Liping Xu,Yu Xin

Sun Yat-sen Business School,Sun Yat-sen University,China

Thornyroses:Themotivationsandeconomic consequences of holding equity stakes in f i nancial institutions for China’s listed nonf i nancial f i rms

Liping Xu,Yu Xin*

Sun Yat-sen Business School,Sun Yat-sen University,China

ARTICLE INFO

Article history:

Mixing of banking and commerce

Holding equity stakes in fi nancial institutions

Motivations

Economic consequences

The reforms of China’s fi nancial system have signi fi cantly changed the country’s fi nancial sector.One noteworthy phenomenon is that many non fi nancial fi rms have obtained equity stakes in fi nancial institutions.This study investigates the motivations behind and economic consequences of this recent proliferation of investments in fi nancial institutions by non fi nancial listed fi rms.We fi nd that the motivations for holding equity stakes in fi nancial institutions include alleviating the pressure of industry competition,reducing transaction costs,and diversi fi cation to reduce risk.These investments,however,have double-edged e ff ects on the performance of the investing fi rms.While their investment income increases,their operating income and overall return on assets decrease,as the investment income cannot compensate for the decrease in other operating income.The investing fi rms’cost of debt also increases,their cash-holding decreases,and stock price performance does not improve after investing in fi nancial institutions.These e ff ects contrast with the enthusiasm non fi nancial listed fi rms have for investing in fi nancial institutions.The empirical fi ndings in this study can inform fi nancial industry regulators and decision-makers in listed fi rms.We advise non fi nancial fi rms to be cautious when considering investing in fi nancial institutions.

?2016 Sun Yat-sen University.Production and hosting by Elsevier B.V.This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1.Introduction

The debate over mixing/separating banking and commerce has carried on for centuries.The activities of banks have been restricted since they f i rst emerged in the Mediterranean city states,and government limitations on the trade investment of banks f i rst appeared in Venice in 1374(Salley,1976),before spreading throughout continental Europe.The powers of England’s banks were restricted in the late 17th century, and the practice was then exported to colonial America.The market collapse of 1929 in the U.S.and the subsequent great depression reinforced restrictive powers of banks with the enactment of the Glass-Steagall Act in 1933(Halpert,1988).Today,f i nancial systems worldwide are generally regulated(Barth et al.,2001).The fear of bank failure and monopoly were previously the main reasons to restrict bank powers,but today the most common concerns include conf l icts of interest,excessive bank powers,and threats to the safety net(Krainer, 2000).There are,however,obvious benef i ts from the mixing of banking and commerce(Wall et al.,2008), such as economies of scale and scope,the fostering of internal capital markets,and diversif i cation.The boundary between banking and commerce has never been clear-cut.Merchant banking was very common among banks in the Italian States of the Middle Ages(Craig,2001),and universal banks in Germany and Japan have long been encouraged.In the U.S.today,there are various ways banking and commerce mix(Haubrich and Santos,2003);commercial f i rms can own banks,for example.In fact,commercial f i rms throughout the world are commonly found to possess equity stakes in banks.

Traditionally,the activities of banks are restricted from two dimensions;f i rst,from carrying out fee-based activities such as securities,insurance,and real estate,and second,from owning commercial f i rms,and/or from restricting commercial fi rms from owning banks.Globally,the divisions between bank and non-bank fi nance have been dismantled since the late 20th century,and increasingly more countries allow commercial fi rms to own banks.Bank ownership of commercial fi rms is permitted in Germany and other countries, but with certain limitations.The e ff ect of bank ownership of fi rms,though restricted throughout the world, has been examined in the literature.But commercial fi rms’ownership in banks,though permitted in many countries,has been largely ignored.In this study,we attempt to fi ll this gap by investigating the motivations and economic consequences of commercial fi rms’equity stakes in banks.We also expand the concept of combining banking and commerce to include the equity stakes in various types of fi nancial fi rms held by commercial businesses.We de fi ne this as the integration of fi nance and commerce,where fi nance represents the broad fi nancial sector including banks1Political economists view the integration of f i nance and commerce as creating f i nance capital.The concept of f i nance capital was f i rst proposed by Hilferding(1910),and then taken up by Lenin in his wartime analysis of the imperialist relations of the great world powers. Hilferding(1910)summarized the development of capitalism and concluded that‘‘the most characteristic features of‘modern’capitalism are those processes of concentration which,on the one hand,‘eliminate free competition’through the formation of cartels and trusts,and on the other,bring bank and industrial capital into an ever more intimate relationship.Through this relationship capital assumes the form of f i nance capital,its supreme and most abstract expression.”,securities,insurance,various funds,trusts,etc.,and commerce represents the non fi nancial sector as a whole.

Durin2In this study,commercial f i rms refer to all nonf i nancial f i rms.Financial institutions include banks,and f i rms dealing in securities, venture capital and private equity,insurance,f i nance,loans,trusts,guarantees,futures,asset management,investment funds,leasing,and pawnshops,etc.g China’s fi nancial system reforms,many commercial fi rms obtained equity stakes in fi nancial institutions.According to the Chinese Entrepreneurs Survey System(2011),20.4%of fi rms surveyed had equity investments in fi nancial institutions,and 27.8%had their own fi nance fi rms.The 2009 report of the International Finance Research Institute of the Bank of China(2010)revealed that non fi nancial business groups actually controlled 24 out of 52 trust fi rms,19 out of the top 50 investment banks,12 out of 25 property insurance fi rms,and 20 out of 39 life insurance fi rms.These represent 46%,38%,48%,and 51%,respectively.Even fi nancial institutions controlled by the government or fi nancial groups were found to be partially held by non fi nancial fi rms.An increasing number of commercial fi rms are interested in investing in the fi nancial sector.For example,in 2010 China Mobile obtained 20%of the equity in Shanghai Pudong Development Bank for RMB39.8 billion.In 2013,Vanke invested RMB2.7 billion in Huishang Bank in exchange for 8.28% ownership,and the Evergrande Group obtained 5%of Huaxia Bank in 2014.Alibaba and Tencent,the two Chinese Internet giants,are currently expanding their fi nancial empire though Alipay and WeChat Wallet.

