Comparing investments of Dutch private firms with Dutch public firms in terms of risk

Dutch public firms in terms of risk

Abstract

This study examines, based on a large dataset of public and private firms in the Netherlands, if there is a difference in terms of risk in the way private and public firms allocate the capital they raise through debt and equity issues. The risk of an investment is measured in terms of relative risk. Capital expenditures are the riskiest investments, while cash holdings are the investments with the least risk. This paper shows that private firms decide to make riskier investments than public firms the year after they increased their outstanding equity. When private firms increase their outstanding long-term debt, they actually take less risk than public firms by investing significantly more in cash holdings.Furthermore, this paper shows that when the credit rating of a firm is taken into account the difference between private and public firms in taking risk changes. Private firms with an investment grade credit rating take less risk than public firms with an investment-grade credit rating when the financing comes from equity. The preference of private firms for taking less risk than their public counterparts when the financing comes from long-term debt disappears when the credit rating of a firm is taken into account.Finally, private firms with an investment grade credit rating spread the risks of their investments more than public firms with an investment grade credit rating by investing significantly more in total assets.

Sander Hemme

Supervisor: Halit Gonenc

University of Groningen

Faculty of Economics and Business

Table of contents

Introduction

3

Literature Review

6

Data & Methodology

9

Data 9

Methodology 12

Results

16

Standard regression 16

Standard regression with financing interaction variables 17

Extended regression 19

Extended regression with financing interaction variables 20

Conclusion

23

References

25

Introduction

This study examines if there is a difference in the way private and public firms allocate the capital they raise through debt and equity issues. There exists some literature about the investment decisions of public firms. DeAngelo, DeAngelo and Stulz (2007) find that firms issue stock primarily to meet a near-term fundamentals-based need for cash. They, however, do not specify to which firm-level investments the equity is allocated. Kim and Weisbach (2008) study the motivations for equity offerings by public firms from 38 countries between 1990 and 2003. They test if firms invest the funds they attracted from an equity issue in their total assets, inventory, capital expenditures, acquisition, research and development, cash holdings, and long-term debt reduction. There exist, however, few papers that compare the differences between public and private firms in terms of investment choices. Brav (2009) is, to my knowledge, the only author who discusses the differences in investments between private and public firms. The fact that attracting equity is relatively more expensive to private firms than for public firms has some implications for the way these firms allocate their attracted capital. That is, Brav (2009) finds that private firms stockpile their cash holdings after they issued new equity and then eat into this cash until the next equity issue. Therefore, I hypothesize that private firms invest relatively more in cash holdings than public firms the year after they issued new equity. Furthermore, Brav (2009) finds that while public firms increase their capital expenditures as soon as their profitability increases, private firms do so with a lag. He argues that if profitability also measures the availability of positive NPV-projects, access to external capital markets allows a firm to profit more from the available investment opportunities. Therefore, I hypothesize that public firms invest relatively more of their attracted capital in capital expenditures than private firms the year after they issued new equity. Since there is no difference between the two firm types in the accessibility of new debt, I hypothesize no difference between investments done by private firms versus investments done by public firms the year after they issued new debt.

types of investments are tested as result of a change in outstanding equity or a change in outstanding long-term debt in Dutch public and private firms from 2004 up and to 2007. The three investment types that are tested differ in terms of risk level. The investment type with the least risk is the firm’s cash holdings. Cash is a relatively safe investment because it rarely loses value, as can stocks or bonds. The investment type considered in this paper that bares the most risk is the capital expenditures. Although capital expenditures exist in many forms, they bare more risk than investing in cash holdings. The third investment type that is considered is the firm’s total assets. The capital expenditures and cash holdings of a firm are both an element of a firm’s total assets. The risk level of investing in total assets lies therefore between the risk of investing in capital expenditures and the risk of investing in cash holdings. The main question of this paper therefore is:

Is there a difference between investments done by Dutch public firms versus investments done by Dutch private firms in terms of risk?

