This Selected Issues paper focuses on some of the key stylized facts of Korean business and export cycles over 1960–2001, and calculates a chronology for the classical cycle in these series by applying a variant of the Bry and Boschan (1971) cycle-doling algorithm. It highlights that the Korean classical business cycle and exports cycles are extremely asymmetric, as they exhibit long-lived expansions and much shorter-lived contractions. The results also indicate that the probability of ending a contraction or expansion phase in Korean industrial production and Korean real exports is independent of their duration.

Abstract

This Selected Issues paper focuses on some of the key stylized facts of Korean business and export cycles over 1960–2001, and calculates a chronology for the classical cycle in these series by applying a variant of the Bry and Boschan (1971) cycle-doling algorithm. It highlights that the Korean classical business cycle and exports cycles are extremely asymmetric, as they exhibit long-lived expansions and much shorter-lived contractions. The results also indicate that the probability of ending a contraction or expansion phase in Korean industrial production and Korean real exports is independent of their duration.

V. Firm Level Analysis of the Korean Corporate Sector: 1996–20001

Previous analyses of the Korean corporate sector have relied mainly on summary statistics to identify broad trends in profitability and capital structure since the crisis. This paper uses firm-level data, drawn from financial statements of 452 companies over a five-year period (1996–2000), to examine the determinants of corporate performance and the changes in the way that markets valued firms since the crisis. A key finding is that despite the improvement in operating performance, overall profitability remained weak because of the large interest burden on debt. Further, the gap between good and bad performers has widened. Markets also changed their perception of the large chaebol after the crisis, assigning lower valuations and higher borrowing charges compared to non-chaebol firms. The concentrated ownership structure of the chaebol was also found to be negatively related to profitability and market valuation.

A. Introduction

1. Since 1997, progress has been made in restructuring the corporate sector: debt-equity ratios have come down from dangerously high levels; the strong recovery has helped to boost cash flow; a number of large distressed groups have either been dismantled or sold; and financial reporting and disclosure have improved. In addition, market discipline now plays a greater role with the emergence of healthier banks after restructuring, greater shareholder activism, and rising foreign investment. Yet, despite these achievements, the corporate sector continues to suffer from high leverage and poor profitability, and the gap between the good and bad performers has widened, indicating that much more restructuring is needed.

2. This paper examines the underlying determinants of corporate performance since the crisis. Previous analyses have relied mainly on summary statistics to identify broad trends in profitability and the capital structure of Korean firms. With almost five years having passed since the crisis, enough time-series data are now available to begin analyzing the factors behind these changes both across time and across firms. The paper uses firm-level data spanning the pre- and post-crisis periods to examine the factors behind changes in corporate structure and profitability. It also looks at how markets valued companies after the crisis, in particular for the Korean chaebol.

B. Data

3. The data used in this analysis comes from the financial statements of Korean listed firms (from Datastream) and includes information on balance sheets, income, cash flow, and retained earnings. To cover both the pre-crisis and post-crisis periods, a five-year (1996–2000) panel data set of various financial indicators was constructed. The size of the dataset varies across years and excludes financial firms. Dropping firms that have missing years provides 452 firms in a consistent data set. Firms are categorized by industry according to the Bank of Korea (BOK) industry classification system. In addition, affiliates of the top-30 chaebol as determined annually by the Korea Fair Trade Commission (FTC) are identified as chaebol affiliated companies. Specifically, among the 452 firms, 97 are identified as chaebol affiliated, and they are typically larger than the non-chaebol firms.

4. One important caveat is that the data set reflects “survivorship” bias as it features only those companies that remained listed and operational during the entire period being examined. Those companies that were removed from the sample typically included firms that fell into bankruptcy, entered workout programs, or were delisted because they failed to meet the listing requirements. However, focusing only on the “survivors” allows a consistent comparison of the performance of companies before and after the crisis.2

5. The remainder of the paper is organized as follows: Section C examines the trends in corporate performance over the five year period, focusing on capital structure, profitability, productivity, and market valuation. Section D looks at the special features of the chaebol structure that have been criticized for causing the financial crisis and the research linking weak corporate governance to the sector’s poor performance. Section E outlines the results of the regression analysis looking at the determinants of corporate profitability and market valuation. Section F concludes with some interpretation of the results and their implications for the chaebol structure.

