Privatesector P U B L I C P O L I C Y F O R T H E Note No. 172 January 1999 Corporate Distress in East Asia The effect of currency and interest rate shocks Stijn Claessens, This Note reports the results of a study that quantified the effect of currency and interest rate Simeon Djankov, shocks in East Asia on the liquidity and solvency of nonfinancial corporations in the region. In a and Giovanni Ferri survey of the five countries most affected by the East Asian financial crisis—Indonesia, the Republic of Korea, Malaysia, the Philippines, and Thailand—the study found that 63 percent of firms are illiquid (with earnings less than their debt service) and 31 percent technically insolvent (with financial obligations exceeding their equity).1 Among solvent firms, about half are illiquid and at risk of insolvency unless their liquidity constraints are relieved. Worst affected is Indonesia, with 77 percent of firms illiquid (table 1). Malaysia has the fewest insolvent firms— 5 percent—but even so, 63 percent of firms are illiquid. Nonperforming loans in the five countries could amount to 7 to 30 percent of GDP, and the share of nonperforming assets could be as high as 64 percent of GDP in Korea, and more than a third of the combined GDP in these countries. The study also sheds light on the high interest rate debate for each country. There is a clear tradeoff: higher interest rates will damage profits but could help to stabilize the currency, thus limiting the burden of external liabilities. TABLE 1 CORPORATE AND ECONOMIC DISTRESS IN EAST ASIA, SEPTEMBER 1998 Illiquid Insolvent Nonperforming Nonperforming firms firms loans assets Country (percent) (percent) (percentage of GDP) (percentage of GDP) Indonesia 77 65 29.9 39.4 Korea, Rep. of 50 41 24.8 64.2 Malaysia 63 5 14.4 25.1 Philippines 57 16 7.1 10.6 Thailand 65 23 27.9 41.2 Average 63 31 20.8 36.1 Source: World Bank staff estimates. The World Bank Group ▪ Finance, Private Sector, and Infrastructure Network 2 Corporate Distress in East Asia Precrisis conditions relatively low, and while leverage in the Philip- pines was rising, it was still much lower than in The riskiness inherent in the liability structures Indonesia and Thailand. Most countries saw of East Asian corporations is evident in the av- some increase in leverage in the past few years, erage debt to equity ratios in the period before most notably Korea, Malaysia, and Thailand. The the crisis, especially when compared with those rise in leverage in the Philippines is probably in two industrial countries, Germany and the the result of its reforms in the mid-1980s, which United States (figure 1).2 Another comparator is revived the corporate sector and improved fi- Taiwan (China), included in the study as a nancing possibilities. benchmark because it has been less affected by the crisis. High external financing, mostly from Short-term debt was high as a share of total debt bank lending, has been a characteristic of East during the entire period in all five East Asian Asian corporations. Korean firms had the high- countries (figure 2). Malaysia and Thailand stand est debt to equity ratios in 1988–96, about five out with shares of about 66 percent; the others times as high as those of Taiwanese firms, which have shares of about 58 percent. Despite all the had the lowest. Leverage in Malaysia was also attention to the role of short-term debt in the FIGURE 1 CORPORATE LEVERAGE, 1988–96 Average debt to equity ratio 4.0 3.5 Republic of Korea 3.0 2.5 Indonesia 2.0 Thailand 1.5 Germany Philippines 1.0 United States Malaysia Taiwan (China) 0.5 0 1988 1989 1990 1991 1992 1993 1994 1995 1996 Source: World Bank staff estimates. The World Bank Group 3 East Asian financial crisis, the data do not sug- debt varied greatly across East Asian econo- gest a massive buildup in short-term debt in the mies in 1996 (figure 3). Korean firms had the region in the past few years. Instead, they indi- largest share of total foreign exchange debt, cate a gradual increase over a longer period, with followed by Malaysian firms. Malaysian firms even some decline in the past several years. Nev- had the largest share of foreign short-term debt, ertheless, these ratios are much higher than those followed by Korean and Thai firms. And firms in many industrial countries. In the United States, in Taiwan (China) and the Philippines had the for example, the share of short-term debt is about largest share of domestic long-term debt. 25 percent, and in Germany it is about 45 percent. Data also suggest large differences across coun- These short-term debt shares do not distinguish tries before the crisis in the share of earnings foreign exchange liabilities from domestic debt, absorbed by interest payments. Taiwanese and in some countries and times the debt com- corporations needed to devote only a small share position may have included more short-term of earnings (before interest and taxes) to inter- foreign exchange debt than in others. Data on est payments, about 14 percent (figure 4). Thai debt composition shows that the structure of and Korean corporations had the highest share FIGURE 2 CORPORATE SHORT-TERM DEBT, 1988–96 Short-term debt as a percentage of total debt 80 70 Malaysia Taiwan (China) 60 Thailand Indonesia Republic of Korea Philippines 50 Germany 40 30 United States 20 0 1988 1989 1990 1991 1992 1993 1994 1995 1996 Source: World Bank staff estimates. 