1
The Incremental Financing Decisions within Firms
LEON SCHOLTE ALBERS
1ABSTRACT
We study the extent to which the modified pecking order theory of capital structure can account for incremental financing decisions of United States’ listed companies for the years 2003 to 2011. First, a multinomial logit model is estimated. The results offer modest support for the modified pecking order theory. Next, we consider ordered logit models. Firms appear to have a unique ordered preference for financing types that is identical to the modified pecking order hierarchy. However, there seems to be no difference between financing hierarchies in a statistical sense. This offers modest support for the modified pecking order theory.
THE MODERN THEORY OF capital structure was established by Modigliani and Miller (1958). They illustrated that in perfect capital markets, without taxes and transaction costs, the value of the firm depends only on the level and risk of future cash flows. This implies that the financing decision does not affect the value of the firm. In this situation firms are indifferent in financing with internal funds, or different forms of external financing. According to Modigliani and Miller there is no optimal capital structure.
Harris and Raviv (1991) survey the theoretical literature that explain the differences in observed capital structures in different economies. This theoretical literature has proven that the irrelevance assumption underlying the Modigliani and Miller does not hold in reality. Hence, in contrast to Modigliani and Miller, the financing decision does affect the value of the firm. Thus, one of the challenges in corporate finance is to provide an explanation as to why some firms finance incremental investments with debt, while others do so with equity. Over the years, several explanations for this empirical fact have been given that, broadly speaking, can be grouped in two schools of thought.
The first is the traditional static trade-off theory. It predicts that an optimal target financial debt ratio exists, which maximizes the value of the firm. This optimal point can be attained when the
1
2
marginal value of the benefits associated with debt issues exactly offsets the increase in the present value of the costs associated with issuing more debt (Myers, 2001). The benefits of debt are the tax deductibility of interest payments. Tax deductibility of corporate interest payments favors the use of debt. If firms are above the target debt ratio the value of the firm is not optimal. Firms face financial distress costs (Modigliani and Miller, 1963) and agency costs triggered by the conflicts between shareholders and debtors.
The second is the pecking order theory, it posits that a firm’s capital structure is established by a financing hierarchy. Due to information asymmetry, firms prefer to finance their activities using internal funds if possible. If internal funds are inadequate, then they turn to the use of debt. Equity financing is only used as a last resort (Frank and Goyal, 2003). The general idea behind the pecking order theory is that if a firm issues a risky security, this security will be undervalued by the market due to information asymmetry (Myers, 1984). Since debt is safer than equity and therefore less exposed to information asymmetry, debt is preferred over equity. Internal funds are not exposed to information asymmetry.
There has been little agreement on whether and when the pecking order is a good descriptor of observed financing behavior. Shyam-Sunder and Myers (1999), for instance conclude that the pecking order is a good descriptor of broad financing patterns; Frank and Goyal (2003) conclude the opposite. According to Lemmon and Zender (2010) earlier study results are mixed because these do not take into account the debt capacity of firms, they conclude that a modified pecking order is a better descriptor of financing behavior; Fama and French (2005) conclude the opposite. Determining what lies behind these contrasting findings is important for our understanding of capital structure and the incremental financing choices of firms. Recent studies show significant external equity issues by high-growth firms (Frank and Goyal, 2003 and Fama and French, 2005). These studies reject the hypotheses of the pecking order theory, as equity should only be issued as a last resort. The extensive use of new equity by these firms may be explained in a pecking order framework by taking into account debt capacity. This is referred to as the modified pecking order theory (Lemmon and Zender, 2010). Myers (1977) defines debt capacity as the point at which additional debt issues would reduce the total market value of a firm’s debt. It also takes into account the possibility of firms repaying debt from their operational cash flows. According to the modified pecking order theory firms prefer to finance their activities using internal funds if possible. In absence of debt capacity concerns, debt is preferred over equity. When firms face severe concerns over debt capacity, they turn to the use of equity.