C. Corporate Performance During 1996–2000

6. Two key findings emerge from this analysis. First, despite an improvement in operating performance since the crisis, overall profitability remained weak because of the large interest burden on debt. Although firms managed to lower their debt-equity ratios, this deleveraging was achieved through the issuance of new equity rather than actual debt reduction, and the burden of servicing this large debt cut deeply into earnings.

7. Second, market perception of the chaebol appeared to have changed after the crisis. Although chaebol started with weaker balance sheets, poorer operating performance, and lower returns than non-chaebol firms, they paid lower borrowing charges and received higher valuations suggesting that markets were not paying attention to corporate fundamentals or that they considered the chaebol “too big to fail.” However, this perception changed after the crisis as the chaebol premium quickly turned into a steep discount. The following sections describe these changes in corporate performance since the crisis as well as possible reasons for the negative change in market perception of the chaebol.

Capital Structure

8. Overall leverage increased sharply prior to the crisis reflecting the pickup in debt-financed investment (Table V.I). The average debt-equity ratio rose from 245 percent in 1996 to 336 percent in 1997, which was quite high by international standards and roughly double the OECD average.3 Because of restrictions that favored debt over equity financing, corporations financed much of their investment with short-term borrowing from banks. Between 1996 and 1997, total liabilities rose by W 70 trillion while the book value of equity remained unchanged.

Table V.1

Korea: Leverage and Debt Structure, 1996–2000 (in percent)

article image

9. After the crisis, average debt-equity ratios fell from their peak in 1997 to 175 percent in 2000. However, almost 90 percent of the decline in the debt-equity ratio was due to the issuance of equity than to actual debt reduction. Taking advantage of the rising stock market, corporations raised equity by W 60 trillion compared to a decline in liabilities of only W 16 trillion (Table V.1). Moreover since 1999, debt levels have remained largely unchanged, and in 2000 were still high compare to pre-crisis levels.

10. The debt-equity ratio of chaebol firms was almost double that of the smaller non-chaebol in 1996, and both groups sharply increased their leverage the following year (Figure V.l). Since then, chaebol firms made greater progress in reducing their leverage than non-chaebol firms, with the gap between the two groups having narrowed steadily during the period.

Figure V.1
Figure V.1

Korea: Debt-Equity Radios, Chaebol vs. Non-Chaebol in percent)

Citation: IMF Staff Country Reports 2002, 020; 10.5089/9781451822069.002.A005

11. The maturity structure of corporate debt has not improved since the crisis. After a marked improvement between 1998 and 1999, the share of short-term borrowing in total debt jumped to 28 percent in 2000. Including the portion of long-term debt that falls due within one year (i.e., on a residual maturity basis), short term debt reached 43 percent in 2000, above the level in 1996. Part of the reason is that corporations issued a large amount of corporate bonds after the crisis with a three-year maturity and began to face large repayments starting in 2000. This bunching of maturities was mainly concentrated among the larger chaebol firms that were able to maintain access to capital markets and rollover their obligations during the crisis. As a result, chaebol firms have made less progress than the non-chaebol in lengthening the maturity structure of their debt (Figure V.2).

Figure V.2
Figure V.2

Korea: Share of Short-term Debt in Total Chaebol vs. Non-Chaebol

Citation: IMF Staff Country Reports 2002, 020; 10.5089/9781451822069.002.A005

12. Firms also became significantly less liquid during the period. The current ratio (the ratio of current assets and current liabilities), which measures a firm’s ability to meet short-term obligations through quick sale of liquid assets, worsened significantly during the period, falling from 99 percent in 1996 to 78 percent in 2000 (Table V.2). Despite the shortening of maturity on the liability side of their balance sheets, firms did not make the associated shift into shorter-term assets. This decline in liquidity suggests that firms became more vulnerable to a cutoff in credit lines or a flight to quality in financial assets. Liquidity is much worse for chaebol firms whose current ratios fell from 94 percent in 1997 to only 70 percent in 2000. Non-chaebol also experienced a slight decline in liquidity, but in 2000, their current ratio was a healthy 97 percent.

