Journal of Asian Economics 16 (2005) 281–297

The effect of financial liberalization on firms’

investments in Korea
Jaewoon Koo *, Kyunghee Maeng 1
Department of Economics, Chonnam National University,
Gwangju 500-757, Republic of Korea
Received 18 October 2004; received in revised form 8 February 2005; accepted 20 February 2005

Abstract

Many studies suggest that firms’ investments depend on the availability of internal funds. Firms
face financial constraints because external funds are more costly than internal funds in an imperfect
capital market. Financial liberalization plays the role of relaxing financial constraints on firms and
thereby reduces the sensitivity of investments to cash flow. This paper empirically examines whether
financial liberalization affects firms’ investment behavior. Using panel data on Korean firms, we find
that cash flow effects on investment decrease as financial markets are liberalized. In particular, small
and non-chaebol firms are found to gain most from liberalization. This implies that large and chaebol
firms are tending to lose their relative advantage in access to credit as liberalization proceeds.
# 2005 Elsevier Inc. All rights reserved.

JEL classification: E22; G31; O16

Keywords: Investment; Financial liberalization; Financial constraints; Chaebol; Korea

1. Introduction

Most developing countries have initiated financial liberalization process in the past
decades. Financial liberalization has been implemented in both domestic and international

* Corresponding author. Tel.: +82 62 530 1551; fax: +82 62 530 1559.
E-mail addresses: [email protected] (J. Koo), [email protected] (K. Maeng).
1
Tel.: +82 16 609 5624; fax: +82 62 530 1559.

1049-0078/$ – see front matter # 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.asieco.2005.02.003
282 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297

scopes. The former includes deregulation of interest rates and a reduction of directed credit
while the latter involves the opening of equity and bond markets.
The effects of liberalization in financial markets on economic growth have received
considerable attention in the literature. McKinnon (1973) and Shaw (1973) propose that
liberalization can promote economic growth. More recently, financial endogenous growth
models emphasize the role of financial development for growth. Among others, Bekaert,
Harvey, and Lundblad (2001) empirically show that equity market opening leads to an
increase in economic growth.
Despite a lot of empirical evidence on the association between financial liberalization
and economic growth, there is scant literature that clarifies the channel of liberalization
effect. In this paper, we focus on the relaxation of financial constraints that firms face as
a potential benefit of liberalization. External funding is more costly than internal
financing because of financial frictions arising from asymmetric information, managerial
agency problems, and transaction costs. Under such constraints to outside financing,
firms’ investments are affected by the availability of internal funds. Financial
liberalization, however, plays the role of relaxing financial constraints. Using various
developing-country data, Laeven (2003) found some evidence that financial liberal-
ization affected favorably the investment behavior of firms. Koo and Shin (2004)
recently argued that liberalization in Korean financial markets succeeded in reducing
financial constraints especially for small and non-chaebol firms that were severely
constrained before liberalization.
Most existing literature splits the sample into pre- and post-liberalization periods
according to an exogenous criterion and examines whether the degree of financial
constraints differs between the two periods. Financial liberalization has, however,
proceeded gradually and in multi-faceted ways. Furthermore, the pace of liberalization
sometimes reversed after a serious shock to economy. Immediately after severe
financial crisis hit the economy in the end of 1997, the Korean government started to
regulate financial markets again. To capture such a nature of financial liberalization,
we attempt to construct a continuous index by adopting a principal component
method.
The dataset consists of 371 listed firms during 1981–2002. We estimate the cash-flow
sensitivity of investment by applying the generalized methods of moments (GMM). The
main findings in this paper are: firstly, financial liberalization significantly reduces the
financial constraints confronted by firms. Secondly, the effect of financial liberalization on
financial constraints is stronger for small and non-chaebol firms than large and chaebol
firms. This suggests that various liberalization policies implemented in financial markets
helped firms to get wider access to external finance.
The rest of the paper is organized as following. Section 2 reviews the main literature on
firms’ investment under financial constraints. We also summarize some existing empirical
studies that investigate the effects of financial liberalization on investments. In Section 3,
we derive investment models in an imperfect financial market and discuss estimation
methods. We attempt to construct a financial liberalization index in Section 4. This
section also contains a description of the data and definitions of relevant variables.
Section 5 presents estimation results and an interpretation. The last section contains a
brief conclusion.
 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297 283

