account. In some cases, the non-interest income is associated with recurring fees that are assessed on customer accounts each month or each year. For instance, investment accounts may apply a semi-annual or annual maintenance fee for as long as the investment account remains open to a specific brokerage. Sometimes, the fees are one-time charges that are applied in return for a type of task or service that is provided by the account issuer.
For banks, more than 50% of the source of non-interest income is the fee associated with the management of customer accounts. For example, a checking account may be structured to allow a small fee to be debited from the client’s account on a monthly basis. These fees can take the form of service fees, where the charge is in exchange for such tasks as posting debits and credits to the account, withdrawals of cash from the account, and supplying the customer with a monthly statement of account.
2.1.2 Risk of commercial banks
Market participants seek the services of financial institutions because of their ability to provide market knowledge, transaction efficiency and funding capability. In performing these roles, commercial banks generally act as a principal in the transaction. They use their own balance sheet to facilitate the transaction and to absorb the risks.
Services include agency and advisory activities, such as (1) trust and investment management, (2) private and public placements through facilitating contracts, (3) standard underwriting, or (4) the packaging, securitizing, distributing and servicing of loans in the areas of consumer and real estate debt. These items are absent from the
8
traditional financial statement because they rely on generally accepted accounting procedures rather than a true economic balance sheet. Nonetheless, the overwhelming majority of the risks faced by banking firms are in on-balance-sheet businesses. Accordingly, it is here that the review of risk management procedures will concentrate.
2.2 Development of non-interest income business in the United States
American commercial banks engaged in non-interest income business since their establishment. Non-interest income business can be traced back at least to the early years of the last century. While the main British and European financial centers were preoccupied with the demands and costs caused by World War I, the US remained far from the battlefield, with the demand for materiel from the belligerent nations enabling its economy to flourish, financial industry to rapidly develop, and non-interest business gain momentum. As the Great Depression began in 1929, all of these developments came to a halt and the US economy experienced the 10-year long Great Depression, with thousands of banks collapsing. In 1933, the US Government promulgated the famous Glass-Steagall Act, which strictly divided investment banking and commercial banking. The banking and Securities sectors ran their businesses separately in the US financial industry. With the hindrance provided by the Glass-Steagall Act, the non-interest business in the US experienced a tough time and, in many areas, stopped developing. As the Western economies recovered from World War II, capital in the international credit markets became active again. Until the late 1970s when the Bretton Woods System collapsed, the financial industry underwent a degree of liberalization. During this period, the non-interest business of American commercial banks undertook some business innovations, including expansion of investment banking businesses, increase of trading and securitization activities and income, and taking increased brokerage commissions. Because nearly all kinds of these innovations have high risk and high volatility, the commercial banks’ development was largely restricted by laws and regulations, and in general the
9
situation allowed for little optimism regarding their development.
In the 1980s, the US conducted a series of reforms within the financial industry. There are four main acts: the 1980 Depository Institutions Deregulation and Monetary Control Act, the Deposit Institution Act 1982, the Banking Industry Equal Competition Act 1987, and the Banking Institutions Reform, Recovery and Enforcement Act of 1988. The loosening-up of measures of control brought American banking into what has been viewed as another golden period. In 1980, the US issued the Depository Institutions Deregulation and Monetary Control Act, which made it possible for the loan interest rate’s ceiling to be canceled. Thus the limits of the business scope of commercial banks were loosened and the differential treatment between different financial organizations disappeared. The deposit interest rate cap regulations (regulation Q) had greatly reduced commercial banks’ ability to take in deposits by strictly regulating the interest rate cap, and thus the overall return that could be generated. In 1994, the US issued the Riegle-Neal Interstate Banking and Branching Efficiency-Act, which allowed commercial banks to operate interstate businesses. American commercial banks could therefore expand the scope of their business all over the country, laying a solid foundation for the development of non-interest business activities.
In 1999, the US Congress passed the Financial Services Modernization Act, the passage of which ended the 66-year restriction separating the banking and security industry under the Glass-Steagall Act and declared the end to the separate operating system. The US financial industry entered a brand-new stage, opening up the new era of development for non-interest business with the loosening of operating controls and restrictions led to the rapid development of non-interest businesses.
At the same time as these legal and regulatory changes were taking place, innovation in information and communication technology enabled non-interest business to use new operating methods, greatly cutting transaction costs. Innovation and improvement in financial technology also broadened the scope of non-interest business. Within a few years, loan securitization and financial derivatives had become the main source of commercial banks’ non-interest revenue.
10
TABLE 1—Non-Interest Income Ratio in American Banks Million US dollar
Year Bank of America Citibank JPMorgan chase PNC Bank US Bancorp Well Fargo Bank average 2007 0.4517 0.3975 0.6300 0.5653 0.5174 0.4551 0.5028 2008 0.3445 -0.1465 0.4209 0.4674 0.4683 0.3877 0.3237 2009 0.6002 0.4323 0.5372 0.4650 0.5010 0.4707 0.5011 2010 0.5219 0.3764 0.5036 0.3922 0.4687 0.4688 0.4553 2011 0.5036 0.3778 0.5086 0.3917 0.4687 0.4656 0.4527 2012 0.5050 0.3758 0.5370 0.3746 0.4664 0.4924 0.4585 2013 0.5154 0.2984 0.5544 0.3785 0.4414 0.4800 0.4447 2014 0.5207 0.2856 0.5412 0.3954 0.4420 0.4689 0.4423
Notes：This table illustrates the non-interest income ration in 6 most important American commercial banks from 2007 to 2014. Data is from bankscope database.