4 Corporate Distress in East Asia of interest expenses, about 37 percent and 39 that firms were very vulnerable to a rise in inter- percent. (Indonesian, Malaysian, and Philippine est rates (or a drop in earnings). corporations averaged about 25 percent.) These ratios are much higher than those typical in in- Currency and interest rate effects dustrial countries. In Germany, for example, interest payments in 1996 represented only 26 The study assessed both the combined and sepa- percent of earnings (before interest and taxes); rate effects of the currency and interest rate for Japanese firms interest payments averaged shocks in the second half of 1997 and in 1998 only 10 percent. The high share for interest pay- on East Asian firms’ liquidity and solvency. The ments in the five East Asian countries meant pros and cons of a high interest rate policy have FIGURE 3 CORPORATE DEBT COMPOSITION, 1996 Percent 100 Domestic long term 80 Domestic short term Foreign 60 long term Foreign 40 short term 20 0 Indonesia Republic of Malaysia Philippines Thailand Taiwan Korea (China) Source: World Bank staff estimates. FIGURE 4 INTEREST COVERAGE, 1996 Interest payments as a percentage of earnings before interest and taxes 40 35 30 25 20 15 10 5 0 Indonesia Republic Malaysia Philippines Thailand Germany Japan Taiwan of Korea (China) Source: World Bank staff estimates. The World Bank Group 5 been much debated in the analysis of the East Singapore) and the onshore three-month inter- Asian financial crisis. One aspect of that debate bank rates for the currency. The change in this is the effect on corporate balance sheets. Here spread indicates how marginal lending rates have a clear tradeoff exists: higher domestic interest changed (as a result of changes in the opportu- rates will damage corporations’ profitability, but nity costs of funds for banks). The study also could help in stabilizing the country’s currency, revised upward the interest rate paid on foreign thus helping corporations by limiting the burden currency debt by a factor proportional to the of high debt service on their external liabilities. increase in the spread between the rate on U.S. The analysis shows the degree of tradeoff be- dollar–denominated sovereign bonds issued by tween these two effects for each country. the country and the rate on U.S. Treasury bonds of similar maturity. For each country it calculated the average ex- change rate for September 1–13, 1998, and com- The study computed the effect of the shocks in pared that with the average exchange rate in terms of corporate financial obligations at the March 1997, taken as the precrisis value. For end of 1996 using 1996 balance sheet figures the interest rate shock, the study compared the and profit and loss statements. It defined firms average domestic currency bank lending rate as technically insolvent when the increase in over the first half of 1998 with the average lend- financial obligations calculated at new exchange ing rate in the first half of 1997. To adjust for and interest rates exceeded equity at the end of the fact that the bank lending rates used are 1996. And it defined firms as illiquid when 1996 average rates, not marginal rates on new loans, earnings (before interest and taxes) fell short of it added to the postcrisis lending rate the change debt service obligations projected at new ex- in the spread between the offshore (quoted in change and interest rates. It used three scenarios: FIGURE 5 INSOLVENT FIRMS AFTER THE CURRENCY AND INTEREST RATE SHOCKS Percentage of total 80 Exchange and interest rate shocks Exchange rate shock 70 Interest rate shock 60 50 40 30 20 10 0 Indonesia Republic Thailand Philippines Malaysia All five Taiwan of Korea countries (China) Source: World Bank staff estimates. 