bs_bs_banner
Asian Economic Journal 2012, Vol. 26 No. 2, 119–135
119
Financial Determinants of Corporate R&D Investment in Korea Sanghoon Lee* Received 13 April 2011; Accepted 13 February 2012
This paper investigates the relationship between corporate investment in R&D and finance. By using panel data on Korean manufacturing firms between 1999 and 2008, we find that R&D investment is sensitive to fluctuations in internal cash flow and that debt is more important than equity in financing R&D expenditure. This empirical result supports the traditional theory of financing hierarchy. In relation to equity ownership, the empirical result shows that ownership concentration and foreign ownership positively affect R&D investment, but institutional ownership does not show any significant effect on investment. This study synthetically analyzes the relation between finance and investment by simultaneously considering corporate finance, equity ownership structure, macroeconomic environments and the institutional setting of Korea. Keywords: R&D investment, internal funds, cash flow, ownership structure, Korea. JEL classification codes: G32, G34, L20. doi: 10.1111/j.1467-8381.2012.02080.x
I.
Introduction
Since Schumpeter (1942), it has been emphasized that technological innovation based on R&D investment is indispensable to economic growth and productivity. There is also empirical evidence to support the role of R&D as a driving force of economic performance (for an extensive empirical analysis, see Guellec and van Pottelsberghe de la Potterie, 2001). Given the importance of R&D in attaining economic goals, it is worthwhile identifying the determinants of firms’ R&D investments. The present study attempts to identify the firm-level financial determinants of R&D investment. We examine the investment–cash flow sensitivity, the effects of external finance, and the relation between ownership structure and R&D investment. The characteristics of R&D investment, such as highly uncertain outcomes, severe information asymmetry and agency problems, make the study more sensitive in detecting the effect of finance on investment, as will be explained below. The paper is constructed as follows. The theoretical background is provided first, followed by a survey of previous empirical studies in Section III. Section IV introduces the sample and methods used in the empirical analysis. Section V discusses the empirical results and Section VI offers a summary of the study. *Department of Economics, Hannam University, 70 Hannamro, Daedeok Gu, Daejeon 306-791 Republic of Korea. Email:
[email protected]. © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
ASIAN ECONOMIC JOURNAL
II.
120
Theoretical Background
Whether or not financial factors such as internal finance play a role in firms’ investment decisions has been a hotly debated issue (for reviews, see Hubbard, 1998; Hall, 2002). Modigliani and Miller (1958) argue that a firm’s financial status does not affect its market value and, as a result, is irrelevant for real investment decisions under perfect capital markets. However, if the assumption of perfect markets is not valid, then the irrelevance theorem does not hold either. Indeed, most firms prefer to use internal resources to finance investment because internal finance is cheaper than external finance in the presence of imperfect capital markets (Hubbard, 1998). Among the factors that put a wedge between the costs of internal and external finance, two have been receiving a great deal of attention: asymmetric information and agency problems. Due to asymmetric information between the firm and outside investors, the investors cannot evaluate investment projects adequately and, therefore, ask for a premium on their funds (Greenwald et al., 1984). Because external finance subjects managers to the discipline of capital markets, agency problems can generate cost disadvantages of external finance (Jensen and Meckling, 1976; Grossman and Hart, 1982). According to the theory of financing hierarchy or pecking order (Myers and Majluf, 1984), these factors lead firms to exhaust internal funds first, and then, turn to external finance. The problems of asymmetric information and agency costs would be more severe for R&D investment than for ordinary investment. Because innovative projects are typically difficult to evaluate, the asymmetric information and agency conflict between outside investors and management would be more problematic in R&D investment. Another question arises: Is there a hierarchy between the two external financing methods, debt and equity? Because information asymmetry tends to be more severe in equity financing than in debt financing, the pecking order theory (Myers and Majluf, 1984) argues that equity financing is more expensive than debt financing. In contrast, agency theorists predict that equity is more important than debt as a source of investment funds, because debt financing has a disciplinary effect on managers and constrains their discretion by limiting the free cash flow available to them (Jensen, 1986). The financing hierarchy between debt and equity is theoretically inconclusive. Thus, it needs to be empirically tested. We also need to look deeper into equity ownership structure because in the corporate governance literature it has been regarded to affect firms’ decisions. The ‘separation of ownership and control’ (Berle and Means, 1932) caused by dispersed ownership leads to an agency problem in the firm. Concentrated ownership can be an appropriate mechanism to solve the agency problem because large shareholders have the incentives and ability to effectively monitor managers (Shleifer and Vishny, 1986). One question remains open: Is the effect of ownership concentration on R&D investment positive or negative? Due to the © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
FINANCIAL DETERMINANTS OF R&D
121
uncertainty of R&D projects and the time delay between R&D spending and the corresponding output, the answer to the question depends on the large shareholders’ risk attitude and time horizon. In general, large shareholders are known to be risk-averse and long-term oriented. Small shareholders can diversify their personal risk by buying stocks in various firms, but large shareholders often cannot easily diversify their risk and, therefore, are conservative in business affairs. Large shareholders are long-term oriented because they hold a large fraction of the shares in the firm; thus, their earnings are dependent upon the survival of the firm. In addition, large shareholders can exert power to influence corporate decision-making and, as a result, they are likely to be involved in the firm’s management. Because high involvement and participation foster a strong sense of ownership of the organization (Denison and Mishra, 1995), large shareholders can have a sense of ownership of the firm, which enhances their commitment to the firm’s long-term success. Although the risk averseness of large shareholders leads to a negative effect on corporate R&D investment, their long-term horizon has a positive influence on R&D investment. If the effect of the risk averseness is dominant, R&D investment is discouraged. In contrast, if the effect of the long-term horizon prevails, large shareholders support R&D investment. This issue also needs to be determined by empirical work. Besides ownership concentration, ownership identity can also affect corporate decisions. The monitoring of management by shareholders may be more effective when shareholders have sufficient knowledge of and experience in business matters and are capable of evaluating the performance of management. For example, foreign investors and institutional investors often control larger blocks of shares than do individuals, can obtain information about investment projects, and serve as active monitors of management. As with the issue of ownership concentration, there are contrasting views on whether risky long-term investment projects should be encouraged: institutional and foreign owners can discourage R&D investment when they seek short-term returns, but can encourage such investment when they are able to access investment information (for a relevant study, see Lee, 2012). III.