Theoretically,by investing in the fi nancial sector,commercial fi rms can obtain high returns,reduce transaction costs,and strategically diversify their operations.A large-scale capital fl ow from comm3Wenzhou is a microcosm of economy instability arising from capital f l ow from commercial to f i nancial sectors.As the birthplace of private economy in China,Wenzhou has millions of small-and medium-sized f i rms.Over the years,these f i rms have lost their competitive advantages.They invest their hot money in real estate,mining,the tertiary industry,and informal loans,resulting in a diminished manufacturing sector.This triggered the large-scale bankruptcies of 2011 in Wenzhou.ercial to fi nancial sectors can,however,give rise to economic instability and resource allocation problems.There are also concerns of contagion e ff ects.In this study,we attempt to discover the reasons behind these capital fl ows and explore their e ff ects on the performance of commercial fi rms.Using hand-collected data on the equity stakes in fi nancial institutions held by Chinese non fi nancial listed fi rms between 1999 and 2012,we fi nd that the more intensive the industry competition,the more likely that a commercial fi rm will invest in fi nancial institutions. This e ff ect is more obvious in non-state controlled listed fi rms,when investee fi rms are non-bank fi nancial institutions,and when investee fi rms are subject to less regulation.Reducing transaction costs is one motivation for commercial fi rms to hold equity stakes in banks.Consistent with the strategy of diversi fi cation,larger fi rms with higher pro fi tability,more debt,and with sufficient cash are more likely to invest in fi nancial institutions.Finally,the ownership type and structure can a ff ect a commercial fi rms’decision on investing in fi nancial institutions.

The economic consequences of investing in fi nancial institutions by non fi nancial listed fi rms are not particularly good.We fi nd that as non fi nancial listed fi rms invest more in fi nancial institutions,their investment income increases,but other operating income decreases and the overall return on assets decreases.Increases in investment income cannot compensate for decreases in other operating income.Investing in fi nancial institutions also increases the investing fi rms’costs of debt,decreases their cash-holdings,and their stock price performance does not improve.Investing in fi nancial institutions does not therefore improve fi rm performance;in fact it deteriorates.Investing in fi nancial institutions is like a thorny rose;it looks beautiful,but it can be dangerous.

The remainder of this paper is organized as follows:Section 2 introduces the evolution of the global trend of integrating of banking and commerce,particularly in China.Section 3 reviews the literature and presents the theoretical analysis.Section 4 describes the research design,Section 5 reports the empirical results,and Section 6 concludes the paper.

2.Institutional background

2.1.The evolution of mixing banking and commerce worldwide

Modern banking developed in the Mediterranean city states in the 13th and 14th centuries from the activities of‘‘money changers”and merchants.4The early upheavals of mixing/separation banking and commerce reviewed here are partly taken from Shull(1999).To prevent banks from undertaking risky activities and monopolizing particular commodities,their activities were at times restricted.For example,in 1374 the Venetian Senate prohibited bankers from dealing in copper,tin,iron,lead,saf f ron,and honey.Regulation alone was,however,not enough to prevent the economic and f i nancial disruptions associated with banking failures, currency problems,and bubbles,so public banks were set up by governments.Established in 1694,the Bank of England was a chartered bank.The activities of the public banks of the European continent and the chartered banks in the U.K.were restricted and various regulations were imposed on them,to address monopoly and public interest concerns.Early banks in the United States were modeled on the Bank of England,and were prevented from engaging in mercantile enterprises.However,by the late 19th century,the bond departments of large national banks in New York and Chicago had begun to undertake investment banking activities,and eventually through securities affiliates they became involved in many types of f i nancial activities.

In October 1929,the New York Stock Exchange crashed,triggering the 1929-1933 global economic crisis. The securities activities of commercial banks were blamed for fueling the crisis.In 1933,the Glass-Steagall Act revoked the powers of commercial banks,preventing them from engaging in securities activities.However, commercial banks could still expand into new activities through bank holding companies until 1956,whenthe Bank Holding Act was passed.Even under this act,commercial banks were able to extend their activities by exploiting various loopholes.In the 1980s,legal and market changes substantially af f ected banks’expansion activities.Sears,a large conglomerate,was able own a retail enterprise,an insurance company,a securities f i rm,a real-estate development company,and a savings and loans company.Securities f i rms and insurance companies could also acquire banks that refrained either from commercial lending or taking demand deposits. The Glass-Steagall Act restrictions eased in the 1980s,and most of the barriers separating commercial banks from nonbank f i nancial services were lifted by the Gramm-Leach-Bliley Act of 1999.

Elsewhere in the world,relationships between banks and commerce are often much closer.Interestingly, there are few explicit legal restrictions on the types of business a bank can undertake in the United Kingdom. For many years they have been able to hold equities of commercial f i rms and commercial f i rms can hold bank equities,but only since the‘‘Big Bang”of 1987,commercial banks have aggressively moved into securities trading and insurance.In the late 19th century,universal banks emerged on the European continent as part of government ef f orts to rapidly industrialize.Universal banks provide short-term bank credit and intermediate and long-term capital,through underwriting and investing in equities.Under the universal banking system, banks and commercial companies maintain close and long-term relationships through ownership,credits, boards of directors,etc.In Japan,after World War II,the Glass-Steagall restrictions were imposed under the Securities Transaction Act of 1948,but banks and companies became associated in keiretsus(groups of enterprises)and since the 1970s banking activities have expanded.The Financial System Reform Law of 1992 permitted Japanese banks to conduct securities business through subsidiaries in which they had a 50% or greater share.

Industrial-f i nancial groups persist and often prosper in many developing countries.During the Soviet regime,for example,the Russian banking system consisted of a single,monolithic bank owned by the state. The f i nancial reforms of 1987 created three regional banks from the former state bank.The reforms of the early 1990s enabled a large number of private banks,over 2000 by 1993,to be established in Russia.The freedom to set up and own banks led to widespread enterprise shareholding.According to a survey in 1994 (Belyanova and Rozinsky,1995),ownership of new banks was dominated by new private companies,while former state banks were in the main held by state institutions,state enterprises,private enterprises,and individuals,each with around 25%of the shares.Bank ownership of enterprises is,however,much less widespread. The banking industry of Taiwan was deregulated in the early 1990s,and before this liberalization most banks were state-owned and banking entry was highly regulated.The Ministry of Finance revised the Banking Law in 1991 to allow for the setting up of private commercial banks.Deregulation provided a means for the entry of private banks into the market,and the number of banks increased from 24 in 1990 to 51 in 2003(Ma,2007).

To summarize,worldwide deregulation has greatly broadened the activities of banks,but there are still substantial variation in the ability of banks to engage in securities,insurance,and real estate activities and in the combining of banking and commerce in dif f erent countries(Barth et al.,2001).More research into these variations is therefore required.

