1
Loans to Distressed Firms:
Political connections, related lending, business group affiliation, and bank governance
Ming Ming CHIU
The Chinese University of Hong Kong
Sung Wook JOH*
Korea University
First draft: July, 2003
This draft: April 7, 2004
*We are grateful to Randall Morck, Jun-Koo Kang, E. Han Kim, Sejik Kim, Hyun Han Shin and participants
at several seminars for their helpful comments on an earlier version of the paper. This paper was written
while the second author was visiting the Center for Economic Institutions at Hitotsubashi University. She
appreciates their hospitality and financial support. Corresponding author. Sung Wook Joh, Business School,
Korea University, 5ga-1, Anam-dong, Sungbuk-gu, Seoul 136-701, Korea. Email: [email protected]
2
Abstract
Financial institutions (FIs) suffered from non-performing loans when debt-ridden firms failed. Nonetheless,
FIs in Korea increased loans to distressed firms in the 1990s. Possible explanations for these loans include
FIs having better inside information on borrowing firms, firms' sharing resources with its business group
affiliated firms, firms' political connections, related lending (FI affiliation), or FIs' moral hazards (poor FI
governance). We examined 6,474 non-financial firms' capital structures and performances during 1990-2000.
Distressed firms had higher leverage ratios and leverage growth rates. Furthermore, firms in distress, with
higher leverage growth rates or political connections tended to show both lower ex-post ability to pay debt
and lower return on assets, suggesting that FIs did not benefit from inside information when making lending
decisions regarding distressed firms. Firms in distress, with political connections, or with FI affiliations all
had higher leverage growth rates. Among distressed firms, those affiliated with business groups had higher
leverage growth. Together, these results support the claims that business group affiliations, political
connections, and related lending affected lending practices. Distressed firms without any business group
affiliations, political connections, or FI affiliations also showed higher leverage growth rates, supporting the
claim of poor FI governance.
3
Introduction
Non-performing loans along with other financial sector problems beset Asian countries during the
economic crisis. When highly indebted firms failed, they caused a cascade of bankruptcies among affiliated
firms, subcontractors, and so on, leading to huge non-performing loans. These non-performing loans
exhausted banks’ debt capacities as foreign investors stopped revolving loans to them, thereby triggering the
financial crisis in 1997 (Corsetti, Pesenti, & Roubini, hereafter CPR, 1999; Diamond & Rajan, 2000). Bank
problems correlated with loans to bankrupt firms, but the question of why banks gave loans to these firms
remains unanswered. This study examines the determinants of loans to Korean firms, especially financially
vulnerable ones, during 1990-2000.
Financial institutions (FIs) might extend loans to distressed firms for various reasons including
government constraints, superior information, external pressure or FI governance problems. Peek and
Rosengren (2003) showed that Japanese firms with lower stock returns borrowed more, and that weaker
banks lent more. Linking these results, they argued that banks were forestalling loan defaults and hid these
losses to maintain their government-required capital ratios on paper. Our study extends their work and
identifies the factors that led to more loans to distressed firms. We examine whether FIs lent more to
distressed firms because of (a) FI's better inside information, (b) firm affiliation with a business group, (c)
firms’ political connections (Dinc 2004, and Faccio, 2004), (d) related lending by affiliated FIs (La Porta,
López-de-Silanes and Zamarripa, 2003, hereafter LLZ, 2003), or (e) FI managers’ moral hazards.
In the most optimistic view, distressed firms were credit-worthy, as they or their affiliated business
group had sufficient resources to repay their loans. A firm might provide sufficient collateral or be likely to
earn sufficient future profits to repay their loans. For example, due to a close historical relationship with a
firm, an FI might have inside information that the firm’s future profits will cover its current losses (see Hoshi,
et al., 1991a, 1991b, Aoki, Patrick and Sheard, 1994). A firm might also use its affiliated firms’ resources
via their business group’s internal capital market. When applying for a loan, a firm can arrange for its
affiliated firms to provide collateral or have another firm within the business group guarantee payment of the
loan (cross debt payment guarantees).
4
On the other hand, firms might pressure FIs for loans through political connections or related lending.
Through political donations and other payments to government officials, firms can have them pressure FIs to
lend more money (Graham, 2003; Kang, 2002; Dinc, 2004; and Faccio, 2004). Indeed, Krugman (1998)
argued that excessive lending and overinvestment driven by crony lending helped cause the Asian economic
crisis. Meanwhile, related lending can occur through firm affiliations with FIs. Firms owning shares in FIs
can get loans from them on favorable terms (e.g., Mexican banking sector, LLZ, 2003). Even without direct
ownership, Thailand firms with FI personnel on its board of directors borrowed more money than other
Thailand firms did (Charumilind, Kali, and Wiwattanakantang, 2003).
Government intervention and managerial reputation incentives create moral hazards for FI managers.
When a government protects FIs (e.g., via deposit insurance, or government recapitalization) or provides
firms with implicit government guarantees, FI managers face a moral hazard and might take excessive risks
by routinely approving loans without evaluating the creditworthiness of a borrower (Krugman, 1998;
Mitchell, 1998). Frequent government interference also reduces the likelihood that FI managers will acquire
or retain the needed expertise to make informed lending decisions. Furthermore, FI managers concerned with
their reputation during their short tenures might focus on avoiding blemishes, such as loan defaults, rather
than maximizing profits (KDI, 2000). In this case, FIs might give short-term loans to a weak firm, thereby
propping it up to avoid loan defaults. Thus, FI managers facing moral hazards might give more loans to
distressed firms than to sound firms, even if these distressed firms do not have high future profit potential,
business group affiliations, political connections or FI affiliations.
We tested the above hypotheses by analyzing the capital structure of Korean firms during 1990–2000.
In the early 1990s, the government reformed the financial sector and reduced their direct intervention in the
corporate sector (Yoo, 1997). Banks were privatized in the early part of the 1980’s, and industrial policy
loans officially ended. Thus, direct government intervention into lending practices became weaker, and the
profit incentive of banks increased. Furthermore, government regulations imposed ceilings on lending
interest rates until 1997 (Graham, 2003), so maximizing bank profits required minimizing non-performing
loans through conservative loans to credit-worthy borrowers. The lower economic growth in Korea
5
throughout most of the 1990’s compared to those in the earlier developmental periods (Bank of Korea) also
supported conservative lending practices.
Specifically, we examined FIs’ lending decisions. The borrower information that FIs used to make
loan decisions is not readily available. So, we tested whether the firm’s capital structure reflected the
available information on its financial stability and its likelihood of repayment. The empirical tests consist of
four parts. First, we tested whether FIs lent more money to financially distressed firms than to other firms.
We examined whether distressed firms showed higher leverage ratios as well as higher leverage growth rates,
for various leverages, such as bank loans, interest bearing borrowing and total debts. As capital structure
depends on asset tangibility, profitability and firm size (Rajan and Zingales; 1995), we controlled for these
effects along with industry effects, yearly effects, crisis period, and stock market listing. Measured as the
ratio of fixed assets over total assets, asset tangibility captured the effect of collateral. For leverage growth,
we also controlled for the lagged leverage to asset ratio.
After showing that distressed firms had higher leverage ratios and leverage growth rates, we
examined whether better FI inside information on firm’s future performance accounted for the high leverage
growth rates of distressed firms. We tested this inside information hypothesis both overall and for each type
of connection: business group affiliation, regional political connection, or FI affiliation. As political
connectedness depends on regionalism in Korea (Siegel 2004), we tested the effect of political connections
by identifying firms whose founders or chief executive officers were born in the same region as the sitting
president of Korea. We tested whether firms receiving loans tended to perform better by examining whether
leverage growth rate, financial distress (losses in last three years), type of connection, or their interactions
affected a firm’s ex-post return on assets and its ex-post ability to pay future loan payments, using similar
control variables.
Next, we tested for the effects of business group affiliation, regional political connection, and FI
affiliation by examining the determinants of firms’ leverage ratios, leverage growth rates, and short-term
leverage over total leverage ratios. Independent variables included distress, business group affiliation,
regional political connection, FI affiliation and their interactions, along with other control variables. If
6
distressed firms without any political connections or FI affiliations received extra loans, then arguably, FIs
were not competently evaluating firms’ loan applications even when FIs were not under outside pressure. In
this case, loans to distressed firms reflected the internal governance problems of FIs. If only distressed firms
with political connections or FI affiliations received additional loans, then arguably, the loan decision was
made because of external pressure. Moreover, if distressed firms with political connections or FI affiliations
received extra short loans, then arguably, FIs might have reduced their risk exposure by offering short term
rather than long term credit.
The empirical results in this study found that distressed firms showed higher leverage ratios and
leverage growth rates of than sound firms, after controlling for the aforementioned variables. The results
also showed that FIs did not benefit from inside information. Instead, chaebol affiliations, political
connections, and FI affiliations, and poor FI governance all affected FI lending. FIs did not benefit from
inside information (if any) as the likelihood of earning sufficient ex-post profits to cover their financial costs
was not higher in distressed firms receiving more loans. Distressed firms, politically connected firms, and FI
affiliated firms generally had higher leverage ratios and higher leverage growth. Among distressed firms, the
leverage ratios of chaebol affiliated firms generally exceeded those of independent firms.
FI governance was also inadequate as FIs lent extra money to distressed firms that were less likely to
influence them, namely firms without any political connections or FI affiliations. Distressed firms,
politically-connected firms, chaebol affiliated firms, and FI-affiliated firms also had higher short-term over
total leverage ratios than other firms, suggesting that FIs gave them more short-term loans to reduce their risk
exposure. The increased lending through short-term loans suggested that FI managers recognized the risk
exposure. However, FI managers made short term loans to distressed firms without political connections or
FI affiliations anyway, suggesting that they were avoiding defaults during their short tenures to preserve their
reputations.