Table V.2

Korea: Liquidity - Current Ratios, 1996–2000 (in percent)

article image

Profitability

13. Operating profitability (excluding financial expenses) has improved since the crisis, due in large part to the strong economic recovery. Operating income to sales, which looks at earnings from the normal course of business, showed steady improvement since 1998 indicating that, at least at the operational level, cash flow and performance have strengthened (Table V.3). Another useful measure is the interest coverage ratio (ICR), defined as the ratio of earnings before interest, tax, depreciation and amortization or EBITDA to interest expense, which measure a firm’s capacity to generate sufficient cash flow to cover its interest payments. After falling sharply in 1998, the ICR recovered strongly, but remained below pre-crisis levels. The share of companies with an ICR of less than 100 fell from its peak of 30 percent in 1998 to 24 percent in 2000.

Table V.3

Korea: Indicators of Profitability, 1996-2000

(in percent)

article image

EBITDA (earnings before interest, taxes, depreciation and amortization divided by interest payments.

14. Despite the improvement in operating performance, overall profitability remained weak because of the large interest burden on debt. Net financial expenses, which include non-operating costs such as interest payments, and losses and gains on foreign currency transactions and assets, remained very high and cut deeply into earnings. In 2000, they took up over 90 percent of operating income. After accounting for these non-operating costs, net earnings (so called “ordinary income”) showed almost no improvement over the period. Ordinary income to sales fell from 1.1 percent in 1996 to 0.6 percent in 2000.

15. Other more comprehensive measures of profitability, such as return on assets (ROA) or on sales (ROS) that net out tax payments and any extraordinary losses on investments (“net income”) paint an even worse picture. In 2000, firms on average made net losses of $1.5 for every $100 of sales in 2000 (minus 1.5 percent return on sales). This is surprising when one considers that the firms in the samples represent those companies that managed to survive the crisis and were in a position to benefit from the rapid recovery.

16. The bleak picture on profitability, however, masked the widening gap between good and bad performers. Separating companies by their return on sales (ROS) reveals that the number of firms at opposite ends of the distribution increased over the period. The proportion of companies with a ROS of below −5 percent or above +5 percent rose from 23 percent in 1996 to 50 percent in 2000 (Figure 3). The proportion of companies with negative ROS increased from one in five in 1996 to almost one in three in 2000.

Figure V.3
Figure V.3

Distribution of ROS

Citation: IMF Staff Country Reports 2002, 020; 10.5089/9781451822069.002.A005

17. Chaebol firms, in general, were less profitable despite having lower financial expenses. Operating income to sales for chaebol firms improved from 4.6 percent in 1996 to 6.1 percent in 2000 while it remained steady for non-chaebol firms, albeit at a higher level. Chaebol also had lower financial expenses, but this was not enough to generate positive net returns. After accounting for financial expenses, taxes, and investment losses, the chaebol recorded negative or zero returns on sale in 1999 and 2000, compared to small but positive returns for the smaller firms (Figure V.4).

Figure V.4
Figure V.4

Korea Return on Sales (ROS) Chaebol vs. Non Chaebol; in Percent

Citation: IMF Staff Country Reports 2002, 020; 10.5089/9781451822069.002.A005

18. Although average borrowing costs on debt have come down steadily from its peak in 1998, they still remain above pre-crisis levels (Table V.4). Part of the reason is that most of the corporate bonds issued after the crisis were at fixed rates and consequently were unaffected by the subsequent decline in interest rates. Prior to the crisis, smaller companies faced higher borrowing costs than the larger chaebol, most likely reflecting their difficulty in accessing directly the capital markets. This pattern, however, reversed starting in 1997 as non-chaebol enjoyed lower interest rates on their borrowings, by as much as 400 basis points in 2000. This premium perhaps reflected their lower credit risk and greater progress in restructuring. Smaller companies that survived the crisis were in general less indebted, more liquid, and more profitable than the larger chaebol.

Table V.4

Korea: Indicators of Financial Costs, 1996–2000

(in percent)

article image

Activity and Productivity

19. After three sluggish years, sales rebounded strongly in 2000, spurred by the rapid recovery of the economy (Table V.5). However, the pickup in sales did not necessarily translate to higher profits as indicated in the previous section. Part of the reason may have been that much of this increase in sales took place internally between chaebol affiliates and was not based on market terms. Facing the threat of a breakup, chaebol groups struggled to support weaker affiliates by transferring resources from their healthy ones. This is partially borne out by the fact that chaebol affiliated firms featured higher asset-turnover ratios but lower profit margins than non-chaebol firms. In 2000, the asset turnover ratio (sales to asset) for chaebol was more than double that for non-chaebol firms. Moreover since 1997, chaebol have increased their turnover ratio while non-chaebol lowered theirs.