2.2.The integration of f i nance and commerce in China

2.2.1.Investing in commercial banks

The f i nancial system of China was highly centralized under the Ministry of Finance after the establishment of the People’s Republic of China in 1949.The People’s Bank of China was the mono-bank and engaged in savings,credit,and money supply.Market entry and f i nancial innovation was suppressed.Decentralization gradually followed with China’s reform and open policy,and in 1979 the People’s Bank of China separated from the Ministry of Finance and became the central bank.Subsequently,the Bank of China,the China Construction Bank,the Agricultural Bank of China,and the Industrial and Commercial Bank of China were established and began functioning as commercial banks.The joint stock commercial banks emerged in the 1980s.Of these,the Bank of Communications was the f i rst national joint stock commercial bank,with 72%of its stocks held by the state and local government,and 28%by commercial f i rms.It was the f i rst time commercial fi rms were allowed to enter the banking sector.Investment in the banking sector by commercial fi rms has since grown rapidly,and they have become important stakeholders in many commercial banks,suchas China Construction Bank,China Minsheng Bank,China Merchants Bank,Huaxia Bank,and Shanghai Pudong Development Bank.

2.2.2.Establishing f i nance f i rms

To facilitate the development of business groups,the State Council issued Provisions on Further Promoting the Horizontal Economic Alliance in 1986,and Opinions on the Formation and Development of Business Groups in 1987.These regulations allowed business groups to set up f i nance f i rms with the approval of the People’s Bank of China.Finance f i rms can arrange borrowing and lending within the business group,and carry out transactions with banks or other f i nancial institutions.Business groups can raise money from the public. The f i rst f i nance company approved was the Dongfeng Motor Finance Company,established in May 1987, and many business groups have since established f i nance subsidiaries,including Petrol China,China Power, the CITIC Group,the China Everbright Group,China Wanxiang,the New Hope Group,the Haier Group, etc.By the end of 2015,186 f i nance f i rms were organized by commercial f i rms in China.5The statistics are obtained from the website of the China Banking Regulatory Commission:http://www.cbrc.gov.cn/chinese/jrjg/index. html

2.2.3.Cross-industry operations and the formation of f i nancial holding groups

The four state-owned banks began multi-operations in 1984,in areas such as securities,leasing,real estate, and investment.The People’s Bank of China was at the time both central bank and regulatory body,in charge of the regulation of banking,investment banking,insurance,and trust f i rms.However,due to the weak legal system,insufficient discipline,and a lack of risk control,the money from the banking sector f l ooded into the stock market and real estate.This dried up the credit funds available for enterprises,producing bubbles in the stock and real estate markets.By the beginning of the 1990s the f i nancial system was seriously chaotic, the inf l ation rate was high,and the economy overheated.The central government then began to rectify and regulate the f i nancial market,and in 1993 the State Council issued the Decision on the Reform of the Financial System,proposing the separation principle for the f i nancial industry.The Law of the People’s Bank of China, the Law of Commercial Banks,and the Law of Insurance were successively issued since 1995.These laws set up the rules of separating banking from commerce,and separating banking,investment banking,and insurance.Banks are prevented from owning equities in commercial f i rms,but commercial f i rms can still invest in banks.The China Securities Regulatory Commission was established in 1992 followed by the setting up of the China Banking Regulatory Commission and the China Insurance Regulatory Commission.These regulatory bodies aided the development of a sound f i nancial market.

A global trend of broadening bank activities has emerged since the 1990s,and China’s separate operation model was also relaxed.Commercial banks were able to set up fund management subsidiaries.And crossindustry operations could be realized through holding companies.For example,the Ping An China Group has insurance,securities,and commercial banking subsidiaries.The separation of Chinese banking and commerce is also a unilateral separation:commercial f i rms can invest in the f i nancial sector.

An increasing number of business groups have entered the f i nancial industry since 1997.The Haier Group invested RMB500 million in Qingdao Bank in 2001,and invested in Changjiang Securities and the Anshan Trust and Investment Co.It established its own f i nance subsidiary in 2002.The Luneng Group has strategically become the largest shareholder of Huaxia Bank,Xiangcai Securities,and Weishen Securities,the fourth largest shareholder of the Bank of Communications,and the controlling shareholder of Jinan Yingda International Trust and Shandong Jinshui Futures.Another example is the New Hope Group,the founding investor of China Minsheng Bank,which has now expanded its investment into insurance,securities,and investment f i rms.The main investing force in the f i nancial industry is in fact the f i rms controlled by SASAC (State-owned Assets Supervision and Administration Commission of the State Council).Most SASAC-controlled f i rms have established f i nance subsidiaries,with some extending to banking,securities,insurance, and futures.In addition,many local governments restructured their banking,securities,insurance,trust,and leasing affiliates into controlling f i nancial groups after 2009.

In July 2013,the State Council issued its Guidance on Financial Support to the Economic Structure Adjustment,Transition,and Upgrading.The Guidance proposed 10 reform policies encouraging private capital to invest in f i nancial institutions.By the end of 2013,36 commercial banks were approved to be sponsored by private capital.

2.2.4.The rise of supply chain fi nance and Internet fi nance

In recent years,certain group companies have started to explore a new fi nancing business model,providing fi nancial services along the group supply chain.This innovation is known as supply chain fi nance,and it provides short-term credit and optimizes working capital for both the buyer and seller.Internet fi nance is another emerging innovation.E-commerce fi rms are able to start up various fi nancial services by leveraging their customer and big data advantages.The businesses involve credit cards,mini-loans,insurance,and asset management.These new fi nancing models have come into being through commercial fi rms rather than traditional fi nancial institutions.

To summarize,China’s fi nancial market is emerging from its preliminary stage.In the process,we witness the unprecedented enthusiasm of commercial fi rms to invest in various sectors of the fi nancial industry.This wave of enthusiasm for the fi nancial sector promises technology and business innovations,but there are also possible hidden risks.

3.Literature review and theoretical underpinning

Theoretically,there are both costs and benef i ts of merging banking and commerce.The often-claimed benef i ts are a reduction of portfolio risk,economics of scale and/or scope,new sources of capital,a reduction of transaction costs,etc.The cost concerns include conf l icts of interest,excessive market power,and risk contagion.The economic perspectives of bank ownership in commercial f i rms have been investigated,particularly in relation to German and Japanese banks.The f i ndings of empirical studies generally support the theoretical arguments that banks’equity stakes in commercial fi rms reduce agency costs and the cost of capital,a ff ect fi rm performance,and lower the cost of fi nancial distress(see the review of Santos,1997).The motivations and economic consequences of commercial fi rms’equity stakes in banks are,however,relatively unexamined. Ma(2007)argues that the investment by Taiwan fi rms during 1990s in the banking sector was used as a strategic commitment to an aggressive output stance,thus moving the industry to an equilibrium that is more favorable to the fi rms.Laeven(2001)and others fi nd that the extensive enterprise ownership of banks in Russia fostered related lending.Lu et al.(2012)investigate the economic consequences of holding 5%or greater equity stakes in banks by non fi nancial listed fi rms from 2006 to 2008 in China.They fi nd that for non-stateowned fi rms,holding signi fi cant bank ownership leads to lower interest expenses and less fi nancial constraints. Combining these theoretical predictions and empirical fi ndings with practices in China,we propose that obtaining high returns in the fi nancial industry,reducing transaction costs,and diversifying risk are the three main economic reasons Chinese non fi nancial fi rms expand their operations into the fi nancial sector.

