Basle Committee on Banking Supervision, 1997, Core Principles for Effective Banking Supervision, Basle: Bank for International Settlements.
Bernanke, Ben, Gertler, M. and Gilchrist, S., 1996, “The Financial Accelerator and the Flight to Quality,” The Review of Economic and Statistics, LXXVIII, pp. 1-15.
Chan-Lau, Jorge, and Chen, Z., 1998, “Financial Crisis and Credit Crunch as a Result of Inefficient Financial Intermediation - with reference to the Asian Financial Crisis,” IMF Working Paper 98/127.
Demigurc-Kunt and E. Detragiache , 1998, “The Determinants of Banking Crisis in Developing and Developed Countries,” IMF Staff Papers, 45, 81-109.
Garcia Herrero, Alicia , 1997, “Monetary Impact of a Banking Crisis and the Conduct of Monetary Policy,” IMF Working Paper 97/124.
Kaminsky, G. and C. Reinhart , 1996, “The Twin Crises: The Causes of Banking and Balance of Payments Problems,” International Finance Discussion Paper 544 (Washington, Federal Reserve Board).
Kihwan and Leipziger , 1997, “Korea: A Case of Government-Led Development,” in Leipziger, D. ed. “Lesson From East Asia,” Ann Arbor: The University of Michigan Press.
Kochhar, Kalpana, Loungani, P. and Stone, M. 1998, “The East Asian Crisis: Macroeconomic Developments and Policy Lessons,” IMF Working Paper 98/128.
Wade, R and F. Veneroso , 1998, “The Asian Financial Crisis: The High Debt Model and the Recognized Risk of the IMF Strategy,” IMF Working Paper 128, Russell Sage Foundation.
We would like to thank Bijan Aghevli, Robert Deckle, Sharmini Coorey, James Gordon, Peter Hayward, Chen-Hong Lim, David Woo, Leslie Teo, and Maxwell Watson for their input and comments. Kiran Sastry provided excellent research assistance.
In 1970, the trust business was assigned exclusively to Korea Trust Bank, which merged with Seoul Bank in 1976. By end-1995, all deposit money banks and development institutions with the exception of Export-Import Bank of Korea and foreign banks were allowed to engage in trust businesses.
In principle, trust account are operated on the client’s own account and not counted when computing a bank’s capital adequacy ratio. In practice, however, a large segment of trust accounts in Korea are economically like deposits, for the bank guarantees both the principal and a predetermined yield. The deposit guarantee issued by the government in 1997 is regarded as covering trust accounts.
There are four specialized banks—the Industrial Bank of Korea and three banks centered on agricultural, fisheries, and livestock cooperatives. Development banks comprise the Korean Development Bank (KDB), Korea Export-Import Bank (KEXIM), and the Long-term Credit Bank.
Although merchant banks were, in principle, supervised by the MOFE, in practice, supervision was minimal; there were no asset classification, capital, or provisioning rules. Merchant banks are, since April 1998, supervised by the Financial Supervisory Commission (FSC).
Credit unions and mutual savings and finance institutions are part of savings and insurance companies. They have a long history, mainly receiving deposits and providing finance for their members. The aggregate assets of these three sectors account for about 5 percent of the total financial market.
Although an accurate measurement of the size of the curb market is difficult, OECD (1994) estimates suggest that in the mid-1990s, total lending in the curb market was between 2 to 5 percent of the total loans of the formal financial sector. In contrast, in the mid-1970s, the curb market was estimated to account for more that one-third of all credit extended in the economy.
Korea’s private savings exceeded 25 percent of GDP for much of the 1980s and early 1990s.
The National Investment Fund (NIF) was created in 1974 for this purpose, and was funded by the compulsory deposit of savings from pensions, savings and postal savings accounts, and by other purchasers of NIF bonds, such as life insurers.
Commercial banks were allowed to discount bills at below-market rates at the BOK’s rediscount facility which had separate windows corresponding to various government priorities (e.g., for loans to import machinery and develop heavy industry).
These were iron and steel, petrochemicals, nonferrous metal, textiles, machinery, electronics, and shipbuilding.
The troubled corporation’s main bank typically assessed the health of the company and the Office of Bank Supervision coordinated the terms of the financial support agreed between the main bank and the company. The MOFE approved the restructuring agreement. All measures related to this exercise have expired (Gobat (1998)).
The preparation for accession to the OECD played a crucial role in accelerating the reform process.
Nevertheless, shareholders remained but one party in the choice of bank authorities. In 1993, the Bank of Korea’s Office of Bank Supervision established a procedure whereby the chairman of a commercial bank would be selected by a committee consisting of representatives of shareholders, corporate clients, general customers, and ex-bank presidents. Once chosen by the committee, the appointment of the chairman needed to be ratified by a general shareholder meeting.