Because of financial globalization, regulatory loosening and technological improvement, American commercial banks enjoyed huge success operating in the non-interest business area until its progress was interrupted by the Global Financial Crisis from 2007 (the average proportion of non-interest income business had increased up to about 50% of total commercial banks income). In 2008, non-interest revenue and its proportion of income underwent a sharp drop, about one-fifth, before recovering in 2009. Since 2009, the banking industry’s non-interest revenue and its proportion of American commercial banks income has remained broadly stable at 45%, although overall non-interest income has actually declined overall. Excessive financial innovation and a lack of proper supervision led to the outbreak of the 2007 financial crisis, which provided the warning that when financial innovation is pursued and non-interest business boosted, attention should also be paid to the accompanying regulatory measures.
2.3 Development of non-interest income business in China
From 1949 until the introduction of the reform and opening-up policy, China followed a development style similar to the Soviet Union’s planned economy system, leading to the phenomenon that there were no modern commercial banks within the country. The People’s Bank of China, the only bank in China, had two functional
11
areas. On the one hand, it functioned as a commercial bank that took in deposits, granted loans and provided financing. On the other hand, it functioned as a central bank. With the implementation of the Deng Xiaoping-1979 reform and opening-up policy onwards, the Chinese economic system started to transform, and the financial system gradually started to change. China established four big nationalized commercial banks – Agricultural Bank of China, China Construction Bank, Bank of Communication and Industrial & Commercial Bank of China. In the early years of their establishment, the four banks were specialized banks with clear divisions in areas of operation between them. In other words, Agricultural Bank of China specializes in dealing with agricultural customers, China Construction Bank provides fund to construction project customers, Bank of Communication deals with road traffic project customers and Industrial & Commercial Bank of China mainly deals with industrial and commercial enterprise customers. Interest revenue was the main income source (more than 85% of total) for the commercial banks while non-interest revenue had no room to develop.
In the process of economic reformation, reformation of nationalized banks was put on the agenda, leading the nationalized banks’ operating goals to transform to increasing profits. Banks became more similar to enterprises that assume sole responsibility for their profits or losses. As the reforms advanced, the clear division between commercial banks was broken up and competition developed between them. In 1986, the Bank of Communications was established and business development concerning non-interest revenue enjoyed a major breakthrough. The establishment of the Bank of Communications provided a diversified and competitive mode of transformation. After the establishment of the Bank of Communications, China embraced the establishment of a large number of joint-stock commercial banks. Nationalized commercial banks started to innovate and develop non-interest business, expanding this area of business revenue. The non-interest income business ratio increased more than 10% from merely 3%. However, China lacked effective financial supervision. In 1993, the State Council began to overhaul the financial industry. In
12
1995, China started to carry out separated banking management3.
However, many countries globally moved into the process of financial liberalization in the 1990s and 2000s onwards, with the universal banking system becoming the mainstream of international financial development. Commercial banks which implemented the separated banking management system had little chance to compete when compared to foreign commercial banks whose business concerning non-interest revenue had developed so well. In the meantime, interest rate liberalization was further enhanced. This liberalization enabled banks to price loans flexibly, and enabled market forces to play a greater role in determining the allocation of credit. Thus, it provided an opportunity for banks to raise their margins and leads to an economy more responsive to changes in rates. In the consequence, this liberalization makes the difference between the deposit interest rate and the credit interest rate smaller (only 3% in 2015). Commercial banks in China avoided regulations brought by separated banking management by taking an active role in financial innovation, expanding the business scope of commercial banks.
In 2003, Bank of Communications, Bank of China and some other commercial banks issued subordinated debt, indicating a marker point in the development of non-interest income business. This is because when banks start to issue subordinated debt, the absolute increase in assets will increase the financing costs, leading banks to hold more low yielding liquid assets. Thus the interest income drops, in order to maintain the requirement of assets adequacy ratio, and banks need to turn to non-interest income business because of its high yield. In 2006, according to the commitment with the World Trade Organization, the banking industry in China would open up more comprehensively to outside competition. When China joined the WTO, it expanded the economic and trade ties between 135 membership countries, leading to the development of further trade diversification. The huge influx of foreign capital made the competition in industry fiercer, placing a huge challenge in front of commercial banks.
3
13
There are various channels of non-interest income business, and each different kind of non-interest income business has a different risk and profit. Commercial banks had to develop their non-interest business actively and broaden their income channel. Commercial banks needed to reinforce innovation or establish an investment banking division, money trading division and some other newly-developing operating divisions, or enlarge their business scope through financial holding insurance companies, trust lease companies and other financial organizations. In 2007, Bank of Communications implemented restructuring of the Hubei Province international trust and set up the international trust of Bank of Communications. In 2008, China Construction Bank became the major shareholder of Hefei Xingtai trust. In 2009, Bank of Communications became the major shareholder of Zhongbaokanglian life insurance limited company, and in 2010, China Construction Bank became the major shareholder of Pacific Aetna Life Insurance limited company. In 2012, China Construction Bank incorporated company purchased Jianxin financial leasing incorporated company, which made Jianxin a wholly-owned subsidiary of the bank.
Via a series of acquisitions and reorganizations, the business scope of commercial banks in China has broadened, and non-interest business has flourished. The proportion that interest revenue and non-interest revenue comprise of total revenue of the major big commercial banks is listed in table 2 below. As can be seen, from 2007 to 2014, the proportion of non-interest revenue kept increasing by more than 10% from 13% in 2007. The sharpest jump was in 2009, when the average ratio of non-interest income business reached more than 20% from 14% in 2008. After that, this ratio kept unchanged with slight increases.