3.1.Obtaining high returns of f i nancial industry

Capital is prof i t driven.Tobin(1969)explains how money and capital can be inter-convertible using q theory.When q is greater than 1,the valuation of existing capital is higher than its replacement cost,causing investment in real capital.However,when q is smaller than 1,the valuation of existing capital is lower than its replacement cost.Selling assets at replacement cost and investment in the money market can reap higher returns on capital.In a similar vein,Porter(1985)points out that when an industry’s rate of return stays at a low level and there is no sign of improvement in the future,f i rms in that industry will look elsewhere for better investment opportunities.The f i nancial industry has traditionally featured high returns and high risk.Over the past ten years,the banking sector has been the most prof i table industry in China,and higher prof i ts have attracted more investment.6For example,in 2011,the average rate of return for the commercial sector was about 8%.For the same year,the return on equity for commercial banks was 20.4%(Yang and Dai,2012).Therefore,we argue that the primary motivation for listed f i rms to invest inthe f i nancial sector is to reap the high prof i tability of the f i nancial industry.This argument is also consistent with life-cycle theory,which states that mature f i rms with abundant cash f l ow will start new businesses to sustain growth.

3.2.Reducing transaction costs

Firms and markets are two substitutable forms of resource allocation.The scope of a f i rm is determined by balancing the costs of organizing within the f i rm with the costs of organizing in another f i rm,or the costs involved in leaving the transaction to be organized by the price mechanism(Coase,1937).Williamson (1979,1985)pointed out that transaction costs include those of search and information,bargaining,and policing and enforcement.Firms weigh the costs of exchanging resources in the environment against the bureaucratic costs of performing activities in-house.Transaction costs related to the exchange of resources with the external environment may be ref l ected by environmental uncertainty,opportunism,risks,bounded rationality, core company assets,etc.For example,if f i rms view the environmental uncertainty as high,they may choose not to outsource or exchange resources with the environment.

Goto(1982)and Diamond(1984)use transaction cost theory to identify the fundamental reason for the existence of business groups and conglomerates.Goto(1982)states that if a f i rm forms or joins a group,it can economizeonthetransactioncostsitwouldhaveincurredifthetransactionhadbeencarriedoutthroughthemarket, and can at the same time avoid the scale diseconomies or control losses that would have occurred if it had expanded internally and performed the transaction within the f i rm.If the net benef i t of forming or joining a group exceeds that of implementing a transaction with the f i rm or through the market,the f i rm has the incentive to form or to join a group.This explains the existence of universal banks in Germany and of Keiretsus in Japan. Diamond(1984)developsatheoryoff i nancialintermediationbasedonminimizingthecostofmonitoringinformation,whichisusefulinresolvingincentiveproblemsbetweenborrowersandlenders.Af i nancingintermediary has a net cost advantage relative to direct lending and borrowing,but intermediaries must bear certain risks for incentive purposes.To diversify the risks,f i nancial intermediaries and f i rms can form conglomerates.

Hoshi et al.(1991)provide empirical evidence that within business groups where banks own large equity stakes in member f i rms and lend considerable capital,the information and incentive costs are low,freerider problems can be reduced,and the costs of f i nancial distress are also lower.In China,bank loans are the main source of f i nancing(Allen et al.,2007).Due to transaction costs,credit quotas,and lending discriminants,many f i rms,particularly private ones,are constrained when obtaining bank loans.To f i nance their projects,f i rms need to maintain sound relationships with banks.They may even directly own equity stakes in banks.7The Law of Commercial Banks released in 2003,article 40,states that commercial banks shall not issue credit loans to related parties; the provisions of collateral debt a commercial bank issued to related parties shall not be superior to those of similar debt issued to other parties.Where related parties include(1)the directors,supervisors,managers,and creditors and their close relatives of a commercial bank; and(2)the corporations,enterprises,and other economic organizations those listed above persons in or serve as top management. However,these regulations do not bar the issuance of credit debt to shareholders by commercial f i rms,as long as the shareholder has not appointed directors,supervisors,or managers to its invested banks.Even if the shareholder of a commercial bank has appointed directors, and/or supervisors,and/or managers to the invested bank,the bank can still issue collateral debt to its shareholder.Therefore,we predict that by investing in f i nancial institutions,f i rms can internalize transaction costs when obtaining f i nance.

3.3.Diversif i cation strategy

Commercial f i rms invest in the f i nancial sector to diversify.There are costs and benef i ts to diversif i cation. Diversif i ed f i rms can utilize the internal capital markets to better allocate f i rm resources(Stein,1997).Diversif i cation also brings synergy ef f ects and reduces risk(Hill and Hoskisson,1987;Amit and Livnat,1988). Diversif i cation can,however,aggregate agency problems(Jensen,1986).Managers use diversif i cation to avoid risks and increase f i rm size uneconomically(Rajan et al.,2000).By investing in the f i nancial sector,commercial f i rms can utilize investment opportunities that dif f er from their own line of business while stabilizing their overall income.For example,the Baosteel Group realized its income even in years when the steel industry as a whole was suf f ering losses,by reaping its prof i ts from the f i nancial sector.

4.Research design

4.1.Models

Based on previous analysis,we construct regression models to investigate the motivations and economic consequences of commercial companies holding equity stakes in f i nancial f i rms.We identify three motivations for commercial f i rms to invest in the f i nancial sector:obtaining high returns,reducing transaction costs,and diversifying to reduce risks.We expect that when an industry’s competition intensi fi es,returns decrease,and a fi rm will seek to invest in a more prosperous and pro fi table industry.We use the Her fi ndahl index(HHI,where a lower HHI indicates a higher level of competition)to measure the extent of industry competition,Q to measure investment opportunities,and ROA to measure pro fi tability.We predict that the lower the HHI,the lower the Q,and the lower the ROA,the more likely a fi rm is to invest in the fi nancial sector.A fi rm needs external fi nance to support its growth.In a perfect world,a fi rm can obtain fi nance without cost,so fi nancing will be determined by the investment opportunity.However,in the real world,transaction costs make fi nancing expensive.Firms are often fi nancially constrained.We use the cost of debt fi nancing to measure the transaction costs for obtaining loans.We expect that the higher the debt cost,the more likely fi rms will be to invest in the fi nancial sector.Based on diversi fi cation theory,larger fi rms,older fi rms,and those with more cash fl ow are more likely to diversify operations and invest in the fi nancial sector.