Table V.5

Korea: Turnover and Productivity, 1996–2000

(in percent)

article image

20. Productivity showed some signs of improvement, as the labor force was reduced. Sales per employee rose steadily over the period due to large layoffs rather than higher profits. The number of employees in the sample companies fell by 22 percent between 1996 and 2000. However, despite this large reduction in labor costs, net income per employee declined over the period and turned negative in 2000.

Market Valuation

21. Prior to the crisis, markets appeared to value chaebol firms more highly than non-chaebol firms. Using a Tobin’s q-measure of market value to book value, chaebol firms on average were valued about 34 percent higher than non-chaebol firms in 1996 (Figure V.5).

Figure V.5
Figure V.5

Korea: Tabin’s Q-Ratio of Market Value to Book Value Chaebol vs. Non-Chaebol; in Percent

Citation: IMF Staff Country Reports 2002, 020; 10.5089/9781451822069.002.A005

However, this chaebol premium peaked in 1997 and then turned negative. By 2000, markets were valuing chaebol firms at a 37 percent discount compared to non-chaebol firms. As described earlier, the same turnaround occurred in the relative interest rates charged to chaebol, but some two years earlier.

D. The Chaebol Factor

22. The chaebol were heavily criticized for causing and contributing to the severity of the crisis. Their over reliance on debt financing, dominant position in several industries, poor corporate governance practices, and complex linkages left the corporate sector weak, uncompetitive, and ultimately vulnerable to a financial crisis. In particular, cross-debt guarantees and shareholdings diluted accountability for poor business decisions by creating soft-budget constraints for weaker affiliates. Cross-shareholdings also allowed a large investor, typically a family owner, to control the company with little of his own capital at risk. Owners took advantage of this to expand into non-core areas or extend control at the expense of minority shareholders. The lack of corporate transparency also deterred outsiders from investing in Korean companies.

23. Empirical research on the chaebol has found evidence that the chaebol structure hindered corporate performance. Shin and Park (1999) compared the financing constraints of the chaebol and non-chaebol firms and found that the soft budget constraint within the chaebol structure permitted chaebol firms to invest more than non-chaebol despite the poorer growth opportunities. Joh (2000) found firm-level evidence that greater control-ownership disparities created conflicts of interest among shareholders and lead to poorer firm performance. More broadly, Johnson et al (2000) have argued that weak corporate governance can leave a country vulnerable to a sudden loss of investor confidence, resulting in a collapse in the exchange rate and a sharp fall in asset prices.

24. The findings in this paper are consistent with the findings of previous studies. In many respects, the chaebol lagged behind their smaller counterparts in improving their capital structure, profitability, and productivity. At the start of the crisis, chaebol had weaker operating performance, lower financial expenses, and lower returns on assets and sales than their smaller counterparts (see Table V.6 for a comparison of the two groups). Yet despite these drawbacks, chaebol firms received more favorable borrowing terms and higher market valuations, suggesting that either equity investors and financial institutions were not paying adequate attention to corporate fundamentals or they considered the larger chaebol “too big to fail.” With just two exceptions—Kukje in 1985 and Woosung Construction in 1996—no large business or nationwide bank was allowed to fail until 1997.

Table V.6

Korea: Comparison of Chaebol vs. Non-Chaebol

(in percent)

article image

25. However, this perception changed after the crisis as both investors and financial institutions began to look more carefully at the weaknesses in the chaebol structure. Starting in 1999, the chaebol premium quickly turned to a discount as chaebol firms faced significantly higher borrowing costs and lower market valuations compared to their smaller counterparts. The string of large bankruptcies, which began in early 1997 followed by the bankruptcy of the Daewoo group in 1999, changed investors’ perception of the large chaebol and reoriented focus to the less risky and better performing non-chaebol firms. In addition, with the restructuring of the banks and improved supervision and loan classification practices, banks became more aggressive in assessing the credit worthiness of their borrowers which was reflected in their loan pricing.