Previous Studies
The sensitivity of investment to the availability of internal cash flow has been a topic of the largest empirical studies in corporate finance. Many studies have found a significant role of internal funds in determining corporate investment decisions. One of the most influential studies on the subject is that by Fazzari et al. (1988), which estimates:
I CF (1) = a + bQ + c +ε, K K where I represents investment, K is the replacement value of the capital stock, Q is Tobin’s Q, and CF is the cash flow. If internal funds affect investment, c is © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
ASIAN ECONOMIC JOURNAL
122
significant. This model has become a standard approach to investigate the relation between internal funds and investment. A number of subsequent studies address some methodological issues (for a review, see Schiantarelli, 1996). There are two main issues in this regard. The first is whether Q is a good proxy for investment opportunities. The expected profitability, or the underlying investment opportunity, is supposed to be measured by marginal Q, but average Q is used in empirical studies because marginal Q is unobservable. The problem is that average Q is an imprecise proxy for marginal Q. The second issue is that cash flow represents internal funds but also indicates future profitability. If this is the case, the coefficient estimate on cash flow is not reliable because it is endogenous in the model. Prior empirical studies have tried various approaches to address these problems (for surveys, see Schiantarelli, 1996; Hubbard, 1998). According to empirical studies of finance and R&D investment, R&D seems to be mainly financed by cash flow (for a survey, see Hall, 2005). However, the evidence is mixed regarding the effect of external financing on investment. Some studies support the superiority of equity over debt: the positive effect of equity on R&D investment (Carpenter and Petersen, 2002; Brown et al. 2009; Brown and Petersen, 2009) and the negative effect of debt on R&D investment (Hall, 1990, 1992) have been reported. Others show that debt financing encourages R&D investment (e.g. Bond et al., 2003). The mixed evidence can imply that the effect of external financing on investment depends on the relative importance of asymmetric information versus agency problems. Gatchev et al. (2009) find that, in financing fixed assets, firms with high asymmetric information use more debt financing, whereas firms with high agency problems use more equity financing. Thus, the effect of external financing on investment depends on the degree of the problems of asymmetric information and agency costs. A few studies, using data on US firms, report a significant positive relationship between ownership concentration and R&D investment (see Baysinger et al., 1991; Hansen and Hill, 1991; Francis and Smith, 1995; Lee and O’Neill, 2003). This evidence might reflect the institutional setting of the USA. According to John et al. (2008), while the median values of the quality of accounting disclosure standards (ASR), the rule of law (RL) and an index of anti-director rights (ADR) for 39 countries are 64, 8.33, and 3, respectively, those for the USA are 71, 10 and 5, respectively. Thus, in the USA, the risk averseness of large shareholders is relatively weak because the country has a better investor protection environment. Indeed, ownership concentration is not significantly related to risktaking (John et al., 2008), and even the presence of large shareholders is likely to increase the willingness of firms to take risks (Amihud and Lev, 1981). This indicates a weak effect of the risk averseness of large shareholders: thus, the positive effect of long-term orientation can be dominant, which leads to a positive relation between ownership concentration and R&D. Comparing the USA and Korea, we find that Korea does not show a strong institutional setting of © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
FINANCIAL DETERMINANTS OF R&D
123
investor protection: its ASR is 62, its RL is 5.35 and its ADR is 2, which are all below the median (John et al., 2008). Therefore, we cannot determine in advance which effect of ownership structure (risk averseness vs long-term orientation) is dominant in Korea.
IV.