3
Most tests of capital structure consist of estimating models that relate firms’ characteristics to their capital, measured by the ratio of debt over assets. Most estimating models rely on financing deficit regressions, initially proposed by Shyam-Sunder and Myers (1999). Their test is based on the pecking order’s prediction concerning the type of external financing chosen to fill the ‘financing deficit’. The financing deficit must be filled by issuing new securities. They argue that, except for firms at or near their debt capacity, the deficit will be filled entirely with new debt issues. However, there are drawbacks to this approach. Firstly, debt ratios represent the proportion that debt takes in all accumulated liabilities since the firm’s birth. In some sense, it is a firm’s complete history of financing choices at a particular point in time. Characteristics of a firm change as it grows, which affects the availability and suitability of different financing options (Berger and Udell, 1998). Secondly, in typical debt ratios no distinction is made between internal equity and external equity. This distinction is essential when considering the effect of asymmetry of information on incremental financing decisions. Chirinko and Singha (2000) and Haan and Hinloopen (2003) argue that studies relying on financing deficit regressions tell us more about the proportion of debt and equity issues in the data, rather than when and why firms are issuing these two securities, and thus has little power to distinguish pecking order behavior from alternative hypotheses.
The purpose of this paper is to study the extent to which the modified pecking order theory of capital structure can account for the incremental financing decisions of United States’ listed companies for the years 2003 to 2011. Most important studies on capital structure rely on samples of United States’ firms (Shyam-Sunder and Myers (1999), Frank and Goyal (2003), Fama and French (2005), Lemmon and Zender (2010) and Leary and Roberts (2010)). Therefore, it is more convenient to directly compare and generalize studies on capital structure.
Following Haan and Hinloopen (2003), a multinomial logit model is estimated to test the modified pecking order theory. A distinction is made between internal financing, external debt financing and external equity financing. Two model specifications are estimated, the first specification directly tests the modified pecking order theory, it contains internal finance variables, debt capacity variables and general control variables. In the second specification static trade-off variables are added to see if the modified pecking order variables hold in an extended model which includes the variables proposed by the static trade-off theory.
4
to determine which financing hierarchy fits the data best.The remainder of the paper proceeds as follows. Section I reviews the literature, discusses the two main theories of capital structure and introduces the hypotheses. Section II constructs the empirical model and describes the testing strategy. Section III discusses the data, sample selection and summary statistics. Sections IV presents and discusses the results. Finally, section V concludes.
I. Literature
The modified pecking order theory serves as the basic theoretical framework for this study. First, the two main theories of capital structure will be discussed; the pecking order theory and the static trade-off theory. This in order to position the modified pecking order theory in existing capital structure theory. Finally, the hypotheses belonging to the modified pecking order theory will be introduced.
A. Pecking Order Theory
The pecking order theory of capital structure is among the most influential theories of corporate finance (Frank and Goyal, 2003). According to this theory a hierarchy exists for financing new projects. Managers choose a type of capital according to the following preference order: (i) internal finance, (ii) debt, (iii) share issues. This theory is based on the principle of information asymmetry between managers and potential new shareholders (Myers, 1984 and Myers and Majluf, 1984).
5
downward impact on stock price than announcement of an equity issue. Issuing debt minimizes the information advantage of managers (Myers, 2001).
Hence, equity is subject to serious information asymmetry problems, while there are only small information asymmetry problems for debt. From the point of an outside investor, equity is riskier than debt. For both ways of external financing a risk premium exist, but this premium is larger on equity. Therefore, an outside investor will demand a higher rate of return on equity than on debt. Hence, from the firm’s perspective retained earnings are a better source of funds than debt is and debt is a better source of funding than equity. The theory predicts to perform best among firms that have large information asymmetries.
B. Static Trade-Off Theory
The static trade-off theory is considered to be the main competitor of the pecking order theory. It predicts that an optimal target financial debt ratio exists, which maximizes the value of the firm. The optimal point can be attained when the marginal value of the benefits associated with debt issues exactly offsets the increase in the present value of the costs associated with issuing more debt (Myers, 2001).
The resulting optimal capital structure is determined by trading off the benefits of debt, especially tax and agency benefits, against the cost of debt, especially bankruptcy and agency costs of debt (Modigliani and Miller (1963), Titman (1984) and Myers (1977)).
The benefits of debt are the tax deductibility of interest payments and agency benefits. Tax deductibility of corporate interest payments favors the use of debt. Managers have the incentive to waste free cash flow on perquisites and bad investment. Debt financing limits the free cash flow available to managers and thereby helps to control this agency problem (Jensen and Meckling, 1976).
6
C. Literature
An interesting discussion has been generated in recent studies designed to detect whether the trade-off theory or the pecking order theory best describes the financing choices of firms. Recent studies have shown contrary and mixed results.