E. Regression Analysis

26. Rather than a firm optimization model, the paper uses regression analysis to examine the correlation between firm profitability and their various determinants. As posited in the literature, these determinants include: firm size, financial structure, turnover, and industry and firm-specific characteristics. Three different measures of profitability—operating income, ordinary income, and net income to sales are used.4 Operating income focuses mainly upon earnings from the regular business operations of the firms; ordinary income accounts for financial expenses/earnings, such as on interest payments; and net income is the final measure after taxes and extraordinary gains and losses.

For the determinants of profitability, the following independent variables are included:

  • Log of sales as a control for firm size.

  • Sum of current maturities and short-term borrowing to total liabilities as an indicator for the maturity structure of debt.

  • Liabilities to assets ratio (or debt-equity ratio) as a measure of leverage or corporate gearing.

  • Net financial expenses over total expense to measure the impact of financial burden affects profitability.

  • Ratio of new fixed assets to total assets as a measure of new investment.

  • Asset turnover ratio (assets / turnover) to examine the correlation between sales and profitability, particularly as they relate to intra-unit trading within chaebol groups.

  • Chaebol dummy variable to account for any chaebol-specific factors.

To control for firm and industry-specific characteristics, a fixed effect regression, with both industry and time dummies, was utilized.

27. In addition to the standard determinants of profitability, the relationship between ownership structure and profitability was also examined. To measure the relative power of major shareholders over minority holders, the Zeno-index (Z-index) measure of ownership was used. The conventional measure of shareholder power is the Herfindal index, which is simply the share of the majority or largest investors. The Z-index is more a more powerful measure of corporate control as it looks not only at the absolute power of a single shareholder, but also at the concentration of voting rights. In other words, the Z-index captures the ability of swing voters (e.g., votes of affiliated firms) to change outcomes and could be considered a more robust indicator of shareholder power in the Korean case.5

Results for Profitability

28. Table V.7 summarizes the results of the regression using the three different measures of profitability. The regression using operating profits yielded the greatest explanatory power and with net income the least. In all three regressions, size of the firm mattered, and year and industry dummies were generally significant. The main findings are as follows:

Table V.7

Korea: Regression Results-Determinants of Profitability, 1996–2000

article image

Significant at 5 percent level

Significant at 10 percent level

  • Leverage was negatively associated with profitability and was significant in all three regressions. Higher financial expenses, as represented by the share of net financial expenses in total expenses, depressed ordinary and net income but not operating profits. This may not be surprising as operating profit is a gross term that does not include financial cost and may have benefited from the higher debt-financed investment.

  • New investment raised operating profits but did not yield significant positive net returns. New investment (as measured by new fixed assets over total assets) was positively associated with operating profits but had no significant link with ordinary or net income, suggesting that new investment failed to generate positive net returns.

  • Higher turnover did not generate more profits. In fact, the turnover ratio (sales over assets) was negatively correlated with operating profit and insignificant with ordinary and net income, indicating that the higher turnover within the chaebol was not profitable. This is consistent with the notion that much of this inter-unit trading was done to support weaker loss-making chaebol affiliates.

  • Chaebol affiliation meant lower profits. Even after accounting for the standard determinants of profitability, being affiliated with a chaebol was negatively associated with profitability, both on an operating and on a net basis.

  • Ownership concentration was negatively associated with profitability. In all three regressions, the Z-index featured a positive coefficient on the linear term and a negative one on the square term, indicating the existence of a non-linear relationship between ownership concentration and profitability. When calculated at the average Z-index level, the combined coefficient was negative, suggesting that higher ownership concentration lowered profitability, perhaps by distorting corporate decision-making and accountability.

Results for Market Valuation

29. Regressions were also run to examine how markets valued firms and whether this perception changed since the crisis (Table V.8). The sample was divided into two periods: a pre-crisis period (1996–97) and a post-crisis period (1998–2000). The dependent variable is a Tobin-q ratio of market to book value. Based on the results of the profitability test, the following independent variables are used: log of sales, current ratio (current liabilities over current assets), debt-equity ratio, financial burden, new investment rate (new fixed assets / assets), liquidity ratio, interest coverage ratio, chaebol dummy, and the Z-index.