14
TABLE 2—Non-Interest Income Ratio in Chinese Banks Million US
dollar
Year ICBC CCB BOC CMBC Bank of
Communications China Citic Bank Average 2007 0.1217 0.1236 0.1692 0.1737 0.1396 0.0639 0.1320 2008 0.1508 0.1647 0.2569 0.1554 0.1466 0.1099 0.1457 2009 0.2056 0.2115 0.2890 0.2161 0.1768 0.1220 0.2035 2010 0.2024 0.2259 0.2725 0.2076 0.1823 0.1459 0.2061 2011 0.2366 0.2364 0.2872 0.2076 0.1924 0.1555 0.2192 2012 0.2218 0.2356 0.2793 0.2209 0.1846 0.1586 0.2168 2013 0.2326 0.2368 0.2833 0.2546 0.2016 0.1474 0.2260 2014 0.2207 0.2360 0.2756 0.3254 0.2337 0.1742 0.2443
Notes：This table illustrates the non-interest income ration in 6 most important Chinese commercial banks from 2007 to 2014. ICBC stands for the Industrial and Commercial Bank of China, CCB stands for the China Construction Bank, BOC stands for the Bank of China, and CMBC stands for China Merchants Bank. Data is from bankscope database.
2.4 Influence of non-interest income in banks’ operating risk in America
Santomero and Chung (1992) assessed the volatility of the asset returns of 62 non-bank financial institutions and 123 bank holding companies. They conclude that when the business scope of commercial banks expanded to non-traditional business, the business risk of these commercial banks would reduce. Furthermore, Hassan (1993) also reached the same conclusion by analyzing the data of American commercial banks. Hassan and Sackley (1994) showed that the expansion of the non-interest income business is an effective means of reducing the risk of American commercial banks and helping to ensure stable operating profits. Gallo and Apilado (1996) analyzed the impact of mutual fund operations, using the financial data of 47 bank holding companies holding mutual funds from 1987 to 1994. They discovered that the diversifications not only increased the profitability of banks but also significantly reduced the systemic risk of banks. They argue that mutual funds had no significant influence on the market risk of banks. This implies that as a kind of non-interest income business, the mutual fund could improve the banks' earnings and
15
significantly reduce the business risk of banks. Rogers and Sinkey (1999) studied the relationship between non-interest income business and bank characteristics, and the relationship between the non-interest income and assets scale of American commercial banks between 1989 and 1993. They examined the profitability, capital adequacy condition, market risk and interest rate risk of banks carefully. Their results indicate that the transformation of commercial banks to non-interest income businesses would increase the profits of banks. In addition, the larger the non-interest income business was, the lower the liquidity and interest rate risk of the banks. The greater the level of capital was, the lower the risk of commercial banks. Using a similar method, Deng and Elyasiani (2005) also drew the analogous conclusion as above. The finding was again that the development of non-interest income business is associated with a considerably lower bank risk.
TABLE 3—Summary of Empirical Studies of The US Banks
Study Risk measure Sample
period
Empirical effect of non-interest income
Hassan, Sackley (1994) Bank loan commitments 1984-1988 Reduce the risk
Gallo, Apilado (1996) Mutual funds operation 1987-1994 Reduce the risk
Rogers, Sinkey (1999) Ratio of adjusted noninterest
income to total bank income
1989-1993 Reduce the risk
Deng, Elyasiani (2005) Z-score 1995-2007 Reduce the risk
De Young, Roland (2001) Revenue volatility 1988-1995 Increase the risk
Stiroh, Rumble (2006) RAROA,RAROE, and Z-score 1997-2002 Increase the risk
Kwan (1998) Variance of return on banking
and security activities
1990-1997 Increase the risk for
large-size banks
Lepetit et al. (2008) SDROA, SDROE, ratio of
loan loss provisions to net loans
1996-2002 Increase the risk for
large-size banks
Notes: This table provides the summary of some important empirical studies.
However, recent studies have generally reached a rather different conclusion. They hold the idea that when the business scope of American commercial banks expands into non-interest income business, levels of higher risk will appear.
De Young and Roland (2001) found that the expansion of fee-based business in America would be associated with increasing the volatility of banks’ profits and the
16
degree of leverage. This increase implies a greater risk for commercial banks. Stiroh (2004) studied the performance of 132 American commercial banks, and concluded that the non-interest income, particularly the trading revenue, is associated with higher risk and lower risk-adjusted profits across commercial banks. One year later, he also showed that increased exposure to non-interest income increases the volatility of equity market returns. As part of their continuous study, Stiroh and Rumble (2006) found that risk-adjusted return is significantly and positively related to diversification of non-interest activities, but that diversification of non-interest activities is associated with a significant overall increase in firm risk.
Manganelli and Marques-Ibáñez (2011) proposed a new opinion, arguing that the impact of non-interest income on the risk of the bank depends on the business model of the bank.
First of all, they divided the banks into two groups, retail banks and the investment banks. They discovered that smaller banks and more retail-oriented banks are significantly more stable if they increase their share of non-interest income, while larger banks and investment banks become significantly riskier. This finding is compatible with the discovery of DeYoung and Torna in 2013. They submitted that the type of non-interest income is more decisive for the risk of the bank than the amount of non-interest income. More specifically, a larger share of income from asset-based non-traditional activities such as investment banking would substantially increase the likelihood of default during a crisis. On the contrary, a larger share of fee-based non-traditional activities such as insurance sales would significantly reduce the probability of failure. Furthermore, the scale of a bank was a crucial factor in determining how non-interest income activities are associated with the levels of bank risk. Higher reliance on non-interest income activities entails a lower level of bank risk for relatively small-sized banks. But higher reliance on non-interest income activities entails a higher level of bank risk for relatively large-sized banks. Kwan (1998), De Young and Roland (2001), Stiroh (2004), and Lepetit et al. (2008) found that non-interest income banking activities increased bank risk for large-sized banks. However, for small-sized banks, the results were less significant.