Research Design
4.1 Sample For the empirical analysis, a large panel dataset of 424 Korean manufacturing firms listed on the Korea Stock Exchange (KSE) during the period 1999 to 2008 is used. The sample data were obtained from the database of the Korea Listed Companies Association, which offers firm-level information based on annual reports, quarterly reports and audit reports of Korean companies. Firms that have a large amount of missing data on the variables required for the empirical test were eliminated from the sample. For example, some firms that had newly entered or exited the dataset in the middle of the time period considered were excluded from the sample. As a result, the sample consists of 424 firms, and the total number of observations is 4240. In this paper, R&D intensity is used as a proxy for R&D investment. Sales, net income, cash flow, debt, equity and ownership are also included in the study as independent variables. The details of the variables are as follows:
R & Di ,t =
R & D spendingi ,t total assetsi ,t
salesi ,t = Δsalesi ,t =
total salesi ,t
(2)
,
(3)
net salesi ,t − net salesi ,t −1 net salesi ,t
roai ,t = cashi ,t =
net salesi ,t
,
net incomei ,t total assetsi ,t
,
total assetsi ,t total debti ,t total assetsi ,t
equityi ,t =
,
total equityi ,t total assetsi ,t
(4)
(5)
operating cash flowi ,t
debti ,t =
,
,
(6)
(7)
,
(8)
© 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
ASIAN ECONOMIC JOURNAL
coni ,t =
shares held by controlling shareholdersi ,t
fori ,t = insi ,t =
124
total sharesi ,t shares held by foreign shareholdersi ,t total sharesi ,t
shares held by institutional shareholdersi ,t total sharesi ,t
,
(9)
,
(10)
,
(11)
where the subscript i refers to the firm and t to time period. The controlling shareholders refer to the shareholders that control the firm, such as those that own a substantial equity stake in the firm, their family members and affiliated entities. Real cash flow is affected by macroeconomic fluctuations, such as changes in interest rates. Thus, following Arza and Espanol (2008), we developed other cash flow variables by using the theory of Minsky (1975, 1986). Minsky introduced three types of financing of positions in assets: hedge, speculative and Ponzi finance. Hedge finance is when the cash flow from operating assets exceeds the payment commitments; speculative finance applies when the cash flow is less than the payment commitments, but the expected income receipts exceed the income (interest) payments; and Ponzi finance occurs when even the expected income receipts cannot exceed the income payments. Applying these Minskian concepts, we use two more cash flow variables as follows:
cash1i ,t =
cfi ,t − cli ,t − rt ∗ cli ,t
cash2i ,t =
total assetsi ,t cfi ,t − rt ∗ cli ,t total assetsi ,t
,
,
(12)
(13)
where
cfi ,t = operating cash flowi ,t , cli ,t = current liabilitiesi ,t , rt = interest ratet . In Minskian terms, hedge finance is when cash1 > 0, speculative finance is when cash1 < 0 and cash2 > 0, and Ponzi finance occurs when cash2 < 0. In addition, the interest rate used to develop these measures of cash flow is expected to play the role of an alternative proxy for the cost of financing. The summary statistics and the correlation matrix for the sample are presented in Table 1. The correlation matrix shows some signs that raise concerns about multicollinearity between the independent variables. The presence of multicollinearity is examined through collinearity diagnostics, such as the variance inflation factor, and the result indicates that multicollinearity does not appear to © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
0.14 0.01 0.02 0.05 -0.09 -0.07 0.51 0.55 0.35 0.09 0.05
R&D (1) sales (2) roa (3) cash (4) cash1 (5) cash2 (6) debt (7) equity (8) con (9) for (10) ins (11)
0.30 0.00 0.11 0.11 0.19 0.17 0.20 0.17 0.19 0.14 0.11
Standard deviation
1 0.04*** 0.02 0.04*** 0.00 0.01 0.04*** -0.11**** -0.02 0.07**** -0.04***
(1)
1 0.15**** 0.15**** 0.01 0.03** -0.15**** -0.24**** 0.04** 0.01 -0.08****
(2)
1 0.38**** 0.30**** 0.33**** 0.27 -0.08**** 0.11**** 0.19**** -0.02
(3)
1 0.53**** 0.61**** 0.14**** 0.03** 0.06**** 0.19**** 0.09****
(4)
1 0.99**** 0.25**** 0.06**** 0.07**** 0.17**** -0.02
(5)
1 0.25**** 0.06**** 0.07**** 0.18**** -0.02
(6)
Summary statistics and correlation matrix
1 0.00 0.20**** 0.20**** -0.12****
(7)
1 0.09**** 0.07**** 0.10****
(8)
1 0.03* -0.11****
(9)
1 0.02
(10)
The table shows the summary statistics and the correlation matrix of the variables used in the study. Figures in the correlation matrix are correlation coefficient estimates. ****, ***, ** and * indicate significance at 0.1, 1, 5, and 10% levels, respectively.
Median
Variable
Table 1
FINANCIAL DETERMINANTS OF R&D 125
© 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
ASIAN ECONOMIC JOURNAL Figure 1
126
R&D intensity and cash flow, 1999–2008
0.09 0.08 0.07
R&D cash
0.06 0.05 0.04 0.03
R&D Cash flow
0.02 0.01 0 1999
2000
2001
2002
2004 2003 Year
2005
2006
2007
2008
present severe estimation problems. In addition, regressions with various combinations of independent variables are conducted, and the results are not significantly different from each other. Thus, we find little evidence of severe multicollinearity in the data. We also examined the time trends of the median values of R&D and cash of the sample and found that R&D intensity and cash flow went hand in hand until 2002 and then moved in opposite directions, as shown in Figure 1. That is, the R&D intensity kept increasing over time, while the ratio of cash flow to assets declined sharply after 2002. This suggests that there is a need to consider the possible impact of macroeconomic factors on corporate R&D activities. We thus include year dummies to capture the trends over time. Note that the aggregate time trend does not imply that a firm’s R&D investment increases as its cash holding goes up or down.