Investment in fi nancial institutions constitutes one part of investment decisions.Therefore we control for other factors that in fl uence investment,including internal cash fl ow(Cash fl ow),the level of debt(LEV),and uncertainty(Risk).In China,government control and institutional environments are important determinants of economic decisions,so we control for government control(GOV,a dummy variable for state-controlled fi rms),ownership concentration(Top1,the ownership of the largest shareholder),and the extent of marketization(Lnmindex).

We use a Logit model to investigate the motivations of holding equity stakes in fi nancial institutions.The model is as follows:

P represents the probability of a non fi nancial listed fi rm investing in the fi nancial sector.The dependent variables are D fi n fi rm,D fi n fi rmb,or D fi n fi rmr.Table 1 gives the de fi nitions of the variables.

To investigate the economic consequences of holding equity stakes in fi nancial fi rms,we test the changes in fi rm performance before and after investing in the fi nancial industry.In particular,we check the ROA and the components of ROA:CROA(operating return on total assets)and IROA(investment income on total assets). We also test for changes in the cost of debt(Debtcost),cash-holdings of investing fi rms(Cashholding),and stock price performance(Rw).

We use the following f i xed-ef f ect panel data regression to evaluate the inf l uence of investing in the f i nancial industry on f i rm performance:

In model(2),the dependent variable,Performance,takes CROA,IROA,ROA,Debtcost,Rw,and Cashholding,where applicable.The main independent variable is the indicator of investing in the f i nancial sector, and takes Df i n,Df i n and Df i nb,Df i n and Df i nr,Df i n and Ratio1,and Df i n and Ratio2,respectively.Size, LEV,Q,Risk,Lnage,GOV,Top1,and Lnindex are control variables.The model also controls for f i rmand year-f i xed ef f ects.The variable def i nitions are given in Table 1.

Table 1Variable def i nitions.

4.2.Sample and data

The China Securities Regulatory Commission(CSRC)has required listed f i rms to publicly release detailed annual reports since 1999,which is therefore when our sample starts.Before 2007,equity investment was reported in the‘‘long-term investment”account and since then,if the invested f i rm was listed,the investments were transferred to‘‘a(chǎn)vailable-for-sale”investments.For equity investment with over 50%ownership,the subsidiary is consolidated and not reported in the‘‘long-term investment”account,but it can be traced in the footnotes of f i nancial statements,where subsidiary information is disclosed.We therefore hand-collect investment totals,and the ratios invested in f i nancial f i rms by nonf i nancial listed companies,from the annual reports.The f i nancial f i rms identif i ed include f i rms providing services of banking,loans,securities,venture capital and private equity,insurance,f i nance,trusts,guarantees,futures,asset management,investment funds, leasing,and pawnshops.The sample period is 1999-2012.The f i nancial and corporate governance data and stock returns are extracted from Wind and CSMAR.

Table 2Research sample.

We start with all nonf i nancial A-share listed f i rms from 1999 to 2012,and exclude extreme observations(for example,if the debt-to-asset ratio is greater than 1)and f i rm-years with missing values.The resulting sample is made up of 15,741 f i rm-year observations.Depending on the status of the equity stakes held in f i nancial institutions,the 15,471 observations are divided into f i ve groups.The f i rst consists of 4778 observations for f i rms that never invest in f i nancial institutions,and the second of 3719 for f i rms consistently reporting investments in fi nancial institutions for all years of the sample period.The third group consists of 4070 observations for fi rms with no equity investment in fi nancial institutions at the beginning of the sample period but that invest in fi nancial institutions later.The fourth group comprises of 1247 observations for fi rms that initially have equity stakes in fi nancial institutions but then sell them,and the last consists of 1927 observations for fi rms that occasionally invest in fi nancial institutions.

To obtain a clean test sample we use the f i rst and third groups totaling 8848 observations to investigate the motivations and the economic consequences of investing in f i nancial institutions.8Typically,the sample for the Logistic regression can be obtained by matching the research sample with a control sample.The matching standards can be industry,f i rm size,prof i tability,...and so on,depending on the research scenario.However,in this study,industry,f i rm size,prof i tability,and other f i rm characteristics are independent variables of interest.If these dif f erences are removed,the regression will become meaningless.,9As lag variables are used,the regression sample is slightly smaller.We use the f i rst group and the early year observations of the third group before f i rms have invested in the f i nancial sector to investigate motivations.The dummy variable Df i nf i rm is coded 1 for f i rms that invest in the f i nancial sector later in the 1999-2012 period and 0 otherwise.Df i nf i rm is a f i rm-level indicator.We use two samples to investigate the performance of investing in the f i nancial sector.The f i rst consists of all observations in the f i rst and third group,and the dummy variable Df i n is coded 1 for f i rm-years with investments in equity stakes of f i nancial institutions and 0 otherwise.This construction results in a dif f erence-in-dif f erence test of the economic consequences when Chinese nonf i nancial listed f i rms integrate f i nance and commerce.The second sample consists of observations from the third group,where the dummy variable Df i n is coded 1 for f i rm-years with equity investment in f i nancial institutions and 0 otherwise.10This sample is used in robustness checks.Using the second sample,we can compare f i rm performance before and after equity stakes in f i nancial institutions are held.Table 2 summarizes the construction process of the test samples.

5.Empirical results

5.1.Descriptive statistics

In empirical tests,all continuous variables are winsorized at 1%and 99%,except the marketization index (Lnmindex).Table 3 reports the descriptive statistics for the research sample.The f i rm-level statistics show that on average,46%of f i rms have equity investment in the f i nancial sector between 1999 and 2012,and 21.21%of f i rms hold equity stakes in banks,with 28.05%holding equity stakes in banking,securities,andinsurance institutions over the same period.At the f i rm-year-level,on average 27.22%have equity investment in the f i nancial sector,11.18%in the banking sector,and 16.67%in banking,securities,and the insurance sector.The depth of investment in the f i nancial sector on average is 0.81%of total assets and 1.71%of net assets, with maximums of 22.16%of total assets and 45.79%of net assets.The investment amount averages 5.1306*107and the maximum value is 1.65*1010.Investing in f i nancial institutions is therefore very attractive for nonf i nancial listed Chinese f i rms,though the level of investment varies greatly.11Note that the statistics here are for Group 1 and 3 only.The remaining three groups of f i rms all occasionally had equity investment in the f i nancial sector.Therefore,the popularity of holding equity stakes in f i nancial institutions is much more common for non-f i nancial listed f i rms as a whole.We conduct correlation checks for independent variables and f i nd that the Pearson correlation coefficients are below 0.4,so multicollinearity is not serious in our research.12To save space,the correlation coefficients are not reported but are available upon request.

Table 3Descriptive statistics.