Shyam-Sunder and Myers (1999) conclude that the pecking order theory is a good descriptor of broad financing patterns. A sample of 157 US firms was used with data over the period 1971 to 1989. Their test is based on the pecking order’s prediction concerning the type of external financing chosen to fill the ‘financing deficit’. The financing deficit must be filled by issuing new securities. They specify the change in debt as a linear function of the financing deficit:
(1) where denotes long-term debt, denotes the financing deficit, denotes the error term, i denotes the individual firm, t denotes year, and are the parameters of the model.
Their testing strategy focuses on the null hypothesis that
, so that the debt changes dollar
for dollar in the financing deficit. They argue that, except for firms at or near their debt capacity, the deficit will be filled entirely with new debt issues. Although some equity was issued during this period, they find that debt issues considerably dominate equity issues.Frank and Goyal (2003) conclude the opposite. They question the conclusions drawn by Shyam-Sunder and Myers on several fronts. The most interesting challenge is whether their findings hold for a broader sample and over a longer time horizon. Frank and Goyal perform a similar test like Shyam-Sunder and Myers on a broad cross-section of publicly traded firms for 1971 to 1988. They also test a model in which conventional leverage factors are included. If the pecking order theory would be correct, the financing deficit should ‘eliminate’ the effect of the conventional variables. However, they conclude that net equity issues track the financing deficit more closely than do net debt issues.
7
external finance and among the external financing types, they prefer bank loans over shares and shares over bonds.
Fama and French (2005) test the trade-off and pecking order predictions about dividends and debt. They use data covering the period 1965 to 1999. They argue that, although motivated by different forces, the trade-off and pecking order models share many predictions about dividends and debt. These shared predictions tend to do well. However, on issues where the two models differ, each suffers one major failure. They find that more than half of small, unprofitable high-growth firms, issue outside equity. This is in line with the pecking order if the net issues of equity are done in extremis. But results show the opposite; the least-levered non-dividend payers make the largest net new issues of equity.
Leary and Roberts (2010) question the ability of the pecking order theory to explain financing decisions. Using a different empirical approach, they find that the classificatory ability of the pecking order varies significantly depending on whether one interprets the pecking order in a strict or liberal (modified) manner. However, the pecking order theory therefore is never able to accurately classify more than half of the observed financing decisions. When the model is expanded to incorporate factors typically attributed to alternative theories, the predictive accuracy of the model increases.
D. Modified Pecking Order Theory
Determining what lies behind these contrasting findings is important for our understanding of capital structure and financing choices of firms. Recent studies show significant external equity issues by high-growth firms (Frank and Goyal (2003), Fama and French (2005)). These studies reject the hypotheses of the pecking order theory, as equity should only be issued as a last resort. Following the pecking order theory, firms prefer internal finance over external finance. This leads to the following hypothesis:
HYPOTHESIS 1: Firms that have more internal funds are less likely to raise additional debt or equity financing.
8
(1984), Shyam-Sunder and Myers (1999), and Chirinko and Singha (2000) define it as ‘sufficiently high’ debt ratios so that costs of financial distress curtail further debt issues. Fama and French (2005) conclude that the least-levered non-dividend payers make the largest net issues of equity. They interpret this finding as invalidating the modified pecking order theory. They argue that firms with low levels of leverage should be able to raise cheaper additional debt. Obviously, highly levered firms have more difficulties to get additional debt financing. Highly levered firms have a higher financial risk, implying less protection for debt holders, because of the smaller equity buffer on which debt holders can rely in case of liquidation.
According to Lemmon and Zender (2010), the combination of debt capacity defined in these terms and the pecking order theory suggests that costs of adverse selection are dominant for “low to moderate” leverage levels but that trade-off-theory-like forces become primary motivators of financing decisions at “high” levels of leverage.
However, even low levered firms may have no or very limited debt capacity. This reflects the traditional view of debt capacity as proposed by Myers (1977). This view also takes into account the possibility of firms repaying debt from their operational cash flows. Accumulating fixed commitments, which increase liquidity risk, will at a certain point reduce the total value of debt. According to the modified pecking order theory, firms with a high debt capacity are more likely to resort to internal financing, as this would point to a substitution between internal and external financing. This leads to the following hypothesis:
HYPOTHESIS 2: Firms with a high debt capacity are more likely to resort to internal financing.
