The impacts of increasing foreign ownership on banking industry

Transcription

1 The Effects of Increased Foreign Ownership 1 on Korean Domestic Banking Industry Yongil Jeon Associate Professor of Economics School of Economics Sungkyunkwan University yjeon@skku.edu Phone: Insill Yi Associate Professor of Economics Graduate School of Economics Sogang University insill723@sogang.ac.kr, Phone: Abstract in Korean: 외환위기이후국내은행산업의판도변화중가장주목할만것은외국계자본의 진입이다. 외환위기이전외국계자본의국내은행산업진입은자국기업의한국진출과 동반한지점운영의형태로이루어져왔으나당시에는국내은행산업내에서주목할 만한영향력을발휘하지못했다. 반면, 외환위기이후에는주식시장에서지분매입과 M&A 등을통한적극적인은행인수의형태를보이면서외국계자본진입의영향력이 점차확대되어가고있다. 원활한은행산업구조조정의일환으로부실은행을정리하는 과정에서몇몇외국계자본이국내은행을인수하게되고주식시장에서포트폴리오 투자로은행지분을늘려감에따라시중은행의외국계자본의지분율이우리은행을 제외하고는 80% 대수준에이를정도로국내은행산업에대한외향적인영향력이 확대되는추세에있다. 본논문에서는외국계자본진입의경제적효과를국내은행산업에대한영향과 주식시장에대한영향으로나누어분석하였다. 그동안외국계자본진입이우리나라 1 The author acknowledges the helpful comments of Dr. Hoon Kim, and other participants at the Bank of Korea workshop. 1

2 은행산업에미친영향에대한연구는은행산업의경영효율성과안정성에대해일관된분석결과를보여주지못하고있다. 외국계금융기관이본격적으로진출한기간이길지않고다양한요인이병존해있기때문이다. 따라서외국계자본진입의경제적효과에대한가설을검증하기위하여국내은행산업에대한영향을대출행태, 수익성, 경영효율성, 선진금융기법의전수등크게 4 가지관점에서분석해보았다 년부터 2006 년까지의국내은행패널자료를분석한결과다음과같은결론을얻었다. 첫째, 외국자본의국내은행산업진입확대로인해기업대출에있어중소기업대출이축소되고대기업대출은확대된다는실증적결과는찾아보기어려웠다. 한편, 소비자대출에대해서는외국자본진입이양의상관관계를보이고있었다. 둘째, 외국계지분율의증가가자산수익률에미치는영향은유의한결과를얻기어려웠으나자본수익률에는부정적인영향을주고있었다. 셋째, 경영효율성측면에서종업원감축및지점정리를통한비용절감에대해서는유의한결과를보이지않았다. 넷째, 총수익중비이자소득비중이증대할것으로예상하였으나유의한결과를보이지않았다. 마지막으로, 외환관리수익및위험관리는양의상관관계를, 파생상품관련수익은음의상관관계를보이고있어국내은행산업의안정성에는기여하는것으로분석되었다. 한국은행의최근논의에의하면 2000 년부터 2004 년까지기간을대상으로외국계자본의진입정도를나타내는변수로외국계자본의영향을받고있는은행의자산을전체은행의자산으로나눈값을사용하여외국계자본의국내은행산업진입에대한효과를분석하였다. 그결과에따르면국내은행의대기업대출비중을줄이고자기자본비율및영업비용을상승시키는결과를초래한것으로나타난반면, 국내은행의수익성, 자산건전성, 그리고성장성에는유의한영향을미치지못하였다. 경영안정성측면에서는외국자본진입이국내은행에선진경영기법을전수하여그결과로써총수익중비이자소득의비중이증대할것으로예상하였으나유의한결과를찾을수없었다. 따라서종합적으로보아외국계자본진입은은행산업의경쟁력을제고하는데일부만을기여하고있는것으로판단된다. 외국자본의국내은행산업진입은외환위기이후정부가도입한여러가지외자유치정책과규제완화및개방화정책의일환으로시작되었다. 하지만최근에는단순한제도개선에따른효과뿐만아니라외국자본입장에서국내은행산업이수익성, 2

3 자산건전성, 안전성차원에서매력적인투자처로부상하고있다. 이에따라은행이나금융지주회사의외국인지분율은 2003 년에서 2006 년에이르기까지 64.8% 가량급증하였으며 2007 년에는우리은행을제외한 4 대은행과금융지주회사의외국인지분율이 80% 를상회하고있다. 이중 2 개의은행은 100% 외국은행으로전환되었다. 본논문에서는또한외국자본의진입확대가주식시장에주는영향을보기위해포트폴리오투자가활발하게이루어지고있는신한금융지주회사, 우리금융지주회사, 국민은행, 하나금융지주회사에대한 2002 년 1 월부터 2007 년 9 월년까지의이용가능한일별시계열자료를분석하였다. 실증분석결과신한과하나금융지주회사는상대적으로주식이분산되어있었으며주가와외국인지분율과의관계가유의하게나타나지않았다. 반면, 국민은행은주가가외국인지분율에영향을주고외국인지분율이다시주가에영향을주는상관관계를나타내었으며우리금융지주회사는주가가외국인지분율에영향을주는것으로나타났다. 이러한분석결과에의하면외국계자본의진입확대가국내은행산업에미치는영향은긍정적효과와부정적효과가병존하는것으로보인다. 금융의세계화및규제완화를통한자율화추세를감안할때국내은행산업에대한외국자본의관심이줄어들기는어려울것이다. 그러나주가상승이외국인지분율과양의상관관계를보이고있고특히국민은행의경우양방향의인과관계를보인것으로보아주가의하락은외국인지분율의하락으로이어질가능성이높다. 따라서, 국내은행의지분율을외국계뿐만아니라국내에서도다양하게분산할필요가있다. 이런차원에서볼때국내금융자본을육성하기위한방편으로국내사모펀드의대형화에관한규제를완화할필요가있으며공정한영업환경을마련해주되외국계자본이국내은행산업에바람직한역할을할수있도록효율성, 수익성및안정성제고차원에서의정책적노력이수반되어야할것이다. 또한외국계은행으로전환된국내은행들의상장폐지예에서참고할수있듯이예금자나투자자의정보획득이나금융감독당국의감독강화차원에서이에대한적절한개선책을마련할필요가있다. Key Words: commercial banks, foreign banks, ownership, global advantage hypothesis JEL Classification: E5, G2 3