2. Literature review

2.1. Firms’ investments in an imperfect capital markets

According to Modigliani and Miller (1958), a firm’s investment depends only on the
profitability of its investment opportunities. A growing body of literature, however, has
found that the firms’ investment depends on the availability of internal funds. Many
researchers attribute the effects of financial condition on investment to imperfections in
financial markets that Modigliani and Miller’s model ignores.
Two streams of literature investigate why investment is sensitive to internal funds in
imperfect financial markets.1 The first focuses on a non-negligible premium for external
finance that firms should pay. Myers and Majluf (1984) and Stiglitz and Weiss (1981)
demonstrate that the cost of external funds is higher than that of internal funds because of
the asymmetry of information between borrowers and lenders.2 Firms face a constraint in
financial markets because of a wedge between costs of internal and external funds. Under
such a financial constraint, firms tend to rely on internal funds to finance investment. The
second stream of literature attributes the importance of internal funds for investment to
managerial agency problems. As noted by Jensen and Meckling (1976), managers who are
not owners may pursue their own interests, not the stockholders’ interest. Jensen (1986)
argues that managerial discretion is likely to cause managers to spend all available funds on
investment projects.
A large body of literature has empirically examined whether imperfections in financial
markets influence firms’ investments.3 Most studies interpret the cash-flow effect on
investment as resulting from financial constraints.4 Fazzari, Hubbard, and Petersen (1988)
initially show that, utilizing the dividend–payout ratio as a measure of the financial
constraints faced by firms, investments of more financially constrained firms are more
sensitive to changes in cash flow. Since then, it has become a basic research methodology to
examine the difference in sensitivities of investment to cash flow between a priori
segmented firms. The existing empirical studies have used various segmenting variables to
identify unobservable degree of financial constraints, for example: group affiliation in
Hoshi, Kashyap, and Scharfstein (1991); firm size and age in Devereux and Schiantarelli
(1990); issuing commercial paper and bond ratings in Whited (1992); exchange listing in
Oliner and Rudebusch (1992); ownership structure in Schaller (1993); and country
characteristics in Bond, Elston, Mairesse, and Lundblad (2003).
Besides cross-sectional categorization, some researchers attempt to examine what
brings about temporal changes in the cash-flow sensitivity of investment, for example:

1
Stein (2001) reviews the main theoretical literature concerning both models of costly external finance and the
agency conflict model.
2
Some researchers pay attention on the extra transaction costs for issuing equity or bonds to explain why
external finance is more costly.
3
Schiantarelli (1996) and Hubbard (1998) provide extensive surveys of empirical literature, mainly about
financial constraints model. As Stein (2001) claims, there have been very few attempts to empirically distinguish
between the two classes of theories.
4
There is an ongoing debate between Kaplan and Zinglaes (1997) and Fazzari, Hubbard, and Petersen (2000)
about the interpretation of the cash flow sensitivities of investment.
284 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297

financial liberalization in Laeven (2003); financial development and business cycles in

Love (2003); foreign direct investment in Harrison, Love, and McMillan (2002); and
restrictions on capital flows in Forbes (2003).
The Korean firma data is exploitable to test for the financial constraint model because
the Korean financial market seems imperfect in many respects. Most interesting findings
from studies using Korean data are about financial constraints faced by Korean unique
business group, chaebol. Because chaebol firms are favored in an imperfect financial
market, financial constraints faced by chaebol firms are expected to be weaker. Cho (1996)
finds that chaebol firms’ investment spending is less affected by the internal funds. Kong
(1998), however, claims that the financial constraints on chaebol firms seemed to increase
as chaebol firms recently grew faster and needed to finance the higher rate of investments.
Koo and Shin (2004) also claim that the chaebol firms appeared to lose preferential access
to credit after liberalization.

2.2. Financial liberalization and investments

Many studies argue that financial liberalization causes the variations over time in the
responsiveness of investment to internal fund. Financial liberalization influence
asymmetric information and agency conflicts between managers and stockholders. In
particular, various liberalization policies play the role of reducing asymmetric information
problems in financial markets by improving banks’ screening ability. Also, managerial
agency problems are attenuated because financial institutions tend to monitor managers’
behavior more intensively.5 Developments in security markets and financial market
opening also result in a reduction of the cost of external financing. We thus expect that
financial liberalization reduces the wedge between the costs of external and internal funds
and thereby decreases the cash-flow effect on investment of financially constrained firms.
There are, however, mixed empirical evidence for the effect of financial liberalization
on firms’ investment. Examining panel data of a large number of firms in 13 developing
countries, Laeven (2003) concludes that financial liberalization relaxes financing
constraints on firms, especially small ones. Several studies report that financial reform
caused a reduction in financial constraints using an individual country data, for instance:
Harris, Schiantarelli, and Siregar (1994) for Indonesia; Gelos and Werner (2002) for
Mexico; Guncavdi, Bleaney, and McKay (1998) for Turkey; and Koo and Shin (2004) for
Korea. Forbes (2003) recently shows that the Chilean capital controls increased financial
constraints for small firms. Jaramillo, Schiantarelli, and Weiss (1996), however, fail to
provide evidence that financial reform in Equador served to ease financial constraints on
small firms. Using Chilean data, Hermes and Lensink (1998) also report that reforms did
not improve access of small and young firms to outside finance. In light of mixed empirical
findings, we need to further investigate the effect of financial liberalization on firms’
investment.

5
Rajan and Zingales (1998) claim that foreign investors have tendency to demand better corporate governance
to protect their investments.
 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297 285

3. Model and estimation method

3.1. Investment model

The investment model in this paper is closely related to Harrison et al. (2002), Love
(2003) and Forbes (2003). We will adopt both the Q-model and Euler model.