17
2.5 Influence of non-interest income in banks’ operating risk in China
TABLE 4—Summary of Empirical Studies of The Chinese Banks
Study Risk measure Sample
period
Empirical effect of non-interest income Yang and Zhong
(2013)
Z-score, Non performing loan ratio 1995-2010 No significantly
increase the risk
Zhou and Li (2011) SDROA, Var(LN operating income) 1997-2008 No significant
relationship
Zhou and Li (2011) SDROA 2005-2010 Increase the risk
Notes: This table provides the summary of some important empirical studies.
Yang and Zhong (2013) found that the increasing concentration of businesses that are using or are used by Chinese commercial banks has no significantly impact on the business risk of bank. This higher concentration of businesses could reduce the unexpected impact of any external factors, which in turn would help to reduce the level of risk faced by the individual bank, while the greater diversification of businesses (such as greater non-interest income business) would help in sharing the business risk. So the concentration of businesses would bring about a double effect, and would ultimately lead to no significantly increase on the business risk of bank. Zhou and Li (2011) analyzed the relationship between the diversification of business and the business risk of 14 Chinese commercial banks between 1997 and 2008 based on the portfolio theory, and this showed that there is no significant relationship between them. They use standard deviation of ROA and the variance of the logarithm value of operating income to value the banks’ credit risk. But when the share of non-interest income business increased in Chinese commercial banks, the volatility of non-interest income also increased. In turn, this increased the overall risk of commercial banks.
Lv (2009) used the financial data of 2l Chinese commercial banks in 2008 to study the effect of non-interest income on the risk of Chinese commercial banks. The empirical results show that a diversification business boosts revenues and reduces risk. Zhe and Shao (2012) analyzed the financial statements of 78 Chinese commercial banks during 2005 and 2010. They use standard deviation of ROA to measure the risk
18
and found that, for unlisted Chinese commercial banks, an increasing share of non-interest income lowered the business risk of banks, and for the listed Chinese commercial banks, the increasing share of non-interest income had no significant impact on the business risk of the banks.
An increase in non-interest income will either significantly reduce or increase the risk of American commercial banks (dependent upon size), while the effect of an increase in non-interest income on the risk of Chinese commercial banks is uncertain. The reason for this uncertainty may be that the Chinese commercial banks, especially large banks, have a government background, ensuing that they would not be allowed to fail: thus they are willing to engage in business that encompasses high-yield and high-risk attributes. Further, the internal risk management mechanism is far away from mature and perfect in the Chinese market. Thus, it could be argued, contrary to some of the studies, it will aggravate the risk. With American commercial banks, their long development means that the judgment and management of the risk mechanism is arguably more mature than that in Chinese commercial banks.
3. Theory analysis
3.1 The portfolio theory
According to the references above, the portfolio theory is commonly used in studying the influence of non-interest income business on the risk of commercial banks.
Portfolio theory was first proposed by Markowit Z. in 1952. He utilized the expectation rate of return and its variance to measure the uncertainty and fluctuation of the excepted revenues of an investment. After that he pointed out that if investors want to reduce risk, they should hold a variety of securities at the same time. Through diversification, the risk of investment will be reduced. According to the portfolio theory of Markowitz, every asset is independent of each other, so when clients add a new asset, the risk of the original asset does not change. This means that adding new
19
assets won't produce additional risks on the original asset, and on the contrary, may help to reduce the overall risk of commercial banks. In fact, Stiroh used the portfolio theory in 2006 when he analysed commercial banks. He took the traditional interest income and non-interest income as an asset combination, as the traditional interest income business and non-interest income business are carried out at the same time in a commercial bank. Given the development of commercial banks, external factors such as economic policy, the financial crisis, the volatility of interest rates etc. may cause unstable effects for commercial banks, while at the same time the portfolio of traditional interest income and non-interest income may demonstrate certain alternative and exclusive attributes. Through developing non-traditional business activities, commercial banks could realize the diversification of their income structure. Therefore, the non-interest income business will assist in diversifying the risks for commercial banks, as long as the relationship between the traditional interest income business and the non-interest business does not have a perfect positive correlation relationship.
3.2 Non-interest income business model
Banking businesses are not completely independent of each other, so the fluctuation of the non-interest income will also have an impact on the interest income.
Firstly, according to Stiroh (2004), the non-interest income business will bring greater management risk to commercial banks. Though the non-interest income business will provide banks with diversification of revenues, bank managers (notably at a senior level) are unlikely to be familiar with the risk characteristics of all the wide variety of businesses. This can lead to increasing management risks for the bank. For example, for the securities and its derivatives’ trading business, the design and management of derivatives require a lot of specialized financial engineering knowledge. In addition, the pricing mechanism of such business is extremely distinct from that of the traditional interest business, and the price fluctuation of such business is more intense. In other words, the pricing mechanism of interest income business is
20
mainly influenced by interest rate spreads, however, the pricing mechanism of non-interest income business is influenced by both internal and external variables. Moreover, the resource allocation for the different businesses also brings management risk. With limited resources, the result may be the allocation of insufficient resources to some departments, or the inappropriate proportion of resource allocation, thus greatly increasing the management risk of banks.