4 I. 서론 The financial structure of Korean banks, both foreign and domestic, experienced dramatic changes since the Asian Financial Crisis (AFC) in After the crisis, the flows of foreign capital played a significant role by injecting foreign financial resources into Korean companies to help alleviate severe liquidity problems. Korean domestic banks have accommodated foreign capital in three different types. First, foreign capital enters through the traditional route of foreign bank branches that deal with corporate loans, which characterized Korea in the twentieth century. Second, Citibank, for example, runs both its retail and wholesale level bank business in Korea, under its own US brand name. Third, and quite differently, foreign buyers directly invest and acquire the stock of Korean domestic banks through the financial market transactions, achieving higher ownerships in Korean banks. Although those shares frequently do not come with voting rights, they do allow the foreign buyers to monitor the Korean domestic banks efficiently and effectively. A recent major change in the Korean banking industry allowed the creation of bank holding companies (BHC), which now encompasses four main Korean BHCs -- Woori, Hana, Shinhan and Dongwon. As an exception in this recent trend, the Kookmin bank, which holds the largest assets and serves the largest customer base, remains an independent bank, not in a BHC. Further, since the Dongwon financial holding company does not hold banks, we exclude it from our analysis. The first three BHCs exhibit different characteristics and backgrounds, stemming from the process of formation. The Korean government through KDIC directly manages and supervises the Woori BHC; whereas US-originated and Japanese-originated capital manages the Hana and the Shinhan BHCs. Each BHC followed a different path to achieve increased foreign ownership and management. In this paper, we examine the historical development of these three BHCs, such as the process of formation and the effects of foreign ownership on their performance. We measure the performance by the return on equity (available quarterly), the return of assets (available quarterly), and the stock price (available daily) for each BHC. Foreign ownership data for these BHCs exist on a daily basis between most of 2000 through much of This research will aid in understanding how the process of formation of these Korean BHCs affects their ultimate performance. This may also provide the Bank of Korea and other regulatory bodies with insight about how and when to regulate foreign ownership of Korean financial institutions. Finally, we 4

5 will explore the optimal size of Korean BHCs, which remains an unknown fact in the literature as well as in practice. 2 Section II reviews the structural changes that occurred in Korean banking industry since the Asian financial crisis in Those changes include the opening of Korea to foreign capital investment to the Korean banking industry and the introduction of bank holding companies. The structural changes enable us to classify Korean banks into three different groups -- foreign-operated banks, mixed groups with both foreign and domestic control of operations, and Korean-operated banks. In the last category, either Koreans or Korean government hold the majority of outstanding stock. But, holding shares of stock does not necessarily keep the voting power in the Board of Directors, which may requires to relax the regulation imposed on foreign investors. The foreign-operated banks include Korean Exchange Bank (KEB), Korean Citi Bank, and Standard Chartered (SC) First Bank. Section III examines the effects of governance change from domestic to foreign ownership on bank performance. In other words, attracting foreign investment may produce different investment outcomes and returns. Section IV considers how changes in the foreign holding share affect on bank s profitability as well as on the stock performance. We hypothesize that foreign investors provide price leadership or perform the function of a leading indicator for domestic investors, who watch the investment strategies of foreign investors. But, this hypothesis may prove false, implying that the information-rich domestic investors become price-makers. The conclusion (Section V) provides the summary of our research with policy suggestions. II. Foreign Ownership of Korean Banks and Bank Holding Companies In this section, we analyze the determinants of foreign entry and its effects on domestic banks performance. The two major aspects we have to pay our attention since the financial crisis are the introduction of the bank holding companies and the open the bank management and stock to foreigners. To resolve the effects of the Asian financial crisis quickly, the Korean government encouraged foreign capital inflows in banking sector. Foreign ownership in the Korean commercial banking sector increased from 33.5 percent in 1999 to 63.1 percent in 2006, on average. Korean commercial banks foreign ownership exceeds 50 percent except for the Woori and Chunbuk banks. This expansion facilitated the bank restructuring undertaken in the wake of 2 The authors thank Mr. NamJin Cho at the Hana BHC for this question. 5

6 Korean financial crisis by selling off a number of ailing domestic banks to domestic and foreign bidders. Table 1. Foreign Capital Participation in the Korean Banking Sector Domestic banks Nationwide domestic banks Regional domestic banks Note: unit - % on average The typical mode of foreign capital participation changed from the opening or enlarging of branches to green-field investment or mergers and acquisitions (M&A), with liberalization of financial markets giving the process momentum. 3 Foreign bank participation in the banking industry can play a positive role in its developments by improving bank management efficiency, delivering advanced financial techniques, and advancing financial supervisory system. Table 2: The Foreign Ownership of Korean Banks Date Woori Shinhan Kookmin Hana KEB SC First Citi Pusan Daegu Since the Agreement on the Liberalization of Trade in Financial Services of 1997, the signatories have relaxed restriction on foreign banks entry into domestic banking sector. 6