These results implied that Korean FIs generally made bad lending decisions. Using publicly available
information on financially distressed firms, FIs could have reduced their lending to them, ceteris paribus, to
minimize their risk of heavy losses, especially under an interest rate ceiling and during periods of declining
7
economic growth. They did not. Their failure and the reasons for their failure suggest several fundamental
reforms of government policies.
The remainder of this paper is organized as follows. First, we discuss the previous literature on FI
lending to firms, followed by background information on the Korean financial sector. After introducing the
methodology and data, we discuss the empirical results. Finally, section 5 concludes with some implications.
2. FI decisions on loans to distressed firms
When a firm defaults on its loan payments, control of the firm can shift from equity holders to debt
holders. Debt holders can either liquidate the firm or let it continue operating. If the firm’s debt and the value
of its fixed assets each exceed its future profits, the debt holder typically liquidates the firm. If the firm’s
future profits exceed either its debt or its fixed assets, the debt holder benefits by letting the firm continue
operating. The debt holder can refinance the existing debt without extending further loans if the default size
is relatively small. Otherwise, the debt holder might give further loans. By doing so, the FI might recover
both its initial and additional loans from the firms’ future profits rather than settling for the typically lower
liquidation value. The debt holder might lend more money to a distressed firm due to (a) better inside
information (b) the firm's affiliated business group's resources, (c) the firms’ political connections, (d) related
lending, or (e) FI managers’ moral hazards.
2.1 Loans to creditworthy firms based on inside information or business group resources
Individual debt holders rely on only publicly available information, but large institutional debt holders,
such as FIs, often have inside information (Fama, 1985, 1990). Compared to public debt holders, FIs can
typically access better firm information more easily and hence have lower monitoring costs, especially if they
have a historical relationship with the firm (Aoki, Patrick, & Sheard, 1994; Gerschenkron, 1962; Hoshi,
Kashyap, & Scharfstein, 1991a, 1991b). With inside information about a firm, a FI can better assess the risks
and expected profits of investment projects. Using this inside information, FIs might believe that a firm’s
financial distress is temporary and that future profits will likely cover its debt. If FIs benefited from inside
information, then distressed firms receiving further credit were more likely to repay their loans than other
8
distressed firms would.
Aside from inside information, FIs might lend more to firms that can use its affiliated business
group's resources. When applying for a loan, a firm can arrange for its affiliated firms to provide collateral
through internal capital market in business groups, or have another firm within the business group guarantee
payment of the loan (cross debt payment guarantees). However, an FI might have difficulty discerning the
true size of a business group's assets or the true extent of a firm's dept payment guarantees. Firms with
interlocking ownership can exaggerate their actual size. For example, if firm A invests its assets in its
affiliated firm B, then the sum of the assets of A and B can exceed the total assets of the group. Also, a firm
need not disclose the full extent of its debt payment guarantees to a FI. If firm C guarantees payment of firm
D's debt and firm D guarantees payment of firm E's debt, FIs cannot easily discern the extent of C's
guaranteed payment on E's debt (Joh, 2004). By exaggerating its business group's size and by limiting
disclosure of its debt payment guarantees, a firm can receive even more loans from an FI. Joh (2004) showed
that the magnitude and extent of cross-debt payment guarantee was large in large business groups, often
exceeding its total equity. Thus, FIs might lend more to firms affiliated with a business group than to
unaffiliated firms.
2.2 Pressuring FIs through political connections or bank affiliations
Rather than having better future prospects or affiliated business group resources, firms might have
used regional political connections or FI affiliations to pressure FIs to give them loans. A firm might use its
political connections to have government officials influence FI lending decisions (Faccio 2004, Siegel 2004).
Governments can influence FIs in several ways, including legislation, regulation, direct subsidies, direct
ownership, and selection of FI managers (directly or indirectly). Dinc (2004) showed that governmentcontrolled banks increased their lending more than privately owned banks did during the election years.
Through legislation or regulation, governments can restrict loans to favored firms. Likewise, governments
can own FIs and appoint its managers (directly or indirectly). Through any of these means, a politically
connected firm might induce FIs to lend it more money (Faccio 2004). Thus, distressed firms with political
connections might have received more loans than distressed firms without such connections.
9
In addition to political connections, a firm might pressure its affiliated FI(s) for loans. Firms affiliated
with FIs might own shares in FIs or place its employees on FIs’ boards of directors. Akerlof and Romer
(1993) argued that when the value of the FI’s capital falls below a threshold value, a firm can divert an
affiliated FI’s resources to itself (looting). LLZ (2003) generalized looting to include a firm defaulting on a
loan from its affiliated FI loan at the cost of foregoing their equity in the FI. A controlling shareholder in the
above firm has an incentive to loot if his or her share of firm profits exceeds his or her share of FI profits (cf.
exploitation of affiliated firms through tunneling, Johnson et al., 2000; and keiretsu exploitation in Japan,
Morck and Nakamura, 1999; Kang and Stulz, 1997). When related lending occurs, distressed firms affiliated
with FIs receive more loans than distressed, unaffiliated firms.
2.3 FI managers' moral hazards
Also, distressed firms may receive extra loans due to FI managers’ moral hazards. These moral
hazards might stem from government protection, government interference or FI manager reputation
incentives. If the government protects the banking system (e.g., deposit insurance or bailouts through
expected recapitalization) without adequate supervision or regulation, then FI managers have more incentives
to take excessive risks or allocate credit according to non-market criteria (CPR, 1999; ;Mitchell, 1998). For
example, FI managers can make loans to their own companies on non-market terms, as the government bears
the costs (CPR, 1999; LLZ, 2003). LLZ (2003) showed that FI loans to affiliated firms in Mexico had higher
default rates and lower recovery rates.
Frequent government direction or interference reduces the likelihood that FI managers will acquire or
retain the needed expertise to make informed lending decisions. As poor FI governance reduces the
incentives to make informed lending decisions, FI managers can minimize their effort and continue lending
to familiar borrowers. As a result, weaker borrower performance does not necessarily induce FIs to lend
more conservatively or reduce the risk in their portfolios (CPR, 1999). In short, under poor FI governance,
many distressed firms without political connections or without FI affiliations are also likely to receive loans.
Without strong FI governance, FI managers might try to preserve their reputations rather than
maximizing FI performance (KDI, 2000). If FI managers have short tenures, they might try to avoid
10
blemishes during their term (such as loan defaults). So, they can try to “prop up” distressed firms with shortterm loans until they become profitable, or at least until their term ends. For example, Japanese FIs avoided
listing potential loan defaults on their balance sheet by giving credit to distressed firms. This practice was
more prevalent in weak FIs whose risk-based capital ratios were dangerously close to the governmentrequired minimum capital ratio (Peek & Rosengren, 2003). By limiting the duration of potential future firm
losses, short-term loans limit FI risk, especially considering the possible dangers of illiquid investment, low
borrower credit quality, low potential for short-term cash flows, and little secondary market value (Diamond
& Rajan, 2000; Rajan, 1992). Therefore, FI managers who are concerned about their reputation and who
recognize these potential problems will tend to give more short-term loans.
3. Korea’s financial sector
Any of the above explanations for loans to distressed firms appear plausible when examining a brief
history of Korea’s financial sector. Through government-controlled FIs' subsidized loans to targeted firms,
Korea developed its heavy and chemical industries in the 1970s (and firms developed strong political
connections). These firms grew into diversified business groups (chaebols) and developed close long-term
relationships with FIs (inside information and FI-affiliation). When these firms failed, the government bailed
them out and recapitalized banks, thereby creating FI moral hazards. Despite gradual reforms, these
conditions generally persisted through the 1997 crisis.
3.1 Industrial policy and development of chaebols
Government control of banks began after Park Chung Hee's coup in 1961. He nationalized the banks
and appointed bank CEOs. These CEOs engaged in bank lending decisions and personally approved large
loans. The government continued appointing bank CEOs until 1993 and approved them until 1997.1 Bank
CEOs often had short appointments (around 3 years), so they only had to prop up failing firms for a few
years to avoid loan defaults during their tenure.
1
Chun Doo Whan privatized the banks in 1981-1983, but the government still appointed bank CEOs. In 1993, banks
introduced CEO recruitment committees whose members were approved by the government's Bank Supervisory Board (later it
became the Financial Supervisory Commission FSC). In 1997, the CEO recruitment committee members consisted of non-chaebol
directors.
11
Park used banks to develop the heavy and chemical industries.2 The government mobilized scarce
capital from foreign investors and domestic depositors and allocated it through banks to targeted firms at
subsidized interest rates until 1981 (Cho and Kim, 1997).3 In addition to powering economic growth, these
government loans helped these firms accumulate capital, expand into many industries to become diversified
business groups (chaebols), 4 and develop long-term banking and political relationships. Park typically chose
firms based in his home province of Taegu, Kyoungsang, beginning a long history of political connections
based on regionalism (Siegel 2004). These firms maintained these political connections by giving large
political donations (and other types of payments) to the president and other top government officials
(Graham; 2003. D. Kang; 2002).
Chaebol access to bank credit declined when the government privatized banks, cut subsidized projects,
introduced prudential regulation, and gradually promoted competition in the 1980's (Yoo, 1997). To ensure
access to credit, business groups invested in and effectively controlled non-commercial banks (such as
merchant banks) and other FIs (such as insurance and securities firms, CPR, 1999; Graham, 2003).5 Kim’s
(1998) study supports the looting view of these chaebol-FI relationships. When a chaebol owned a FI, firms
affiliated with that chaebol had higher debt-equity ratios and lower profitability than other firms.
3.2 Moral hazards
Moreover, these subsidized loans created moral hazards and distorted firm and FI incentives. As the
government directed banks’ lending, they did not evaluate whether borrowers could pay back their loans.