Furthermore, this paper will investigate if credit ratings influence the difference between investments done by Dutch public firms versus investments done by Dutch private firms. Firms rebalance their capital structure to stay in the optimal range (Leary and Roberts, 2005). When firms are out of this optimal range, due to, for instance, a below-investment grade credit rating, they want to issue equity to regain the optimal leverage (see for instance Leary and Roberts, 2005; Graham and Harvey, 2001). Firms with investment grade credit ratings should be able to raise new external financing easier and less costly than firms with non-investment grade credit ratings. Since attracting equity is already more difficult for private firms than for public firms, I hypothesize that credit ratings influence the difference between investments done by Dutch public firms and investments done by Dutch private firms in terms of risk.Therefore, the sub-question is:

Literature review

There are several theories in the economic literature that discuss the motivations firms have to raise certain forms of capital. In their 1958 article, Modigliani and Miller conclude that under the ideal conditions of perfect capital markets, no taxes and no bankruptcy or financial distress costs, the total value of a firm is independent of its debt level. According to Modigliani and Miller (1958) an increase in debt dilutes the debt holders’ claim and benefits the shareholders, but the gain of the shareholders exactly offsets the loss of the debt holders. This proves that under perfect capital market assumptions corporate financing and investment decisions are completely separable. The Modigliani and Miller debt irrelevance proposition holds because of the assumed ideal conditions. In practice, however, the conditions are not perfect. Therefore, there are several theories, like the tradeoff theory, the pecking order theory, and the market timing theory, developed that take these imperfect conditions into account. These theories show under which conditions it can be optimal to rather issue debt than equity or vice versa.

Although, the above mentioned theories offer several predictions with respect to the financing policies and capital budgeting decisions of public firms, they do not elaborate on the possible differences in financing policies between public and private firms. Brav (2009) classifies the predictions with respect to the relative differences in the financial policies and capital structures between private and public firms into two groups: the level effect and the sensitivity effect. The level effect refers to “consequences that

arise from the fact that private firms’ relative cost of equity to debt capital is higher than that of public firms” (Brav, 2009). Since the relative cost of equity to debt for private

firms is higher than the relative cost of equity to debt for public firms, the frequency with which debt versus equity financing is chosen is higher for private firms than for public firms. This leads to a higher leverage ratio for private firms vis-à-vis public firms. The sensitivity effect refers to “consequences that arise from the fact that private firms’

absolute cost of accessing the external capital markets is higher than that of public firms”

arguments that explain the difference in capital structure and financing policy between public and privately held firms. Amihud, Lev and Travlos (1990) and Stulz (1988) argue that firms controlled by a major shareholder should be reluctant to use equity financing when doing so causes the controlling shareholder the risk of losing control. Private firms are held by, at most, a few shareholders who all have significant control over the firm, where public firms are mostly held by many shareholders without significant control over the firm. Therefore, the cost for private firms to issue equity is higher than for public firms, since shareholders of private firms have to give up some of their control (Brav, 2009). Furthermore, Morellec (2004) shows that given the separation between management and ownership, which is typical for public, but not for private firms, managers of public firms may decide to diminish the control of a single shareholder by issuing more equity. Finally, Brav (2009) argues that private firms are more opaque to outsiders, and since the value of equity is more sensitive to information asymmetry than the value of debt, the cost of equity relative to debt will be much higher for private firms than for public firms.

Triantis (1994) therefore analyze the interaction between financing and investment decisions in a multi-period contingent claim model where the firm has the flexibility to dynamically manage both decisions over time. They find that the impact of debt financing on the firm’s investment and operating decisions is economically insignificant.

Brav (2009) is the first to link the differences between private and public firms with respect to financing to the differences between public and private firms with respect to investments. He finds that private firms increase their cash holdings after an equity issue and then eat into their cash holdings until the next issue. Furthermore, Brav (2009) finds that while public firms increase their capital expenditures as soon as their profitability increases, private firms do so with a lag. He argues that “if profitability also

measures the availability of positive NPV-projects, this result suggests that access to public capital markets allows a firm to better exploit the available investment opportunities”.

Brav’s (2009) finding that public firms invest relatively more in capital expenditures, while private firms invest relatively more in cash holdings after they increased their outstanding debt or equity can be explained with the use of Jensen and Meckling’s (1976) extension on Berle and Means’ (1932) agency theory concerning the shareholders and the managers as well. This theory predicts that managers are rewarded for firm performance measures, such as accounting earnings and stock returns, in order to align the interests of the shareholders with those of the managers. This is consistent with the central belief in this approach that there are conflict of interest between a firm’s owners and managers. Owners want their manager to maximize the value of their firm, whereas managers can act in their own interest. This can lead them to choose those courses of action that are in their self-interest, even if they conflict with the welfare of the owners (Demski, Patell, and Wolfson, 1984). Managers can choose to invest in projects with higher risks which will result in higher pay on the short-term, whereas it would be optimal for the firm value to invest in projects which result in a higher net present value.