7 Note: Gwangju Jeju Jeonbuk Kyongnam KyungNam and Gwangju merged with Woori BHC on March 2001, so we use the data on Woori BHC. SC First merged on April 2005, but the previous data on the First bank are available from its banking management announced data. Hana BHC is used after 2005 while the previous information comes from Hana bank s bank management announcement. For the Seoul Citi Bank, we use Hanmi data before 2002 and after 2002, we use the merged bank data, available from Seoul Citi Bank management announcement. For the Hanmi bank, the foreign ownership before 2002 come from the Korea Financial Yearbook, with the summed data from the foreign names (KAI 15.46%, CHADWICK 3.42%, FREEWAY 3.42%, SCARLET 3.18%, KAB INVEST 2.46%, EAGLE 2.35%, JPMCB-CAP RE EM GTHFD 1.89%, GABLE, MADDEN 1.11%, MSCO-BLUE RIDGE 1.09%, which total 36.13%). From the early 1990s, the Korean financial authority significantly started to relax its control over the financial sector, launching five year financial liberalization blue prints in In April 1994, regulators significantly lowered foreign entry barriers. Regulators also abolished the economic needs test previously mandated for foreign bank investment. In May 1995 regulators further eliminated the requirement to establish a representative office prior to opening other branches. These deregulatory measures ushered in a rapid increase in foreign entry through the opening of branches. Foreign direct investment (FDI) in financial sector through the opening of branches increased from $27.1 million in 1994, to $536.1 million in 1995, and to $447.2 million in The foreign entry mode in banking sector, however, met the great converting point, because of 1997 Korean financial crisis. The Foreign Investment Promotion Act of 1998 opened up the vast majority of corporations and financial institutions to foreign investors. By offering tax and other incentives, this Act created a more transparent and open business environment. As a result, a rapid increase of FDI in financial sector occurred in The FDI in financial sector increased from $341 million in 1997, to $2,580 million in 1999, and to $1,925 million in The turning point of foreign entry to the Korean banking sector occurred in Opening branches provided the most important organizational form of foreign entry before the Asian financial crisis. But, green field investment and M&As replaced branch openings since The Ministry of Commerce, Industry, and Energy s (MOCIE s) foreign investment data 7

8 for financial sector reports that around 75 percent of foreign investment in financial sector occur as green field investment and M&As, on average. 4 The Korean government eased long standing legal restrictions on the foreign ownership of domestic banks, including 100 percent foreign ownership. Two reasons explain why the Korean government took this action. First, the Korean government wanted to attract foreign capital into the Korean financial sector to facilitate restructuring in the financial sector. Second, the government believed that the entry of foreign equity should strengthen the financial and domestic banking systems. Actually, foreign banks began to play a substantially greater role in the domestic banking sector by 1998 and In early 1999, a group led by Goldman Sachs invested $500 million to acquire a 17 percent stake in Kookmin Bank, Korea s largest retail bank. Commerzbank invested $167 million to acquire one-third ownership in the Korean Exchange Bank. Moreover, New Bridge Capital invested $417 million in fall 1999 to acquire a major stake in the Korea First Bank. These transactions generated a widespread expectation during and just after the Asian financial crisis that foreign investors would acquire a large part of the Korean banking sector, as these nationalized banks were restructured and sold off. These expectations were not realized by 2001, however. The proposed purchase of Seoul Bank by Hong Kong and Shanghai Banking Corporation (HSBC) collapsed in the second half of Moreover, the government s announced plan to sell its majority stakes in Chohung Bank and Hanvit Bank and its minority stakes in Korea First Bank and Korea Exchange Bank was abandoned in December 2000 in favor of injecting additional public funds and moving several of the financial institutions into a financial holding company. The role that foreign capital contributed to the recovery and restructuring of the Korean banking system fell well below its initial expectation and substantially below that in parts of Latin America. For example, foreign banks controlled 45% of all banking assets in Mexico by the end of 2000, just five years after the Mexican financial crisis. Foreign financial institutions played a important role in providing capital and, thereby, boosting the economy. But, similar to other Asian countries experience with the financial crisis, the Korean economy first plummeted and then recovered quickly, tracing out a V-shaved pattern. Thus, foreign capital played a less 4 Due to the lack of disaggregated data on the mode of foreign entry to the banking sector, the MOCIE data provide foreign direct investment in the financial and insurance industries. 5 The Korean Financial Supervisory Commission hired Deutsche Bank to restructure the bank, and it is possible that Deutsche Bank may eventually acquire a significant stake in Seoul Bank. 8

9 important role in rehabilitating the Korean banking system and boosting the Korean economy. Furthermore, the government incurred massive increases in domestic debt, associated with protecting depositors from losing their saving. After injecting the second 40 trillion Korean won of public funds, restructuring of the banking sector drew to a close. The financial performance of domestic banks improved dramatically. A remarkable increase in foreign participation through green field investment and M&As established a better business environment. This increase in foreign entry into the financial sector through green field investment and M&As after the Asian financial crisis created a high degree of foreign ownership and foreign management control of domestic banks. Of the seven major domestic banks -- Kookmin Bank, Woori Bank, Hana Bank, Shinhan Bank, Korean Exchange Bank, Korean Citi Bank, Standard Chartered First Bank, six are foreign owned, where the foreign holding share exceeds 50 percent in August This increase in foreign ownership of major domestic Korean banks compares to the pre-asian crisis environment, where foreign owners held less than 10 percent of major domestic banks shares. Whether foreign participation takes the form of wholly owned branch banks or an ownership role in Korean domestic banks, foreign financial firms provide substantial competitive pressure. Table 3: Financial Performance of Domestic Banks BIS Capital Ratio Non-Performing Loans to Total Loans ROA ROE Source: Korean Financial Supervisory Commission (unit :%) The improved business environment within Korean banking sector facilitated a remarkable increase in foreign participation through green field investment and M&As. Moreover, this process of foreign penetration led to a high degree of foreign ownership with an increasing stake in and foreign management of domestic banks. Of the seven major domestic 9