Unlike failing small firms that exited, politically-connected chaebol firms facing defaults received subsidized
loans (e.g., 1972 debt crisis, 1979-1983 depression, 1984-88 firm insolvencies). This government behavior
led to the belief that some firms were “too big to fail” (Cho & Kim, 1997; Joh, 2001; Kim 1991; S. Lee,
1995; Yang, 1997).
2
The share of heavy and chemical industry in GDP was 11.9% in 1970, 26.3% in 1980 and 31.3% in 1988. (Source: Bank
of Korea, Input-output Tables)
3
These rates were lower than time deposit rates or inflation rates (Cho and Kim, 1997).
4
According to Yoo (1997), the number of subsidiaries of the largest 30 chaebols grew from 126 in 1970 to 429 in 1979.
5
Industrial groups (chaebols) owned shares in Korea's commercial banks, but did not control them. To reduce chaebol
influence, government regulations required either low chaebol ownership stakes in banks or multiple chaebol owners. Thus,
commercial banks were not affiliated with any chaebol. Although chaebols did not control commercial banks, they influenced
many other FIs, including merchant banks, insurance firms, securities etc. (CITE XXX)
12
Even when the government did not bail out distressed firms, government supervision and bankruptcy
proceedings aggravated moral hazards. For example, firms successfully lobbied the government for favorable
treatment from FIs. Bank supervisory authority members' close relationships with these FIs also created an
incentive against disciplining the FIs' for their unprofitable lending practices. Furthermore, supervisory
authority members often moved to high positions at these FIs, similar to the ‘amakudari,’ in Japan. (Hanazaki
and Horiuchi, 2000).
3.3 Financial sector at the time of the crisis
These conditions generally persisted through the 1990's. Poor firm performance continued (Joh,
2003), non-performing loans rose, and FIs weakened (Hahm, 2003). Government ceilings on lending interest
rates continued until 1997 (Graham, 2003). So, maximizing FI profits required minimizing losses to nonperforming loans and lending to the most credit-worthy borrowers.
Nevertheless, FIs lending to firms increased, especially to financially distressed firms (Joh, 2004). In
1997, the average debt to equity ratio of Korean firms (396%) was much higher than that of their
counterparts in Taiwan (86%), the US (154%), or Japan (193%).6 Lee and Lee (1998) showed that, ceteris
paribus, chaebols had higher debt-equity ratios than non-chaebol firms. Financially distressed firms with
equity losses had very high debt equity ratios, often reaching 1,000%. (Joh, 2004)
Beginning in January 1997, a series of chaebols defaulted on their loans and failed, causing a cascade
of bankruptcies among affiliated firms. Joh (2004) showed that the total debt of the six conglomerates that
went bankrupt before the agreement with the IMF in 1997 was 24.02 trillion won, 35.5% of the government
budget and 5.3% of GNP. She argued that banks were too weak to absorb the huge losses incurred by the
non-performing loans. They exhausted their own debt capacity as foreign investors stopped revolving loans
to them, thereby triggering the 1997 financial crisis.
The government responded in part by passing a law requiring the Supervisory Regulatory Authority
(SRA) to evaluate each FI’s capital adequacy ratio and take appropriate action (recommendation, request or
order depending on the degree of the problem). This law applied to commercial banks, merchant banks and
6
For Taiwanese firms, the figure is based on 1996 data.
13
securities firms in April 1998, to insurance firms in June 1998 and to mutual savings banks in December
1999.
4. Data and Methods
We examined how FIs made lending decisions. The borrower information that FIs used to make loan
decisions was not readily available. So, we tested whether the firm’s capital structure reflected the available
information on its financial stability and its likelihood of repayment. Firms with higher debt-equity ratios
were more likely to face bankruptcy (Kang, et al., 2000) than firms with lower debt-equity ratios. But debt
ratio alone is not sufficient for us to evaluate FIs’ lending decisions. Profitable firms can have high debt
ratios because they have high growth potential or temporary financial distress. Instead of simply relying on
debt-equity ratios, we examined firms' capital structures and ex-post profits. Specifically, we tested (a)
whether a firm's capital structure reflected its financial distress, (b) the inside information hypothesis, and (c)
the political lending, related lending, business group connection and FI governance hypotheses. As noted
earlier, government imposed ceilings on interest rates, so FIs could not charge higher interest rates to riskier
borrowers. Instead, FIs had an economic incentive to lend conservatively to the most credit-worthy
borrowers, especially as economic growth slowed during the 1990's.
4.1 Data
We used firms’ financial statements from the National Information and Credit Evaluation's (NICE)
database during 1990 -2000.7 Information on firm’s political connectedness was collected from NICE and
Korea Information Service. Each firm submitted a financial statement to the Korea Securities Supervisory
Board. Then, NICE checked the integrity of the data. To reduce the likelihood of problems associated with
accounting standards, this study only used firms subject to outside auditing. About 29% of the firms in this
data were publicly traded.8 All the firms used in the analyses had at least 6 billion won in assets in 1997. FIs
and state-controlled firms were not included. We identified a firm as a "chaebol-affiliated" if it belonged to
one of the 70 largest business groups in 1995 (as measured by size of assets). About 7% of all firms in the
7
Financial statements of Korean firms are available from two major credit-evaluating firms, NICE (National Information
Credit Evaluation) and KIS (Korea Information Service).
14
data belonged to one of these 70 chaebols.
FIs' revenues depended on firm payment of its debt service. So we used a dummy variable for expost firm’s ability to pay its debt. We measured a firm’s ex-post performance in two ways. First, using logit
and probit models, we examined whether a distressed firm’s future ability to pay its debt improved or not. We
created dummy variables to capture (a) whether a firm's operating income covers the financial cost and (b)
whether ordinary income controlling for financial expenditure/revenue was positive. Second, using a within
unit-panel data analysis with two-way (industry and time) fixed effects, we tested whether financial distress
in the current period affected firms’ future returns on assets.
We measured financial distress in two ways, with ordinary income and with operating income. We
assigned a value of one to Distress, a dummy variable, when a firm had negative ordinary income at times t,
t-1 and t-2. When we analyzed the data using a distress dummy which takes 1 if financial expenses exceeded
operating income at times t, t-1 and t-2, we have the same results. Because the results are fundamentally
same, we did not report these results in the paper.
Accounting profitability was likely a better performance measure than stock market-based measures
for three reasons. First, a firm's accounting profitability was more directly related to its financial survivability
than its stock market value was. Although a firm with low stock market return reflected low investor
expectations, it does not necessarily financial distress as the firm’s operating profits could still exceed its debt
payments. Second, accounting measures allowed us to evaluate the performance of privately held firms as
well as that of publicly traded firms. Third, stock prices were less likely to reflect all available information in
inefficient stock markets.
We also defined liabilities in three ways: total debt, interest-bearing borrowing, and FI loans. Total
debt included some components that did not bear interest such as provisions for employee retirement and
pension plans, promissory notes for future payment of goods received, and so on. Borrowing was more
closely related to the firm's financial distress because its failure to pay interest could cause bankruptcy.
8
In 1997, there were 1,135 publicly traded firms; 776 were listed on the Korea Stock Exchange and the rest were
registered with KOSDAQ.
15
However, interest-bearing borrowing included bonds issued by firms. Since publicly issued bonds did not
necessarily reflect FI lending decisions, we also examined loans specifically from FIs. Debt growth rate
t+2 / t-2, borrowing growth rate t+2 / t-2, and loan growth rate t+2 / t-2 measured the changes in debt,
borrowing, and loan levels between t-2 and t+2, respectively.
We measured chaebol affiliation and political connections as follows. Chaebol affiliated firms were
tightly connected to one another through an internal capital market, cross-debt payment guarantees, and
ownership of one another's shares. As noted above, we defined a firm as chaebol affiliated if it belonged to
one of the top 70 chaebols. For 239 business groups, we identified group affiliated FIs, including as banks,
insurance, securities, mutual savings banks and so on. Then, we use a log value of accumulated assets of
affiliated FIs, which captured the size of an FI's available loans. When a firm does not have an affiliated FI,
we assigned a minimum value to the variable. We assigned a value of one to FI-Affiliated, a dummy variable,
when a firm belonged to a business group that had at least one FI. Regional connection is a dummy variable
that has a value of one when a firm has a top level manager from the same region as the sitting president.
Other explanatory variables included the log value of firm size (log value of total assets), fixed asset
over total asset (proxy for collateral), stock exchange listing dummy, year dummies, and industry dummies.
Stock exchange listing dummy takes one when a firm is listed on the Korea stock exchange or on KOSDAQ.
Publicly traded firms are subject to regulations on their capital structure, and they are subject outside scrutiny.
So, the capital structures of listed firms and unlisted firms might differ. We controlled time- and industryfixed effects using year dummies and industry dummies (constructed at the 3-digit level). We deflated all
monetary values using the producer price index in 2000.
Table 1 summarizes the variables used in the study both overall and for each year.
<Insert Table 1: Overall and yearly summary statistics >
4.2 Methods
We test the above hypotheses using three sets of regressions. In the first set of regressions, we
examined whether financial distress affected a firm’s leverage ratio or its leverage growth rate.
16
yit   0  1  Sizeit   2  ROAit   3  Fixed Asset Ratio it   4  Distress
 5  Control Variables it   it
(1)
The dependent variable, yit, was the leverage ratio of firm ‘i’ at time ‘t’ or growth rate of leverage
between t-2 and t+2, i.e. leveraget+2/leveraget-2. We used several leverage ratios such as FI loans over assets,
borrowing over assets and total debt over assets. We tested for the effects of financial distress on firm's
leverage ratio while controlling for log of firm size, return on asset (ROA), fixed asset over total asset, lagged
leverage (t-1), crisis period, stock market listing, industry dummies, and year dummies. Crisis period
referred to years 1998-2000. We used within unit analysis with two-way fixed effects (industry and year),
where  it is an identical and independently distributed (i.i.d.) error term. The regressions on growth rates of
leverages were identical except for the extra control variable, lagged leverage to asset ratio (t-2).