On the opposite, under the static trade-off theory firms with a high debt capacity are supposed to turn to external debt financing, as this would allow firms to benefit from tax advantages of additional interest payments. Obviously, firms with a high debt capacity, low levels of debt and high levels of cash flow, should be able to raise cheaper additional debt. This leads to the following hypothesis:
HYPOTHESIS 3: Firms with a high debt capacity are more likely to resort to external debt financing.
9
reduce the total value of debt. As a result, firms with low levels of cash flow may find that issuing additional debt is not advantageous compared to outside equity, or is even impossible to raise (Helwege and Liang, 1996). Therefore, firms are forced to issue outside equity as a last resort. This leads to the following hypothesis:
HYPOTHESIS 4: Firms with limited debt capacity are more likely to raise additional external equity financing.
Lemmon and Zender (2010) conclude that the modified pecking order is a good descriptor of financing behavior. They examine the impact of explicitly incorporating a measure of debt capacity in recent tests of competing theories of capital structure. Their main indicator of debt capacity is whether the firm has, based on its underlying characteristics, a high likelihood of being able to access the public debt markets. A modified version of the Shyam-Sunder and Myers (1999) is tested on a broad cross section of United States’ firms for the period 1971-2001. Their main results are that if external funds are required, in absence of debt capacity concerns, debt appears to be preferred to equity. Concerns over debt capacity largely explain the use of new external equity financing by publicly traded firms.
However, Chirinko and Singha (2000) and Haan and Hinloopen (2003) argue that studies relying on financing deficit regressions tell us more about the proportion of debt and equity issues in the data, rather than when and why firms are issuing these two securities, and thus has little power to distinguish pecking order behavior from alternative hypotheses.
Debt ratios represent the proportion that debt takes in all accumulated liabilities since the firm’s birth. In some sense, it is a firm’s complete history of financing choices at a particular point in time. Characteristics of a firm change as it grows, which affects the availability and suitability of different financing options (Berger and Udell, 1998). Second, in typical debt ratios no distinction is made between internal equity and external equity. This distinction is essential when considering the effect of asymmetry of information on incremental financing decisions (Haan and Hinloopen, 2003).
II. Methodology
A. Multinomial Logit
10
internal financing, (2) external debt financing and (3) external equity financing. The multinomial logit model assumes that the probability of observing each category in Y is given by:
∑ (2)
Where, j=1,2,3, are the different financing alternatives, X is a vector of independent variables and βj is the vector of coefficients pertaining to finance alternative j. The parameters β are not all identified unless a normalization is imposed. The coefficients for the reference choice are set equal to zero.
Two model specifications are estimated, the first specification contains internal finance variables, debt capacity variables and general control variables. In the second specification static-trade-off variables are added. Following Frank and Goyal (2003), it is interesting to include these variables in the model and see if the modified pecking order variables remain relevant in an extended model which includes the variables proposed by the static trade-off theory. If the modified pecking order theory is correct, the variables of the first specification should eliminate the effect of the static trade-off variables. Otherwise, it is just one factor among others that firms trade off (Frank and Goyal, 2003).
B. Ordered Logit
Although the multinomial logit provides valuable information as to the determinants of firms’ incremental financing decisions, it does not capture all information potentially present in the data. In particular, it does not test for the presence of a most preferred hierarchy of financing types as suggested by the modified pecking order theory. Hence, an ordered logit model is estimated using the same set of variables as in the second specification of the multinomial logit model:
(3)
11
external equity financing. Accordingly, internal finance, external debt financing and external equity financing will be coded [1,2,3] respectively:
{
(4)
where are the threshold parameters to be estimated. Following Haan and Hinloopen (2003), ordered logit models are estimated for all possible financing hierarchies. Every ordering has a twin ordering that has a perfect inverse correlation with the original ordering. Therefore, only 3 financing hierarchies have to be considered. The ordered logit models will then be compared by means of a likelihood ratio test (LR), thus revealing the hierarchy that best fits the data.
C. Dependent Variables
According to the modified pecking order theory there are several ways in which activities can be financed. The dependent incremental financing event variables are constructed following Marsh (1982), Hovakimian, Opler and Titman (2001) and Haan and Hinloopen (2003).