10 banks -- Kookmin Bank, Woori Bank, Hana Bank, Shinhan Bank, Korean Exchange Bank, Korean Citi Bank, Standard Chartered First Bank, six are foreign owned banks, where foreign ownership exceeds 50 percent of shares in August This increase in foreign ownership of major domestic Korean banks compares to the pre-asian crisis environment, where foreign owners held less than 10 percent of major domestic banks shares. The foreign participation in banks combines ownership and management participation. The foreign participation in other financial institutions such as life insurance, non-life insurance, and securities companies focuses more on management and business control, rather than a significant ownership stake. Currently, most major banks have adopted the financial holding company system, except for the largest bank, Kookmin Bank. The Korean government introduced the financial holding company system when it passed the Financial Holding Company Law of The goal of the legislation was to expedite restructuring of the financial sector, to promote universal banking and large-sized banks, and to accommodate the globalization of international financial markets. The subsidiary company system ran parallel with the financial holding company system. Due to more inconvenient regulation on the subsidiary company system, the financial holding company system became the main focus in banking sector. The Korean government planned to further expand and open the Korean financial markets and Korean financial institutions to promote the Korean financial markets as a North-East-Asian financial hub. Recently, the Korean government executed several important deregulatory changes. 6 Whether foreign participation takes the form of wholly owned branch banks or a direct ownership role in domestic banks, foreign financial firms provided substantial competitive pressure. This trend will continue. III. Foreign Ownership and Bank Management Thirteen Korean banks existed at the end of 2006, seven nationwide banks (i.e., Shinhan, Woori, SC First, Hana, Korean Exchange, Korea Citi, and Kookmin Banks), and six regional banks (Daegu, Pusan, Jeju, Jeongbuk, Kyungnam, and Gwangju). Also, five government-supported special banks also existed and were nationwide managed. The important M&As include the following: in 2001 Kookmin Bank merged housing and commercial bank and Hanvit Bank 6 These changes included the additional abolition of requiring permission in capital transactions (January 2006), zero-based financial deregulation (the first step in November 2005 and second step in February 2006), deregulation of the asset management businesses (June 2005), and the early execution of the plan for foreign exchange liberalization (moved up to 2009 from 2011). 10

11 merged Peace Bank. Hanvit bank changed its name to Woori Bank, Hana Bank merged Seoul bank in 2002, Citi Bank s Seoul Branch merged Hanmi Bank in 2004, SC Seoul branch merged SC First in 2005 and Shinhan Bank merged Choheung Bank in Although KyungNam and Gwangju Banks merged with Woori BHC on March 2001, their stock still trades independently. The Woori BHC started in April 2001; Shinhan BHC, in September 2001; and Hana BHC, in December Kookmin Bank was listed in the stock market on December 19, 2001 with an IPO of 42,000 won and became fully privatized on December 17, The corporate governance, expansion process, and business styles of each BHC are examined in Keonbeom Lee, Woojin Kim (2005), which provides a policy guideline on how to regulate BHC. Table 4: Board of Directors in Each Bank (2006) Total Proportion of Foreign-holding Resident Outside number of Korean Foreigner Bank director director directors directors directors Foreigner directors Share percentage Kookmin % 82.70% Woori % 9.52% Hana % 80.21% Shinhan % 58.90% KEB % 77.06% SC First % % Citi % % Daegu % 65.72% Pusan % 56.10% Jeonbuk % 28.11% Jeju % Shinhan BHC Gwangju % Woori BHC Kyongnam % Woori BHC Although Korea experienced a huge entry of the foreign capital, its effects on the Korean banking system and on the Korean economy are not fully realized. Using bank data from Bank Management Statistics and Korean macroeconomic data enable us to examine the empirical effects. The related literature includes Taekyu Lee(2006), Kyungseo Park, Eunjeong Lee (2006), Jongkoo Kang, Hyuneui Kim (2006), Hyungchan Jeong, Myeongcheol Lee(2006). Foreign capital favors profitability over stability, thus aiming for more aggressive management. That is, 11

12 we propose and test the following three hypotheses about foreign ownership in the Korean banking industry using panel data analysis. Hypothesis 1. The higher the foreign holding share is, the lower is the ratio of loans to medium and small enterprises to total loans. Also, loans to households increase more than loans to enterprises. Hypothesis 2. A higher share of foreign capital enables Korean banks to achieve business efficiencies by reducing human resources and the number of branches. That is, banks hire fewer employees with efficient local branch management. But the number of top managers remains the same in order to keep overall management continuity. Hypothesis 3. A higher share of foreign capital enables the introduction of advanced marketing strategies that generates a higher ratio of non-interest income to interest income. In addition, the returns from foreign exchange and financial derivatives increase in importance with a higher ratio to non-interest income. We note that Bank Management Statistics classifies banks into nationwide and regional categories. Our focus differentiates banks into ownership classifications of foreign-, mixed-, and domestic-owned banks. As a consequence, we must confine the analysis to only nationwide banks, because of data availability. We assign the foreign dummy variable to 1 for the Korean Exchange Bank since 2003, when the Loanstar, a private equity fund, injected new capital into KEB. We also assign the foreign dummy to 1 for Citibank since 2004, and 1 for SC First since 2005 Table 5a: Test Results for Hypotheses 1-2 Hypothesis 1 Hypothesis 2 small business house Number number of number of loans loans of to total loans to total loans employees executives Branches Constant