IV.2 Econometric methods This study employs a sales accelerator investment model as a base model:
R & Di ,t = β1Δsalesi ,t + β 2 salesi ,t −1 + ε i ,t .
(14)
In this model, we do not need to include Tobin’s Q variable because we use the sales variables as a proxy for the investment fundamentals, which is based on the Keynesian idea that investment decisions depend on past performance. Thus, we can avoid the problem about average Q as a proxy for marginal Q; that is, the © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
FINANCIAL DETERMINANTS OF R&D
127
profitability of investment. Moreover, because Tobin’s Q is the ratio between the market value and the replacement value of a physical asset, it is questionable whether Tobin’s Q is applicable to R&D investment. We add a cash flow term to the base model to investigate the effect of cash flow on R&D investment while controlling for the profitability of investment:
R & Di ,t = β1Δsalesi ,t + β 2 salesi ,t −1 + β3cashi ,t + ε i ,t .
(15)
Then, we include external finance variables as follows:
R & Di ,t = β1Δsalesi ,t + β 2 salesi ,t −1 + β3cashi ,t + β 4 debti ,t + β5equityi ,t + β6 equityi ,t ∗ ownershipi ,t + ε i ,t ,
(16)
where the ownership variable (ownership) includes ownership concentration (con), foreign ownership (for) and institutional ownership (ins). Here, an interaction model is used to investigate the relationship between equity ownership structure and investment. Because we predict that the effect of ownership structure on investment depends on the weight of equity, the interaction variables of equity and ownership structure are appropriate to capture the effect of ownership structure. The classical assumptions of homoskedasticity and no serial correlation might be too restrictive for panel data. We performed many statistical tests, and the results show the presence of serial correlation and heteroskedasticity. To control the heteroskedasticity and serial correlation problems, a feasible generalized least squares (FGLS) regression is employed, and we use a fixed effects model, which has been shown to be appropriate for the sample data by the Hausman test. Year dummies are included in all regressions to control for economic fluctuations over different periods. Industry dummies are not used because the fixed effects model controls for firm-level characteristics. The twostage least squares (2SLS) technique is used to check the endogeneity problem of cash flow variables. The 2SLS regression is a method of calculating the instrumental variable estimates in the first stage, which are later used in the regression. In this study, the lagged cash flow variables for lag orders one to three are used to obtain the instrumental variable estimates in the first stage. In addition to the typical linear regression analysis, a quadratic regression is used to examine possible nonlinear relationships between finance and R&D because nonlinear relationships are often reported in similar studies (for example, Kim et al., 2008; Lee, 2008). To avoid the endogeneity problem of cash flow variables, many studies have proposed a comparative approach between groups of firms. They split the sample into subsamples based on firm characteristics to examine the investment–cash flow sensitivity in each of the subgroups. If the profitability of investment is reflected in the cash flow variable, there is no reason to believe that the profitability is different across the groups of firms. Thus, an observation of different sensitivities would indicate the pure effect of cash flow in explaining investment while controlling for profitability. It is expected that firms © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
ASIAN ECONOMIC JOURNAL
128
with more financial stress are likely to be more sensitive to fluctuations in cash flow. We split the sample based on firm maturity and on whether or not the firm belongs to chaebol. Firm maturity can affect the relationship between investment and cash flow. In contrast to young firms, mature firms often have internal funds in excess of the demand for investment or have easy access to external finance; thus, they could be less sensitive to cash flow. An empirical study by Brown et al. (2009) found that the effect of cash flow on investment exists for young, but not for mature, firms. We use firm age as a proxy for firm maturity stage. We divide the sample into two groups: old and young firms. Firms above the median age are regarded as old firms and those below the median age as young firms. The second criterion based on which the sample is divided into subgroups is whether or not the firm belongs to a chaebol group. It has been argued that firms can benefit from belonging to a business group because the business group’s internal capital market provides a financial cushion to absorb fluctuations in cash flows (Leff, 1978; Khanna and Palepu, 1997). Hoshi et al. (1991) used keiretsu, a large industrial group in Japan, as a sorting device and showed that the cash flow effect on investment is more important for non-keiretsu firms than for keiretsu firms. In another study by Chirinko and Schaller (1995), a similar result was obtained for Canadian firms. For Korean firms, Shin and Park (1999) found insignificant investment–cash flow sensitivity for chaebol firms and significant sensitivity for non-chaebol firms. In this study, we examine the Korea Fair Trade Commission annual reports to determine whether or not a firm is a chaebolaffiliated member. We then divide the sample into two groups: chaebol and non-chaebol firms. In addition to the regression analysis, the Granger causality test (Granger, 1969) is performed as a robust method. The Granger causality test compares the restricted model that y is only explained by the lags of y and the unrestricted model that y is explained by the lags (up to the order) of y and x. In the study, the two regression equations below are compared using the Wald test in order to determine Granger causality: N
M
n =1
i =1
yt = ∑ β y ,t −n yt −n + ∑ N
M
N
N
∑β n =1
a
ai ,t − n i ,t − n
,
N
yt = ∑ β y ,t −n yt −n + ∑ ∑ β ai ,t −n ai ,t −n + ∑ β x ,t −n xt −n , n =1
i =1 n =1
(17)
(18)
n =1
where yt is a dependent variable at time t, xt is an independent variable at time t and ats are the remaining independent and control variables at time t. In the system, Equation 17 is the restricted model and Equation 18 is the unrestricted model. Note that although the Granger causality test is used to detect the causal direction between variables, we need to recognize the limitedness of the test in examining a causal direction because the Granger causality test is not a true causality test but rather a test of precedence. © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
FINANCIAL DETERMINANTS OF R&D
V.