3.1.1. The Q-model

Each firm is assumed to maximize its present value subject to the capital accumulation
constraint. The optimization problem is:
X1
VðKt ; jt Þ ¼ max1 Dt þ Et btþs1 Dtþs
fItþs gs¼0
Q s¼1 ; (1)
subject to : Dt ¼ ðKt ; jt Þ  CðIt ; Kt Þ  It
Ktþ1 ¼ ð1  dÞKt þ It
where Kt is the capital stock at the start of period t; jt is a technology shock; DQt is the
dividend; Et is the expectation operator conditional on information available at t; ( ) is
the profit function; C( ) is the adjustment cost function; It is the investment expenditure
over the period t; and d is the depreciation rate.
Then the first-order condition for investment is:
      
@V @C @V
¼ 1 þ btþ1 Et ¼ 0: (2)
@I t @I t @K tþ1
Now we define marginal Q as the increase in firms’ value from one additional unit of
capital. That is,
 
@V
Qt ¼ : (3)
@K tþ1
We specify the adjustment cost function as follows:
 2
a It It1
CðIt ; Kt Þ ¼ g  n Kt : (4)
2 Kt Kt1

This specification assumes linear homogeneity in capital and investment and allows for
the persistence in investment ratio.
We assume rational expectations, so that the expectation operator is replaced with
realized values plus an expectational error. Substituting (3) and marginal adjustment cost
function derived from (4) into Eq. (2), we obtain the final equation as follows:
   
‘ I I
¼ c þ b1 þb Qit þ fi þ dt þ eit ; (5)
K it K it1 2

where f i is the firm-specific effect; dt is the time-specific effect; and eit is white noise.
According to this standard Q-model, a firm’s financial status does not affect investment. As
Modigliani and Miller (1958) state, only Q is a determinant of investment.
286 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297

We include a term measuring firms’ financial condition to Eq. (5) in order to test
whether financial constraints affect investment.
     
I I CF
¼ c þ b1 þb2 Qit þ b3 þ fi þ dt þ eit ; (6)
K it K it1 K it

where CF is the cash flow that captures firm’s internal financial position. Thus, a test for the
existence of financing constraints amounts to a test for the null hypothesis that b3 = 0 in
Eq. (6). If financial theory of investment is valid, we expect b3 is significantly positive.
To test whether financial liberalization reduces financial constraints, we add an
interaction term between cash flow and a measure of financial liberalization.
       
I I CF CF
¼ c þ b1 þb Qit þ b3 þb FLIt þ fi þ dt þ eit ;
K it K it1 2 K it 4 K it
(7)

where FLI is a financial liberalization index. If b3 > 0 and b4 < 0, it implies that, as the
degree of financial liberalization increases, the cash-flow sensitivity of investment
decreases. This implies that financial liberalization mitigates financial constraints on firms.
Next, we test whether the effect of financial liberalization on the cash-flow sensitivity
depends on the size of firms by estimating the following equation:
       
I I CF CF
¼ c þ b1 þb2 Qit þ b3 þb4 SMALLi FLIt
K it K it1 K it K it
 
CF
þ b5 LARGEi FLIt þ fi þ dt þ eit ; (8)
K it

where SMALL is a dummy variable whose value is one for small firms and zero for large
firms. Similarly, the value of LARGE is one for large firms and zero for small firms. We
expect that b3 > 0, b4 < 0 and b5 < 0. Comparing the absolute value of b4 and b5, we can
examine whether the effect of financial liberalization of the financial constraint is stronger
for small firms than large firms.
Similarly, we will estimate the following equation to investigate whether the affiliation
with business groups affects the effect of financial liberalization on the cash-flow
sensitivity of investment:
       
I I CF CF
¼ c þ b1 þb2 Qit þ b3 þb4 INDEPi FLIt
K it K it1 K it K it
 
CF
þ b5 GROUPi FLIt þ fi þ dt þ eit ; (9)
K it

where INDEP is a dummy variable for independent firms; and GROUP is a dummy
variable for chaebol firms. Comparing the absolute values of b4 and b5, we can examine
whether financial liberalization has stronger influence on the cash-flow sensitivity of
investment for non-chaebol firms than chaebol firms.
 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297 287

3.1.2. The Euler-equation model

Since the Q-model has certain limitations, a number of studies directly estimate the
Euler equation.6 We can derive the Euler equation from the same maximization model used
to derive the Q-model. It is assumed that a firm faces the same optimization problem
described in Eq. (1) subject to another constraint that dividends must be non-negative. The
non-negative binding constraint on dividends implies that there is a shadow cost related
with equity or debt financing due to imperfections in financial markets.
Dt 0: (10)
Let l be the multiplier for the constraint (10). Then l can be interpreted as the shadow
cost of external funds. The first-order condition for maximization is:
       
@V @C @V
¼ ð1 þ lt Þ þ1 þ btþ1 Et ¼0 (11)
@I t @I t @K tþ1
and the envelope theorem is:
   
@V @D
¼ Qt1 ¼ ð1 þ lt Þ þð1  dÞbtþ1 Et Qt : (12)
@K t @K t
Then, combining (11) and (12) yields the following Euler equation:
  Q
@CðIt ; Kt Þ 1 þ ltþ1 @ ðKtþ1 ; jtþ1 Þ
1þ ¼ bt E t
@It 1 þ lt @Ktþ1
 
@CðItþ1 ; Ktþ1 Þ
þ ð1  dÞ 1 þ ; (13)
@Itþ1
Q
where @C/@I is the marginal adjustment cost of investment; @ /@K is the marginal profit of
capital; and (1 + lt+1)/(1 + lt) is the relative shadow cost of external finance in periods t
and t + 1, which represents a factor of financial constraint. In a perfect capital market,
lt = lt+1. If the shadow cost of external funds is higher at t than at t + 1, then the firm is
financially constrained.
As Gilchrist and Himmelberg (1999) show, if a production function has a Cobb-Douglas
form, then we can obtain the following relation:
 Q  
@ S
¼ ’0 þ ’1 ; (14)
@K it K it

where S represents sales.