13 [2.66] [3.01] [2.16] [0.01] [1.58] Foreign Ownership [-0.73] [2.51] [0.84] [-0.27] [-0.44] unemployment [3.66] [0.38] [0.42] [0.19] [0.04] GDP per capita [-1.33] [-2.18] [-3.43] [0.72] [-1.81] Total Assets [-0.95] [2.95] [14.4] [5.03] [19.93] Foreign bank Dummy [-0.58] [1.68] [-0.18] [0.64] [-0.94] Note: We transform the original data by multiplying ownership by 100, income by 1000, and total assets by 1,000,000. While the main interest goes to the coefficients of foreign ownership variable on other policyoriented dependent variables, the macroeconomic influence and bank size were controlled by including the unemployment rate, GDP per capita and total assets. Other variables were included as robust checks, but the coefficients of the foreign ownership are stable and thus other results are omitted (available upon request) Table 5b: Test Results for Hypothesis 3 net noninterest Foreign foreign risk averserelated derivatives revenues currency currency To net interest And derivatives net revenues to NIR to NIR revenues to NIR to NIR Constant [2.06] [1.45] [1.57] [-0.94] [-1.09] Foreign Ownership [-0.05] [3.79] [3.59] [-2.43] [1.71] Unemployment [0.59] [-0.87] [-1.07] [0.72] [0.84] GDP per capita [-3.49] [-1.38] [-1.47] [0.9] [0.93] Total Assets [2.12] [-1.33] [-2.22] [2.07] [-0.47] Foreign bank Dummy [0.54] [-0.79] [-1.67] [2.16] [0.23] Note: NIR equals net non-interest revenues. Table 5a suggests that the foreign ownership does not significantly affect loans to medium and small enterprises, but does significantly and positively affect household loans. Since the household loans are repaid to loan providers with a higher probability, they represent the secure loans. Thus, banks with a higher foreign holding share prefer a stable loan strategy rather than a more profitable, but higher risk, loans. Also, although foreign investment alters the ownership and management of Korean banks, this foreign investment does not significantly affect the bank infrastructures such as the number of employees, executives, and branches. Thus, 13

14 we conclude that foreign ownership of Korea banks does not affect management decision making, yet. Table 5b indicates that the foreign holding share does not change the bank s management patterns, such as interest-related activities versus non-interest related one, although we anticipated that foreign investment would provide pressures to improve the business. We do find that the higher the foreign holding share is, the higher are the foreign-related activities, and the lower are the derivative-related activities. Thus, strong evidence exists that a higher foreign holding share produces more stable and safer bank management, rather than more aggressive management practices such as derivatives. In summary, the domestic banks in Korea did not experience much pressure from foreign owners to aggressively pursue riskier sources of revenue, but did see safer management strategies and not the more aggressive strategies. This environment provides higher protection to the investors, but it may not improve the efficiency of banks and also welfare of the Korean economy. IV. Foreign Ownership and Profitability of bank management and in stock market This section examines the relationship between the shareholding by foreign investors and profitability. First, using banking industry using panel data analysis, we test the following hypothesis. Hypothesis 4. The higher the foreign holding share is, the higher is bank profitability, using measures, such as ROA and ROE. Table 6: Hypothesis Testing on bank s profitability Hypothesis 4 ROA ROE Constant [-0.95] [-1.64] Foreign Ownership [-0.76] [-2.3] unemployment [0.2] [5.02] 14

15 GDP per capita [2.73] [-2.39] Total Assets [0.84] [0.1] Foreign bank Dummy [0.21] [0.64] Note: We transform the original data by multiplying ownership by 100, income by 1000, and total assets by 1,000,000. The hypothesis test in Table 6 offer the following findings: the foreign holding share does not significantly affect ROA, but does significantly and negatively affects ROE. Since the equity more closely relates to the capital held by stock holders, the ROE more closely captures the stockholders interest, than ROA. Therefore, the profitability of banks was worsened in a way that contrasts the interest of stockholders, including foreign investors. Second, we examine the effect of foreign ownership on the stock prices in each bank in Korea. We omit those banks that merged with other banks or that withdrew their listing on the stock market. Thus, we can fully check the bank performance through the stock market. Those banks announce their bank performance officially through publication and internet announcement. We examine the four major banks for the causality of foreign ownership and the stock return. Thus, rather than emphasizing the commonly used contemporaneous relationships, we employ the Granger non-causality concept with dynamic structure. 15

16 Figure 1. Foreign Ownership and Stock Price for Kookmin Bank 90 KOOKMIN_SHARE KOOKMIN_STOCK KOOKMIN_SHARE_DIF.12 KOOKMIN_RETURN The Kookmin bank, the largest Korean bank by asset size, received much attention from foreign investors. Figure 1 indicates upward trends in both foreign ownership and the stock price. Foreign ownership ranges between 66 and 89 percent, staying between 80 and 85 percent since The stock price of Kookmin bank increased in 2005 from roughly 40,000 Korean Won per share at the beginning of 2005 to 80,000 Korean won at the end of 2005, as shown in Figure 1. 16