Table V.8

Korea: Regression Results-Market Value to Book Value, 1996–2000

article image

Significant at 5 percent level

Significant at 10 percent level

30. In general, it appears that markets did focus on corporate fundamentals, such as liquidity, leverage, and profitability in valuing firm’s equity, in addition to other factors such as ownership and chaebol affiliation.

  • Before the crisis, markets assigned higher valuation to chaebol than to non-chaebol firms. In the 1996–97 regression, the coefficient on the chaebol dummy was positive and significant at the 10 percent confidence level, suggesting that chaebol affiliated firms received a premium by the markets over non-chaebol firms.

  • After the crisis, the chaebol premium disappeared and turned negative. Even after controlling for other factors, chaebol firms were still assigned lower values by the market than non-chaebol firms. Possible reasons include the undoing of the “too big to fail” perception following the collapse of several large chaebol and greater investor emphasis on corporate governance.

  • After the crisis, a significant link between ownership concentration and market valuation appeared. The combined coefficient for the Z-index (evaluated at the mean level) turned strongly negative and significant in 1998–2000, suggesting that markets were more focused on the ownership structure of firms after the crisis. One possibility is that investors, particularly foreigners, began paying more attention to the treatment of minority shareholders by major shareholders in their investment decision. Foreign ownership in the Korean stock market rose rapidly during this period, from 13 percent in 1996 to over 30 percent in 2000 following the removal of limits on foreign ownership of Korean equity.

  • Liquidity also appeared to matter more after the crisis. The coefficient on the current assets to current liabilities ratio turned significantly positive in 1998–2000.

F. Conclusion

31. The paper identifies factors that contributed to the poor corporate performance since the crisis. Excessive leverage depressed profitability, mainly through the large interest burden, and lowered market valuation. New investments on average did not generate significant positive returns, suggesting that companies that chose to expand instead of consolidate during the period fared poorly. In particular, the increase in turnover among the chaebol firms was not profit-generating, offering evidence that intra-chaebol trading may have been directed to support weaker affiliates.

32. In addition, financial markets and institutions appeared to have signaled that the old chaebol structure no longer provided significant advantages. The premium once enjoyed by the chaebol in terms of lower borrowing costs and higher market valuations disappeared in 1999 and switched to a large discount. Strengthened supervision, the emergence of sound banks and a new credit culture, and the undoing of the important lesson of “too big to fail” might have contributed to souring market perception on the chaebol. The negative view may also reflect the slow progress by the large chaebol since the crisis in reducing leverage and improving profitability compared to their non-chaebol counterparts.

References

  • Crama, Yves 1999, “Corporate Governance Structures, Control and Performance in European Markets: A Tale of Two Systems,” CORE Discussion Paper 9942 (July).

    • Search Google Scholar
    • Export Citation
  • Joh, Sung Wook, 2000, “Control, Ownership, and Firm Performance: The Case of Korea,Korea Development Institute.

  • Johnson, Simon 2000, “Corporate Governance in the Asian Financial Crisis,Journal of Financial Economics, 58, October-November 2000, pp. 141186.

    • Search Google Scholar
    • Export Citation
  • Shin, Hyun-Han, and Young S., Park, 1999, “Financing Constraints and Internal Capital Markets: Evidence from Korean Chaebols,Journal of Corporate Finance: Contracting, Governance and Organization, pp. 169191 (June).

    • Search Google Scholar
    • Export Citation
1

This paper was prepared by Sungbin Cho (Summer Intern, APD) and Kenneth H. Kang (APD).

2

The dataset also avoids having to account for the impact of changes in the institutional setting, such as for companies entering court receivership or out-of-court workout programs.

3

In addition to the rise in debt-financed investment, the spike in the debt-equity ratio in 1997 also reflected the overshooting of the exchange rate.

4

Dividing by sales rather than by assets is more applicable to the Korean case since chaebol had higher asset turnover (sales / assets) than non-chaebol. In order to account for the large amount of intra-unit trading among chaebol affiliates, dividing by sales rather than assets provides a more accurate picture of profit margins.

5

For more details, see Crama et al. (1999, 2001).

Republic of Korea: Selected Issues
Author: International Monetary Fund