Secondly, the development of the diversification of non-interest income business will cause the risk of rising fixed costs for a commercial bank. Generally speaking, the traditional interest income business will not generate additional costs once the relationship between savings and loan is set up. But the operation of a non-interest income business needs to pay more fixed costs (e.g. various management costs and labour costs), leading to a substantial increase of fixed costs. For example, the principal non-interest income of Chinese commercial banks at present is the commission fee, which includes credit card business, commitment business, payment and settlement business, financial management business, financial advisory business and other businesses. All of these businesses would have prompted the growth of fixed costs. Once fixed costs increase, it can be hard to change them due to their specific characteristics. Only when the bank's business scale reaches a certain level, can it dilute fixed costs and realise economies of scale. Taking the credit card business as an example, in order to promote the success rate of those applying for cards (and in theory the ultimate income generated), the bank will add more salesmen and increase the corresponding publicity. If sales are unable to achieve the expected goals, because of the costs of salary and publicity expenses increase in proportion to income, the bank’s performance will decline and the risk increases.
Thirdly, the expansion of non-interest income business will increase liquidity risk, credit risk, and market risk. At the same time, it will also affect the stability of interest income, the decrease of liquidity will lead to an increase of risk in the traditional loan business. For the credit risk, it is produced in the credit card business, commitment business, and trading business. Credit default is viewed as not conducive to the operation and management of the banks. The commercial banks will undertake
21
more risks because customers are more likely to default at these businesses. For the transactional risk, innovative financial derivatives are easily affected by interest rate changes and exchange rate fluctuations, leading to a violent volatility in price, creating a market risk. For some banks, in order to seek more profits, there is a tendency to carry out entrusted loans, trust loans, and other off-balance-sheet financing. Although this way can help banks to generate more non-interest income, the quality of the entrusted loans and trust loans is not guaranteed, which brings a new risk to the bank.
Expanding the non-interest income business may lead to the increase in management risk and credit risk, although in many real cases the relationship between the non-interest income business and the traditional interest income business demonstrates a low correlation relationship, reducing the dispersion effect of bank risk. Thus based on the positive and negative effect of the non-interest business on the risk of commercial banks, the effect of diversification business on the risk of the bank is uncertain.
4. Empirical method and Data
4.1 Empirical method
In this section, the article analyzes the impact of non-interest income on commercial banks' operating risk. As is mentioned earlier, this article adopts and expands the definition of Z-score according to Lown (2000) and Stiroh (2004) and defines Z-score as the sum of ROA and E/A divided by the standard deviation of ROA4, and built a panel regression model as follow：
Z, , = c + β NIIR, , + β LN(Asset), , + β (E A)⁄ , , + β LN(LLR), , + β LN(GDP), , + β US + β US × NIIR , , + δ + γ + ε, , (1)
4
22
Formula (1) is used to inspect the influences that commercial banks' non-interest revenue has on operating risks.
Variable Z, , is the Z-score of the jth commercial bank in country i during
period of t. The smaller the Z-score is, the bigger the risk banks are confronted with. Variable NIIR, , is the proportion that non-interest revenue takes up in the jth
commercial bank’s operating revenue in i country during the period of t and is used to measure the size of non-interest revenue. The key coefficient of interest is β , which measures the link of non-interest income and risk. This coefficient provides comparison clues between Chinese commercial banks and US commercial banks. Positive β indicate that a 1% increase in non-interest income will lead to β % increase in Z-score, while negative α1 means that 1% increase in non-interest income will lead to β % decrease in Z-score. If the coefficient of NIIR is positive and significant, then the diversification of business will not raise the operating risk.
The first control variable is LN(Asset), , .This is the logarithm of the jth commercial bank’s asset in country i during period of t, representing the size of a bank. It is used to analyze the influence the size of a bank has on a bank’s operating profit and operating risk. According to the study of Lepetit etal.(2008), bigger commercial banks have scale effect and scope economy, thus facing with lower risks.
A further control variable, (E A)⁄ , , , is the proportion equity capital takes up in the total assets of the jth bank in country i during period of t. It can roughly reflect conditions of capital adequacy, and also reflect the operating and management conditions. De Yong and Roland find that for a commercial bank with high Capital Adequacy Ratio, the adequate capital will act as a powerful buffer to external shocks, thus lowering the overall risks the bank is faced with.
Further, LN(LLR), , is the logarithm of the loan loss reserve of the jth
commercial bank in country i during the period of t. The loan loss reserve can lower the risk and serve as a buffer. It mirrors the degree of risk of a bank’s loans. The bigger the loan loss reserve, the bigger the risk the bank judges that it is confronted
23
with. The loan loss reserve represents external credit risk (Stiroh, 2004).
LN(GDP), , is the logarithm of Gross Domestic Production. The context of a
country’s macro-economy and economic development directly influence the operating conditions for commercial banks. When the economy is experiencing an upward period, the need for funds and financial services will increase, thus increasing profits and promoting the operating performance of commercial banks. GDP broadly mirrors the context of the macro-economy that commercial banks have around them.
In this formula, a dummy variable (“US”) is added. When the sample is an US commercial bank, then the value of the dummy equals 1. Otherwise, the dummy equals 0. In this equation, the interest coefficient is β , which indicates the influence of country type on the level of risk. Positive β indicates that a developed country has lower risks when it undertakes more non-interest businesses. Negative β means that a developing country has lower risks when it undertakes more non-interest businesses.
Besides, this article also includes a firm fixed effects δ , as well as year fixed effects γ , designed to capture aggregate specific shocks, for example, the fluctuations in the global economy.
4.2 Data
4.2.1 Data sources and sample selections
In order to guarantee the reliability of the data, this article comprehensively compared the Bankscope5 database, WIND database, and China Financial Yearbook etc, and ultimately uses the Bankscope database to gather data.