Now we assume that the degree of financial constraints depends on the cash flow. That
is,
 
1 þ ltþ1 CF
¼ f0 þ f1 : (15)
1 þ lt K it

6
Hayashi (1982) shows that the average Q that most studies adopt equals the marginal Q under very restrictive
conditions. Therefore, using the average Q instead of the marginal Q might cause measurement errors.
288 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297

In addition, we need to change the Euler equation in the linear form and eliminate the
expectation operator. Using a first-order Taylor approximation around the means to
linearize the term with Lagrange multipliers, we obtain:
       
I I S CF
¼ c þ b1 þb2 þb3 þ fi þ dt þ eit ; (16)
K it K it1 K it K it

where f i is the firm-specific effect; dt is the time-specific effect; and eit is white noise.
Eq. (16) is a baseline Euler equation model.
If a measure of financial liberalization as well as the cash flow affects the degree of
financial constraints, then (15) is modified into:
 
1 þ ltþ1 CF
¼ f0 þ ðf1 þ f2 FLIt Þ : (17)
1 þ lt K it

The investment equation for this case is:

         
I I S CF CF
¼ c þ b1 þb þb þb FLIt þ fi þ dt þ eit :
K it K it1 2 K it 3 K it 4 K it
(18)
If the financial constraint is affected by firms’ size and affiliation with chaebols, then
Eq. (18) is modified as follows:
         
I I S CF CF
¼ c þ b1 þb2 þb3 þb4 SMALLi FLIt
K it K it1 K it K it K it
 
CF
þ b5 LARGEi FLIt þ fi þ dt þ eit (19)
K it
and  I   
I
 
S
 
CF
 
CF
¼ c þ b1 þb2 þb3 þb4 INDEPi FLIt
K it K K it K it K it
  it1
CF
þ b5 GROUPi FLIt þ fi þ dt þ eit : (20)
K it
Comparing the absolute value of b4 and b5 in Eqs. (19) and (20), we can test for which
class of firms are more strongly affected by financial liberalization.

3.2. Estimation method

Dynamic investment models that we derive are likely to suffer from both endogeneity
and heterogeneity problems. The error term in investment model generally captures a
technology shock to the profit function. It is, however, known that many explanatory
variables such as sales and cash flow also depend on the technology shock. It means that
explanatory variables in the investment function may be correlated with the error term.7 In
7
Hayashi and Inoue (1991) argued that many explanatory variables of investment such as output and cash flow
depend on the technology shock, and are thus endogenous as well.
 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297 289

addition, the presence of the lagged investment-to-capital ratio as an explanatory variable

may bias coefficient estimates from the ordinary least squares (OLS) estimation.
Substantial differences across firms in their investment behavior may also result in a
heterogeneity problem.
The GMM estimation is widely used for dynamic panel data models. The success of the
GMM technique depends on the appropriate adoption of instruments and elimination of
unobserved firm effects. We will rely on the two specification tests developed by Arellano
and Bond (1991). A Sargan test for over-identifying restrictions is used to test for the
validity of instruments, and a test of serial correlation of error terms is used to detect the
presence of unobserved individual effects. If there are no strong unobserved individual
effects, we can apply the GMM technique to investment equation in levels.
If specification test results suggest that unobserved firm effects remain, we can use the
GMM technique in a difference model, following Arellano and Bover (1995). This
estimation method first-differences each variable in order to eliminate the firm-specific
effects and uses the lagged levels of variables as instruments. If there is no evidence of
strong firm-specific effects, it is more efficient to estimate the equation in levels.

Fig. 1. Financial liberalizations in Korea. Note: X1: interest rates liberalization, X2: foreign exchange liberal-
ization, X3: lowering of legal reserve ratio, X4: reduction of policy loans, X5: new banks entry, X6: capital
liberalization, and X7: the privatization of banks.
290 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297

4. Data and variables

4.1. Financial liberalization index

To investigate the effect of financial liberalization on investment, we need to construct a