17 Figure 2: Summary Statistics for Foreign Ownership and Stock Price in Kookmin Bank Series: KOOKMIN_SHARE Observations 1428 Mean Median Maximum Minimum Std. Dev Skewness Kurtosis Jarque-Bera Probability Series: KOOKMIN_STOCK Observations 1441 Mean Median Maximum Minimum Std. Dev Skewness Kurtosis Jarque-Bera Probability The levels for both the stock price and foreign stock share percentage exhibit trend, while the differenced series do not indicate upward trend. Thus, we incorporate these properties in the unit root tests. That is, the test of the level series includes the linear trend term, and the test of the differenced series includes a linear term. Both the augmented Dickey-Fuller test and the Phillips- Perron test indicate nonstationarity of the level series and stationarity of the differenced series. Table 6: Unit-Root Tests for Kookmin Bank Null Hypothesis: KOOKMIN_SHARE has a unit root Exogenous: Constant, Linear Trend 17

18 t-statistic Prob.* Augmented Dickey-Fuller test statistic Phillips-Perron test statistic Test critical values: 1% level % level % level Note: *MacKinnon (1996) one-sided p-values. ADF Test Lag Length: 2 (Automatic based on SIC, MAXLAG=23) Null Hypothesis: KOOKMIN_STOCK has a unit root Exogenous: Constant, Linear Trend t-statistic Prob.* Augmented Dickey-Fuller test statistic Phillips-Perron test statistic Test critical values: 1% level % level % level Note: *MacKinnon (1996) one-sided p-values. ADF Test Lag Length: 2 (Automatic based on SIC, MAXLAG=23) The Johansen cointegration test does not find any statistical evidence for a long-term relationship between Kookmin s stock price and its foreign holding share. Thus, we find no common factor between the two series. Thus, no long-term equilibrium exists. Table 7: Cointegration Root Tests for Kookmin Bank Included observations: 1425 after adjustments Trend assumption: Linear deterministic trend Series: KOOKMIN_STOCK KOOKMIN_SHARE Lags interval (in first differences): 1 to 2 Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No. of CE(s) Eigenvalue Statistic Critical Value Prob.** None At most

19 Trace test indicates no cointegration at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values The short-run Granger-causality test indicates simultaneous feedback within one to two days between Kookmin Bank s stock price and the foreign holding share, but over three to five days, one-direction Granger causality emerges from the stock price to the foreign holding share. That is, when the stock price changes in the one to two day window, foreign investors respond to the stock price. Thus, the foreign investors follow the market operation as well as domestic investors in Kookmin Bank s stock. But, for a shorter period of one to two days, they actively interact with each other. Table 8: Granger-Causality Tests for Kookmin Bank F-Statistic (p-value) Null Hypothesis: Obs with lag 1 lag 2 lag 3 lag 4 lag5 KOOKMIN_SHARE_DIF does not Granger Cause KOOKMIN_RETURN KOOKMIN_RETURN does not Granger Cause KOOKMIN_SHARE_DIF E E E E E-13 The Shinhan Bank, a major part of the Shinhan Bank Holding Company, exhibits different time-series trend from those of Kookmin Bank. As shown in Figure 3, the foreign holding share stabilized and fluctuated between 55 and 65 percent since 2004, but the stock price continued to increase. Thus, the formal unit-root tests incorporate different assumptions, only a constant term in foreign ownership share and a linear term in the stock price. 19

20 Figure 3. Foreign Ownership and Stock Price for Shinhan Bank 68 SHINHAN_SHARE SHINHAN_STOCK SHINHAN_SHARE_LOGDIF.15 SHINHAN_RETURN Table 9: Unit-Root Tests for Shinhan Bank Null Hypothesis: SHINHAN_STOCK has a unit root Exogenous: Constant, Linear Trend t-statistic Prob.* Augmented Dickey-Fuller test statistic Phillips-Perron test statistic Test critical values: 1% level % level % level *MacKinnon (1996) one-sided p-values. 20

21 Null Hypothesis: SHINHAN_SHARE has a unit root Exogenous: Constant t-statistic Prob.* Augmented Dickey-Fuller test statistic Phillips-Perron test statistic Test critical values: 1% level % level % level No cointegration relationship exists between the foreign ownership share and the stock price, which seems natural since one series does not indicate any obvious trend while the other series includes an upward time trend. Thus, no long run relation exists between the two series for Shinhan Bank. Table 10: Cointegration Tests for Shinhan Bank Included observations: 1423 Series: SHINHAN_SHARE SHINHAN_STOCK Lags interval: 1 to 4 Selected (0.05 level*) Number of Cointegrating Relations by Model Data Trend: None None Linear Linear Quadratic Test Type No Intercept Intercept Intercept Intercept Intercept No Trend No Trend No Trend Trend Trend Trace Max-Eig *Critical values based on MacKinnon-Haug-Michelis (1999) Also, no evidence exists of a short-run relationship between the share price and the foreign ownership share for Shinhan Bank, which is a part of the second largest BHC in Korea. This result may reflect the fact that the bank possesses diversified stockholders. The bank started as a subsidiary bank from the Japan based Korean businessman. Therefore, the shares do not exhibit any specific pattern of holding by ethnic groups. 21

22 Table 11: Granger-Causality Tests for Shinhan Bank Null Hypothesis: Obs F-Statistic (p-value) with lag 1 lag 2 lag 3 lag 4 lag5 ` SHINHAN_RETURN does not Granger Cause SHINHAN_SHARE_DIF SHINHAN_SHARE_DIF does not Granger Cause SHINHAN_RETURN Both the foreign ownership share and the stock price for the Woori Bank exhibit upward time trends, which we incorporate in the unit-root tests. Figure 4. Foreign Ownership and Stock Price for Woori Bank WOORI_SHARE WOORI_STOCK WOORI_SHARE_DIF WOORI_RETURN