Firstly, the selection of sample size should be as large as possible. Considering the analysis is at the national level, the analysis should cover all kinds of banks,
5
The Bankscope database is the world's leading financial professional empirical database provided by BvD and FitchRatings and provides up to 16 years of data of more than 12,800 of the world's major banks and world's major financial institutions and organizations. It is the most comprehensive, global database of banks' financial statements, ratings, and intelligence
24
including large commercial banks, medium commercial banks, and small commercial banks. However, in the Bankscope database, the total number of Chinese commercial banks is considerably smaller (only around 250 banks) to that in the US, (more than 4000 banks). The differences in the number of banks between the countries may lead to unconvincing results; thus, this article keeps the ratio of large banks consistent. The large bank6 ratio in China is 24%, so this paper uses the same ratio in the US, leaving 1600 American commercial banks and 260 Chinese commercial banks.
Secondly, in order to make the regression result more convincing, the time span for the period studied should be as long as possible. However, the fact is that there is only limited data on China's commercial banks before 2007, and therefore, it is impossible to collect complete and comparable data before 2007. Considering the availability and integrity of the data, therefore, this article studies annual data between 2007 and 2014. As for the processing of raw data, only banks that appear for at least three consecutive years in the sample are included, and banks with missing values in more than three variables are also excluded, resulting in a final sample of 1500 US commercial banks and 252 Chinese commercial banks.
4.2.2 Explanatory variables
This paper uses the CAMEL credit rating indicator system, which mirrors its emphasis on assets' safety. Capital Adequacy, Asset Quality, Management, Earnings and Liquidity are the five evaluation indexes of the CAMEL rating system.
Measures of risk-taking
There are some classical studying in the related field, for example, Kwan (1998) uses variance of return on banking and security activities to measure the banks’ operating risk, Stiroh and Rumble （2006） test the RAROA, RAROE and Z-score to value the risk. The measure of risk in this paper is the Z-score. American professor Edward Altman (1968) observed the experiences of bankruptcy and non-bankruptcy
6
25
enterprises and applied mathematical statistics approaches to set up the Z-score model. After this, this model has been widely used by researchers such as Lown, Stiroh, Laeven and Levine, and they argue that the Z-score model can measure risks effectively. The Z-score measures banks' bankruptcy risk (Stiroh (2004), Matthias & KKÖhler (2013)). The definition of Z-score is as follow:
Z =( ROA + E A)⁄ σ
The higher the Z-score is, the smaller the possibility of the bank going bankrupt: hence the bank confronts smaller risks and is thus more stable.
The level of non-interest income business
The independent variable is the non-interest income ratio. It is the ratio of the non-interest income to net operating income and reflects the size of non-interest income in relation to the overall income of the bank. This article takes operating revenue minus net interest revenue as the non-interest revenue of commercial banks.
Size
This paper includes the size of the banks in order to control for the effect that a bank's size may have on its risk-taking, as the size of commercial banks has a vital influence on operating risks. The larger the bank, the more able is the bank to raise external funds. Generally, the enlargement of commercial banks' sizes can bring them scale economy and scope economy, leading the decrease of risks. That is to say, larger banks are too big to fail and may be able to hold lower levels of capital and chase higher levels of risk. Total assets are used as a measure of the size of a bank.
Leverage ratio
The leverage ratio can also reflect the operating and managing conditions inside commercial banks. In this article, leverage ratio equals book common stock equity divided by book value of total assets.
26 Loan loss reserves
Commercial banks are faced with the risk that the money lent out cannot be returned anytime in their daily operations, and the level of loan loss reserve can measure this credit risk. Loan loss reserve is the reserve that can be drawn against to compensate for the certain possibility of loss when there's demonstrable evidence showing loans will be or can suffer a decrease in value.
GDP
GDP growth may affect both risk-taking as well as the level of capital held by banks. Risk-taking tends to be low during economic booms and tends to be high during economic contractions (Borio et al. 2012). The economic cycle may also change the amount of capital a bank holds (Angora et al. 2009, Behr et al. 2010). The macroeconomic environment, which can be measured by GDP, is an important external influencing factor on commercial banks' operating conditions.
4.2.3 Descriptive statistics
Table 5 summarizes the descriptive statistics of all the variables used in the regression analysis. It shows averages, medians, standard deviations, 25th percentiles and 75th percentiles of variables for Chinese banks and US banks between 2007-2014.
TABLE 5 — SUMMARY STATISTICS, 2007—2014
Mean Median SD 25th percentile 75th percentile Obs.
China The US China The
US
China The US China The
US China The US China The US Z-score 4.297 4.461 2.069 3.903 6.272 2.651 1.503 3.433 3.584 4.593 1172 8200 NIIR 0.256 0.330 0.155 0.326 0.297 5.794 0.065 0.146 0.334 0.426 1172 9430 Asset 1175 1381 1359 1411 1132 815.724 807 815 1980 1231 1184 9835 E/A 16.870 11.575 7.880 9.930 23.379 9.275 6.119 8.285 12.151 12.224 1184 9835 LLR 375.242 206.281 312.12 603.7 1231 1733 78.58 15.88. 927.21 239.52 986 8206 GDP 5.784 15.541 5.648 15.239 2.735 1.03 3.490 14.657 7.886 16.296 2016 11999
Notes：This table presents summary statistics for the main variables used in the study from 2007 to 2014. The table can be divided into two parts and sorted by Chinese banks and American banks. The definition of these variables are given as follows: Z-score is defined as average ROA plus E/A，and
27
then divided by standard deviation of ROA. NIIR is defined as the non-interest income ratio. Asset is the value of bank asset. E/A is the ratio of equity and assets. LLR is the value of loan loss reserves. And GDP is the value of GDP.