129
Empirical Results and Discussion
The total sample regression results on the relation between R&D investment and internal funds are shown in Table 2. The sales accelerator model regression shows extremely high coefficients and t-scores for the previous sales term and relatively high R2 values. Comparing this result with those of studies that use ordinary investment, such as Guariglia (2008), we confirm that the sales accelerator model is very useful in explaining R&D investment. The significant effect of cash flow on R&D investment is also confirmed, and the 2SLS regression shows that the effect is robust to controlling for the cash flow endogeneity problem. The other cash flow variables (cash1 and cash2) also confirm the significant investment– cash flow sensitivity. However, we could not find any nonlinear relationship, such as an inverse U-shaped relationship between internal funds and investment, from the quadratic regression result. In short, the regression results shown in Table 2 support previous findings that internal funds measured by cash flow have an influence on investment.] The subsample regression results shown in Table 3 are consistent with the total sample regression results that R&D investment is affected by internal funds. The subsample results show that while investment is not sensitive to fluctuations in cash flow among old firms, it is positively sensitive in young firms. Moreover, in contrast to non-chaebol firms, in which cash flow has a positive effect on investment, the relation between investment and cash flow in chaebol firms is even negative. Overall, these two results indicate that investment–cash flow sensitivities are higher for firms a priori classified as more financially distressed. One
Table 2 Accelerator Dsalest salest-1
0.00*** (2.85) 7.30**** (11.03)
casht
Regression results: Internal cash flow
Linear
2SLS
Quadratic
Cash1
Cash2
0.00*** (2.78) 7.28**** (11.00) 0.06**** (4.05)
0.00*** (3.59) 7.90**** (8.93) 0.09**** (4.90)
0.00*** (2.10) 6.67**** (10.10) 0.10**** (6.65) 3.59**** (11.10)
0.00**** (3.20) 7.43**** (11.22)
0.00** (3.14) 7.41**** (11.20)
cash2t
0.06**** (5.06)
cash1t cash2t R2
0.5334
0.5334
0.6199
0.5338
0.5337
0.07**** (5.34) 0.5337
The table shows the results of fixed effects feasible generalized least squares regression. Figures are regression coefficient estimates, and t values are shown in parentheses below coefficient estimates. ****, *** and ** indicate significance at 0.1, 1 and 5% levels, respectively. 2SLS, two-stage least squares. © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
ASIAN ECONOMIC JOURNAL Table 3
130
Subsample regression results: Internal cash flow Maturity
Dsalest salest-1 casht R2
Business group
Old
Young
Chaebol
0.00**** (6.87) 23.41**** (25.55) 0.01 (0.86) 0.7353
0.00 (0.78) -2.86**** (-3.58) 0.14**** (7.95) 0.4619
0.00**** (7.88) 5.18**** (7.82) -0.17**** (-7.50) 0.8580
non-chaebol 0.00 (1.19) 4.90**** (5.77) 0.10**** (5.83) 0.4928
The table shows the results of fixed effects feasible generalized least squares regression. Figures are regression coefficient estimates, and t values are shown in parentheses below coefficient estimates. **** indicates significance at 0.1% level. Year dummies are included in all regressions.