Secondly, we tested the inside information hypothesis both overall and for each type of connection:
chaebol affiliation, political connection or FI affiliation. We analyzed the determinants of borrowing firm’s
ex-post ability to pay its debt in two ways. First, as a debt holder’s payoff did not increase beyond the debt
payment, we tested whether the borrowing firm’s performance was large enough to cover its debt payment
using two dummy variables.
yi ,t 2   0  1  Distress it   2  Leverage Growthit   3  Distress it  Leverage Growthit
  4  Connection Type   5  Distress it  Connection Type
(2)
  6  Distress it  Leverage Growthit  Connection Type   7  Control Variables it   it
We measured the dummy variable y in two ways: (a) whether a firm’s ordinary income before tax at time t+2
was non-negative and (b) whether a firm’s operating income at time t+2 equaled or exceeded its interest
payment. Using a logit model, we explored whether financial distress, leverage growth (t+1 / t-2),
connection type, or interactions among them affected a firm’s ex-post performance. We controlled for log of
firm size, log leverage, debt over asset ratio at t-1, crisis period, stock market listing, industry dummies, and
year dummies.
In addition, we examined the firm’s rate of return on asset (ROA) ex-post, traditionally used to
measure the firm’s accounting performance. We measured a firm's rate or return in two ways: (a) ordinary
17
income, and (b) operating income minus interest payment. Using the same predictors in the logit model, we
did a within unit analysis with two-way fixed effects (industry and year).
ROAi ,t  2   0  1  Distress it   2  Leverage Growthit   3  Distress it  Leverage Growthit
  4  Connection Type   5  Distress it  Connection Type
(3)
  6  Distress it  Leverage Growthit  Connection Type   7  Control Variables it   it
Thirdly, we examined how political connections, bank affiliation, or FI manager moral hazards
affected FI lending using the same predictors as those in the ROA model but with the control variables from
the first set of regressions along with stock market listing.
yt   0  1  Distress it   2  Leverage Growthit   3  Distress it  Leverage Growthit
  4  Connection Type  5  Distress it  Connection Type
(4)
  6  Distress it  Leverage Growthit  Connection Type   7  Control Variables it   it
The dependent variable was measured in three ways: (a) leverage ratio at time t, (b) growth rate of leverage
between time t-2 and t+2, namely Leverage t+2/Leveraget-2, or (c) ratio of short term leverage over total
leverage. The predictors for short term leverage ratio were the same as those for leverage ratio.
5. Results and Discussion
Korea FIs generally made bad lending decisions. Distressed firms received more loans. Furthermore,
firms in distress, receiving more loans or both were more likely to have ex-post losses.
This article only includes the main results. All tests for robustness are available upon request from
the authors. We also tested combinations of different accounting performance measures, different measures
of distress, probit rather than logit, different FI affiliation measures, and interaction terms among predictors.
All of these results were similar to the main results.
5.1 Leverage ratio of distressed firms
ROA and distress affected both leverage ratios and leverage growth rates (see table 2). Firms with
lower ROA typically had higher leverage ratios, controlling for firm size, past leverage, fixed asset ratio,
stock market listing, crisis period, industry-specific effects and yearly effects. Controlling for ROA as well,
the loan over asset ratios, borrowing over asset ratios, and debt over asset ratios of distressed firms exceeded
18
those of sound firms by an average of 7.4%, 9.5%, and 3.7%, respectively. Likewise, firms with lower ROA
also tended to show higher leverage growth rates, controlling for total debt over asset ratio as well as the
above control variables. Controlling for ROA as well, the growth rates of loan, borrowing and total debt in
distressed firms exceeded those of sound firms by an average of 40.6%, 42.5% and 24.5% respectively.
These results showed that distressed Korean firms managed to increase their leverage, aggravating their
financial vulnerability.
<Insert Table 2: loan/asset, borrowing/asset, total debt/asset, and debt growth rate>
5.2 Ex-post firm ability to pay debt
Distress, leverage growth, and FI-affiliation affected both ex-post ROA and the likelihood of ex-post
profits, controlling for size, leverage, debt over asset ratio, crisis periods, stock market listing, industry
effects and yearly effects (see tables 3 and 4). The ROA of sound firms exceeded that of distressed firms by
an average of 5% (for all measures of leverage). Likewise, the likelihood of ex-post profits in distressed
firms was only 27-29% of that in sound firms (computed by the antilogs of the logit coefficients of distress
when measuring leverage by loan, borrowing and total debt, respectively; antilog [-1.31] = 0.27; antilog
[-1.26] = 0.29; antilog [-1.28] = 0.28). Firms with higher leverage growth also had lower ex-post ROA and
lower likelihood of ex-post profits.
<Insert Table 3: ex-post debt payment ability, using logistic models>
<Insert Table 4: ex-post debt payment ability, using a within unit estimation with two way fixed
effects>
Distressed firms with higher leverage growth were not more likely than other firms to cover its debt
payment. Likewise, no specified subset of distressed firms with higher leverage growth (* political
connection, *chaebol, or *FI-affiliation) were more likely to do so either. Indeed, distressed firms affiliated
with FIs that had higher leverage growth generally showed lower ROA and were less likely to have ex-post
profits (though the effect was not significant for ROA when measuring leverage by loans and not significant
for likelihood of ex-post profits when measuring leverage by borrowing). Thus, FIs did not benefit from
inside information when making loans to these distressed firms.
19
5.3 Factors that affect FI lending
Distress, chaebol affiliation, political connections, and FI-affiliation affected firms' leverage ratios,
controlling for size, past leverage, fixed asset ratio, crisis, and stock exchange listing (see table 5). The
leverage ratios of distressed firms, politically-connected firms, and FI-affiliated firms exceeded those of other
firms on average, though rpolitically-connected or FI-affiliated firms did not have significantly higher debtasset ratios. Among distressed firms, the loan over asset and borrowing over asset ratios of chaebol-affiliated
firms tended to exceed those of independent firms. The results showed no other interaction effects. A positive
effect of political connection on firm’s loan ratio, and borrowing over asset showed that politically connected
firms borrowed more, consistent with Faccio (2004). In addition, a positive relation between leverage ratio
and FI affiliation is consistent with LLZ (2003 ). Likewise, a positive effect of chaebol affiliation is also
consistent with Lee and Lee (1998) .
<Insert Table 5 leverage ratios of loan, borrowing, total debt>
Distress, political connection, and FI affiliation also affected firms' leverage growth, controlling for size,
past leverage, past leverage ratio, fixed asset ratio, crisis time periods, and stock exchange listing (see table
6). The leverage growth of distressed firms, politically-connected firms, and FI affiliated firms exceeded
those of other firms, though politically-connected firms of firms did not have significantly higher loan growth.
The results showed no interaction effects.
<Insert Table 6: growth rate of loan, borrowing, total debt>
These two sets of results showed that political connections, chaebol affiliation, and FI affiliation can
help firms increase their leverage. Controlling for these connections and affiliations however, the leverage
ratios and leverage growths of distressed firms still exceeded those of sound firms, indicating that these links
were not necessary for distressed firms to increase their leverage.
5.4 Factors affecting short term leverage over total leverage ratios
ROA, distress, chaebol affiliation, political connection, and FI affiliation all affected short-term
leverage over total leverage ratios, controlling for size, past leverage, fixed assets, the crisis period, industry
effects and yearly effects (see table 7). Firms with higher ROA had lower short-term leverage over total
20
leverage ratios. Likewise, the short-term loan ratios of distressed firms exceeded those of sound firms by
3.9% on average. Those of politically connected firms exceeded those of firms without political connections
by an average of 1.7%. The short-term loan ratios of chaebol-affiliated firms averaged 1.8% higher than those
of independent firms. Lastly, those of FI-affiliated firm typically exceeded those of firms unaffiliated with
FIs.
<Insert Table 7: short term debt>
These results suggest that FIs were aware of firms in financial distress as well as those with chaebol
affiliations, political connections or FI affiliations. As overall FI lending to distressed firms increased, a
higher short-term loan ratio suggests that FIs recognized borrowing firms’ distress by reducing their exposure
to risk with additional short-term loans. The higher short-term leverage ratios of chaebol affiliated firms also
suggest that FI managers recognized these firms' possible asset exaggerations and possible undeclared extent
of debt payment guarantees. Likewise, these results suggest FI manager awareness of the greater risk
exposure of lending to politically connected or FI affiliated firms. Nonetheless, FI managers still gave loans
to distressed firms without political connections or FI affiliations. These results support the argument that FI
managers used short-term loans to avoid blemishes on their reputation rather than accepting responsibility for
poor lending decisions.
6. Conclusion and implications
Korean FIs generally made poor lending decisions during the 1990’s. Despite publicly available
information, they increased lending to financially distressed firms more than to sound firms. These
distressed firms were not credit-worthy as they typically had ex-post losses, so FIs did not generally benefit
from inside information. Among distressed firms, firms with business group affiliations received more loans
than independent firms did, suggesting that these firms capitalized on their business group's resources.
Political connections and FI affiliations both affected lending decisions, supporting the claims of harmful
political lending and related lending effects. Among firms without high future profit potential, without
business group affiliations, without political connections and without FI affiliations, lending to distressed
firms also tended to exceed lending to sound firms, showing that distressed firms did not need political
21
connections or FI affiliations to obtain further credit. Lastly, the short term loan ratios of distressed firms,
business group-affiliated firms, politically-connected firms, and FI-affiliated firms typically exceeded those
of other firms, suggesting that FIs were reducing their exposure to the greater risk of lending to these firms.