The mean value and the median value of Z-score of Chinese commercial banks is smaller than those of American banks, indicating that the risk faced by Chinese commercial banks is a bit higher than for US banks.
As can be seen from the table, American banks have even higher non-interest income ratio than their Chinese counterparts: the median value of non-interest income ratio in American commercial banks is two times that in China (32.6% versus 15.5%). This matter is attributable to the inconsistency of non-interest income business development in both countries, as described earlier and notably linked to the 2000s-era boom in asset securitization business, securities underwriting and distribution business, venture capital business and private banking business in the US to reach historically high rations of over 40% before falling back during to the Global Financial Crisis.
Even though the median value and the average value of non-interest income ratio are lower in China than in America, the overall business appears more stable. In the US, the standard deviation of NIIR is 5.794, while in China it is 0.297. This is the result of the relatively simple structure of China's commercial banks' non-interest business. It mainly includes fees, exchange gains & losses, and investment gains & losses. Importantly, fees and commission income have the lowest volatility among all non-interest businesses. Compared to the sources of income for Chinese banks, US commercial banks' non-interest income business is diversified, including four main categories: trust business, deposit account service business, trading account profits & losses, and additional non-interest income business. The volatility of these businesses is very high. In short, the ability of US banks to operate in a wider range of business space has led to an increase in volatility.
The value of GDP of the two countries is quite different. The average value and median value in the United States are almost triple that in China. The Chinese financial industry obeys separate operation regulations and there are more stringent
28
control limits among banking, securities and insurance industries, while the US financial sector is more liberalized.
5 Results
Table 6 below presents the empirical results that were gained following the methodologies explained above and reports estimates of various specifications of equation (1) in China and the US respectively between 2007 and 2014.
TABLE 6—OPERATING RISK AND NON-INTEREST INCOME RATIO Simple OLS Simple OLS with Interaction OLS with FE OLS with FE OLS with FE OLS with FE OLS with FE (1) (2) (3) (4) (5) (6) (7) NIIR 0.006 4.219*** 0.854 4.110*** 3.311*** 1.817*** 2.182** (0.00) (0.54) (1.54) (0.54) (0.32) (0.30) (0.98) US*NIIR -4.214*** -0.849 -4.104*** 3.314*** 1.822* 2.186*** (0.54) (1.54) (0.54) (0.32) (1.26) (0.29) LN(Assets) -0.322*** 0.422*** 1.542*** 1.478*** (0.05) (0.02) (0.08) (0.01) E/A 0.397*** 0.423*** 0.421*** (0.00) (0.02) (0.01) LN(LLR) -0.343*** -0.339*** (0.02) (0.02) LN(GDP) 0.286** (0.42) US 2.107*** 1.843*** 3.266*** 2.726* 1.161* (0.35) (0.35) (0.16) (1.20) (1.12) Constant 5.66*** 3.697*** 5.557*** 8.857*** -8.484*** -7.913*** -10.139*** (0.12) (0.33) (0.04) (0.85) (0.44) (0.34) (0.37) Year fixed effects
No No Yes Yes Yes Yes Yes
Firm fixed effects
No No Yes Yes Yes Yes Yes
Adjusted R
0.102 0.375 0.002 0.201 0.480 0.654 0.742
Obs. 1600 1600 1600 1600 1600 1600 1493 1493
Notes: This table tests the equation (1) and reports the link between the commercial banks’ operating
29
the log value of asset. E/A is the leverage ratio. LN(LLR) is the log of loan loss reserves. LN(GDP) is the log value of GDP. Column 1 reports the simplest OLS regression results without the interaction of NIIR and country dummy, and also without fixed effects. Column 2 adds the interaction variable and country dummy. But it still excludes the fixed effects. After that, all regressions include both time fixed effects and firm fixed effects. Column 3 runs the same regression in Column 1, but includes fixed effect. Column 4 adds the first control variable—log value of assets to run the regression. Column 5-7 add leverage ratio, log value of loan loss reserves and log value of GDP respectively.
***Significant at the 1 percent level **Significant at the 5 percent level *Significant at the 10 percent level.
As can be seen from table 6, column (1) and column (2) don’t include the fixed effect, and adjusted R squares are below 0.4, which means that the fact can’t be explained well by this equation because of possible omitted variables in this equation. So including firm fixed effect, time fixed effect and control variables could solve this problem. When we add fixed effects, the value of adjusted R square drops to 0.002 in column 3. However, this value keeps increasing and goes up to 0.742 with the adding of different control variables.
First of all, the level of non-interest income ratio has a positive effect on the value of Z-score in all seven regressions. Nearly all coefficients of non-interest income ratio are significant at the 1 percent confidence level, except column 1. It implies that changes in non-interest income ratio will significant influence the banks’ operating risk. For example, the coefficient in column 7 is 2.189, which means that 1% increase in the proportion of non-interest income will lead to double increase in the value of Z-score. All coefficients are positive, implying that when the level of the non-interest income ratio is increased, it will also cause the value of Z-score to rise. This can be viewed as evidence that increasing banking diversification will lead to a decrease in the level of operating risks. In addition to this, it is also not hard to find that when we add interaction of US and NIIR into the original equation, the coefficient of NIIR changes a lot, for instance, it jumps from 0.006 in column 1 to 4.219 in column 2, and goes up from 0.854 in column 3 to 4.110 in column 4.