6 Corporate Distress in East Asia FIGURE 6 ILLIQUID FIRMS AFTER THE CURRENCY AND INTEREST RATE SHOCKS Percentage of total 80 Exchange and interest rate shocks Exchange rate shock Interest rate shock 70 60 50 40 30 20 10 0 Indonesia Republic Malaysia Philippines Thailand All five Taiwan of Korea countries (China) Source: World Bank staff estimates. FIGURE 7 ILLIQUID AND INSOLVENT FIRMS AFTER THE CURRENCY AND INTEREST RATE SHOCKS Percentage of total 100 Illiquid and insolvent Illiquid and solvent Liquid and solvent Liquid and insolvent 80 60 40 20 0 Indonesia Republic Malaysia Philippines Thailand Taiwan of Korea (China) Source: World Bank staff estimates. The World Bank Group 7 both exchange rate and interest rate shocks, in- hit, with 65 percent of firms estimated to be terest rate shock only, and currency shock only. insolvent. Here, as expected, the damage to the corporate sector stems mainly from the large The study calculated the effect of the exchange depreciation of the currency, much less so from rate shock as the increase in the value of for- the increase in interest rates. In Korea 41 per- eign currency debt—assumed to be denomi- cent of corporations are technically insolvent. nated entirely in U.S. dollars—due to the Here the effect of the currency depreciation still domestic currency devaluation.3 It estimated the dominates, but the interest rate shock is also effect of the interest rate shock by applying the important, and half the firms that are insolvent estimated increase in domestic corporate bor- are so because of the combination of the inter- rowing rates to debt denominated in domestic est rate and currency shocks. In Thailand about currency,4 and applying the estimated increase a quarter of firms are insolvent, most because in corporate foreign currency borrowing rates of the exchange rate shock. No Taiwanese firm to debt denominated in foreign currency. in the sample is technically insolvent. The results show that after the two shocks the There can, of course, be both overlap and inter- share of insolvent firms is 31 percent on aver- action between the two shocks. In Indonesia age for the five most affected East Asian coun- the share of insolvent firms in the scenario with tries (figure 5). For all five countries the currency both shocks is smaller than the sum of the shares shock is more important than the interest rate in the scenarios with single shocks because the shock. If only the currency shock had happened, same firms are insolvent under both shocks. In 22 percent of East Asian firms would be techni- Korea the two shocks interact, and the share of cally insolvent. Indonesia is by far the hardest insolvent firms in the scenario with both shocks TABLE 2 ESTIMATED NONPERFORMING LOANS AND NONPERFORMING ASSETS AS A SHARE OF GDP, SEPTEMBER 1998 Sample firms Sample firms’ Bank credit Total Total with debt nonperforming to the nonperforming Domestic nonperforming servicing loans private sector loans credit assets difficulties (percentage (percentage (percentage (percentage (percentage Country (percent) of total loans) of GDP) of GDP)a of GDP)b of GDP)c Indonesia 64.8 52.6 57.0 29.9 75.0 39.4 Korea, Rep. of 40.8 38.9 64.0 24.8 165.0 64.2 Malaysia 4.6 15.2 95.0 14.4 165.0 25.1 Philippines 16.3 13.6 52.0 7.1 78.0 10.6 Thailand 22.9 26.6 105.0 27.9 155.0 41.2 Simple average 30.2 29.4 74.6 20.8 127.0 36.1 a. Calculated by multiplying sample firms’ nonperforming loans as a percentage of total loans by total bank credit to the private sector as a percentage of GDP. b. Domestic credit includes credit from nonbank financial institutions (such as insurance companies) and capital market instruments. c. Calculated by multiplying sample firms’ nonperforming loans as a percentage of total loans by domestic credit as a percentage of GDP. Source: World Bank staff estimates except for bank credit to the private sector (Bank for International Settlements 1998) and domestic credit as a share of GDP (J.P. Morgan). 8 Corporate Distress in East Asia exceeds the sum of the shares in the scenarios cent), followed by Thailand (41.2 percent) and with single shocks. Indonesia (39.4 percent). The Philippines has the smallest amount of nonperforming loans Illiquidity is much more pervasive than insol- relative to GDP (10.6 percent). vency. Even before the crisis 11 percent of Thai firms, 15 percent of Malaysian firms, and 27 per- A related paper (Claessens, Djankov, and Lang 1998) documents the cent of Philippine firms were illiquid (that is, performance and financing structures of East Asian corporations over their earnings did not cover interest payments). the past decade. Using a database of 5,550 East Asian firms in nine countries over the period 1988–96, it identifies some important pat- After the crisis 77 percent of Indonesian firms, terns across countries. Leverage was high and increasing in Indonesia, 65 percent of Thai firms, 63 percent of Malay- Japan, Korea, and Thailand. Foreign exchange short-term borrowing sian firms, 57 percent of Philippine firms, and became important in the past few years, especially in Korea, Malaysia, and Thailand. These vulnerabilities in corporate financial structures 50 percent of Korean firms were illiquid (figure have played an important part in triggering and aggravating East Asia’s 6). Even in Taiwan (China) about 20 percent of financial crisis, and in leading many corporations into bankruptcy (Ferri, firms face liquidity problems. Interestingly, in Hahm, and Bongini 1998). 