23 Table 12: Unit-Root Tests for Woori Bank Table: Unit root tests for Woori bank Null Hypothesis: WOORI_SHARE has a unit root Exogenous: Constant, Linear Trend t-statistic Prob.* Augmented Dickey-Fuller test statistic Phillips-Perron test statistic Test critical values: 1% level % level % level *MacKinnon (1996) one-sided p-values. Null Hypothesis: WOORI_STOCK has a unit root Exogenous: Constant, Linear Trend Lag Length: 0 (Automatic based on SIC, MAXLAG=22) t-statistic Prob.* Augmented Dickey-Fuller test statistic Phillips-Perron test statistic Test critical values: 1% level % level % level *MacKinnon (1996) one-sided p-values. Once again, no long-term relationship exists between the two variables, since we do not find cointegration under any lag structures. Table 13: Cointegration Tests for Woori Bank Lags interval (in first differences): 1 to 1 Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No. of CE(s) Eigenvalue Statistic Critical Value Prob.** None

24 At most Trace test indicates no cointegration at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values Lags interval (in first differences): 1 to 2 Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No. of CE(s) Eigenvalue Statistic Critical Value Prob.** None At most Trace test indicates no cointegration at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values Lags interval (in first differences): 1 to 3 Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No. of CE(s) Eigenvalue Statistic Critical Value Prob.** None At most Trace test indicates no cointegration at the 0.05 level * denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values Now, we consider the short-run causality. We reject the non-granger causality only for one direction. Woori Bank s stock price leads the Woori Bank s foreign holding share, only in the short term (one to two days). The causality test proves statistically significant within models having one lag at the 5-percent level and with two lags at the 10-percent level. 24

25 Table 14: Granger-Causality Tests for Woori Bank Null Hypothesis: Obs F-Statistic (p-value) with lag 1 lag 2 lag 3 lag 4 lag5 WOORI_RETURN does not Granger Cause WOORI_SHARE_DIF WOORI_SHARE_DIF does not Granger Cause WOORI_RETURN The stock price for the Hana BHC does not show any trend, so we perform the unit-root test with a constant term. This observation differs from three previous banks experiences. Interestingly, both the augmented Dickey-Fuller test, with the optimal lag selected by the Bayesian information criteria, and the Phillips-Perron test statistically rejects the null hypothesis of nonstationarity. But, we cannot reject the null hypothesis for the foreign ownership share in the Hana bank. That is, the long-term relation proves impossible in theory. Thus, we consider, instead, short term Granger noncausality. With any lag structures, we find no causality in each direction at the 5-percent significance level. 25

26 Figure 5. Foreign Ownership and Stock Price for Hana Bank 82 HANA_SHARE HANA_STOCK M01 06M04 06M07 06M10 07M01 07M04 07M M01 06M04 06M07 06M10 07M01 07M04 07M07 8 HANA_SHARE_DIF.08 HANA_RETURN M01 06M04 06M07 06M10 07M01 07M04 07M M01 06M04 06M07 06M10 07M01 07M04 07M07 Table 15: Unit-Root Tests for Hana Bank Null Hypothesis: HANA_STOCK has a unit root Exogenous: Constant Adj. t-stat Prob.* Augmented Dickey-Fuller test statistic Phillips-Perron test statistic Test critical values: 1% level % level % level *MacKinnon (1996) one-sided p-values. 26

27 Null Hypothesis: HANA_SHARE has a unit root Exogenous: Constant t-statistic Prob.* Augmented Dickey-Fuller test statistic Phillips-Perron test statistic Test critical values: 1% level % level % level *MacKinnon (1996) one-sided p-values. Table 16: Granger-Causality Tests for Hana Bank Null Hypothesis: Obs F-Statistic (p-value) with lag 1 lag 2 lag 3 lag 4 lag5 HANA_STOCK does not Granger Cause HANA_SHARE_DIF HANA_SHARE_DIF does not Granger Cause HANA_STOCK In summary, no long-run trend relationship exists between the stock price and the foreign ownership share for the four banks considered in this section. That is, the stock markets for the shares of these four banks prove efficient with respect to the foreign holding share, implying that no one, foreigners or Koreans, possesses superior information about the market value of these banks. Further, limited evidence exists for two of the banks Kookmin and Woori Banks whereby stock price movements lead changes in the foreign ownership share. In other words, limited evidence exists that foreign investors respond to stock price movements, and not the reverse. The exception occurs for Kookmin bank at the short one-to-two-day window where twoway causality exists between the stock price and the foreign holding share Conclusion and policy suggestions 7 The time series analysis using dynamic structure indicates some possible causalities between the foreign ownership and profitability under limited cases, while the panel data analysis focusing with contemporaneous relationship provides a negative relation between ownership and ROE. These are not necessary conflicted in the policy implications. 27