measure of financial liberalization. Most previous literature attempts to identify the
liberalization date by examining a shift in various variables related to liberalization in
financial markets. Then researchers divide the whole sample period into two sub-periods of
before and after liberalization and investigate whether the financial constraint is
significantly different between pre- and after-liberalized periods.8
In the actual economy, financial liberalization takes place in various dimensions and in
stages. The conventional method, however, fails to properly reflect the gradual nature of the
financial liberalization process. Furthermore, if financial liberalization involves reversals,
it is a futile attempt to investigate the effect of financial liberalization on investment with
the ‘before and after’ method. Recent studies tend to construct a continuous measure of
financial liberalization to cure the limitation of a discrete index.9 We also need a continuous
measure of financial liberalization to properly allow for the non-linear and gradual nature
of liberalization in Korean financial markets.
Updating the data by Chun (2002), we constructed a financial liberalization index as
follows.10 First, we derived quarterly data on the implementation of reform packages
related to seven different measures. The seven reform variables include interest rate
deregulation, liberalization of foreign exchange market, reduction of reserve requirements,
reduction of policy loans, entry of new banks, capital market liberalization, and
privatization of state banks.11 The initial value of each variable is set at zero. We add one at
each step of the incremental process of liberalization and subtract one at each step of
decremental step of liberalization. In Fig. 1, we present plots of seven variables. To derive a
single measure of financial liberalization from seven variables, we apply a principal
component method. Because the first principal component explains more than 80% of
variance, we will regard the first principal component as a financial liberalization index.
Finally, we construct annual data of financial liberalization index (FLI) by averaging
quarterly data.

8
For example, Koo and Shin (2004) investigate the liberalization date of seven important variables in Korean
financial markets and identify the date when all seven variables were liberalized. Most country studies employ the
similar method to find the liberalization date.
9
Laeven (2003) uses the sum of six dummy variables that capture financial reform as a financial liberalization
index. His index takes values between 0 and 6. Love (2003) also devises a continuous financial development
index, which is an average of five standardized indices of financial variables. Bandiera, Caprio, Honohan, and
Schiantarelli (2000) apply a principal component method to dummy variables that reflect the liberalization date of
eight variables to derive a single measure of financial liberalization.
10
Chun (2002) devises a measure of financial liberalization index and examines the effect of financial
liberalization on the aggregate consumption in Korea. The authors are grateful to Byung Chul Chun for providing
his data.
11
The variables considered in this paper are similar to the ones adopted by the previous literature. We excluded
prudential regulation considered in Koo and Shin (2004) and Bandiera et al. (2000) and included the liberalization
of foreign exchange market, instead. We did not allow for prudential regulation because it is difficult to obtain a
reliable data.
 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297 291

Fig. 2. Financial liberalization index.

Fig. 2 presents the plot of the FLI. As expected, the FLI increases gradually over time
but decreases in the period of last few years. The main reason for a recent decrease in the
FLI is an increase in government share of several private insolvent banks. Regulation in
financial markets was reinforced because of a strong restructuring program that was
implemented to overcome the currency crisis in the end of 1997. In 2002, the FLI increases
again since some government-owned banks were privatized through a merger with large
private banks.

4.2. Data and variables

We constructed a firm-level panel dataset from Korea Investors Service-Financial

Analysis System (KIS-FAS) database. We derived an unbalanced panel dataset of the
manufacturing firms that were listed longer than 5 years during 1980–2002. The data
consists of the 5084 observations of 371 firms. We focus on the listed firms only since we
need the stock price data to calculate Tobin’s Q.12 Variables are defined according to the
conventional method. The detailed description of relevant variables is presented in Table 1.
We split the sample according to two exogenous characteristics of firms to test for a
difference in the effect of financial liberalization on financing constraints between firms.
As a measure of firm size, we use the number of employees. A small size dummy variable,
SMALL, takes the value one if the number of employees is less than 300, and zero
otherwise. Similarly we construct a large size dummy variable, LARGE, which indicates
large firms having more than 300 employees. Another classification criterion is the
affiliation with a business group, chaebol. The group firms that belong to the 30 largest
chaebols are represented as a dummy variable, GROUP, and the remaining firms as INDEP.
12
Although some critics point to the difficulty of finding substantial differences in the asymmetry of information
between listed firms, many researchers still prefer to exclude non-listed firms to utilize Tobin’s Q.
292 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297

Table 1
Variable definitions
Abbreviation Description
Kt Capital at the beginning of period t (=tangible fixed assets at the end of the period t1  1)
It Investment during period t (=Kt+1 + DEPRt  Kt)
DEPRt Depreciation during period t
Qt Average Q at the beginning of period t (=(Bt + Et)/TAt)
Bt Book value of debt at the beginning of the period t
Et Market value of equity at the beginning of period t
TAt Total assets at the beginning of period t
CFt Cash flow during the period t (=net profit after-tax + depreciation during period t1  1)
St Sales during period t
FLIt Financial liberalization index at t
SMALLi =1 for small firms, =0 for large firms
LARGEi =1 for large firms, =0 for small firms
GROUPi =1 for chaebol firms, =0 for non-chaebol firms
INDEPi =1 for non-chaebol firms, =0 for chaebol firms

Table 2
Descriptive statistics for key variables (1981–2002)
I/K CF/K Q S/K Number of firms
All firms 0.280 (0.481) 0.335 (0.363) 0.573 (0.727) 3.642 (3.276) 371
Large firms 0.286 (0.478) 0.340 (0.346) 0.573 (0.748) 3.478 (2.913) 270
Small firms 0.257 (0.492) 0.318 (0.421) 0.572 (0.645) 4.253 (4.318) 101
Chaebol firms 0.307 (0.469) 0.282 (0.194) 0.571 (0.866) 2.796 (1.993) 36
Non-chaebol firms 0.276 (0.482) 0.342 (0.379) 0.573 (0.780) 3.752 (3.392) 335
Note. Numbers are the mean (S.D.) of variables.