Previously, the BOK had extended loans automatically in amounts corresponding to the proportion in which banks had provided loans to SMEs and the export sector. Under the new system, the BOK set individual credit ceilings for each bank, within an aggregate ceiling for the entire banking system.
Policy loans are estimated to have declined from 60 percent of total bank loans in 1979 to 3 percent by end-1997.
Access to foreign loans was restricted to selected public enterprises, and, on a limited basis, to companies with foreign ownership. In mid-1997, the range of approved borrowers was expanded to include all enterprises, provided the borrowing was used to finance imports related to export activities.
For example, practices such as consolidated accounting and marking to market were absent in Korea.
In fact, in many cases only bad loans were reported as nonperforming, giving an even more misleading image of the soundness of the Korean banking system. BIS (1997) reports that nonperforming loans as a percentage of total loans are only 0.8 percent in 1996, the figure corresponding to bad loans; however, the “real” figure, according to the Korean definition of NPLs, was 4.1 percent (see Table 2)
In fact, supervisors sometimes waived the full application of regulations, such as provisioning rates, to avoid weakening the earnings reports of banks.
The debt ratio of most chaebols exceeded 400 percent during the 1990s, compared to an average of 150 percent in the US, 210 percent in Japan and 90 percent in Taiwan. In 1997, 15 percent of the top 30 chaebols’ affiliates had leverage ratios in excess of 500 percent. In addition, low profits decreased the ability to service this debt, and the operating cashflow as a percentage of interest payments was only 80 percent in 1996.
At end-1997, the 30 largest chaebols accounted for half, and the 5 largest chaebols for one-third of the corporate debt outstanding. The top 30 chaebols were responsible for about 30 percent, and the top 5 chaebols for about 18 percent, of commercial bank loans at end- 1997.
In particular merchant banks which, despite the increasing riskiness of their operations, decreased their provisions for loan losses during the period (see Table 2).
Bail-outs began in the 1969-70, when some of the firms whose foreign borrowing had been guaranteed by the government were in financial difficulties. In order to induce another business group to get involved, the government usually offered incentives in the form of new loans to be used as “seed money” and special tax treatment. This has been described as the “implicit socialized or shared risk between government and business on borrowed funds” (Kiwhan and Leipziger, 1997).
In mid-1997, IBCA awarded a legal rating of 2 (next to highest) to 11 of the 16 major commercial banks, explicitly considering these banks as “too big to fail.”
Because of this market share strategy, chaebols’ net profit margins had been declining since the late 1980s, a trend interrupted only by occasional peaks in the price of memory chips, Korea’s most profitable export. As an example, in 1996, the average return on equity of Korean chaebols was 2.5 percent.
Bank credit to the private sector increased during 1990-97 at an annual real rate of 12 percent, compared to 0.5 percent in the US and 4 percent in G-10 Europe (see BIS 1998)
One of the reasons frequently suggested for this reliance in bank finance is the underdevelopment of the Korean stock market. In fact, in 1996, stock market capitalization was only 25 percent of GDP, compared with 67 percent in Japan, 108 percent in the US, 151 in the UK or 280 percent in Hong Kong (see BIS 1997).
Foreign borrowing was obtained at increasingly low spreads, despite the increasing economic imbalances (see Figure 2). See McKinnon and Pil (1996), Corseti, Pesenti and Roubini (1998), Krugman (1998) and Levy Yeyati (1998) for a theoretical discussion of the causes of the “overborrowing” and “overlending” syndromes in the context of the recent crisis.
Foreign debt of domestic financial institutions, including overseas subsidiaries and foreign branches, increased from US$ 40 billion at end-1993 to US$106 billion by end-September 1997. At end-September, 1997, short-term liabilities accounted for about 60 percent of the foreign debt of domestic financial institutions.
Regulations limited international issuance of securities to corporations with international credit rating of BBB or higher, and to obtain loans at spreads higher than 100 over LIBOR. These regulations heightened the role of the major Korean banks (whose ratings benefited from implicit government support) as the conduits of external finance to domestic corporations.
The KDB issue, for example, was trading at end-September at 80 basis points spread over US treasuries for three year maturity, 100 basis points for five-year and 110 basis points for ten-year maturity (see Figure 2).
International interbank lending to Korea soared, accelerating from an annual growth rate of about $15 billion in 1994 to about US$25 billion in 1995-96, and amounting to US$ 108.5 billion at end-1996 (see BIS 1997). Of this amount, about 70 percent had a maturity of less than a year. One of the reasons behind this boost was the more favorable capital ratio requirement associated with the country’s entry into the OECD, which reduced the risk-weight for loans to Korean banks from 100 to 20 percent, thus raising international banks’ returns on risk-adjusted capital and lowering spreads for loans to Korean banks.