28 One of the most striking structural changes in Korean banking sector is a remarkable increase in foreign ownership after 1997 financial crisis. The mode of foreign participation changed from branch opening for its own country s corporate financing to Greenfield, M&A, and portfolio investment. As the Korean government eased the long standing legal restrictions on the foreign ownership of domestic banks, it served as a momentum to enlarge the foreign ownership. The Korean government wanted to attract foreign capital into the Korean financial sector to facilitate restructuring in the financial sector right after the financial crisis. The entry of foreign equity, however, has been expected to strengthen the financial and domestic banking systems, since then. As a result, the increase in foreign entry into the financial sector through green field investment and M&A after the Asian financial crisis created a high degree of foreign ownership and foreign management control of domestic banks. Six out of seven major domestic banks are foreign owned, where the foreigner share ratios are around 80 percent currently. In this paper we analyze the economic effect of foreign ownership on Korean banking industry and stock exchange. The effects of foreign ownership on banking industry are resolved into four parts, such as loan market behavior, profitability, management efficiency, transmission of advanced financial techniques. We use annual banking panel data from 2001 to 2006 to analyze whether loans to corporate increase. Increases in foreign bank ownership are not directly related to increase in large sized firms. There is a positive empirical results that loan and reduction of small and medium sized firms loans. Consumer loans increases as the level of foreign ownership increases. It is not statistically significant that increase in foreign ownership affect the ROA positively. There is, however, statistically significant negative relation between increasing foreign ownership and ROE. The increase in foreign ownership does not show statistically significant cost saving results, like layoff and shutdown of branches. Non interest profits are expected to increase as the foreign ownership increase. We can not find any statistically significant relationship. We do find that the higher the foreign holding share is, the higher are the foreign-related activities, and the lower are the derivative-related activities. The foreign investors prefer a more stable and safer bank management particularly with foreign exchange control, rather than more aggressive management practices such as derivatives. Thus, our expectation on adopting the advanced banking activities has been fully achieved. 28

29 The four major Korean banks who are active in portfolio investment are considered to examine the entry effect of foreign capital, including Shinhan, Woori, Hana BHC and Kookmin bank. Their stock shares is efficient with respect to the foreign holding share, implying that no one, foreigners or Koreans, possesses superior information about the market value of these banks. It is true particularly for Shinhan and Hana BHC since there is no single dominant group holding their stocks. Further, limited evidence exists for Woori Bank where foreign investors respond to stock price movements, and not the reverse. The exception occurs for Kookmin bank at the short one-to-two-day window, indicating contemporaneous causalities between the stock return and the foreign holding share. Higher foreign participation in the Korean banking sector induces some cost efficiency gains and service improvements for domestic banks, presumably by intensifying competitive pressures. This higher foreign participation, however, also evoked criticism about the public good characteristic of the banking industry and the extension of loans to the corporate sector, especially reductions in lending to medium and small businesses. Questions also emerge on whether foreign ownership of domestic banks may contribute to greater soundness of the Korean banking system. In general, the optimal level of foreign ownership of domestic banks depends on the profitability of the bank industry. Absent entry and exit barriers, then participation of foreign capital will continue to increase, as long as the Korean banking industry continues to exhibit sustainable growth. References: 강종구, 김현의 (2006), 외국금융기관의진입이국내은행산업에미친영향 한국은행금융경제연구원 working paper. 김우진 (2005) : 외국자본진출증가에따른이슈및과제, 한국금융연구원주간금융브리프, Vol.14, No.35, pp.3~7 김우진, 이건범 (2002) : 금융지주회사의설립과운영, 한국금융연구원금융조사보고서, Vol.2002, No.6, pp.1~105 김상환 (2001) : 은행위기이론과시사점, 한국금융연구원 KIF 은행경영브리프 : 은행경영일반, 기타보고서, Vol.2002, No.0, pp.3~7 29

30 김희경 (2006) : The Effect of Foreign Capital Entry on the Business Performance of Korean Banks, 국제통상학회한국국제통상연구, Vol.11, No.2, pp.169~190 민상기, 강신애, 설원식 (2007) : Foreign Bank Entry and the Change of Performance after Asian Financial Crisis, 국제지역학회국제지역연구, Vol.11, No.1, pp.308~328 박경서, 이은정 (2006) : The Role of Foreign Investors on the Management and Corporate Governance of Korean Companies, 한국금융연구원금융연구, Vol.20, No.2, pp.73~113, 외국인투자자가한국기업의경영및지배구조에미치는영향, (in Korea) 손상호 (2005) : Characteristics of Financial Businesses and the Prospects for Universal Banking in Korea, 한국금융연구원금융조사보고서 No.11, pp. 1~174 윤창현 (2005) : 외국자본유입에따른문제점과과제, 한국경제학회한국경제학회소식, 81, 9~10. 이건범 (2005) : 금융지주회사제도개선방안, 한국금융연구원, 주간금융브리프, 14 (39), 3~9. 이건범, 김우진 (2005), 금융지주회시의효율적운용을위한제도개선방안의연구, 한국금융연구원정책보고서, Issues on Financial Holding Company Regulation. 이병윤 (2004) : 외국자본의국내은행산업진출현황및과제, 한국금융연구원주간금융동향, Vol.13, No.2, pp.2~7 이병윤 (2005) : Effects of the Increased Presence of Foreign Capital in Korea, and Related Policies - Effects on the Dividends and Investments, and Hostile Takeover, 한국금융연구원금융조사보고서, Vol.2005, No.14, pp.1~81 이태규 (2006), 은행민영화와소유규제개선, 금융연구 20, Bank Privatization and Improving Bank Ownership Regulation. 임성학 (2005) : Foreign Capital Entry in the Domestic Banking Market of Korea: Bitter Medicine or Poison, 한국정치학회한국정치학회보, Vol.39, No.4, pp.189~ 210 정형찬, 이명철 (2006) : Foreign Acquisitions of Korean Commercial Banks and Restructuring of Korean Banking Industry: The Case of Korea First Bank, 한국금융학회, 금융학회지, Vol.11, No.1, pp.81~128 외국자본의국내은행인수와은행산업의구조변화 : 뉴브리지캐피털의 제일은행인수사례, (in Korea) 최재헌, 김원경 (2002) : The Change of Financial Space According to the Investment of the Foreign Financial Capital in Korea, 한국도시지리학회지제 5 권 1 호, pp.49~63 한상일 (2001) : 은행지주회사설립의경제적유인에대한고찰, 한국금융연구원주간금융동향, Vol.10, No.45, pp.2~5 30