Data & Methodology

Data

A dataset is retrieved from the Amadeus Database, including all Dutch private and public firms from 2004 up and to 2007. All financial institutions and utilities are excluded from the sample due to their different balance sheets. In total 37014 observations are found in the given time period. These observations include all firms from which there is at least one data point available.

From all firms the following characteristics are retrieved from the AMADEUS database: total assets, cash and cash equivalents, tangibles, long-term debt, shareholders funds, depreciation, earnings before interest and taxes (EBIT), interest paid, MORE credit ratings, and the NACE Rev. 2 code. All the values of these variables are inflation-adjusted with the use of a GDP-deflator (World Bank, World Development Indicators, ERS estimates and ERS Baseline Regional Aggregations). The year 2000 is used as the base year for the inflation-adjustment. After removing 12872 observations due to the lack of data availability, 22420 observations remain present in the sample. The final sample consists of 480 observations from Dutch public firms and 21940 observations from Dutch private firms.

2004 ∆EQUITY ∆DEBT SIZE CASHFLOW TA CASH CAPEX PUBLIC Mean 0.0310 -0.0179 12.1405 0.0295 0.0104 0.0149 0.0468 Median 0.0186 -0.0007 12.1944 0.0135 0.0148 0 0.0415 117 obs. PRIVATE Mean 0.0538 0.0029 9.7475 0.0188 0.1085 0.0184 0.0655 Median 0.0229 0 9.7300 0 0.0371 0 0.0181 5142 obs. TOTAL mean 0.0533 0.0024 9.8007 0.0190 0.1063 0.0184 0.0650 median 0.0228 0 9.7570 0 0.0367 0 0.0184 5259 obs.

2005 ∆EQUITY ∆DEBT SIZE CASHFLOW TA CASH CAPEX PUBLIC Mean 0.0862 0.0230 12.0794 0.0403 0.1760 0.0328 0.1447 Median 0.0435 0 12.2057 0.0122 0.1110 0.0016 0.1000 120 obs. PRIVATE Mean 0.0403 0.0024 9.8292 0.0196 0.1119 0.0136 0.0674 Median 0.0212 0 9.7804 0 0.0542 0 0.0200 5701 obs. TOTAL mean 0.0412 0.0028 9.8756 0.0201 0.1132 0.0140 0.0690 median 0.0216 0 9.8097 0 0.0560 0 0.0210 5821 obs.

2006 ∆EQUITY ∆DEBT SIZE CASHFLOW TA CASH CAPEX PUBLIC Mean 0.0897 0.0267 12.3384 0.0247 0.1539 0.0317 0.1479 Median 0.0403 0 12.3788 0.0262 0.0804 0.0021 0.0727 120 obs. PRIVATE Mean 0.0593 0.0002 9.9211 0.0157 0.1248 0.0162 0.0721 Median 0.0263 0 9.8560 0.0021 0.0677 0 0.0188 5711 obs. TOTAL mean 0.0599 0.0007 9.9709 0.0159 0.1254 0.0165 0.0736 median 0.0267 0 9.8799 0.0023 0.0678 0 0.0193 5831 obs.

2007 ∆EQUITY ∆DEBT SIZE CASHFLOW TA CASH CAPEX PUBLIC Mean 0.0572 -0.0058 12.4421 -0.0016 0.0573 0.0063 0.0839 Median 0.0296 0 12.6889 0.0163 0.0289 -0.0010 0.0412 123 obs. PRIVATE Mean 0.0817 -0.0009 10.0046 0.0217 0.1441 0.0237 0.0910 Median 0.0323 0 9.9255 0.0041 0.0752 0 0.0224 5386 obs. TOTAL mean 0.0812 -0.0011 10.059 0.0212 0.1422 0.0234 0.0909 median 0.0322 0 9.9484 0.0044 0.0746 0 0.0228 5509 obs.

Table 1: Descriptive statistics of the sample The table above shows the mean and median of the four

Second, the median of the variable describing the change in debt is zero for almost all years for both types of firms, while the mean of this variable is never zero. This implies that most firms do not change their outstanding debt and that, on average, the debt issues are larger than the debt repurchases. Third, the estimated means and medians of the size variable show that public firms are, as expected, larger than private firms. Furthermore, the large number of positive means and medians of the variable representing the cash flow of a firm show that most firms have funds originating from their previous’ year net income and depreciation to invest as well. Finally, almost all means and medians of the variables representing the investments in total assets, cash holdings and capital expenditures are positive, implying that most firms invest in one of these items.