drawback of the subsample regression is that it does not include a formal test of the difference between groups. That is, it does not test whether the effect of internal funds on R&D investment differs between chaebol firms and non-chaebol firms. In this case, a dummy interaction model can be used to test whether the coefficient estimates of internal funds differ across groups. We ran the regression using the interaction variable of chaebol dummy and the cash flow term, and found that the regression confirms the difference across groups. The results are not reported in detail here for brevity. The negative relation in chaebol firms needs an explanation. First of all, it might be a reflection of the severe risk-averse attitude of chaebol firms. Another possible explanation is that at a lower level of internal funds, the firm that invests less faces a higher risk of default. Thus, a drop in cash flow leads to an increase in investment due to the expected revenue from the investment. That is, with lower internal funds, investment increases as internal funds decrease (Cleary et al., 2007). Furthermore, chaebol firms do much of their lending and borrowing within their group using an internal capital market; thus, they are able to invest in strategically important projects even when the cash shortfall is severe (Shin and Park, 1999). The regression results on external finance, presented in Table 4, show the significant and positive effect of debt finance and the insignificant effect of equity finance on investment. This result supports the traditional theory of financing hierarchy called the pecking order theory. We should be cautious when interpreting the effect of external finance. The positive effect of debt might point to the institutional setting in Korea, in which banks are important in corporate financing. During the last several decades of Korean development, the government and banks have played an important role in allocating investment activities through incentives or regulations. In the early 1960s, given the lack of industry infrastructure, the newly established Korean government had to mobilize and allocate © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
FINANCIAL DETERMINANTS OF R&D Table 4
Regression results
Finance
Dsalest salest-1 casht debtt equityt
131
Equity ownership
Linear
Quadratic
Concentration
0.00**** (3.87) 7.50**** (11.31) 0.05**** (3.55) 0.17**** (12.45) 0.02 (1.33)
0.00**** (4.54) 7.28**** (10.97) 0.06**** (4.25) 0.16**** (11.45) -0.83**** (-11.01) 0.79**** (11.86)
0.00**** (4.04) 7.38**** (11.13) 0.06**** (3.77) 0.16**** (11.39) -0.02 (-1.12)
equity2t equityt*cont
Foreign 0.00**** (3.44) 7.41**** (11.17) 0.05**** (3.70) 0.16**** (11.75) 0.00 (0.41)
Institutional 0.00**** (3.72) 7.55**** (11.37) 0.05**** (3.54) 0.17**** (12.42) 0.02 (1.23)
0.13**** (6.35)
equityt*fort
0.30**** (7.51)
equityt*inst R2
All
0.5335
0.5335
0.5335
0.5335
0.06 (1.12) 0.5336
0.00**** (4.04) 7.14**** (10.75) 0.07**** (4.58) 0.14**** (9.80) -0.89**** (-11.69) 0.78**** (11.58) 0.11**** (5.41) 0.33**** (8.14) 0.09 (1.57) 0.5335
The table shows the results of fixed effects FGLS regression. Figures are regression coefficient estimates, and t–values are shown in parentheses below coefficient estimates. **** indicates significance at 0.1% level. Year dummies are included in all regressions.
scarce financial resources. The government owned and controlled all major banks and directed policy loans to strategically targeted sectors, such as heavy and chemical industries, which can realize economies of scale and scope, but involves substantial risk. This practice has been institutionalized and transformed into a generally accepted practice. Thus, large firms have been relying on bank financing for investment. The quadratic regression shows a U-shaped relationship between R&D investment and equity financing. Equity financing has a negative effect on R&D investment at a low level of equity, but it has a positive effect at a high level of equity. This result indicates that equity financing suffers from the problem of asymmetric information at a low level of equity, but it can give managers an incentive to take more risk at a high equity level. The U-shaped relationship may need more detailed investigation, which is left to future work. The results of the quadratic regression of debt financing are not reported here because the quadratic term is not significant. Based on the results, the relationship between ownership concentration and R&D investment is positive and significant. This implies that, for large shareholders in Korea, the long-term orientation effect dominates over the risk-averseness effect. Regarding ownership identity issues, foreign ownership is shown to © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
ASIAN ECONOMIC JOURNAL Table 5
Dsales fi R&D fi sales fi R&D fi cash fi R&D fi debt fi R&D fi equity fi R&D fi con fi R&D fi for fi R&D fi
R&D Dsales R&D sales R&D cash R&D debt R&D equity R&D con R&D for
132 Granger causality test results Order 1
Order 2
Order 3
151.24**** 12.03**** 27.74**** 6.81*** 103.03**** 0.27 12.27**** 3.60′ 116.92**** 9.88*** 0.02 1.18 7.54*** 39.86****
230.22**** 3.28** 120.55**** 11.46**** 12.51**** 6.30*** 20.65**** 29.34**** 53.86**** 4.74*** 26.58**** 2.73 19.51**** 66.46****
933.44**** 3.66** 75.67**** 9.63**** 9.13**** 2.96* 26.18**** 33.21**** 30.58**** 8.78**** 28.87**** 1.99 21.19**** 41.79****
The table shows the result of the Granger causality test. Arrows indicate the direction of granger causality. Order refers to Nth order of lags. Figures are F-values. ****, *** and ** indicate significance at 0.1, 1 and 5% levels, respectively.