5.2.Motivations for investing in fi nancial institutions

Here we examine the results of logistic regressions of the determinants or motivations of investing in the fi nancial sector.Government and regulation are important in fl uencers of economic life in China,so we therefore investigate state-controlled fi rms and non-state controlled fi rms separately.The fi nancial sector is highly regulated,and banking,securities,and insurance industries are subject to the strictest regulations.For non financial fi rms,the goal of investing in the strictly regulated fi nancial sector may be to obtain permits rather than pro fi ts.We therefore separately investigate the motivations of investing in the strictly regulated fi nancial sector and in the relatively less strictly regulated sector.As the debate on separating banking and commercehas continued for many years,we also test the motivations of equity investments in banks and non-bank f i rms separately.Tables 4a and 4b report the results.

Table 4aDeterminants of equity investment in f i nancial institutions.

Tables 4a and 4b shows that the coefficient on the extent of industry competition(HHI)is negative in six out of seven regressions,indicating that the lower the HHI(i.e.,the higher the extent of industry competition), the more likely nonf i nancial f i rms are to invest in the f i nancial sector.The negative coefficients are,however, signif i cant only for the whole sample,the non-state controlled f i rm sample,the non-bank equity investment sample,and for the sample of investment in relatively less strictly regulated industries.The extent of industry competition is not a consideration for state-controlled f i rms,for those investing in the banking sector,and those investing in the strictly regulated sector.Government inf l uence,rather than market forces,may drive the investment decisions of state-controlled f i rms.We also f i nd that obtaining permits,rather than industry competition pressure,is the main concern when investing in the banking sector and the highly regulated f i nancial sector.

The coefficients on the cost of interest-bearing debt(Debtcost)are signif i cant only in the non-state controlled sample and in the investment in the banking sector regression,which indicates that non-state controlled fi rms are more fi nancially constrained and aim to reduce transaction costs by investing in the fi nancial sector. Firms with higher debt costs are more likely to invest in banks,in the hope of reducing their debt costs.

The coefficients on pro fi tability(ROA),leverage(LEV), fi rm size(Size),and cash holding(Cashholding) are statistically positive across the seven regressions.More pro fi table and larger fi rms,and those with ample debt fi nancing and abundant cash,are therefore more likely to investment in the fi nancial sector,as they are less fi nancially constrained.Their motivations for holding equity investment in fi nancial institutions are more consistent with the diversi fi cation strategy.The coefficient on listing age(Lnage)is,however,negative and signi fi cant,which seems inconsistent with diversi fi cation theory,but the unique IPO market in China,with its high o ff ering prices,high pricing in terms of P/E ratio,and high over-raised funds,causes newly listed fi rms to over-invest,including investing in the fi nancial sector.

Table 5aEf f ect of holding equity stakes in f i nancial institutions on operating performance(dependent variable:CROA).

The explanatory power of the major determinants of investment,investment opportunity(Q),internal cash fl ow(Cash fl ow),and uncertainty(Risk)varies.The coefficients on investment opportunity are not signi fi cant in any regressions.Therefore,listed f i rms do not wait until investment opportunities in their own industry are exhausted before entering the f i nancial sector.For state-controlled f i rms,and for determinants of equity investment in banks,the coefficients on internal cash-f l ow are positive and signif i cant,possibly indicating a free cash-f l ow problem.The coefficients on uncertainty are signif i cant for the state-controlled f i rms sample and on the determinants of equity investment in banks and in the strictly regulated f i nancial sector,which is consistent with the real stock option theory of investment.

Finally,the explanatory power of the corporate governance variables—the type of controlling shareholder (GOV),ownership concentration(Top1),and the marketization index(Lnmindex)—also varies.First,government control is only important in determining equity investment in the strictly regulated f i nancial sector, which may indicate that connection with the state is useful in obtaining entry permits in the highly regulated sector.Second,for determinants of equity investment in banks,in the highly regulated f i nancial sector,and for the state-controlled sample,the coefficients on the ownership of the largest shareholder are negative,but positive for the non-state controlled sample.Therefore,the higher the ownership concentration,the less willing state-controlled f i rms are to invest in the f i nancial sector,but the more willing non-state-controlled f i rms are to invest.Finally,the overall explanatory power of the extent of marketization is low.

5.3.Economic consequences of holding equity stakes in f i nancial institutions:Ef f ect on operating returns (CROA)

Tables 5a and 5b report f i rm-and year-f i xed ef f ect estimates of the ef f ect on operating returns of holding equity stakes in f i nancial institutions.The coefficients on indicators of equity investment in f i nancial institutions and in banks are insignif i cant,but this does not mean there is no ef f ect on operating returns. The coefficients on the indicator of equity investment in strictly regulated f i nancial institutions and on the depth of investment are negative and statistically signif i cant.Investing in strictly regulated f i nancialinstitutions is therefore negatively related to a f i rm’s ability to acquire operating earnings.As the amount of investment in f i nancial institutions increases,operating return on assets decrease.

Table 5bEf f ect of holding equity stakes in f i nancial institutions on operating performance:The ef f ects of government control(dependent variable: CROA).

Table 5b demonstrates the ef f ect of government control when including the interactions of GOV and the depth variables of holding equity stakes in f i nancial institutions.The results in Column(8)show that government control inf l uences the ef f ect of operating earnings when investing in strictly regulated f i nancial institutions.For non-state-controlled listed f i rms,investing in strictly regulated f i nancial institutions is associated with a decrease in operating performance.The coefficient on Df i nr is-0.0161,which is signif i cant at less than 1%.For state-controlled listed f i rms,this investment is not associated with decreased operating performance and the coefficient on Df i nr*GOV is 0.0149,signif i cant at less than 5%.The aggregated ef f ect of holding equity stakes in strictly regulated fi nancial institutions is-0.0012(-0.0161+0.0149).Non-state-controlled fi rms may place more importance on obtaining entry permits than on short-run economic returns.

5.4.Economic consequences of holding equity stakes in f i nancial institutions:Ef f ect on investment income (IROA)

Here,we investigate the ef f ect of holding equity stakes in f i nancial institutions on the performance of external expansion.As one component of external expansion,investment in f i nancial institutions can contribute to investment income,which we therefore expect to increase.The f i rm-and year-f i xed ef f ect estimates are reported in Tables 6a and 6b.In regression(11)of Table 6a,the coefficient on Df i n is positive and signif i cant, indicating that holding equity stakes is positively related with investment income.In regression(12),the coef-fi cient on Df i n becomes insignif i cant,but the coefficient on Df i nb is signif i cant.Therefore,only investment in banks can bring more investment income,which is also found in regression(13),where the coefficient on Df i nr is not signif i cant.Nevertheless,as regressions(14)and(15)show,investment income increases with the amount of investment in f i nancial institutions.