The coefficients of interaction are negative in column 2, column 3 and column 4, but are positive from column 5 to column 7. This may because of omitted variables in
30
former regressions. In column 5 and column 7, the coefficient of interaction is significant at 1% confidence level, and significant at 10% confidence level in column 6. This illustrates that whether it is a developed country will influence the operating risk when a bank starts to expand the non-interest income business. If it is a US bank, then engaging in more non-interest income business will decrease more the bank’s operating risks. In other words, if it is in the US, a 1% increase in non-interest business would decrease operating risks by more than 2% than in China. A possible explanation for this finding may be that non-interest businesses have been developed for a very long time in the US, and the US financial market has already developed complete regulatory system, thus the increase of the ratio of these kinds of businesses has more influence in the US banks than Chinese banks.
As for the coefficient of control variables, they are nearly all significant at 1% confidence level. The coefficient of LN(Asset) is negative in the forth regression, but it turns to positive in next regressions. For example, in column 6, Z-score increases 1.5 times together with LN(Asset). That is to say, the larger the bank is, the less risk possibility. This result seems reasonable and is consistent with common sense. In the whole world, larger banks tend to be less risky than smaller banks (Behr et al. 2010). An explanation for this might be that it is easier for larger banks to diversify their loan portfolios, thus enabling them to diversify their risk exposure as well.
The coefficients of E/A from column 5 to column 7 are positive and significant at 1 percent confidence level, which possibly reflects the fact that the leverage has positive influences in operating risk: in other words, that the increase in leverage ratio will lead to an increase in risk. Leverage ratio has an outstanding influence on commercial banks' operating risk (both coefficients are significant at 1 percent confident level). Including E/A ratio changes the coefficient of the interaction from negative to positive, so it has differences between both countries. Commercial banks in China have reached the supervisory standards of the Basel Accord, due to government-leading exfoliations of non-performing assets from banks' assets and asset injection to banks. The lift of CAR in Chinese commercial banks is not the result of the promotion of professional ability or operating level, but the result of
31
government assistance. The improvement of capital adequacy will make Chinese commercial banks more stable when facing external risk shocks. In the US, by comparison, the prosperity of the financial industry promotes financial liberalization, more effective market regulation, self-developed and well-functioned capital, and more effective internal management for banks in America.
Looking at the last two regressions, the coefficient of LN(LLR) is negative. The reason may be that too much loan loss reservations will harm the investment channels. And the narrowed investment channels will drop the bank’s profits, thus leading banks to choose higher profit and higher risk businesses. In column 7, the context of the macro economy is considered. Its coefficient is 0.286 and significant at 5 confidence level. In addition, all coefficients in column 7 are significant and more importantly, it has the highest adjusted R square (0.742).
To sum up, in the above-mentioned regression of all sample banks, the main conclusions drawn are as follows: (1) no matter whether in China or the US, the diversification of banks' revenue is beneficial to raising the risk-adjusted return and reducing banks' bankruptcy risk; (2) all control variables have a significant influence on banks' operating risk.
6. Robustness check
This section covers a number of additional specification and further robustness checks on the previous results. Column 1 is the original regression model in the column 7 in Table 6. Column 2 uses RAROA (ROA divided by standard deviation of ROA) as a dependent variable instead of Z-score. Column 3 adds ROA as a new control variable to measure the possible influence of banks’ performance. Column 4 includes NIIR as a new dependent variable to test whether it is a non-linear relationship between non-interest income ratio and banks’ operating risks. Column 5 and 6 report regression results for two sub-periods: 2011-2014 in column 5 and 2007-2010 in column 6. Column 7 restricts the sample only to top 100 commercial
32
banks in 2015. Below are the regression results of the six robustness tests.
This paper uses RAROA as a dependent variable instead of Z-score. This alternative measure of risk is used in order to make sure of the reliability of the link between non-interest income and operating risks. Based on Lown (2000) and Stiroh (2004), the Sharpe ratio is another way to examine the performance of an investment by adjusting for its risk and can be used to assess the operating risk of commercial banks instead of using Z-score. For this reason, RAROA is used, defined as ROA to the standard deviation of ROA, as a new dependent variable. Thus the higher the ratio is, the lower risk a bank has taken. The definition of Return on Assets (RAROA) and Risk Adjusted Return on Equity (RAROE) are defined as follows:
RAROA = , RAROE =
Diversification of revenue in large commercial banks has a greater influence on banks' risks than medium or small banks (Elyasiani et al, 2012). As for the choice of larger banks, this paper refers to the ranks in The Banker7. Based on the top 100 commercial banks according to total assets, 11 US banks and 15 Chinese banks have been selected. In China, these are Industrial & Commercial Bank of China, China Construction Bank, Agricultural Bank of China, Bank of China Limited, Postal Saving Bank of China, Bank of Communications, China CITIC Bank Corporation Limited, China Merchants Bank, China CITIC Bank, Industrial Bank Co Ltd, CITIC Group Corporation, Shanghai Pudong Development Bank, China Minsheng Banking Corporation, China Everbright Bank and Huaxia Bank. In the US, these include: Freddie Mac, JPMorgan Chase, Bank of America, Wells Fargo Bank, Citibank NA, Federal Home Loan Banks, Goldman Sachs & Co, US Bank National Association, Citigroup Global Markets Holdings Inc, PNC Bank, Credit Suisse, Bank of New York Mellon and Morgan Stanley.
7
The Banker was founded in 1926. It combines in-depth regional and country coverage with reports on capital markets, structured finance, risk management, working capital management and securities services, etc. (www.thebankerdatabase.com). The earnings performance, capital adequacy and stability rate and some other financial conditions of over 4000 commercial banks are all displayed in a database. The Banker ranks commercial banks all over the world annually according to different indicators such as brand, total assets, etc.