According to these classifications, the number of large (small) firms is 270 (101); and that
of group (independent) firms is 36 (335).13
Table 2 provides the mean and standard deviation of each variable for 1981–2002 on
371 firms. Large and chaebol firms have higher investment-to-capital ratio than small and
non-chaebol firms. However, we find that sales-to-capital ratio for non-chaebol and small
firms is slightly higher than for large and chaebol firms. Non-chaebol and large firms
appear to hold more cash flow relative to capital.

5. Estimation results

Columns (1) and (2) in Table 3 present the basic GMM estimation results of investment
function for the whole sample period. We applied the GMM techniques to Eqs. (6) and (16)
in levels using the t  2, t  3, t  4, and t  5 lagged right-hand side variables as
instruments.14 Sargan test results for over-identifying restrictions indicate that the
13
The 35 out of 36 group firms are classified as large firms.
14
It is noted that Laeven (2002) and Koo and Shin (2004) also adopt the GMM-level estimation technique
because the individual firm effect is found to be weak in the Korean firm data.
 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297 293

Table 3
Estimation results of investment model (baseline model)
Variables (1) Q-model (2) Euler equation (3) Q-model (4) Euler equation
*
Constant 0.065 (0.89) 0.218 (0.29) 0.177 (1.64) 0.096 (0.94)
(I/K)it1 0.115** (1.92) 0.130*** (2.39) 0.084* (1.60) 0.102** (2.02)
Qi 0.066*** (2.67) – 0.054*** (2.80) –
(S/K)it – 0.013*** (3.00) – 0.013*** (3.00)
(CF/K)it 0.304*** (4.92) 0.184*** (3.58) 0.387*** (4.89) 0.265*** (3.51)
FLIt (CF/K)it   0.057** (2.10) 0.051** (2.16)
m1 0.221 0.137 0.435 0.302
m2 0.058 0.096 0.074 0.139
Wald test 0.000*** 0.000*** 0.000*** 0.000***
Sargan test 0.151 0.292 0.285 0.411
Notes. (1) The t values are in parentheses. (2) Time dummies are included, but are not reported. (3) The
instruments include t  2, t  3, t  4, and t  5 lagged variables. (4) The m1 and m2 are p values of the test for
first- and second-order serial correlation. of residuals, respectively. (5) Wald test results present p value of test for
joint significance of coefficients. (6) Sargan test results present p value of the test for over-identification.
*
indicates significance at the 10% level.
**
indicates significance at the 5% level.
***
indicates significance at the 1% level.

instruments we adopted are valid. The p value of m1 and m2 test statistics also indicates that
little unobserved individual effects remain in the GMM-level estimation results. The joint
test result for the significance of explanatory variables shows that the model is satisfactory.
Firms are found financially constrained since the coefficient in CF/K is statistically
significant at the conventional level for both models. We also find persistence in firms’

Table 4
Estimation results of investment model (by categories)
Variables (1) Q-model (2) Euler equation (3) Q-model (4) Euler equation
Constant 0.164 (1.47) 0.086 (0.83) 0.207** (1.89) 0.134 (1.30)
(I/K)it1 0.085* (1.64) 0.102** (2.06) 0.085** (1.62) 0.100** (2.01)
Qit 0.056*** (2.87)  0.052*** (2.67) –
(S/K)it – 0.014*** (3.08) – 0.013*** (2.92)
(CF/K)it 0.391*** (4.80) 0.262*** (3.45) 0.394*** (4.97) 0.276*** (3.69)
LARGEi FLIt (CF/K)it 0.041 (1.37) 0.040 (1.51) – –
SMALLi FLIt (CF/K)it 0.097** (2.29) 0.070*** (2.42) – –
GROUPi FLIt (CF/K)it – – 0.166 (1.14) 0.177 (1.38)
INDEPi FLIt (CF/K)it – – 0.060** (2.20) 0.053** (2.26)
m1 0.426 0.291 0.458 0.346
m2 0.070 0.134 0.140 0.252
Wald test 0.000*** 0.000*** 0.000*** 0.000***
Sargan test 0.379 0.426 0.274 0.359
Notes. (1) The t values are in parentheses. (2) Time dummies are included, but are not reported. (3) The
instruments include t  2, t  3, t  4, and t  5 lagged variables. (4) The m1 and m2 are p values of the test for
first- and second-order serial correlation. of residuals, respectively. (5) Wald test results present p value of test for
joint significance of coefficients. (6) Sargan test results present p value of the test for over-identification.
**
indicates significance at the 5% level.
***
indicates significance at the 1% level.
294 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297

investment because the lagged investment-to-capital ratio is statistically significant.