The latter two results suggest that FI managers recognized the financial distress of these firms but gave them
short-term loans due to manager reputation incentives.
The results of this study suggest the need to address these harmful effects, specifically by reducing
external influences on lending decisions and improving FI governance. As FI links to government and
borrower firms can affect lending decisions, weakening or severing these links can reduce their effects. The
government influences FIs through FI ownership by government agencies, selection of bank CEOs, and so on.
The government can weaken these links and increase the independence of the financial sector by privatizing
more banks, and remaining silent during the selection of bank CEOs. Firms are affiliated to FIs through
ownership, cross-debt payment guarantees, interlocking personnel, etc. To reduce these lender-borrower
links, the government can ban firm ownership of FIs, ban cross-debt payment guarantees, and require
declaration of employees’ conflicts of interest. Greater transparency through enhanced reporting
requirements for both FIs and firms (e.g., terms of loans, borrowing firm’s financial stability) can reduce the
effects of firm affiliations with FIs. Lastly, improving FI governance is needed. Prudence regulation,
protection of shareholder rights, greater transparency of FI management and financial sector competition can
help improve FI governance by disciplining FI managers and aligning their incentives with those of FI
shareholders.
FI lending affects an economy’s efficiency, as FIs supply resources to the corporate sector. Therefore,
enhancing independence and improving FI governance might improve both FIs’ own performances and the
economy’s efficiency through better allocation of resources.
22
References
Aghion, P. and P. Bolton (1992) “An incomplete contracts approach to financial contracting” Review of
Financial Studies, 59, 473-494
Allen, Franklin (1993) “Stock markets and resource allocation” in C. Mayer and Xavier Vives (eds) Capital
markets and financial intermediation, Cambridge University Press
Allen, Franklin and D. Gale (1999) “Diversity of opinion and financing of new technologies,” Journal of
Financial Intermediation 8
Aoki, M. H. Patrick and P. Sheard (1994) “The Japanese main bank system: An Introductory overview” in
Aoki, M. and H. Patrick (eds.), The Japanese Main bank system, Oxford University Press
Bank of Korea (various years) “Economic Statistics Year Book” Seoul, Bank of Korea
Charumilind, Chutatong, Raja Kali, and Yupana Wiwattanakantang (2003) “Connected lending: Thailand
before the financial crisis” Center for Economic Institutions working paper 2003-19, Hitotsubashi
University.
Cho, Yoon Je and Joon Kyung Kim (1997) Credit Policies and Industrialization of Korea, Seoul, Korea
Development Institute
Corsetti, Giancarlo, Paolo Pesenti, and Nouriel Roubini. (1999) “What Caused the Asian Currency and
Financial Crisis?” Japan and the World Economy, 11, 305-373.
Dewatripont, M. and Jean Tirole (1994) “A Theory of Debt and Equity: Diversity of Securities and ManagerShareholder Congruence,” Quarterly Journal of Economics, 439, 1027-1054
Diamond, Douglas W., and Raghuram Rajan (2002). “Banks, short term debt and financial crises: theory,
policy implications and applications”, NBER Working paper # 7764
Dinc, I. Serdar (2004) “Government Ownership of Banks and Political Lending in Emerging Markets”,
working paper, Michigan University
Faccio, Mara (2004) “Politically Connected Firms” Working paper, Vanderbilt University
Fama, Eugene (1985) “What's different about banks,” Journal of Monetary Economics, 29-36
Fama, Eugene (1990) “Contract costs and financing decisions” Journal of Business, s7-91
Graham, Edward (2003) Reforming Korea’s Industrial Conglomerates, Institute for International Economics,
Washington DC.
Hall, Brian and David Weistein (2000) “Main banks, Creditor Concentration, and the Resolution of Financial
Distress in Japan” in Aoki and Saxonhouse (eds) Finance, Governance and Competitiveness in Japan,
Oxford University Press
Holmstrom, Bengt and Jean Tirole (1993) “Market liquidity and performance monitoring” Journal of Political
Economy 101
Holmstrom, Bengt and Jean Tirole (1997) “Financial intermediation, loanable fund, an the real sector”
Quarterly Journal of Economics CXII,
Hanazaki Masaharu and Akiyoshi Horiuchi (2000) “Is Japan’s Financial System Efficient” Oxford Review of
Economic Policy, Vol 16, No. 2 pp 61-73
Hart, Oliver and John Moore (1989) “Default and Renegotiation: A dynamic model of debt” working paper
23
520, MIT
Hoshi, Takao, Anil Kashyap and David Scharfstein (1991a) “Corporate Structure, Liquidity and Investment:
Evidence from Japanese Industrial Groups,” Quarterly Journal of Economics 106,
Hoshi, Takao, Anil Kashyap and David Scharfstein (1991b), “The role of banks in reducing the costs of
financial distress in Japan” Journal of Financial Economics
Joh, Sung Wook, (2001), "The Korean Corporate Sector: Crisis and Reform," in Kwon and Shepherd (eds.)
Korea’s Economic Prospects: from Financial Crisis to Prosperity, Edward Elgar Publishing.
Joh, Sung Wook, (2003), “Corporate Governance and Firm Profitability: Evidence from Korea before the
Crisis, Journal of Financial Economics.
Joh, Sung Wook, (2004), “Korean Economic Crisis and Corporate Governance System” in Ito and Krueger
(eds) Governance, Regulation and Privatization in the Asia Pacific Region, NBER-EASE, Vol. 12,
Chicago University
Kang, David, (2002) “Bad loans to good friends: Money politics and the developmental state in Korea”
forthcoming in International Organization. Mimeo, Dartmouth College.
Kang, Dong Soo. Youngjae. Lim, Chinhee. Hahn, Jinwon Kang, and Jeongsam Yang (2000) firm failure and
corporate restructuring, KDI-policy series 200-01, KDI
Kang, Jun-Koo. (1998) "Bank oriented corporate governance system: cost and benefit of bank ownership in
commerce, " economic analysis V. 4, N 4, BOK
Krugman, Paul (1998) “What happened Asia” mimeo
KDI (2000) Policy recommendations on corporate exit mechanism, Seoul, KDI
Kim, Joon Kyung, 1999, “Chaebol’s Ownership of NBFIs and Related Problems,” KDI Economic Outlook
16 (1)
La Porta, Lopez-de-Silanes, Shleifer and Vishny (1997) “Legal Determinants of external finance”, Journal of
Finance
La Porta, Rafael, Florencio Lopez-de-Silanes and Andrei Shleifer (2000) "Government Ownership of Banks,"
NBER Working paper No. 7620
La Porta Rafael, Florencio López-de-Silanes Guillermo Zamarripa. (2003). “Related Lending” Quarterly
Journal of Economics V. 118, N. 1, 231-268
Lee, Byong Ki, 1998, Debt Payment Guarantees of Korean Chaebols, Korea Economic Research Institute.
Lee, J. and Y. Lee(1998) “Capital structure of Korean firms: Comparison of Chaebol firms and non-chaebol
firms 1981-96” paper presented at the Korean Meeting of Econometric Society
Lee, Soon Woo (1995), Reorganization of distressed firms, Jilritamgoo, Seoul
Lim, Youngjae, 2002, “corporate bankruptcy system and economic crisis in Korea” forthcoming in S.
Haggard, W. Lim and E. Kim (eds) Chaebol reform and corporate restructuring, Cambridge
University
Manove, Michael, A. Jorge Padilla and Marco Pagano (2001) “Collateral versus project screening: a model of
lazy banks,” Rand Journal of Economics 32(4) 726-744
Mitchell Janet (1998) "Strategic Creditor Passivity, Regulation and Bank Bailouts", Discussion Paper No.
24
1780, Center for economic policy research
Nam, Il Chong and Soogeun Oh (2000) “Bankruptcy of Large Firms and Exit Mechanisms in Korea”
Research Monograph 2000-01, Korea Development Institute
Peek, Joe and Eric S. Rosengren (2003) “Unnatural Selection: perverse incentives and the misallocation of
credit in Japan” NBER working paper, # 9643
Rajan, Raghuram G. (1992) “Insiders and Outsiders: the choice between informed and arm's length debt”
Journal of Finance vol 47, 4, 1367-1400
Rajan, Raghuram G and Zingales (1995) “What do we know about capital structure? Some evidence from
international data”, Journal of Finance 50, 1421-1460
Shin, D. (1999) “Capital structure of distressed firms and bankruptcy prediction model” Hanil Kyongsang
nonchong, 131-156
Song, H. (2000) “Risk and moral hazard of Korean banking sector in the 1990” KDIC Financial study 1(3)
3-22,
Smith, Clifford and Jerold Warner (1979) “On financial contracting: an analysis of bond covenants” Journal
of Financial Economics, 111-161
Weinstein, D. and Y. Yafeh (1998) “On the Cost of a bank-Centered Financial System: Evidence from the
Changing Main Bank Relations in Japan” Journal of Finance 635-672
25
Table 1: Summary Statistics of the variables used in the analysis.
Chaebol is a dummy variable that takes 1 when a firm belongs to top 70 largest chaebols measured in assets
in 1995. Distress is a financial distress dummy variable that takes 1 when a firm’s operating income falls
short of its interest payment. Panel A shows the summary statistics over 1992-1998 and Panel B shows the
mean value for each year.