First, firms prefer to finance activities using retained earnings. When the net increase of retained earnings within a year exceeds 5% of total assets, it is coded as an internal finance event. Second, if retained earnings are inadequate, firms turn to the use of debt. Firms are coded as using external debt financing when the yearly net increase of outstanding financial debt exceeds 5% of total assets. Finally, firms use equity financing as a last resort. Firms are coded as using external equity financing when the yearly net increase of external equity exceeds 5% of total assets. The threshold value of 5% is used to assure that the focus of the analyses is on relatively substantial financing events and guarantees consistency with previous studies. Financing events are measured on an annual basis.
D. Independent Variables
D.1. Internal Financing
12
finance. Following the modified pecking order theory, profitability and liquidity should be positively associated with internal finance and negatively associated with external debt financing and external equity financing. The opposite is true for the pay-out ratio.
Following the static trade-off theory, firms with a lot of internal funds are expected to rebalance their capital structure and issue additional outside debt financing. Firms with a lot of internal funds are less likely to fail, reducing the bankruptcy costs associated with debt financing. It buffers companies from shocks in the environment and allows them to survive during turbulent times (Sharfmann, Wolf, Chase and Tansik, 1988). Additional outside debt financing may mitigate potential agency conflicts resulting from abundant internal funds (Jensen, 1986). Hence, the static trade-off theory indicates that internal financing and outside debt financing are complements rather than substitutes (Vanacker and Manigart, 2010). Under the static trade-off theory, the excepted signs of profitability and liquidity are negatively related to internal finance, positively related to external debt financing and also positively related to external equity financing.
D.2. Debt Capacity
We adopt the traditional view of debt capacity as proposed by Myers (1997). This view not only takes a firm’s debt level into account but also includes the possibility of firms repaying debt. Hence, debt capacity is captured by a firms leverage and interest coverage ratio. Leverage is measured as long term debt on total assets. Interest coverage ratio is measured as EBIT on interest expense. Firms with a high debt capacity, a low debt ratio and a high interest coverage ratio, are more likely to resort to internal financing, as this would point to a substitution between internal and external financing. In contrast, under the static trade-off theory, firms with a high debt capacity are expected to turn to external debt financing, as this would allow firms to benefit from tax advantages of additional interest payments.
Obviously, firms with a high debt capacity, low levels debt and high levels of cash flow, should be able to raise cheaper additional debt. However, sufficiently high debt ratios make it more difficult to attract additional debt financing. Accumulating fixed commitments, which increase liquidity risk, will at a certain point reduce the total value of debt. As a result, firms with low levels of may find that issuing additional debt is not advantageous compared to outside equity, or is even impossible to raise (Helwege and Liang 1996). Therefore, firms are forced to issue outside equity as a last resort.
13
Table I Proxy Variables and Expected Influence Under the Modified Pecking Order Theory
Proxy variables and expected signs under the modified pecking order theory. +/- indicates a positive/negative expected association of the proxy variable with the probability of choosing a particular type of financing.
Variable Operationalization Expected sign
Internal financing Debt financing Equity financing Internal financing
Profitability ratio Earnings/total assets + - -
Liquidity ratio Cash and equivalents/total assets + - -
Payout ratio Dividends/total assets - + +
Debt Capacity
Debt ratio Financial debt/total assets - - +
Cash flow ratio EBIT/interest expense + + -
E. General Control Variables
Under the modified pecking order theory, large firms are expected to acquire external finance at lower cost for two reasons. Firstly, large firms are better known by market participants, which reduces information asymmetry between insiders and outsiders. Secondly, flotation costs of public issues are relatively less burdensome for larger firms. Firm size is measured as the natural logarithm of total assets. Accordingly, firm size is expected to be positively related with both external debt and external equity financing, while negatively related to internal financing.
Alternatively, under the static trade-off theory, firm size may be used as a proxy for business risk, as larger firms have a better diversification of possibilities. If this is true, larger firms are considered to be less risky and therefore can have higher optimal debt ratios. Firm size is expected to be positively related to external debt financing, while being negatively related to internal financing and external equity financing.
14
F. Static Trade-off Control variables
A number of control variables related to the static trade-off theory are included. Under the static trade-off theory tax shields, financial distress and agency costs are expected to determine financing decisions. Two types of tax shields are included, debt tax shields, operationalized as interest on total assets, and non-debt tax shields, operationalized as depreciation on total assets. Financial distress is operationalized as the ratio of property, plant and equipment to total assets, where a lower ratio indicates a higher cost of financial distress. Agency costs are particularly prevalent in a setting characterized by considerable future growth options (Myers, 1977). Hence, agency costs are operationalized as intangible assets on total assets.