Tobin’s Q and the marginal productivity of capital represented by sales-to-capital ratio play
the role of increasing the investment–capital ratio as predicted by the theory.
In columns (3) and (4) of Table 3, we investigate the effect of financial liberalization on
financial constraints by estimating Eqs. (7) and (18). The interaction term of FLI and CF/K
is included in both investment models. The GMM-level estimation successfully passed
both specification tests. Neither the m1 nor the m2 statistics is significant, implying that
little individual firms effect remains. Sargan test statistics is also insignificant. All
estimates have the expected sign and the statistical significance. The estimates in the
interaction terms are negative and significant at the 5% level. This implies that the financial
constraints are significantly relaxed as the measure of financial liberalization increases.
Next, we attempt to test whether financial liberalization has a different impact on the
cash-flow sensitivity of investment across firm characteristics. The columns (1) and (2) in
Table 4 present the estimation results of the model with the interaction terms between size
variable and the financial liberalization index. The LARGE and SMALL dummies capture
large and small firms, respectively. It is evident that small firms are more influenced by
financial liberalization. The sign of both coefficients in SMALL FLI (CF/K) and
LARGE FLI (CF/K) is negative, implying that the cash-flow sensitivity of investment
decreases with financial liberalization. However, the statistical significance for large firms
is quite low. In line with Laeven (2002) and Koo and Shin (2004), liberalization in Korean
financial markets seems to result in a reduction in financial constraints confronted by firms,
especially small ones.15
The second classification of firms is based on the affiliation of business groups. Group
captures chaebol firms while INDEP represents non-chaebol firms in columns (3) and (4)
of Table 4. Many studies claim that non-chaebol firms were more liquidity-constrained in
their investment than chaebol firms. As financial market is liberalized, the financial
constraints faced by non-chaebol firms are significantly mitigated. The estimation results
of the Q-model indicate that the coefficient in the interaction between financial
liberalization index and cash flow is 0.166 for chaebol firms and 0.060 for non-chaebol
firms. The effect of financial liberalization on the cash-flow sensitivity for chaebol firms is,
however, not statistically significant while that for non-chaebol firms is significant at 5%
level. This implies that non-chaebol firms have better access to external financing as
liberalization proceeds. In line with Borensztein and Lee (2002), chaebol-affiliated firms
appear to have lost the preferential access to credit that they enjoyed for the pre-
liberalization period.
We find strong evidence that financial liberalization affects the financial constraints
faced by firms. In particular, small and non-chaebol firms that were severely constrained in
financing seem to be more influenced by liberalization. As a measure of financial
liberalization increases, the cash-flow sensitivity of investment for small and non-chaebol
firms significantly decreases. This result is considered evidence for a benefit of financial
liberalization.

15
Laeven (2002) claims that financing constraints faced by small and medium-sized enterprises (SMEs) were
eroded in the 1990s because of Korean government policy to favor SMEs. Koo and Shin (2004) also show that
small firms got much wider access to the outside credit after liberalization.
 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297 295

6. Concluding remarks

The impact of financial liberalization, which recently has been implemented in many
developing countries, has received much attention in the existing literature. This paper
investigates whether financial liberalization affects firms’ investments in Korea. We tested
for the hypothesis that financial liberalization had an impact on firms’ investment behavior.
The main empirical finding is that the financial constraint estimated by the cash-flow
sensitivity of investment decreases with financial liberalization. This implies that financial
liberalization improves firms’ accessibility to external finance. In particular, small and non-
chaebol firms that were more severely constrained seem to gain more from liberalization
than large and chaebol firms.
Numerous empirical studies have so far revealed that financial liberalization promotes
economic growth. However, the channels of financial liberalization on economic growth
should be further investigated. The empirical results of this paper shed light on potential
benefits of financial liberalization. We find that financial liberalization plays the role of
reducing the financial constraints on firms, and thus improving external financing for
investment.

Acknowledgements

This study was financially supported by Chonnam National University in Jaewoon

Koo’s sabbatical year of 2003. The author wrote the paper during his visit at the
Department of Business Economics and Public Policy, Indiana University. We
acknowledge participants at the 11th International Conference offered by the Korean
Economic Association. We are also grateful for an anonymous referee for comments.

References

Arellano, M., & Bond, S. R. (1991). Some tests of specification for panel data: Monte Carlo evidence and an
application to employment equations. Review of Economic Studies, 58, 277–297.
Arellano, M., & Bover, O. (1995). Another look at the instrumental-variable estimation of error components
models. Journal of Econometrics, 68, 29–51.
Bandiera, O., Caprio, G., Honohan, P., & Schiantarelli, F. (2000). Does financial reform raise or reduce saving?
Review of Economics and Statistics, 82, 239–263.
Bekaert, G., Harvey, C. R., & Lundblad, C. (2001). Does financial market liberalization spur growth (NBER
Working Paper 8245)?
Bond, S., Elston, J. A., Mairesse, J., & Mulkay, B. (2003). Financial factors and investment in Belgium, France,
Germany, and the United Kingdom: A comparison using company panel data. Review of Economics and
Statistics, 85, 153–165.
Borensztein, E., & Lee, J. (2002). Financial crisis and credit crunch in Korea: Evidence from firm-level data.
Journal of Monetary Economics, 49, 853–875.
Cho, Y.-D. (1996). Financial factors and corporate investment: An empirical analysis of Korean manufacturing
firms. Avebury: Ashgate Publishing.
Chun, B. C. (2002). Effects of financial liberalization on consumption in Korea. Economic Papers, 6(1), 24–41
(Bank of Korea).
296 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297