1 The data come from Worldscope and Extel databases and are com- most countries the interest shock becomes more piled from annual reports of the nonfinancial companies listed on important when illiquidity is considered. the major stock exchanges in the region. The sample consists of 463 firms: 82 for Indonesia, 79 for Korea, 79 for Malaysia, 62 for the Philippines, 99 for Thailand, and 62 for Taiwan (China). The ex- The differences between liquidity and solvency change rates used are from the International Monetary Fund’s In- problems reveal important insights. The share ternational Financial Statistics and Bloomberg. The lending rates of firms that are illiquid but technically solvent are from national sources, mostly the central bank, and the off- shore interest rates from J.P. Morgan, Singapore. Viewpoint is an open is large, averaging 35 percent for the five East 2 As used in this Note, the debt to equity ratio is total debt (long forum intended to Asian countries (figure 7). Thus many solvent term, short term, trade credits) over the nominal value of common encourage dissemina- firms, unable to cover current debt service pay- equity. 3 tion of and debate on Data availability problems prevented the study from computing the ideas, innovations, and ments from earnings, are at risk of insolvency. positive effect of the devaluation in terms of the increased value of best practices for ex- These firms may have difficulty obtaining out- assets denominated in foreign currency. Neglecting this effect could panding the private side financing for working capital to maintain lead to overestimates of insolvency. But the analysis also neglects sector. The views pub- the negative demand shock triggered by the crisis, a shock that lished are those of the ongoing operations, and may be operating at likely outweighs any positive effect stemming from revaluation of authors and should not too low a level. Restoring credit flows to these foreign currency assets. 4 be attributed to the solvent firms is important to restart growth in To account for the repricing of loans, the study applied the entire World Bank or any of increase to short-term domestic currency debt but only a third of its affiliated organiza- these countries. the increase to long-term domestic currency debt. This amounts to tions. Nor do any of the assuming that long-term domestic currency debt pays fixed rates conclusions represent Finally, the study estimated nonperforming loans and has a three-year maturity. This process was repeated for for- official policy of the eign currency debt. World Bank or of its and nonperforming assets as a share of GDP Executive Directors for the five most affected countries, extrapolat- or the countries they References ing from data for the firms in the sample that represent. cannot meet their debt service requirements Bank for International Settlements. 1998. 68th Annual Report. Basle, To order additional under the current interest and exchange rates Switzerland. copies please call Claessens, Stijn, Simeon Djankov, and Larry Lang. 1998. “East Asian (table 2). Because the estimates are based on Corporations: Growth, Financing, and Risks over the Last Decade.” 202-458-1111 or contact Suzanne Smith, editor, large firms, which typically have higher debt to Policy Research Working Paper 2021. World Bank, Washington, D.C. Room F11K-208, equity ratios, they may overstate the actual share Ferri, Giovanni, Hongjoo Hahm, and Paola Bongini. 1998. “Corporate The World Bank, Bankruptcy in Korea: Only the Strong Survive.” World Bank, Wash- of nonperforming loans. ington, D.C. 1818 H Street, NW, Washington, D.C. 20433, World Bank. 1998. East Asia: The Road to Recovery. Washington, D.C. or Internet address The results show that nonperforming loans as ssmith7@worldbank.org. a share of GDP are largest in Indonesia (29.9 Stijn Claessens (sclaessens@worldbank.org) and The series is also available on-line percent), Korea (24.8 percent), and Thailand Simeon Djankov (sdjankov@worldbank.org), (www.worldbank.org/ (27.9 percent), and smallest in the Philippines Financial Economics Group, and Giovanni html/fpd/notes/). (7.1 percent). The average for the five coun- Ferri (gferri@worldbank.org), East Asia and Printed on recycled tries is 20.8 percent. The ratio of nonperforming Pacific Region, Financial Sector Development paper. assets to GDP is largest in Korea (64.2 per- Sector Unit