Because of already high returns available domestically, Korean banks, particularly merchant banks, ventured abroad to invest heavily in high-risk high-yield instruments in emerging markets. For example, Korean banks are reported to have accounted for about 40 percent of the debut Eurobond issue of the Russian Republic in 1996. They also invested extensively in the region, especially in Thailand and Indonesia.
At end-June 1997, accumulated loan loss provision accounted only for 83 percent of expected losses, and provisions against securities losses accounted only for 37 percent of expected losses. These ratios were in compliance with OBS rules.
Nonperforming loans are normally a lagging indicator of the soundness of the banking sector, and more so when loans are only classified as non performing after having been in arrears six months, rather than the usual three months. A more contemporaneous indicator is the ratio of dishonored bills and checks, which more than doubled during the same period and increased fivefold in the last quarter of 1997 (see Figure 3)
One measure of balance sheet deterioration is the shortfall in capital adequacy represented by the amount of funding needed to bring a bank’s ratio of capital to risk weighted assets to the minimum of 8 percent recommended by the Bank of International Settlements. Estimates based on end-September 1997 balance sheet data showed, under Korean provisioning and loan classification rules, a shortfall of some W 11.3 trillion (3.0 percent of 1997 GDP) for commercial banks, merchant banks, development and specialized banks.
This was decided in November 1997, as part of an emergency package prepared by the Korean authorities but which failed to restore market confidence and had to be substantially strengthened the next month.
The resolution of distressed financial institutions through closure, merger, or the injection of government resources was facilitated by legislation passed in December 1997. However, the legislation fell short with respect to rules by which the supervisory authorities can write down the equity of failed banks. While the amended legislation authorizes the write-down of existing shareholder equity, it does not permit a reduction in capital below W 100 billion for commercial banks and W 50 billion for merchant banks, even if the institutions have zero or negative net worth. The authorities corrected this by reforming the legislation in August, 1998.
The number of bank employees decreased by 34 percent as of end-1998, compared to end-1997, and the number of branches decreased by 17 percent during the same period.
The five closed banks had end-March 1998 capital adequacy ratios between -4 and -11 percent and a negative net worth totaling W920 billion, and represented 7 percent of total assets of the banking sector.
One small regional bank was given conditional approval despite receiving a negative evaluation from the Evaluation Committee because the Korean regulation does not allow for the closure of a bank with positive net worth. The authorities amended this legislation in August, 1998.
The bridge bank (called Hanaerum Merchant Bank) was established at the end of December 1997, financed by the KDIC. In January, it took over the deposits of the suspended merchant banks along with most of their performing assets. After a due diligence process, the value of assets and liabilities transferred is W8.7 trillion and W12.1 trillion respectively. Shortly upon intervention, depositors were offered cash reimbursement, with households being compensated first, followed by enterprises and then other financial institutions. As of end-June 1998, 95 percent of private and institutional depositors had been repaid, as well as all financial institutions call money deposits, for a total amount of W8.2 trillion. A further W4.2 trillion remains to be repaid, mainly deposits of financial institutions. Loans have been rolled-over at market rates until a strategy for the disposal of assets is devised. The two suspended merchant banks that reopened will repay the bridge bank the deposits repaid to their customers by January 1999.
The assumption was that 20 percent of precautionary loans, 40 percent of substandard loans and 100 percent of doubtful loans will become a loss.
These number do not include loans that were already sold to KAMCO in December 1997. If this is taken into account, nonperforming loans of commercial banks would be about 10.7 percent of total loans, or 8.5 percent of 1997 GDP.
This figure was perceived by the market as a gross underestimate of the true amount of nonperforming loans. For example, the Korea Institute of Finance, a government think tank, estimated the amount of nonperforming loans of commercial banks as of June 1998 at about 18 percent of total loans, or 13.6 percent of GDP.
More pessimistic unofficial estimates point to a final amount of public funds reaching about W140 trillion, or 33 percent of 1997 GDP.
The assumption for the calculation of the interest cost is that bond issues are spread evenly over the second half of 1998 and 1999 and the annual interest rate is 16 percent.
The initial maturity of BOK’s support was three months, but it has been rolled-over because the Bridge Bank has no funding to repay commercial banks their claim on the closed merchant banks, and has not been fully repaid. BOK charges commercial banks the average call rate minus 100 basis points, but commercial banks receive interest payments from KDIC at the average call rate plus 50 basis points. Hence, this liquidity support has an element of subsidy to commercial banks amounting to 150 basis points.