Table 2 compares the means and medians of the relevant variables for public and private firms. The results in table 2 show that public firms increase their outstanding equity significantly more than private firms. It is remarkable to notice that most firms, public and private, do not change their outstanding debt. When there is a change in outstanding debt, public firms increase their outstanding debt significantly more than private firms. Furthermore, public firms are, as one may expect, significantly larger than private firms. Finally, table 2 shows no significant difference in the increase of total assets and cash holdings between public and private firms, but it does show that public firms increase their capital expenditures significantly more than private firms.

∆EQUITY ∆DEBT SIZE CASHFLOW TA CASH CAPEX MEAN Public 0.0662 0.0066 12.2520 0.0230 0.0997 0.0214 0.1060 Private 0.0586 0.0011 9.8770 0.0189 0.1224 0.0179 0.0740 Difference 0.0076* 0.0054*** 2.3750*** 0.0041 -0.0227 0.0035 0.0321*** MEDIAN Public 0.0346 0 12.3060 0.0161 0.0597 0.0004 0.0637 Private 0.0254 0 9.8317 0.0010 0.0586 0 0.0199 Difference 0.0092*** 0 2.4743*** 0.0152*** 0.0011 0.0004 0.0438***

Table 2: Comparison of means and medians of relevant variables for public and private firms The

Methodology

This study examines if there is a difference in the way private and public firms allocate the capital they raise through debt and equity issues. To identify where firms invest this capital, this study uses a methodology that is based on the methodology used in the Kim and Weisbach (2008) study. In particular, the change in total assets, capital expenditures, and cash holdings are measured the year following an increase in outstanding long-term debt or outstanding equity. The increase in total assets and capital expenditures are, according to Kim and Weisbach (2008), meant to capture possible ways in which the capital could be invested in real assets, while changes in cash holdings measures possible uses of the funds other than direct investment. To minimize the impact of outliers this study focuses, in accordance with Kim and Weisbach (2008), on a logarithmic transformation of each variable. This results for each variable in the natural logarithm of one plus the change in each variable normalized by total assets prior to the change in debt or equity is calculated:ln[((Vt Vt1)/totalassetst1)1], where V is the variable being measured, year t-1 is the fiscal year-end just prior to the change in debt or equity, and t is the year in which the debt or equity issue takes place.

To estimate the investments resulting from an increase in outstanding long-term debt or outstanding equity both sources of funds are normalized by total assets and the natural logarithm of one plus the normalized cash flow is taken as a way of minimizing the effects of outliers. Furthermore, the natural logarithm of the total assets is included as a control for size, the natural logarithm of the sum of a firm’s net income and depreciation is included to control the level of cash a firm already holds, a private dummy is included to control for the nature (private or public) of a firm, ten industry dummies are included to control for industry effects, and four year dummies are included to control for fixed year effects. This leads to the following equation to test if there is a difference in investments as a result of a change in equity versus a change in debt:

where ] 1 ) / ) ln[((  _{1 1}   Vt Vt totalassetst

Y for V = total assets, cash holdings, or CAPEX.

t t t TANGIBLES DEPRECIATION TANGIBLES CAPEX   1  ] 1 ) ln[( 1 1       t t t assets total debt longterm debt longterm DEBT ) ln( 1  totalassets_t SIZE ] 1 ) ln[( 1 1      t t t assets total CashFlow CashFlow CASHFLOW

Cash Flow = (Net Income + Depreciation)

PRIVATE = This dummy variable has the value 1 if the firm is private and the value 0 if

the firm is public.

From IND0 up and to IND9 = These variables are dummy variables for industries. The first number of a firm’s NACE Rev. 2 code is used to determine in which industry the firm belongs.

From YEAR04 up and to YEAR07 = These variables are year dummy variables for the years 2004 up and to 2007. The year that belongs to the observation receives the value 1, while the other years receive the value 0.