positively affect R&D investment, but the coefficient of institutional ownership is insignificant. The following explanation may account for the empirical result on institutional ownership. Some researchers argue that different categories of institutional investors may have varying time horizons. Eng (1999) found that the quarterly changes in the holdings of banks, insurance companies, investment companies and investment advisors are more significant than those in colleges and universities, private foundations, and private and public pensions. The more frequent trading by the former suggests that they may have a shorter time horizon in managing their portfolios than do the latter. However, the latter institutions are expected to make investments that will provide long-term returns for their organizations. Thus, aggregating these two types of institutional investors may lead to a statistically insignificant result. Due to a problem with data availability, this study could not discriminate between the two types of institutional investors; therefore, this task is also left to future work. Another possible explanation for institutional ownership in Korea considers the current position of institutional investors in the country, who are known to be very passive in the corporate governance process. Shareholder activism has been spearheaded primarily by nongovernmental organizations, such as the People’s Solidarity for Participatory Democracy (Choi and Cho, 2003). Institutional investors in Korea may act more as dispersed shareholders concerned predominantly with stock prices. Thus, it might not be surprising that the regression result does not show a meaningful link between institutional ownership and R&D investment. The results of the Granger causality test shown in Table 5 indicate that: (i) sales, internal funds, and ownership concentration clearly affect R&D investment, © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
FINANCIAL DETERMINANTS OF R&D
133
but not vice versa; (ii) debt and equity also seem to affect R&D, but not vice versa; and (iii) foreign ownership and R&D affect each other. Although caution should be taken when interpreting the results of the Granger causality test because the time series of the data is not long enough, it seems clear that internal cash flow Granger-causes R&D investment, not the other way around. Similar conclusions can be drawn from the results on external finance and ownership variables. One point should be mentioned about the interacting effects of foreign ownership and R&D investment. The result shows that R&D investment can simultaneously be an outcome of foreign ownership as well as a cause of the foreign ownership. This implies that foreign investors encourage R&D investment and, at the same time, are attracted to firms with high R&D investment. In addition, the interacting effects between foreign ownership and R&D investment can give rise to an overestimation of the coefficient of the foreign ownership variable.
VI.
Conclusion
This paper investigates the relationship between corporate investment in R&D and finance. In particular, we focus on two issues: (i) the sensitivity of R&D investment to the availability of internal funds; and (ii) the effects of external finance and ownership structure on R&D investment. Corporate finance theories based on transaction costs, asymmetric information and agency problems claim that investment is determined by the availability of the firm’s internal funds as well as the firm’s expected future profitability. The traditional financing hierarchy theory, called the pecking order theory, goes on to argue that internal finance is preferred over external finance, and debt is preferred over equity when external finance is needed. Corporate governance theories usually analyze ownership structure through an agency framework and point out that ownership structure, including ownership concentration and ownership identity, is an important determinant of firm investment decisions. To examine the relation between finance and investment, we conduct an empirical analysis using a sales accelerator model of investment as base model and employing fixed effects FGLS regression, subsample regression, 2SLS regression, quadratic and interaction models, and the Granger causality test. By investigating panel data on Korean manufacturing firms between 1999 and 2008, we find that R&D investment is sensitive to fluctuations in internal cash flow and that debt is more important than equity in financing R&D expenditure. The results of the empirical analysis seem to support the traditional theory of financing hierarchy and the consideration of the institutional setting in Korea. Regarding equity ownership, the empirical results show that ownership concentration and foreign ownership positively affect R&D investment, but institutional ownership does not have any significant effect on investment. We thus conclude that in Korea the positive impacts of large shareholders and foreign investors on R&D investment dominate over their negative effects. © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
ASIAN ECONOMIC JOURNAL
134
The novelty of this study is that it synthetically analyzes the relation between finance and investment by considering corporate finance, equity ownership structure, macroeconomic environments and the institutional setting in Korea. In the empirical analysis, we examine both financial and ownership factors as simultaneous determinants of investment. The empirical analysis takes the effect of interest rate changes and the Korean business/industrial environments into account by applying different cash flow variables and considering the chaebol system and investor characteristics. In addition, we use various empirical methods to present robust econometric evidence. References Amihud, Y. and B. Lev, 1981, Risk reduction as a managerial motive for conglomerate mergers. Bell Journal of Economic, 12, 605–17. Arza, V. and P. Espanol, 2008, Les liaisons dangereuses: A Minskyan approach to the relation of credit and investment in Argentina during the 1990s. Cambridge Journal of Economics, 32, 739–59. Baysinger, B. D., R. D. Kosnik and T. A. Turk, 1991, Effects of board and ownership structure on corporate R&D