All independent and control variables are lagged one year to avoid problems of endogeneity due to reverse causality. Table AI is attached in appendix A and reports the correlations for the independent and control variables. All correlations are sufficiently low and do not indicate potential multicollinearity problems.
III.
Data and Summary Statistics
The empirical evidence of this paper is based on a database containing all relevant information of United States’ listed companies covering the years 2003 to 2011. The data are obtained from the financial statements and extracted from the Orbis database. Following standard practices, financial firms are excluded to avoid capital structures governed by regulation. Also excluded are firms with missing book value of total assets, retained earnings, long term debt and share capital for two consecutive years. This yields a sample of 6,292 companies covering all major industries of the United States. The first column of table II shows the sector distribution of the companies included in the sample.
A particular finance event is included in the sample if data on all variables from the previous year are available. This leads to a significant reduction in observed financing events, however this is necessary for ease of comparison.2
The financial statement variables as a percentage of assets are trimmed to remove the most extreme 1% in either tail of the distribution. This removes outliers and the most extremely misrecorded data. Initially the sample consists of 7,578 observations on financing events. However, hybrid forms of financing, involving two or more financing types by the same firm in
2 If only data on internal finance variables and debt capacity variables are included 9,205 financing
15
Table II Sample Classified by Industry
This table classifies the sample by industry. Column (1) presents the distribution of sample firms by industry. Financial firms, firms with missing book values of total assets, retained earnings, long term debt and share capital for two consecutive years are excluded. Column (2) presents the number of financing events by industry. A particular finance event is included in the sample if data on all variables from the previous year are available. Hybrid forms of financing, involving two or more financing types by the same firm in the same year, are removed from the sample. Column (3) presents the fraction of a particular type of financing by industry.
Distribution of sample firms by industry (1)
Number of financing events by industry (2)
Fraction of particular type of financing by industry (3)
Industry Frequency Fraction Frequency Fraction
Internal financing Debt financing Equity financing Other services 1,821 28.94 1,553 23.37 18.41 29.16 31.39
Machinery & equipment 1,321 20.99 1,805 27.16 30.73 21.54 27.27
Chemicals & non-metal 591 9.39 652 9.81 10.19 8.84 11.32
Wholesale & retail trade 444 7.06 565 8.50 10.43 6.18 5.66
Primary sector 441 7.01 254 3.82 3.65 4.40 2.57
Metals & metal products 377 5.99 206 3.10 3.57 2.58 2.23
Publishing, printing 303 4.82 298 4.48 4.46 4.10 6.17
Food, beverages, tobacco 156 2.48 228 3.43 4.05 2.75 2.23
Education, Health 151 2.40 230 3.46 3.62 3.43 2.57
Gas, Water, Electricity 137 2.18 106 1.59 0.46 3.13 2.57
Post & telecommunications 111 1.76 137 2.06 0.70 3.81 3.60
Hotels & restaurants 104 1.65 174 2.62 2.78 2.96 0.17
Transport 94 1.49 172 2.59 2.49 3.17 0.86
Construction 84 1.34 98 1.47 1.51 1.65 0.51
Textiles, wearing apparel, leather 66 1.05 98 1.47 1.97 0.97 0.34
Wood, cork, paper 55 0.87 68 1.02 0.97 1.23 0.51
Public administration & defense 36 0.57 2 0.03 0.00 0.08 0.00
16
the same year, are removed from the sample. Coding a combination of financing types as one particular form of financing is arbitrary. Moreover, interpreting the sign and size of the explanatory variables for these hybrid financing categories is difficult from an economic point of view. This has the advantage of estimating a model that relates to mutually exclusive financing choices only. This results in a total of 6,646 financing events. Column 2 in table II shows the number of financing events by industry. Table III shows the number of observations by financing type and year.3 Figure 1 shows the fraction of a particular financing type by year. It shows that internal financing is the most common financing route, accounting for almost 55% of financing needs. 36% of the financing events are related to external debt financing. Finally, only 9% of financing events are related to external equity issues. At first glance, this seems to be in compliance with the modified pecking order theory.
Table III also shows the average size of financing events, both mean and median are provided, scaled by total assets. The mean and median external equity issue per firm are 52% and 24% respectively, compared to only 13% and 10% for internal finance and 20% and 14% for external debt financing. Although equity is issued less often, the average equity issue size is larger than of external debt issues and internal financing events. This may indicate that the current emphasis on external equity finance in the literature may be somewhat overstated, at least for the period 2004 through 2011.