To test the difference between investments done by private firms versus investments done by public firms, two financing interaction variables are added to equation (1). The equation with the financing interaction variables is as follows:

where

INTERACTION1 = [ EQUITY ] x [PRIVATE]

INTERACTION2 = [DEBT] x [PRIVATE]

To test if a firm’s credit rating influences its investment decisions a credit rating dummy is added to the equations above. The credit rating dummy is assigned a value of 1 when the firm has an investment-grade rating (BBB or higher) during the year the change in debt or equity took place and a value of 0 when the firm has a non-investment grade rating (lower than BBB) during the year the change in debt or equity took place. The MORE credit ratings from the AMADEUS database are used to determine a firm’s credit rating. The addition of a credit rating dummy to the equations results in adjusted interaction variables as well. There is a credit rating interaction variable added to equation (1), resulting in the following equation:

                     





  ] 3 [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ 21 20 19 16 15 6 5 4 3 2 1 N INTERACTIO RATING YEARi i INDi i PRIVATE CASHFLOW SIZE DEBT EQUITY Y i i (3)

Furthermore, a credit rating dummy and two financing interaction variables are added to equation (2) to examine the influence of a firm’s credit rating on its investment decisions as well. This results in the following equation:

                   







   ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ 22 21 20 19 16 15 6 5 4 3 2 1 RATING Ni INTERACTIO i YEARi i INDi i PRIVATE CASHFLOW SIZE DEBT EQUITY Y i i i (4)

RATING = A dummy variable which is assigned a value of 1 when the firm has an

investment grade rating (BBB or higher), and a value of 0 when the firm has a non-investment grade rating.

INTERACTION3 = [RATING] x [PRIVATE]

INTERACTION4 = [RATING] x [PRIVATE] x [EQUITY] INTERACTION5 = [RATING] x [PRIVATE] x [DEBT]

Before the regressions are actually estimated, several aspects of the used sample have to be tested. To avoid the problem of heteroscedasticity in the error-terms all regressions are estimated with the use of heteroscedasticity-robust standard errors Furthermore, if the explanatory variables are correlated with one another, this is known as multicollinearity. With the use of a correlation matrix the degree of multicollinearity is tested. Appendix A shows the correlation between the independent variables. The correlation matrix shows no indication of removing one of the independent variables, since the highest correlation between two of the independent variables is a correlation of 0.356.

Results

This section discusses the results of several regressions. First, the standard regression, without the interaction variables is discussed. Second, the results of the standard regression with the inclusion of financing interaction variables are described. The financing interaction variables are composed of a financing component and the private dummy. The financing component is the variable describing the increase in outstanding equity or the variable describing the increase in outstanding debt. Third, an extended regression without the inclusion of the financing interaction variables is discussed. The extended regression is the standard regression with the inclusion of a credit rating dummy and a credit rating interaction variable. The credit rating interaction variable consists of the private dummy and the credit rating dummy. Finally, this section describes the results of the extended regression with the inclusion of financing interaction variables.

Standard regression TA CASH CAPEX ∆EQUITY 0.9877 (143.25***) 0.1705 (46.98***) 0.6175 (105.08***) ∆DEBT 0.4791 (16.50***) 0.0968 (5.24***) 0.5090 (17.02***) SIZE -0.0476 (-28.37***) -0.0134 (-11.75***) -0.0313 (-17.08***) CASHFLOW 0.0941 (7.71***) 0.1952 (24.19***) -0.0029 (-0.22) PRIVATE -0.0792 (-4.14***) -0.0285 (-2.13**) -0.1069 (-4.94***) R-squared 0.5754 0.1659 0.3799 Adj. R-squared 0.5750 0.1653 0.3795

Table 3: The relative difference in investments resulting from a change in long-term debt versus a change in equity For all regressions the values of all coefficients are given, with their accompanying

t-statistics in parentheses. When a t-statistic is labeled with *, ** or *** the value of the coefficient is statistically significant on a 10%, 5% or 1% significance level, respectively. ∆EQUITY stands for the change in outstanding equity, ∆DEBT stands for the change in outstanding long-term debt, SIZE stands for the natural logarithm of the total assets, CASHFLOW stands for the change in a firm’s cash flow, and PRIVATE stands for the dummy variable denoting 1 if the firm is private and 0 if the firm is public. The remaining three variables are the dependent variables. TA stands for the change in total assets, CASH stands for the change in a firm’s cash holdings, and CAPEX stands for the change in a firm’s capital expenditures.

influence of a change in outstanding equity on the total assets and the cash holdings of a firm is almost twice as large as the influence of a change in outstanding long-term debt on these two independent variables. The difference between the estimated coefficients of both variables with respect to the capital expenditures of a firm is rather small. The significant negative coefficient of the private dummy variable for all three dependent variables shows that private firms invest significantly less in total assets, cash holdings and capital expenditures than public firms.