Panel A:
name of variables
Return on Asset (%)
Return on Asset at t+2 (%)
bank loan/asset (%)
borrowing/asset (%)
total debt/asset (%)
Short term loan/loan (%)
fixed asset/asset (%)
growth of loan t+2/t-2
growth of borrowing t+2/t-2
growth of total debt t+2/t-2
Log value of size
Log Asset of affiliated firms
Debt payment ability dummy at t+2
Distress dummy
bank affiliated dummy
Chaebol dummy
connection dummy
publicly traded firm dummy
Distress * Chaebol
Distress * log value of FI affiliated
Distress * Connection
number
29415
29415
29415
29415
29415
28327
29415
29415
29415
29415
29415
29415
29415
29415
29415
29415
29415
29415
29415
29415
29415
mean
0.975
0.026
35.4
43.8
80.0
68.7
35.9
1.506
1.285
0.788
10.319
-12.259
0.554
0.234
0.078
0.067
0.076
0.292
0.018
-2.569
0.018
std. dev
12.4
19.2
27.6
30.2
30.5
26.4
22.9
5.635
4.857
2.784
1.270
5.401
0.497
0.423
0.268
0.250
0.265
0.455
0.135
5.622
0.135
minimum
-800.2
-973.5
0.0
0.0
0.3
0.0
0.0
-1
-1
-0.996
4.840
-13.816
0
0
0
0
0
0
0
-13.816
0
maximum
72.7
160.2
930.7
930.7
932.3
100.0
163.6
98.0
98.0
92.9
17.8
10.8
1
1
1
1
1
1
1
10.1
1
26
PanelB:
name of variables
Return on Asset (%)
Return on Asset at t+2 (%)
bank loan/asset (%)
borrowing/asset (%)
total debt/asset (%)
Short term loan/loan (%)
fixed asset/asset (%)
growth of loan t+2/t-2
growth of borrowing t+2/t-2
growth of total debt t+2/t-2
Log value of size
Log Asset of affiliated firms
Debt payment ability dummy at t+2
Distress dummy
bank affiliated dummy
Chaebol dummy
connection dummy
publicly traded firm dummy
Distress * Chaebol
Distress * Connection
Distress * FI affiliated
Number of observations
1992
mean
1.939
2.378
35.2
43.3
78.1
71.0
36.4
1.532
1.490
0.951
10.27
-12.06
0.593
0.221
0.090
0.079
0.074
0.330
0.018
0.018
-2.569
3414
1993
mean
1.912
1.481
34.1
42.3
78.6
70.5
36.6
1.674
1.479
0.955
10.22
-12.12
0.538
0.228
0.086
0.075
0.080
0.300
0.018
0.024
-2.669
3947
1994
mean
2.431
0.980
34.1
42.3
79.6
70.2
35.5
2.052
1.706
1.091
10.25
-12.23
0.533
0.233
0.080
0.070
0.077
0.278
0.018
0.025
-2.713
4396
1995
mean
1.446
-2.714
35.7
43.8
80.8
69.4
34.6
2.065
1.808
1.138
10.29
-12.29
0.501
0.238
0.076
0.064
0.077
0.272
0.019
0.025
-2.788
4669
1996
mean
1.322
-2.185
35.4
43.2
80.2
68.3
35.2
1.188
1.012
0.564
10.37
-12.32
0.471
0.239
0.073
0.061
0.075
0.280
0.019
0.023
-2.830
4749
1997
mean
-1.728
0.898
37.9
47.1
83.9
65.8
34.9
1.011
0.816
0.475
10.43
-12.33
0.618
0.242
0.073
0.061
0.083
0.284
0.017
0.024
-2.885
4394
1998
mean
-0.416
0.417
35.6
44.5
77.8
66.2
38.7
0.970
0.665
0.334
10.39
-12.42
0.665
0.233
0.069
0.062
0.064
0.314
0.017
0.014
-2.871
3846
27
28
<Table 2> This Table shows determinants of various leverage ratios at time t in Panel A and growth rates between t-2 and t+2, namely Leverage t+2/ Leverage
t-2 in Panel B. Leverage are measured by bank loan over asset, borrowing over asset, and total debt over asset. Financial distress is a dummy that takes 1
when a firm’s operating income falls short of its financial cost at time t, t-1 and t-2. Numbers in parentheses are standard errors.
PANEL A:
Leverage is measured by
Bank loan
Predictor
LOGSIZE
Eqn
Eqn
-3.550 ***
(0.131)
Log Leverage at t-1
2.692 ***
(0.054)
ROA
-0.984 ***
(0.011)
Fixed Asset/Asset
0.024 **
(0.008)
Crisis
Interest bearing borrowing
-3.388 ***
(0.545)
Distress
Eqn
-3.636 ***
(0.130)
2.607 ***
(0.054)
-0.903 ***
(0.012)
0.015 *
(0.008)
-3.267 ***
(0.541)
Eqn
-2.994 ***
(0.138)
3.365 ***
(0.064)
-1.218 ***
(0.012)
0.063 ***
(0.008)
-3.619 ***
(0.557)
7.437 ***
Total debt
Eqn
-3.057 ***
(0.136)
3.214 ***
(0.063)
-1.116 ***
(0.012)
0.052 ***
(0.008)
-3.448 ***
(0.550)
Eqn
-36.496 ***
-36.038 ***
(0.309)
(0.311)
36.718 ***
36.149 ***
(0.303)
(0.306)
-1.015 ***
(0.010)
-0.023 ***
(0.011)
-0.027 ***
(0.007)
(0.007)
-12.162 ***
-11.980 ***
(0.469)
(0.469)
9.501 ***
(0.350)
-0.980 ***
3.665 ***
(0.357)
(0.303)
Industry dummies
Yes
Yes
Yes
Yes
Yes
Yes
Year dummies
Yes
Yes
Yes
Yes
Yes
Yes
0.328
0.339
0.41
0.424
0.596
0.598
R-square
29
PANEL B:
Debt is measured by
Bank loan
Predictor
LOGSIZE
Eqn
2.947 ***
(0.041)
Log Leverage at t-1
-3.151 ***
(0.034)
Total debt over
asset at t-2
ROA
Fixed Asset/Asset
Crisis
2.990 ***
(0.126)
-0.040 ***
Interest bearing borrowing
Eqn
2.953 ***
(0.041)
-3.169 ***
(0.034)
2.913 ***
(0.127)
-0.037 ***
Eqn
2.810 ***
(0.038)
-2.838 ***
(0.032)
2.407 ***
(0.112)
-0.042 ***
Eqn
2.820 ***
(0.038)
-2.861 ***
(0.032)
2.333 ***
(0.113)
-0.038 ***
Total debt
Eqn
3.327 ***
(0.035)
-3.513 ***
(0.035)
2.573 ***
(0.067)
-0.027 ***
Eqn
3.332 ***
(0.035)
-3.526 ***
(0.035)
2.526 ***
(0.068)
-0.025 ***
(0.002)
(0.003)
(0.002)
(0.002)
(0.001)
(0.001)
0.000
0.000
0.000
0.000
-0.001
-0.001
(0.002)
(0.002)
(0.001)
(0.001)
(0.001)
(0.001)
-0.017
-0.007
(0.098)
(0.098)
Distress
-0.195 *
(0.085)
0.406 ***
-0.183 *
(0.085)
0.163 ***
(0.047)
0.425 ***
(0.075)
0.170 ***
(0.047)
0.245 ***
(0.066)
(0.036)
Industry dummies
Yes
Yes
Yes
Yes
Yes
Yes
Year dummies
Yes
Yes
Yes
Yes
Yes
Yes
0.258
0.259
0.244
0.246
0.299
0.300
R-square
*p < .05, **p < .01, ***p < .001
30
<Table 3> This table summarizes the logit regression results on firms’ ex-post ability to make their debt payments. The dependent variable is the binary
variable that takes 1 when the ordinary income of the firm at time t+2 is non-negative. Numbers in parentheses are standard errors.