The fractions of the several financing types seem to be relatively stable over time. For the years 2007 and 2008 a small relative decrease in internal financing and a small increase in external financing is observed. This probably has to do with the financial crisis. However, not enough years are observed to draw inferences on a stable pattern.
It seems that the distribution of financing events per industry is pretty similar to the sector distribution of firms included in the sample. However, some differences are observed between sectors. It seems that relatively little financing events take place within the other services, primary and metals sector, while the opposite is true for the machinery and equipment sector. For example, the other services sector comprises 29% of sample firms, while only 24% of financing events are observed in this sector. 27% of financing events are observed within the machinery sector, while this sector only comprises 21% of sample firms.
3
17
Table III Sample Split According to Financing Type
Year Internal financing Debt financing Equity financing Total
2004 394 221 63 678 2005 422 263 72 757 2006 470 310 68 848 2007 481 361 67 909 2008 394 351 84 829 2009 434 237 72 743 2010 552 289 87 928 2011 554 343 57 954 Total 3,701 2,375 570 6,646
Issue size/total assets
Mean 0.13 0.20 0.52 Median 0.10 0.14 0.24 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 2004 2005 2006 2007 2008 2009 2010 2011 R atio Year
Internal financing Debt financing Equity financing
18
Column 3 in table II presents the fraction of a particular type of financing by industry. For example, 18% of internal financing events are observed within the other services sector, while roughly 30% of external debt and equity financing events are observed within this sector. On the other hand, 31% of internal financing events, 27% of external equity financing events and only 22% of external debt financing events are observed within the machinery, equipment, furniture and recycling sector.
19
Table IV Variables by Type of Financing
This table presents the mean and median (parentheses) values of the explanatory variables split up according to type of financing. Where * and ** denote significance at 5% and 1% level, respectively (two group mean comparison t-tests).
Variables Internal finance (IF) Debt financing (DF) Equity Financing (EF) Significance of t-test between:
IF-DF IF-EF DF-EF
Internal finance Profitability ratio 0.11 -0.36 -2.45 ** ** ** (0.07) (0.02) (-0.44) Liquidity ratio 0.17 0.13 0.22 ** ** ** (0.12) (0.06) (0.13) Payout ratio -0.01 -0.01 0.00 ** ** ** (0.00) (0.00) (0.00) Debt capacity Debt ratio 0.13 0.43 0.13 ** ** (0.11) (0.21) (0.04) Interest coverage ratio 0.09 0.00 -0.03 ** ** ** (0.02) (0.00) (0.00) Static trade-off
Debt tax shields 0.02 0.07 0.31 ** ** **
(0.01) (0.02) (0.03)
Non debt tax
20
IV. Results
A. Multinomial Logit Results
First, a multinomial logit model is estimated to test the modified pecking order theory. Two model specifications are estimated, the first specification contains internal finance variables, debt capacity variables and general control variables.4 In the second specification static trade-off variables are added.
It is reasonable to assume that observations on incremental financing decisions are not independent within firms. The homoscedasticity assumption should be relaxed, as covariance structures may vary by firm cluster. This may affect the estimated standard errors and covariance matrix of the estimates. In order to allow for more flexibility in the variance-covariance matrix, cluster robust standard errors are used. The standard errors are clustered on firm level and robust to heteroscedasticity. Year and industry dummies are included in both model specifications to control for time and industry effects.
Table V presents the estimation results of the multinomial logit model. For ease of interpretation, the marginal effects are given as these are directly interpretable in terms of the implied effect on the respective issuance probabilities. The results on internal finance variables and debt capacity variables in both specifications are broadly similar.
Profitability is positively associated to the probability of using internal finance, while negatively associated to the probability of external debt financing or external equity financing. This suggests the existence of a possible substitution between internal and external financing. Liquidity is positively related to internal financing, negatively associated to external debt financing and weakly positively associated to external equity financing. Although the latter is not statistically significant. So there is some evidence for a substitution between internal and external financing. The payout ratio never shows a significant coefficient. These findings offer reasonable support for hypothesis 1, implying that firms that have more internal funds are less likely to raise additional debt or equity financing.