Standard regression with financing interaction variables

TA CASH CAPEX ∆EQUITY 0.9747 (18.13***) 0.1227 (3.27***) 0.4054 (6.67***) ∆DEBT 0.5423 (4.97***) -0.0979 (-1.28) 0.5078 (4.10***) SIZE -0.0476 (-28.37***) -0.0133 (-11.73***) -0.0313 (-17.07***) CASHFLOW 0.0942 (7.71***) 0.1957 (24.22***) -0.0005 (-0.04) PRIVATE -0.0796 (-4.01***) -0.0326 (-2.40**) -0.1204 (-5.47***) INTERACTION1 0.0132 (0.24) 0.0480 (1.28) 0.2135 (3.51***) INTERACTION2 -0.0681 (-0.60) 0.2066 (2.62***) 0.0002 (0.00) R-squared 0.5754 0.1662 0.3802 Adj. R-squared 0.5750 0.1655 0.3797

Table 4: The difference in investments between Dutch public and private firms For all regressions the

values of all coefficients are given, with their accompanying t-statistics in parentheses. When a t-statistic is labeled with *, ** or *** the value of the coefficient is statistically significant on a 10%, 5% or 1% significance level, respectively. ∆EQUITY stands for the change in outstanding equity, ∆DEBT stands for the change in outstanding long-term debt, SIZE stands for the natural logarithm of the total assets, CASHFLOW stands for the change in a firm’s cash flow, PRIVATE stands for the dummy variable denoting 1 if the firm is private and 0 if the firm is public, INTERACTION1 stands for the interaction variable between the PRIVATE dummy and ∆EQUITY variable, and INTERACTION2 stands for the interaction variable between the PRIVATE dummy and the ∆DEBT variable. The remaining three variables are the dependent variables. TA stands for the change in total assets, CASH stands for the change in a firm’s cash holdings, and CAPEX stands for the change in a firm’s capital expenditures.

contradiction is presumably caused by the addition of several variables to the regression underlying the results presented in table 4.

The results in table 4 show no statistical significant relation between a change of outstanding long-term debt during the previous year and the amount of cash a firm holds in the next year. The negative sign of the statistically significant coefficient of the dummy for private firms indicates that private firms invest significantly less in their total assets, cash holdings and capital expenditures than public firms the year following an increase in their outstanding long-term debt or equity. Although private firms invest on average less in these three items than their public counterparts, there is, judging by the interaction variables, a difference in investments when the type of attracted capital is split up in outstanding equity and long-term debt. The coefficient of the second interaction variable with respect to capital expenditures implies that a private firm that increased its outstanding long-term debt is more likely to invest in cash holdings than a public firm that increased its outstanding long-term debt. It is remarkable to notice that the coefficient of this second interaction variable, which describes the interaction between the change of outstanding long-term debt and being a private firm, is statistically significant, while the coefficient of the variable describing the increase of outstanding long-term debt has no statistical significant influence on the amount of cash is held by a firm. This means that the increase of outstanding long-term debt only has a statistical significant influence on the amount of cash held by a firm in the case of a private firm.

show that there is no difference between private and public firms with respect to their investments in total assets as a result of an increase in outstanding long-term debt or outstanding equity.

This means that private firms decide to make riskier investments than public firms the year after they increased their outstanding equity. This contradicts the results found by Brav (2009) that public firms invest more in capital expenditures, while private firms are more likely to invest in the less-riskier cash holdings. When private firms increase their outstanding long-term debt, they take less risk than public firms by investing significantly more in cash holdings. This result contradicts the hypothesis that there is no difference between investments done by public and private firms as a result of a debt issue.

Extended regression

This interaction variable shows what the influence of a private firm with an investment grade-credit rating is on the allocation of the available funds. The lack of statistical significance of the interaction variable shows that a credit rating does not affect the investment decision of a firm.

TA CASH CAPEX ∆EQUITY 0.9870 (141.86***) 0.1721 (47.29***) 0.6133 (104.42***) ∆DEBT 0.4749 (16.34***) 0.0923 (5.01***) 0.5025 (16.92***) SIZE -0.0468 (-27.70***) -0.0131 (-11.48***) -0.0291 (-15.89***) CASHFLOW 0.1053 (8.46***) 0.2170 (26.47***) -0.0109 (-0.82) PRIVATE -0.0676 (-2.61***) -0.0309 (-1.71*) -0.1234 (-4.25***) RATING 0.0535 (1.45) 0.0038 (0.15) 0.0238 (0.57) INTERACTION3 -0.0189 (-0.51) 0.0075 (0.29) 0.0454 (1.08) R-squared 0.5690 0.1714 0.3752 Adj. R-squared 0.5686 0.1707 0.3747