Devereux, M., & Schiantarelli, F. (1990). Investment, financial factors, and cash flow: Evidence from U.K. panel
data. In R. G. Hubbard (Ed.), Asymmetric information, corporate finance, and investment. Chicago: University
of Chicago Press.
Fazzari, S. M., Hubbard, R. G., & Petersen, B. C. (1988). Financing constraints and corporate investment.
Brookings Papers on Economic Activity, 1, 141–195.
Fazzari, S. M., Hubbard, R. G., & Petersen, B. C. (2000). Investment-cash flow sensitivities are useful: A
comment. Quarterly Journal of Economics, 115, 695–705.
Forbes, K. J. (2003). One cost of the Chilean capital controls: Increased financial constraints for smaller trade
firms (NBER Working Paper 9777).
Gelos, R. G., & Werner, A. M. (2002). Financial liberalization, credit constraints, and collateral: Investment in the
Mexican manufacturing sector. Journal of Development Economics, 67, 1–27.
Gilchrist, S., & Himmelberg, C. (1999). Investment, fundamentals, and finance. NBER Macroeconomics Annual .
Guncavdi, O., Bleaney, M., & McKay, A. (1998). Financial liberalization and private investment: Evidence from
Turkey. Journal of Development Economics, 57, 443–455.
Harris, J., Schiantarelli, F., & Siregar, M. (1994). The effect of financial liberalization on firms’ capital structure
and investment decisions: Evidence from a panel of Indonesian manufacturing establishments, 1981–1988.
World Bank Economic Review, 8, 17–47.
Harrison, A. E., Love, I., & McMillan, M. S. (2002). Global capital flows and financing constraints (NBER
Working Paper 8887).
Hayashi, F. (1982). Tobin’s marginal q and average q: A neoclassical interpretation. Econometrica, 50, 213–224.
Hayashi, F., & Inoue, T. (1991). The relation between firm growth and Q with multiple capital goods: Theory and
evidence from panel data on Japanese firms. Econometrica, 59, 731–754.
Hermes, N., & Lensink, R. (1998). Financial reform and informational problems in capital markets: An empirical
analysis of the Chilean experience, 1983–1992. Journal of Development Studies, 34, 27–43.
Hoshi, T., Kashyap, A., & Scharfstein, D. (1991). Corporate structure, liquidity, and investment: Evidence from
Japanese panel data. Quarterly Journal of Economics, 106, 33–60.
Hubbard, R. G. (1998). Capital-market imperfections and investment. Journal of Economic Literature, 36, 193–
225.
Jaramillo, F., Schiantarelli, F., & Weiss, A. (1996). Capital market imperfections before and after financial
liberalization: An Euler equation approach to panel data for Ecuadorian firms. Journal of Development
Economics, 51, 367–386.
Jensen, M. (1986). Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review,
76, 323–329.
Jensen, M., & Meckling, W. (1976). Theory of the firms: Managerial behavior, agency costs, and ownership
structure. Journal of Financial Economics, 1, 305–360.
Kaplan, S. N., & Zinglaes, L. (1997). Do investment-cash flow sensitivities provide useful measures of financing
constraints? Quarterly Journal of Economics, 112, 169–215.
Kong, M. (1998). Stock market development, liquidity constraint, and investment: A case of Korean Jaebol and
non-Jaebol manufacturing firms in the 1980s. Asian Economic Journal, 12, 1–22.
Koo, J., & Shin, S. (2004). Financial liberalization and corporate investments: Evidence from Korean firm data.
Asian Economic Journal, 18, 277–292.
Laeven, L. (2002). Financial constraints on investments and credit policy in Korea. Journal of Asian Economics,
13, 251–269.
Laeven, L. (2003). Does financial liberalization reduce financial constraints? Financial Management, 32, 5–34.
Love, I. (2003). Financial development and financing constraints: International evidence from the structural
investment model. Review of Financial Studies, 16, 765–791.
McKinnon, R. I. (1973). Money and capital in economic development. Washington, DC: Brookings Institutions.
Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance, and the theory of investment.
American Economic Review, 48, 261–297.
Myers, S. C., & Majluf, N. S. (1984). Corporate financing and investment decisions when firms have information
that investors do not have. Journal of Financial Economics, 13, 187–221.
Oliner, S. D., & Rudebusch, G. D. (1992). Sources of the financing hierarchy for business investment. Review of
Economics and Statistics, 74, 643–654.
 J. Koo, K. Maeng / Journal of Asian Economics 16 (2005) 281–297 297

Rajan, R., & Zingales, L. (1998). Financial dependence and growth. American Economic Review, 88, 559–586.
Schaller, H. (1993). Asymmetric information, liquidity constraints, and Canadian investment. Canadian Journal
of Economics, 26, 552–574.
Schiantarelli, F. (1996). Financial constraints and investment: Methodological issues and international evidence.
Oxford Review of Economic Policy, 12, 70–89.
Shaw, E. S. (1973). Financial deepening in economic development. New York: Oxford University Press.
Stein, J. (2001). Agency, information and corporate investment (NBER Working Paper 8342).
Stiglitz, J., & Weiss, A. (1981). Credit rationing in markets with imperfect information. American Economic
Review, 71, 393–410.
Whited, T. M. (1992). Debt, liquidity constraints, and corporate investment: Evidence from panel data. Journal of
Finance, 47, 1425–1460.