Table 6aEf f ect of holding equity stakes in f i nancial institutions on investment performance(dependent variable:IROA).

The ef f ect of government control is also considered,and the results are reported in Table 6b.As regressions (16)and(17)show,though non-state-controlled f i rms earn higher investment income by investing in f i nancial institutions,particularly banks,the investment returns of state-controlled f i rms do not increase.The coefficients on Df i n*GOV in regression(16)and on Df i nb*GOV in regression(17)are negative and signif i cant. Combined with the negatively signif i cant coefficients on Df i n and Df i nb,the overall results become insignificant.However,as the amount of investment grows,state-controlled f i rms also increase their investment income,as shown in regression(19).

5.5.Economic consequences of holding equity stakes in f i nancial institutions:Ef f ect on net income(ROA)

We then investigate the economic consequences of holding equity stakes in f i nancial institutions in terms of net income,which is the bottom line of the operating results reported in Tables 7a and 7b.The coefficients on Df i n,Df i nb,and Df i nr in Table 7a are not signif i cant,so investing in f i nancial institutions does not increase or decrease return on assets.However,the coefficients on Ratio1 and Ratio2 are both negative and signif i cant at less than 1%.Therefore,as the amount of the investments increases,return on assets signif i cantly decrease.To summarize the results across Tables 5a,6a,and 7a,we f i nd that as the depth of investing in f i nancial institutions increases,performance from external investment improves.The gains in investment income cannot,however,make up for the losses in other operating earnings.The aggregated result is a decrease in the overall return on assets.

We again consider the interaction ef f ects of government control and of holding equity stakes in f i nancial institutions.Table 7b shows that the only signif i cant ef f ect is found in regression(28),where for non-state controlled listed f i rms,investment in strictly regulated f i nancial institutions result in lower return on assets,but this is not the case for state controlled listed fi rms,where the coefficient on D fi nr is insigni fi cant at-0.0015 (-0.0133+0.0118).

Table 6bEf f ect of holding equity stakes in f i nancial institutions on investment performance:The ef f ects of government control(dependent variable: IROA).

In summary,holding equity stakes in fi nancial institutions does not improve fi rm performance.As these fi rms are larger,more pro fi table,and have abundant cash before becoming involved in the fi nancial sector, their actual performance deteriorates.

5.6.Additional tests

5.6.1.Other dimensions of f i rm performance

We further investigate the economic consequences of holding equity stakes in f i nancial institutions by testing the ef f ect on other dimensions of f i rm performance apart from prof i tability;transaction costs(Debtcost), market performance(stock returns),and cash-holdings.13To save space,the empirical results are not reported but are available upon request.

First,we test changes in transaction costs around investing in the f i nancial sector and f i nd the investment to be associated with a higher cost of debt,as the coefficient on Df i n is positive and signif i cant.Investing in f i nancial institutions raises leverage and overdrawing f i nancial capacity,which may increase the cost of debt.The coefficient on Ratio1*GOV is also positive and signif i cant,so the ef f ects on the cost of debt for statecontrolled f i rms are therefore even higher.

Second,by using the mean of weekly market adjusted idiosyncratic stock returns during a year,Rw,to represent the stock price performance,we f i nd that this is not inf l uenced by holding equity stakes in f i nancialinstitutions.In regressions of Model 2,the coefficients on indicators of holding equity stakes in f i nancial institutions are all insignif i cant at a 5%level.Decreases in the accounting performance are therefore not ref l ected or identif i ed by the market.

Table 7aEf f ect of holding equity stakes in f i nancial institutions on f i rm performance(dependent variable:ROA).

Third,we f i nd that as its holding of equity stakes in the f i nancial sector deepens,a f i rm’s cash-holding decreases,so investing in f i nancial institutions consumes cash reserves.The coefficient on Ration1*GOV is also negative and signif i cant,indicating that this ef f ect is even greater for state-controlled f i rms.This of course can be a mixed blessing,as it also reduces free cash f l ow.

5.6.2.Robustness checks

To reinforce our empirical results we conduct four types of robustness checks.14To save space,the empirical results are not reported but are available upon request.

First,we check the robustness of the inf l uence of government control through the interaction of GOV and the determining variables of exploring the motivations.In the main tests,we conduct regressions for statecontrolled and non-state-controlled samples separately,and when using the interactions of GOV and the determinants(HHI,Q,ROA,Debtcost,Size,Lnage,and Cashholding),the results are consistent with those in Table 5a.

Second,we repeat the tests using only Group 3 f i rms to examine the inf l uence on prof i tability of holding equity stakes in f i nancial institutions.These f i rms did not invest in f i nancial institutions at the beginning of the research period,but made later investments,which they held to the end of the research period.The results are essentially the same as those in the main tests.

Third,we redo the tests after excluding equity investments in f i nance f i rms,which are established to serve the f i nancial matters within a group,and therefore internalize transaction costs.The results are consistent with those including f i nance f i rms.

Table 7bEf f ect of holding equity stakes in f i nancial institutions on f i rm performance:ef f ects of government control(dependent variable:ROA).

Finally,we explore any possible non-linear relationship between investing in the f i nancial sector and f i rm performance by introducing the square terms of Ratio1 and Ratio2.No non-linear relationships are found in regressions on CROA,ROA,Debtcost,Rw,and Cashholding,but may exist for regressions of IROA.The square terms of Ratio1 and Ratio2 are positive and signif i cant at a 1%level,so as the investment level in f i nancial institutions increases,the investment income may f i rst decrease and then increase.The relation is U-shaped.However,to conform to other regressions,we do not include the square item in the main tests.

6.Conclusions and future research

An increasing number of commercial f i rms have become involved in the f i nancial sector during the process of establishing multi-layered capital markets in China.These f i rms are keen to obtain equity stakes in banks and f i rms dealing in securities,venture capital and private equity,insurance,f i nance,investment and trusts, guarantees,futures,asset management,investment funds,and pawnshops,etc.Integrating banking(f i nance) and commerce has been the subject of debate in both practice and in theory for many years,but empirical evidence on the commercial ownership of banks(and/or f i nancial f i rms)is scarce.In this study,we provide evidence by comprehensively investigating the motivations and economic consequences of commercial f i rms entering the f i nancial sector.

From a sample of Chinese nonf i nancial listed A-share f i rms from 1999 to 2012,we f i nd that there are numerous motivations for them to hold equity stakes in f i nancial f i rms.They may be alleviating the competition pressure in the commercial sector,reducing transaction costs,diversifying operations,or obtaining precious permits.

We also f i nd that investment income can increase after a f i rm holds equity stakes in f i nancial institutions, but this is only the case for non-state-controlled f i rms,and overall operating income decreases,which cannot be of f set by the increase in investment income.The return on assets declines as a result.Furthermore,after investing in f i nancial institutions the cost of debt rises,cash-holding falls,and stock returns do not improve.