Table 5: The relative difference in investments resulting from a change in long-term debt versus a change in equity with the rating dummy included in the regression For all regressions the values of all

coefficients are given, with their accompanying t-statistics in parentheses. When a t-statistic is labeled with *, ** or *** the value of the coefficient is statistically significant on a 10%, 5% or 1% significance level, respectively. ∆EQUITY stands for the change in outstanding equity, ∆DEBT stands for the change in outstanding long-term debt, SIZE stands for the natural logarithm of the total assets, CASHFLOW stands for the change in a firm’s cash flow, PRIVATE stands for the dummy variable denoting 1 if the firm is private and 0 if the firm is public, RATING stands for the dummy variable denoting 1 if the firm has an investment-grade rating and 0 if the firm has a non-investment-grade rating, and INTERACTION3 stands for the interaction variable between the RATING dummy variable and the PRIVATE dummy variable. The remaining three variables are the dependent variables. TA stands for the change in total assets, CASH stands for the change in a firm’s cash holdings, and CAPEX stands for the change in a firm’s capital expenditures.

Extended regression with financing interaction variables

assets than public firms, regardless of the form of capital they attracted in the previous year.The results in table 4 show, however, no difference in the investments in total assets between private firms and public firms.The difference in this result, therefore, is caused by the inclusion of the credit rating dummy. Fourth, table 6 shows no statistical significant difference between private and public firms with respect to the investments in cash holdings. This contradicts the results in table 4 as well. The results in table 4 show that private firms that increased their outstanding long-term debt in the previous year are, compared with public firms, more likely to invest in cash holdings. Fifth, the negative statistical significant coefficient of the fourth interaction variable with respect to investments in capital expenditures shows that private firms with an investment grade credit rating that increased their outstanding equity in the previous year are less likely to invest in capital expenditures compared to public firms with an investment grade credit rating. This contradicts the results presented in table 4 as well, since the results in table 4 show that private firms that increased their outstanding equity in the previous year are even more likely to invest in capital expenditures compared to their public counterparts.

risks of their investments more than public firms with an investment grade credit rating by investing significantly more in total assets. Private firms invest significantly more in total assets regardless if the attracted capital results from an equity issue or a debt issue. This contradicts the hypothesis that there is no difference in the investments of public firms and the investments of private firms the year after a debt issue.

TA CASH CAPEX ∆EQUITY 0.9400 (93.01***) 0.1723 (29.20***) 0.7432 (77.20***) ∆DEBT 0.4071 (12.85***) 0.0995 (5.01***) 0.4853 (15.23***) SIZE -0.0469 (-27.86***) -0.0131(-11.50***) -0.0278 (-15.26***) CASHFLOW 0.1057 (8.51***) 0.2171 (26.50***) -0.0191 (-1.45) PRIVATE -0.0808(-4.23***) -0.0274 (-2.05**) -0.0938 (-4.39***) RATING 0.0293 (5.30***) 0.0114 (3.06***) 0.0718 (12.01***) INTERACTION4 0.0828 (6.51***) -0.0003 (-0.04) -0.1946 (-16.99***) INTERACTION5 0.4833 (6.09***) -0.0518 (-0.98) -0.0158 (-0.19) R-squared 0.5705 0.1715 0.3825 Adj. R-squared 0.5701 0.1708 0.3820

Table 6: The difference in investments between Dutch public and private firms with the credit rating dummy included in the regression For all regressions the values of all coefficients are given, with their

Conclusion

This paper focuses on the difference in the riskiness of investments between Dutch public and private firms. In the period from 2004 up and to 2007 there were 22424 public or private firms that changed their outstanding long-term debt or equity in the Netherlands. This paper provides evidence regarding the difference in investments in terms of risk between Dutch public and private firms. Furthermore, evidence is provided regarding the influence of credit ratings on the difference in investments in terms of risk between Dutch public and private firms.

An increase of a firm’s outstanding equity or long-term debt in the previous year results in a significant increase of a firm’s total assets, cash holdings and capital expenditures in the next year. The influence of an increase in outstanding equity on the total assets and the cash holdings of a firm is almost twice as large as the influence of an increase in outstanding long-term debt on these two independent variables. The difference between the estimated coefficients of both variables with respect to the capital expenditures of a firm is rather small. These results are robust when a credit rating dummy is included in the regression.