Leverage is measured by
Loan
Predictor
Constant
equation
4.015
***
(0.246)
log of size
-0.056
-0.07
***
-0.825
at t-1
(0.065)
Crisis Period
0.501
***
-0.114
***
-0.07
***
-0.756
***
0.51
-0.114
-0.07
***
-0.758
***
0.509
4.42
***
-0.113
***
-0.07
***
-0.757
***
0.511
***
0.235
-0.347
***
-0.424
***
0.468
3.801
***
0.177
***
-0.352
***
-0.345
***
0.479
3.799
0.176
-0.352
***
-0.345
***
0.478
(0.046)
Distress
-1.298
dummy (A)
(0.035)
Leverage growth
-0.018
t+1/t-2(B)
(0.003)
(0.003)
(0.003)
(0.003)
(0.004)
(0.004)
(0.004)
A*B
-0.008
-0.005
-0.002
-0.007
0.016
0.018
0.021
(0.008)
(0.009)
(0.008)
(0.009)
(0.009)
(0.009)
(0.009)
***
-1.294
(0.047)
***
(0.036)
***
-0.017
-1.293
(0.047)
***
(0.036)
***
-0.017
-1.299
(0.047)
***
(0.036)
***
-0.017
-1.268
(0.047)
***
(0.035)
***
-0.035
-1.264
(0.036)
***
-0.034
***
-1.261
***
-0.034
***
0.176
-0.352
***
-0.345
***
0.479
***
***
***
(0.047)
***
-1.266
*
-0.034
***
0.867
-0.992
***
***
3.448
***
0.777
***
-0.957
equation
3.441
***
(0.265)
***
(0.077)
0.777
-0.957
3.446
***
0.778
***
-0.957
(0.075)
(0.075)
0.161
0.18
0.177
0.18
(0.096)
(0.094)
(0.094)
(0.094)
***
-1.283
-0.032
0.433
***
(0.047)
***
-1.282
-0.031
***
(0.047)
***
(0.036)
***
0.432
-1.278
-0.031
0.435
***
***
(0.047)
***
(0.037)
***
***
(0.076)
(0.075)
0.422
***
(0.265)
(0.077)
***
equation
(0.076)
(0.035)
***
equation
(0.265)
(0.047)
(0.036)
***
3.013
(0.077)
(0.069)
***
equation
(0.258)
(0.021)
(0.036)
***
3.801
(0.026)
(0.047)
***
equation
(0.257)
(0.069)
Dummy
(0.047)
***
(0.021)
(0.069)
***
equation
(0.026)
(0.021)
***
Debt
(0.257)
(0.026)
(0.069)
***
equation
(0.257)
(0.021)
(0.065)
***
3.365
(0.025)
(0.008)
***
equation
(0.25)
(0.016)
(0.065)
***
equation
(0.253)
(0.008)
(0.065)
***
***
(0.016)
(0.008)
***
4.417
(0.253)
(0.016)
(0.008
Debt/Asset
4.42
equation
(0.253)
(0.014)
log Leverage
equation
Borrowing
-1.286
***
(0.037)
***
-0.031
(0.004)
(0.008)
(0.008)
(0.008)
(0.008)
0.017
0.008
0.014
0.016
0.008
(0.009)
(0.017)
(0.016)
(0.016)
(0.017)
***
31
CHAEBOL
0.078
0.081
0.081
0.055
0.06
0.062
0.064
0.064
0.069
(0.075)
(0.068)
(0.068)
(0.075)
(0.068)
(0.068)
(0.075)
(0.068)
(0.068)
0.048
0.054
0.047
0.079
0.088
0.079
0.059
0.069
0.058
dummy (D)
(0.064)
(0.072)
(0.064)
(0.065)
(0.073)
(0.065)
(0.064)
(0.072)
(0.064)
connection
-0.193
dummy (E)
(0.051)
(C)
Bank affiliated
publicly traded
0.313
firm dummy
(0.036)
A*C
A*B*C
***
-0.194
***
(0.051)
***
0.313
-0.211
-0.171
***
(0.058)
***
(0.036)
0.313
***
(0.052)
0.336
***
(0.036)
(0.036)
-0.172
***
(0.052)
***
0.336
(0.036)
-0.184
-0.189
**
(0.058)
***
0.336
(0.052)
0.3
***
(0.036)
(0.036)
0.031
0.045
0.108
(0.127)
(0.127)
(0.132)
-0.023
-0.021
-0.133
(0.028)
(0.031)
(0.077)
A*D
A*B*D
***
-0.19
(0.052)
***
0.301
0.037
0.093
(0.117)
(0.118)
(0.122)
-0.068
-0.17
(0.041)
(0.073)
*
(0.048)
A*E
A*B*E
-0.212
***
0.301
(0.036)
*
0.083
0.062
0.094
(0.117)
(0.119)
(0.121)
-0.006
-0.008
-0.001
(0.04)
(0.046)
(0.067)
Industry
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
0.149
0.152
0.152
0.151
0.157
0.160
0.160
0.160
0.151
0.154
0.154
0.154
dummies
Year dummies
R-square
*p < .05, **p < .01, ***p < .001
***
(0.058)
(0.036)
0.065
-0.111
***
***
32
<Table 4> Ex-post firm performance when within unit analysis with controlling for two way (industry and time) fixed effects. The dependant variable is
ordinary income over total asset ratio at time t+2. A firm is considered financially distressed if its operating income falls below the financial cost for three
years. Numbers in parentheses are standard errors
Leverage is measured by
Loan
Borrowing
Debt
Predictor
equation
equation
equation
equation
equation
equation
equation
log of size
-0.681
-0.955
-0.957
-0.955
1.662
1.366
1.352
***
(0.118)
log Leverage
-0.459
(0.133)
***
(0.068)
Debt/Asset
-11.553
at t-1
(0.465)
Crisis Period
-0.871
Dummy
(0.378)
Distress
-6.549
dummy(A)
***
-0.455
(0.133)
***
(0.068)
***
-11.302
-0.836
***
(0.393)
-6.313
*
-11.316
-0.841
***
-6.280
***
***
-11.298
*
-0.842
***
-6.343
*
-8.744
-1.123
***
-5.892
***
***
-8.458
**
-1.075
***
-5.653
**
-8.489
-1.101
***
-5.712
***
-2.700
-8.456
**
-1.085
***
-5.649
***
***
-7.281
-5.101
**
-1.497
***
-6.229
***
***
-7.119
-5.036
***
-1.454
***
-6.022
***
***
-7.108
-5.066
***
-1.492
***
-5.987
***
-7.123
-5.032
***
-1.462
***
-6.045
t+1/t-2(B)
(0.027)
(0.027)
(0.027)
(0.027)
(0.031)
(0.031)
(0.031)
(0.031)
(0.061)
(0.061)
(0.061)
(0.061)
A*B
-0.010
-0.021
0.000
-0.011
-0.062
-0.078
0.053
-0.055
0.005
0.000
0.049
0.008
(0.067)
(0.069)
(0.068)
(0.069)
(0.073)
(0.074)
(0.075)
(0.073)
(0.120)
(0.122)
(0.121)
(0.121)
1.082
1.125
0.900
0.927
1.291
0.968
1.021
(0.592)
(0.559)
(0.558)
(0.589)
(0.556)
(0.555)
(0.591)
(0.558)
(0.556)
Bank affiliated
-0.587
-0.225
-0.522
-0.321
0.046
-0.252
-0.486
-0.106
-0.424
dummy (D)
(0.529)
(0.564)
(0.528)
(0.527)
(0.561)
(0.526)
(0.528)
(0.563)
(0.527)
connection
-0.837
-0.833
-0.657
-0.659
-0.666
-0.422
-0.820
-0.802
-0.639
dummy(E)
(0.426)
(0.426)
(0.454)
(0.424)
(0.424)
(0.452)
(0.425)
(0.425)
(0.453)
CHAEBOL (C)
1.403
*
*
***
***
*
***
1.200
***
*
***
***
***
***
*
***
***
***
(0.408)
-0.103
***
***
(0.380)
Leverage growth
***
***
(0.716)
(0.412)
-0.268
5.757
(0.568)
(0.380)
***
equation
(0.584)
(0.716)
(0.408)
-0.266
5.743
(0.568)
(0.380)
***
equation
(0.584)
(0.716)
(0.394)
-0.269
5.755
(0.568)
(0.380)
***
equation
(0.584)
(0.716)
(0.407)
-0.170
6.155
(0.567)
(0.377)
***
equation
(0.577)
(0.502)
(0.410)
-0.171
***
(0.159)
(0.376)
(0.407)
-0.170
***
(0.501)
(0.377)
***
-2.689
1.366
(0.199)
(0.159)
(0.502)
(0.392)
-0.174
-2.695
***
(0.199)
(0.159)
(0.377)
(0.408)
-0.099
***
(0.499)
(0.378)
***
-2.687
***
(0.199)
(0.159)
(0.468)
(0.412)
-0.100
-0.456
***
(0.191)
(0.068)
(0.378)
(0.408)
-0.099
***
(0.467)
(0.378)
***
-0.455
***
(0.133)
(0.068)
(0.467)
*
***
equation
***
-0.267
***
33
publicly traded
1.514
***
1.507
***
1.515
1.637
***
***
1.628
***
1.641
1.328
***
firm dummy
(0.299)
A*C
-1.889
-1.969
-1.736
(1.302)
(1.299)
(1.341)
0.147
0.273
0.060
(0.238)
(0.273)
(0.560)
A*B*C
A*D
A*B*D
(0.299)
(0.299)
(0.297)
(0.297)
(0.297)
(0.299)
***
1.301
(0.298)
-1.551
-0.213
-0.330
(1.184)
(1.173)
(1.227)
-0.348
-1.666
(0.282)
(0.248)
A*E
A*B*E
-2.358
***
***
1.330
(0.299)
***
(0.6205)
-1.422
-1.346
-1.201
(1.259)
(1.282)
(1.295)
-0.056
-0.838
-0.414
(0.266)
(0.534)
(0.6362)
Industry dummies
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Year dummies
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
0.087
0.088
0.088
0.088
0.096
0.097
0.098
0.097
0.092
0.092
0.093
0.092
R-square
*p < .05, **p < .01, ***p < .001
***
34
<Table 5> This Table shows determinants of various leverage ratios at time t when leverage is measured by bank loan, interest bearing borrowing,
and total debt. Financial distress is a dummy variable that takes 1 when a firm’s operating income falls short of its financial cost at time t, t-1 and t-2.
Numbers in parentheses are standard errors.