strategy. Academy of Management Journal, 34, 205–14. Berle, A. and G. Means, 1932, The Modern Corporation and Private Property. Macmillan, New York. Bond, S., D. Harhoff and J. Van Reenen, 2003, Investment, R&D and financial constraints in Britain and Germany. Working Paper. Institute for Fiscal Studies. Brown, J. R., S. M. Fazzari and B. C. Petersen, 2009, Financing innovation and growth: Cash flow, external equity, and the 1990s R&D boom. Journal of Finance, 64, 141–85. Brown, J. R. and B. C. Petersen, 2009, Why has the investment-cash flow sensitivity declined so sharply? Rising R&D and equity market developments. Journal of Banking & Finance, 33, 971–84. Carpenter, R. E. and B. C. Petersen, 2002, Capital market imperfections, high-tech investment, and new equity financing. Economic Journal, 112, F54–72. Chirinko, R. S. and H. Schaller, 1995, Why does liquidity matter in investment equations? Journal of Money, Credit and Banking, 27, 527–48. Choi, W. Y. and S. H. Cho, 2003, Shareholder activism in Korea: An analysis of PSPD’s activities. Pacific-Basin Finance Journal, 11, 349–63. Cleary, S., P. Povel and M. Raith, 2007, The U-shaped investment curve: Theory and evidence. Journal of Financial and Quantitative Analysis, 42, 1–39. Denison, D. R. and A. K. Mishra, 1995, Toward a theory of organizational culture and effectiveness. Organization Science, 6, 204–23. Eng, L. L., 1999, Comparing changes in stockholdings of different institutional investors. Journal of Investing, 8, 46–50. Fazzari, S. M., R. G. Hubbard and B. C. Petersen, 1988, Financing constraints and corporate investment. Brookings Papers on Economic Activity, 1988, 141–206. Francis, J. and A. Smith, 1995, Agency costs and innovation: Some empirical evidence. Journal of Accounting and Economics, 19, 383–409. Gatchev, V. A., P. A. Spindt and V. Tarhan, 2009, How do firms finance their investments? The relative importance of equity issuance and debt contracting costs. Journal of Corporate Finance, 15, 179–95. Granger, C. W. J., 1969, Investigating causal relations by econometric models and cross-spectral methods. Econometrica, 37, 424–38. Greenwald, B., J. E. Stiglitz and A. Weiss, 1984, Informational imperfections in the capital market and macroeconomic fluctuations. American Economic Review, 74, 194–9. Grossman, S. and O. Hart, 1982, Corporate financial structure and managerial incentives. In: The Economics of Information and Uncertainty (eds McCall J. J.), pp. 107–40. University of Chicago Press, Chicago, IL. © 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
FINANCIAL DETERMINANTS OF R&D
135
Guariglia, A., 2008, Internal financial constraints, external financial constraints, and investment choice: Evidence from a panel of UK firms. Journal of Banking & Finance, 32, 1795–809. Guellec, D. and B. van Pottelsberghe de la Potterie, 2001, R&D and productivity growth: Panel data analysis of 16 OECD countries. Working Paper 2001/3, OECD, Paris. Hall, B. H., 1990, The impact of corporate restructuring on industrial research and development. Brookings Papers on Economic Activity, 1990, 85–124. Hall, B. H., 1992, Investment and research and development at the firm level: Does the source of financing matter? NBER Working Paper, NBER. Cambridge, MA. Hall, B. H., 2002, The financing of research and development. Oxford Review of Economic Policy 18, 35–51. Hall, B. H., 2005, The financing of innovation. In: Blackwell Handbook of Technology and Innovation Management (ed. Shane S), pp. 409–30. Blackwell Publishers, Oxford. Hansen, G. S. and C. W. L. Hill, 1991, Are institutional investors myopic? A time-series study of four technology-driven industries. Strategic Management Journal, 12, 1–16. Hoshi, T., A. Kashyap, and D. Scharfstein, 1991, Corporate structure, liquidity, and investment: Evidence from Japanese industrial groups. Quarterly Journal of Economics, 106, 33–60. Hubbard, R. G., 1998, Capital-market imperfections and investment. Journal of Economic Literature, 36, 193–225. Jensen, M. C., 1986, Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review, 76, 323–29. Jensen, M. C. and W. H. Meckling, 1976, Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305–60. John, K., L. Litov and B. Yeung, 2008, Corporate governance and risk-taking. Journal of Finance, 63, 1679–728. Khanna, T. and K. G. Palepu, 1997, Why focused strategies may be wrong for emerging markets. Harvard Business Review 75, 41–51. Kim, H., H. Kim, and P. M. Lee, 2008, Ownership structure and the relationship between financial slack and R&D investments: Evidence from Korean firms. Organization Science, 19, 404–18. Lee, P. M. and H. M. O’Neill, 2003, Ownership structures and R&D investments of U.S. and Japanese firms: Agency and stewardship perspectives. Academy of Management Journal, 46, 212–25. Lee, S., 2008, Ownership structure and financial performance: Evidence from panel data of South Korea. Corporate Ownership and Control, 6, 254–67. Lee, S., 2012, The effect of ownership concentration on R&D decisions in Korean firms. Seoul Journal of Economics, 25, 89–110. Leff, N. H., 1978, Industrial organization and entrepreneurship in the developing countries: The economic groups. Economic Development and Cultural Change, 26, 661–75. Minsky, H. P., 1975, John Maynard Keynes. Colombia University Press, New York. Minsky, H. P., 1986, Stabilizing an Unstable Economy. McGraw-Hill, New York. Modigliani, F. and M. H. Miller, 1958, The cost of capital, corporation finance and the theory of investment. American Economic Review, 48, 261–97. Myers, S. C. and N. S. Majluf, 1984, Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13, 187–221. Schiantarelli, F., 1996, Financial constraints and investment: Methodological issues and international evidence. Oxford Review of Economic Policy, 12, 70–89. Schumpeter, J. A., 1942, Capitalism, Socialism and Democracy. Harper, New York. Shin, H. H. and Y. S. Park, 1999, Financing constraints and internal capital markets: Evidence from Korean ‘chaebols’. Journal of Corporate Finance, 5, 169–91. Shleifer, A. and R. Vishny, 1986, Large shareholders and corporate control. Journal of Political Economy, 94, 461–88.
© 2012 The Author Asian Economic Journal © 2012 East Asian Economic Association and Blackwell Publishing Pty Ltd