The results in this paper show that private firms decide to make riskier investments than public firms the year after they increased their outstanding equity. This contradicts the results found by Brav (2009) that public firms invest more in capital expenditures, while private firms are more likely to invest in the less-riskier cash holdings. When private firms increase their outstanding long-term debt, they actually take less risk than public firms by investing significantly more in cash holdings. This contradicts the hypothesis that there is no difference in the investments of public firms and the investments of private firms the year after a debt issue.

less risk with their investments than public firms with an investment-grade credit rating when the financing comes from equity. This is in accordance with the results found by Brav (2009). Furthermore, the preference of private firms for taking less risk than their public counterparts when the financing comes from long-term debt disappears. Private firms with an investment grade credit rating spread the risks of their investments more than public firms with an investment grade credit rating by investing significantly more in total assets. Private firms invest significantly more in total assets regardless if the attracted capital results from an equity issue or a debt issue. This contradicts the hypothesis that there is no difference in the investments of public firms and the investments of private firms the year after a debt issue.

References

Amihud, Y., Lev, B. and Travlos, N.G. (1990) “Corporate Control and the Choice of

Investment Financing: The Case of Corporate Acquisitions”, The Journal of Finance, Vol.

45 No. 2, p. 603-616

Berle, A.A. Jr. and Means, G.C. “The modern corporation and private property”, Macmillan, New York, 1932.

Bradley, M., Jarrell, G.A., and Kim, E.H. (1984) “On the Existence of an Optimal

Capital Structure: Theory and Evidence”, The Journal of Finance, Vol. 39 No. 3, p.

857-878.

Brav, O. (2009) “Access to Capital, Capital Structure, and the Funding of the Firm”, The Journal of Finance, Vol. 64 No. 1, p. 263-308

Brooks, C., “Introductory Econometrics for Finance”, Second edition, New York: Cambridge University Press, 2008

DeAngelo, H., DeAngelo, L. and Stulz, R.M. (2007) “Fundamentals, Market Timing, and

Seasoned Equity Offerings”, Working Paper, Ohio State University

Demski, J.S., Patell, J.M. and Wolfson, M.A., “Decentralized choice of monitoring

systems”, Accounting Review, 1984, Vol. 14 p. 16-34

Graham, J.R. and Harvey, C.R. (2001) “The Theory and Practice of Corporate Finance:

Evidence from the Field”, Journal of Financial Economics, Vol. 60, No. 2-3, p. 187-243.

Jensen, M.C. and Meckling, W.H. (1976) “Theory of the Firm: Managerial Behavior,

Agency Costs and Ownership Structure”, Journal of Financial Economics, Vol. 3 No. 4, p.

Kim, W. and Weisbach, M.S. (2008) “Motivations for public equity offers: An

international perspective”, Journal of Financial Economics, Vol. 87 No. 2, p. 281-307

Mauer, D.C., and Triantis, A.J. (1994) “Interactions of Corporate Financing and

Investment Decisions: A Dynamic Framework”, Journal of Finance, Vol. 49 No. 4, p.

1253-1277.

Leary, M.T. and Roberts, M.R. (2005) “Do Firms Rebalance Their Capital Structures?”, The Journal of Finance, Vol. 60 No. 6, p. 2575-2619.

Modigliani, F. and Miller, M.H. (1958) “The cost of capital, corporation finance and the

theory of investment”, The American Economic Review, Vol. 48 No. 3, p. 261-297

Morellec, E. (2004) “Can Managerial Discretion Explain Observed Leverage Ratios?”, Review of Financial Studies, Vol. 17 No. 1, p. 257-294

Myers, S.C. (1974) “Interactions of Corporate Financing and Investment Decisions –

Implications for Capital Budgeting”, Journal of Finance, Vol. 29 No. 1, p. 1-25.

Myers, S.C. (1984) “The capital structure puzzle”, Journal of Finance, Vol. 39 No. 3, p. 573-592.

Stulz, R.M. (1988) “Managerial control of voting rights: Financing policies and the

market for corporate control”, Journal of Financial Economics, Vol. 20 No. 1, p. 25-54

Titman, S., and Wessels, R. (1988) “The Determinants of Capital Structure Choice”, Journal of Finance, Vol. 43 No. 1, p. 1-19.

Tirole, J. “The theory of corporate finance”, Princeton: Princeton University Press, 2006

Appendix A

∆DEBT ∆EQUITY SIZE CASHFLOW ∆DEBT 1 0.038 -0.031 0.051

∆EQUITY 1 -0.178 0.356

SIZE 1 -0.117

CASHFLOW 1

Table 7: Correlation matrix This table shows the correlation between the several relevant independent