Leverage as measured by
loan
Predictor
log of size
Eqn
-3.410
Eqn
***
(0.164)
log Leverage at t-1
3.018
0.067
***
dummy
Distress dummy
(A)
-1.629
***
3.018
0.068
**
-1.636
***
15.311
***
3.018
***
0.067
***
-1.626
***
16.165
3.018
0.067
**
-1.627
***
15.506
***
3.880
***
0.116
***
-1.354
***
19.589
3.880
0.116
*
-1.360
19.356
(0.363)
(0.376)
(0.807)
(0.379)
(0.385)
(0.399)
CHAEBOL
-0.258
-1.238
-0.203
-0.254
-1.027
-1.931
dummy (B)
(0.753)
(0.846)
(0.756)
(0.753)
(0.796)
(0.895)
0.065
0.081
(0.040)
(0.035)
Log Asset of Fis
0.081
affiliated ©
(0.035)
Connection
1.710
dummy (D)
(0.571)
publicly traded
firm dummy
-5.685
(0.399)
*
0.086
*
(0.035)
**
1.720
**
(0.571)
***
-5.687
(0.399)
1.713
**
(0.571)
***
-5.679
(0.399)
1.523
*
-5.686
(0.399)
***
(0.037)
*
(0.670)
***
0.137
2.326
-3.997
(0.422)
0.142
***
2.336
***
-3.999
(0.422)
3.880
0.116
*
-1.353
***
*
***
19.814
***
*
***
0.116
-1.353
***
19.559
***
43.887
***
0.022
***
-12.032
**
11.017
43.887
0.022
***
-12.032
***
11.029
***
43.879
**
0.022
***
-12.037
**
9.685
43.892
0.022
***
-12.037
***
11.110
(0.402)
(0.331)
(0.343)
(0.722)
(0.345)
-1.006
-1.025
-0.411
-0.362
-0.533
-0.417
(0.799)
(0.796)
(0.671)
(0.754)
(0.673)
(0.671)
0.040
0.040
0.077
(0.032)
(0.032)
(0.036)
(0.032)
0.971
0.970
0.965
1.274
(0.509)
(0.509)
(0.509)
(0.597)
**
2.327
-3.995
(0.422)
0.137
***
(0.037)
***
2.231
**
(0.709)
***
-3.998
(0.4220
***
-4.566
(0.356)
***
-4.565
(0.356)
***
-4.579
(0.356)
**
***
(0.532)
(0.854)
0.131
***
(0.007)
(0.532)
***
***
(0.334)
(0.007)
***
-42.929
(0.351)
(0.334)
(0.532)
***
-42.911
Eqn
(0.351)
(0.007)
(0.532)
***
***
(0.334)
(0.007)
*
-42.922
Eqn
(0.351)
(0.334)
(0.624)
(0.604)
***
3.880
-42.921
Eqn
(0.351)
(0.009)
(0.043)
***
***
(0.071)
(0.624)
***
-3.531
Eqn
(0.177)
(0.009)
(0.604)
***
***
(0.071)
(0.038)
(0.604)
***
***
(0.624)
***
-3.533
debt
Eqn
(0.177)
(0.009)
(0.624)
***
***
(0.071)
(0.009)
**
-3.528
Eqn
(0.177)
(0.071)
(0.590)
***
-3.532
Eqn
(0.177)
(0.008)
(0.589)
***
***
(0.058)
(0.008)
**
-3.408
Eqn
(0.164)
(0.058)
(0.589)
***
-3.411
Eqn
(0.164)
(0.008)
(0.589)
15.564
***
(0.058)
(0.008)
crisis period
Eqn
(0.164)
(0.058)
Fixed Asset/Asset
-3.406
interest bearing borrowing
*
***
***
0.040
-4.563
(0.356)
*
***
35
Distress * Chaebol
3.386
(A*B)
*
3.122
(1.331)
Distress * log asset of
FIs affiliated (A*C)
-0.170
*
(1.408)
(1.187)
0.050
0.019
-0.111
(0.060)
(0.064)
(0.054)
Distress * Connection
(A*D)
*
0.655
0.335
-1.064
(1.229)
(1.300)
(1.095)
Industry dummies
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Year dummies
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
0.214
0.214
0.214
0.214
0.26
0.26
0.26
0.26
0.482
0.482
0.482
0.482
R-square
*p < .05, **p < .01, ***p < .001
36
<Table 6> This Table shows determinants of various leverage growth rates between t-2 and t+2, namely Leverage t+2/ Leverage t-2 when leverage is
measured by bank loan, interest bearing borrowing, and total debt. Financial distress is a dummy variable that takes 1 when a firm’s operating income falls
short of its financial cost at time t, t-1 and t-2. Numbers in parentheses are standard errors.
Leverage as measured by
loan
Predictor
log of size
Eqn
2.902
Eqn
***
(0.043)
log Leverage at t-2
-3.114
Fixed Asset/Asset
Crisis Period dummy
Distress dummy (A)
3.121
Eqn
***
(0.043)
***
(0.034)
Debt/Asset at t-2
2.902
-3.114
3.120
2.902
Eqn
***
(0.043)
***
(0.034)
***
interest bearing borrowing
-3.113
3.119
***
(0.043)
***
(0.034)
***
2.902
Eqn
-3.114
3.121
***
(0.040)
***
(0.034)
***
2.692
Eqn
-2.796
2.575
***
(0.040)
***
(0.032)
***
2.692
Eqn
-2.795
2.577
Eqn
***
(0.040)
***
(0.032)
***
2.692
-2.795
2.571
2.691
Eqn
***
(0.040)
***
(0.032)
***
debt
-2.796
2.577
***
(0.035)
***
(0.032)
***
3.210
Eqn
-3.440
2.662
***
-3.441
***
2.665
***
***
-3.442
2.661
***
-3.440
***
2.662
(0.127)
(0.127)
(0.113)
(0.113)
(0.113)
(0.113)
(0.068)
(0.068)
(0.068)
(0.068)
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
-0.001
-0.001
-0.001
-0.001
(0.002)
(0.002)
(0.002)
(0.002)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
(0.001)
0.070
0.070
0.069
0.070
-0.092
-0.091
-0.093
-0.093
0.218
(0.098)
(0.098)
(0.098)
(0.098)
(0.086)
(0.086)
(0.086)
(0.086)
(0.048)
0.697
***
0.694
***
0.603
***
0.696
***
0.731
***
0.753
***
0.468
***
0.765
***
0.457
0.218
(0.048)
***
0.475
(0.159)
(0.076)
(0.064)
(0.066)
(0.139)
(0.066)
(0.035)
(0.036)
CHAEBOL
0.020
0.007
0.011
0.020
0.032
0.119
0.008
0.030
0.116
0.185
dummy(B)
(0.147)
(0.166)
(0.148)
(0.147)
(0.129)
(0.145)
(0.129)
(0.129)
(0.071)
0.017
(0.007)
*
0.017
(0.007)
*
0.020
(0.008)
*
0.017
(0.007)
*
0.025
(0.006)
***
0.025
(0.006)
***
0.032
(0.007)
***
0.025
(0.006)
***
0.007
(0.003)
***
*
0.217
***
(0.048)
***
0.243
0.217
***
***
(0.048)
**
0.464
(0.076)
(0.036)
0.096
0.116
(0.080)
(0.071)
(0.071)
0.006
0.013
(0.003)
(0.004)
*
***
(0.035)
(0.127)
***
***
(0.035)
(0.035)
***
3.209
(0.127)
((0.075)
affiliated (C)
3.211
Eqn
(0.035)
(0.035)
(0.073)
Log Asset of Fis
***
(0.035)
(0.035)
***
3.211
Eqn
**
0.007
(0.003)
***
*
37
Connection dummy
0.201
0.201
0.200
0.198
0.265
(D)
(0.112)
(0.112)
(0.112)
(0.131)
(0.098)
(0.098)
(0.098)
(0.115)
(0.054)
publicly traded firm
-0.183
0.112
0.112
0.109
0.113
0.118
Dummy
(0.079)
(0.069)
(0.069)
(0.069)
(0.069)
(0.038)
Distress *
Chaebol(A*B)
*
-0.183
(0.079)
*
-0.184
(0.079)
*
-0.183
(0.079)
*
**
0.264
**
0.263
**
0.374
**
0.108
*
0.108
(0.054)
**
0.118
(0.038)
0.046
-0.298
-0.239
(0.261)
(0.228)
(0.126)
Distress * log asset of
-0.008
-0.022
FIs affiliated (A*C)
(0.012)
(0.010)
Distress * Connection
(A*D)
*
0.107
(0.054)
**
0.116
0.131
**
0.118
(0.038)
**
(0.006)
0.009
-0.383
-0.080
(0.241)
(0.210)
(0.116)
Industry dummies
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Year dummies
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
0.254
0.254
0.254
0.254
0.239
0.239
0.239
0.239
0.291
0.291
0.291
0.291
R-square
*p < .05, **p < .01, ***p < .001
*
(0.063)
(0.038)
-0.018
*
*
**
38
<Table 7> The Table summarizes the results on determinants of firms’ short term loan ratio over total loan.
Numbers in parentheses are standard errors.
Predictor
Eqn
log of size
log Leverage at t-1
Return on Asset
Eqn
0.657 ***
Eqn
0.537 ***
Eqn
0.536 ***
Eqn
0.538 ***
0.536 ***
(0.146)
(0.146)
(0.146)
(0.146)
(0.146)
1.028
0.741
0.747
0.737
0.744
(0.580)
(0.571)
(0.571)
(0.571)
(0.571)
-0.096 ***
(0.013)
Fixed Asset/Asset
-0.373 ***
(0.008)
Crisis period dummy
-3.318 ***
(0.574)
Distress dummy (A)
-0.373 ***
(0.008)
-3.134 ***
(0.573)
3.911 ***
(0.365)
CHAEBOL dummy (B)
1.797 *
(0.735)
Log Asset of Fis affiliated
(C)
connection dummy (D)
0.111 **
(0.035)
1.672 **
(0.556)
publicly traded firm dummy
1.876 *
(0.734)
0.108 **
(0.034)
1.687 **
(0.556)
-0.373 ***
(0.008)
-3.132 ***
(0.573)
3.984 ***
(0.377)
2.163 **
(0.827)
0.106 **
(0.035)
1.684 **
(0.556)
-0.373 ***
(0.008)
-3.135 ***
(0.573)
3.544 ***
(0.791)
1.842 *
(0.737)
0.118 **
(0.039)
1.686 **
(0.556)
-0.373 ***
(0.008)
-3.136 ***
(0.573)
3.958 ***
(0.380)
1.872 *
(0.734)
0.108 **
(0.034)
1.839 **
(0.652)
-0.687
-0.557
-0.556
-0.561
-0.556
(0.390)
(0.390)
(0.390)
(0.390)
(0.390)
Distress * Chaebol
-0.983
(A*B)
(1.297)
Distress * log asset of FI
-0.031
affiliated (A*C)
(0.059)
Distress * Connection
-0.530
(A*D)
(1.194)
Industry dummies
Yes
Yes
Yes
Yes
Year dummies
Yes
Yes
Yes
Yes
Yes
0.182
0.184
0.184
0.184
0.183
R-square
*p < .05, **p < .01, ***p < .001
Yes