The BOK made a profit with these interventions. Considering the opportunity cost of funds to be the yield on MSBs for domestic currency and LIBOR for foreign currency, the BOK made an accumulated profit of about W6 trillion in the period December 1997-June 1998.
The agreement consisted in the exchange of short term debt into government guaranteed transferable loan certificates maturing in one, two and three years at floating interest rates of 2.25, 2.25 and 2.75 over six month LIBOR respectively. The agreement includes a plan to enable Korean banks to renegotiate the loans at lower rates if the country’s international credit ratings improve, and a call option on loans with extended maturities of two and three years exercisable on July 1998.
The law does not apply to specialized and development banks. However, the responsibility for examining these institutions has been delegated by the MOFE to the FSC, although the legal responsibility will be retained by the MOFE.
In particular, all trust accounts with guarantee will be regarded as on-balance sheet for supervisory and accounting purposes. For capital adequacy calculations, assets in such accounts will be weighted at 50 percent as of January 1999 and fully as of January 2000. Rules will be introduced to provide full disclosure of trust beneficiaries and preclude any possibility of payment by managing banks to make good or guarantee any loss, and to ensure segregation of management and accounting.
Under the current arrangement, the supervision of foreign exchange activities is under the jurisdiction of MOFE, with some aspects delegated to the BOK. The government will introduce amendments to the Foreign Exchange Act to entrust all responsibility for the setting of foreign exchange limits and the supervision of foreign exchange risk to the FSC, with information provided regularly by the BOK.
This will require banks to report maturity mismatches for different time buckets (sight to 7 days, 7 days to 1 month, 1 to 3 months, 3 to 6 months, 6 months to 1 year and over 1 year), and maintain positive mismatches for the first period. From sight to 1 month, any negative mismatch should not exceed 10 percent of total foreign currency assets, and from sight to 3 months it should not exceed 20 percent
This measure is expected to induce some increase in turnover in the foreign exchange market, one of the lowest among developed countries, and enhance the pricing of the forward market so as to reflect interest rate differentials. It will also encourage higher swap market activity, so as to provide a closer link between won and dollar money markets.
Excesses in aggregate exposures and connected lending will be published regularly.
Korea followed a model similar to the “London Approach” to corporate restructuring. A Corporate Restructuring Agreement (CRA) has been signed by 200 financial institutions, under which financial institutions agree to follow specific procedures for debt workouts and to subject themselves to binding arbitration by a private agency specially set up for the purpose, the Corporate Restructuring Coordinating Committee (CRCC). These procedures include the creation of creditor committees to deal with the restructuring of individual corporations or conglomerates. Lead banks or groups of institutions holding more then 25 percent of a corporation’s debt can call a creditor’s committee meeting. An automatic standstill on debt payments applies while the committee negotiates. Upon agreement upon banks, the Lead Bank will negotiate with the debtor corporation. In all of the cases, arbitration by the CRCC will solve bottlenecks in the negotiations.
See Kochhar, Loungani and Stone (1998) for a review of macroeconomic developments in east Asia during the first year of the crisis.
This phenomenon, namely the amplification of an adverse shock by worsening credit market conditions, has been referred to in the literature as the “financial accelerator”, see Bernanke, Gertler and Gilchirts (1996)
The credit crunch dimension of the crisis spurred the development of new theoretical models of financial crises (see, among others, Caballero and Krishnamurthy (1998) and Chan-Lau and Chen (1998)) where inefficient financial intermediation and lack of adequate international collateral play a fundamental role in understanding the developments that follow the unraveling of the crisis.
These figures, however, must be interpreted with caution because of the impact of KAMCO purchases and debt write-offs on banks’ portfolios, especially during the periods December 1997-January 1998 and September-October 1998.
Korean commercial banks posted losses of about W4 trillion in 1997 and W67 trillion (about 3 times bank’s total end-1997 capital) in the first half of 1998.
The spreads of call interest rates between strong and weak banks reached 9 percentage points in late December 1997- early January 1998.
Part of the difficulties of SMEs in obtaining credit arose from the closure of 16 of the 30 merchant banks, which had been an important source of credit for them. In addition, banks extended an important amount of rescue loans to big conglomerates (about 15 percent of new lending in the first quarter of 1998).
The modification was as follows: If a financial institution defaults before end-2000, the depositors with deposits of more than W20 million will be paid the amount of the principal, whereas depositors with deposits of W20 million or less will receive the principal plus a payment calculated on the basis of market interest rates (rather than the deposit’s contractual interest rate).
Credit to banks were a major expansionary source in December 1997 and January 1998, at the height of the crisis.