In summary,investing in f i nancial institutions does not result in improvements in operating performance, nor does it reduce transaction costs.Given that these f i rms are larger,more prof i table,and possess abundant cash reserves before becoming involved in the f i nancial industry,their overall performance does in fact deteriorate.This contrasts with the view that the large-scale investment of capital from the commercial sector is chasing opportunities in the f i nancial sector.Our empirical results caution regulators in the f i nancial sector and decision-makers in the commercial sector when considering or allowing entry into the f i nancial sector.

The empirical tests in this study are comprehensive but general.Research can further examine the integration of f i nance and commerce.Deeper insights can be gained on the ef f ects on both the f i nancial and the commercial sectors.The ef f ects of holding equity stakes in f i nancial institutions on investment and f i nancing decision-making processes of f i rms in the commercial sector can,for example,be investigated further.In general,more studies on commercial f i rms’ownership in f i nancial f i rms would be of benef i t,as the current evidence is slim.

Acknowledgement

Liping Xu and Yu Xin are grateful for the valuable comments from the referee and for the f i nancial support from the National Natural Science Foundation of China(Project Nos.:71372151 and 71272201).

Allen,F.,Qian,J.,Qian,M.,2007.China’s Financial System:Past,Present,and Future.Available at SSRN:＜http://ssrn.com/abstract= 978485＞.

Amit,R.,Livnat,J.,1988.Diversi fi cation strategies,business cycles and economic performance.Strategic Manage.J.9(2),99-110.

Barth,J.R.,Caprio Jr.,G.,Levine,R.,2001.Banking systems around the globe:do regulations and ownership a ff ect performance and stability?In:Mishkin,F.S.(Ed.),Prudential Supervision:What Works and What Doesn’t.Univ.of Chicago Press,pp.31-88.

Belyanova,E.,Rozinsky,I.,1995.Evolution of commercial banking in Russia and the implications for corporate governance.In:Aoki, Masahiko,Kim,Hyung-Li(Eds.),Corporate Governance in Transitional Economies:Insider Control and the Role of Banks.World Bank,Washington,DC,pp.185-214.

China Entrepreneurs Survey System,2011.Capital markets and the foster of China’s entrepreneurs:present and future,problems and suggestions.Manage.World 6,76-90(In Chinese).

Coase,R.H.,1937.The nature of the fi rm.Economica 4(16),386-405.

Craig,V.V.,2001.Merchant banking:past and present.FDIC Bank.Rev.14(1),29-36.

Diamond,D.W.,1984.Financial intermediation and delegated monitoring.Rev.Econ.Stud.51(3),393-414.

Fan,G.,Wang,X.,Zhu,H.,2011.NERI Index of Marketization of China’s Provinces 2011 Report.Economic Science Press,Beijing(In Chinese).

Goto,A.,1982.Business groups in a market economy.Eur.Econ.Rev.19(1),53-70.

Halpert,S.K.,1988.The separation of banking and commerce reconsidered.J.Corporat.Law 13,481-533.

Haubrich,J.G.,Santos,J.A.C.,2003.Alternative forms of mixing banking with commerce:evidence from American history.Financ. Markets,Inst.,Instrum.12(2),121-164.

Hill,C.W.L.,Hoskisson,R.E.,1987.Strategy and structure in the multiproduct fi rm.Acad.Manage.Rev.12(2),331-341.

Hilferding,R.,1910.Finance Capital:A Study of the Latest Phase of Capitalist Development.Translated by Morris Watnick,Routledge &Kegan Paul,London,1981.

Hoshi,T.,Kashyap,A.,Schafstain,D.,1991.Corporate structure,liquidity,and investment:evidence from Japanese Industrial Groups. Quart.J.Econ.106(1),33-60.

International Finance Research Institute of the Bank of China,2010.The trends and strategies of China’s industrial capital investing in fi nancial sector.Res.Glob.Econ.Finance 16,1-36.

Jensen,M.C.,1986.Agency cost of free cash fl ow,corporate fi nance and takeovers.Am.Econ.Rev.76(2),323-369.

Krainer,J.,2000.The separation of banking and commerce.FRBSF Econ.Rev.,15-25

Laeven,L.,2001.Insider lending and bank ownership:the case of Russia.J.Comp.Econ.29(2),207-229.

Lu,Z.,Zhu,J.,Zhang,N.,2012.Bank discrimination,holding bank ownership,and economic consequences:evidence from China.J. Bank.Finance 36(2),341-354.

Ma,T.,2007.Bank- fi rm relationship as a strategic commitment in a Duopolistic environment.J.Competit.Law Econ.3(2),233-241.

Porter,M.E.,1985.Competitive Advantage:Creating and Sustaining Superior Performance.Free Press,New York,NY.

Rajan,R.,Servaes,H.,Zingales,L.,2000.The cost of diversity:the diversif i cation discount and inefficient investment.J.Finance 55(1), 35-80.

Salley,C.D.,1976.Origins of the regulatory separation of banking and commerce.Bank.Law J.93,196-206.

Santos,J.A.C.,1997.Banking and commerce:how does the United States compare to other countries?Fed.Reserve Bank Cleveland Econ. Rev.34(4),14-26.

Shull,B.,1999.The separation of banking and commerce in the United States:an examination of principal issues.Financ.Markets,Inst., Instrum.8(3),1-55.

Stein,J.C.,1997.Internal capital markets and the competition for corporate resources.J.Finance 52(1),111-133.

Tobin,J.,1969.A general equilibrium approach to monetary theory.J.Money,Credit Bank.1(1),15-29.

Wall,L.D.,Reichert,A.K.,Liang,H.-Y.,2008.The f i nal frontier:the integration of banking and commerce.Part 1:The likely outcome of eliminating the barrier.Fed.Reserve Bank Atlanta Econ.Rev.93(1),1-16.

Williamson,O.E.,1979.Transaction-cost economics:the governance of contractual relations.J.Law Econ.22(2),233-261.

Williamson,O.E.,1985.The Economic Institutions of Capitalism:Firms,Markets,Relational Contracting.Free Press,New York,NY.

Yang,Q.,Dai,S.,2012.A survey on the reverse changes in returns of banks and that of real economies.China Finance 20,34-35(In Chinese).

13 November 2014

*Corresponding author.Tel.:+86 20 84115578.

E-mail address:mnsxy@mail.sysu.edu.cn(Y.Xin).

http://dx.doi.org/10.1016/j.cjar.2016.04.001

1755-3091/?2016 Sun Yat-sen University.Production and hosting by Elsevier B.V.

This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Accepted 6 April 2016

Available online 6 May 2016