The debt ratio is negatively associated to internal finance, while the interest coverage ratio is positively associated to internal finance. This offers support for hypothesis 2, implying that firms with a high debt capacity are more likely to resort to internal financing. This indicates a substitution between internal and external financing.
4If only data on internal finance variables and debt capacity variables are included 9,205 mutually
21
Table V Multinomial Regression Results
This table presents the results of the multinomial logit model explaining financing choices. The figures reported are the marginal effects, being the partial derivatives of the probabilities with respect to the explanatory variables evaluated at their respective means. The first model specification contains internal finance, debt capacity and general control variables, in the second specification static trade-off variables are added. The sample consists of 6,646 financing events. Standard errors are robust to heteroscedasticity and clustered at the firm level. Both specifications include year and industry dummies. Standard errors * and ** denote significance at 5% and 1% level, respectively.
Variable (1) Internal finance Debt financing Equity financing (2) Internal finance Debt financing Equity financing
Internal finance Profitabilityt-1 0.42** -0.37** -0.05** 0.40** -0.35** -0.04** Liquidity t-1 0.32** -0.35** 0.03 0.11* -0.17** 0.05** Payout ratio t-1 0.45 -0.62 0.17 0.11 -0.19 0.07 Debt capacity Debt ratio t-1 -0.13** 0.13** 0.00 -0.09** 0.09** 0.00
Interest coverage ratio t-1 0.38** -0.31** -0.07** 0.37** -0.30** -0.07**
Static trade-off
Debt tax shields t-1 0.00 0.00 0.01
Non debt tax shields t-1 -1.07** 0.99** 0.09
Financial distress t-1 -0.18** 0.12** 0.05**
Agency costs t-1 -0.49** 0.46** 0.03
General control
Size t-1 0.00 0.01** -0.01** 0.01* 0.00 -0.01**
Previous debt financing -0.04** 0.06** -0.01 -0.02* 0.04** -0.01
Previous equity financing -0.13** 0.00 0.13** -0.11** -0.01 0.13**
N Obs. 6,646 6,646
Prob. 0.00 0.00
22
Noteworthy is the positive association between the debt ratio and external debt financing and the negative correlation between the interest coverage ratio and external debt financing. There is no evidence supporting hypothesis 3. The results imply that firms with a low debt capacity will be more likely to resort to external debt financing. In theory, firms with a high interest coverage ratio and a low debt ratio are better able to pay off fixed debt related payments and therefore are able to issue more external debt.
According to the modified pecking order theory, firms with limited debt capacity will be more likely to raise external equity financing. Implying that firms with a high debt ratio and a low interest coverage ratio are more likely to issue equity. The interest coverage ratio is negatively correlated to external equity financing, however there is no significant positive correlation between the debt ratio and external equity financing. These findings only offer modest support for hypothesis 4.
Looking at the general control variables, we find no association between firm size and the modified pecking order theory. Firms that previously issued debt or equity financing are more likely to issue outside financing again in the future, as both previous financing dummies are positive and significant. This could indicate a ‘learning effect’ does exist, or it may indicate that external financing needs are correlated from one year to the next.
In the second specification static trade-off variables are added. Following Frank and Goyal (2003), it is interesting to include these variables in the model and see if the modified pecking order variables remain relevant in a model which includes the variables proposed by the static trade-off theory. Non-debt tax shields and agency costs have a significant impact on internal financing and external debt financing, however no significant impact on external equity financing is distinguished. Financial distress has a significant impact on all types of financing. Firms with a high risk of financial distress are more likely to raise external debt or equity financing. This implies that the static trade-off theory should not be ruled out.
Additional analyses on smaller sub-samples have been conducted to test for robustness of the main results. As shown in table III and figure 1 of section II a small decrease in internal financing and a small increase in external financing is observed. As financing patterns seem to be slightly altered, it is interesting to examine sub-samples before and during the financial crisis.5 Results appear to be broadly similar to the main sample.
5 Two subsamples have been examined; the first subsample covers the period 2004 to 2006 and
23
Column 2 and 3 in table II in section IV show that there are notable differences in financing events between industries. Equation (1) is also tested on a subsample of high equity industries and on a subsample of low equity industries.6 Results for high equity industries are similar to the main results. Results for low equity industries appear to be slightly different, as the coefficients on debt capacity become insignificant. This implies that less evidence is found for the modified pecking order theory.
B. Ordered Logit Results
The multinomial logit results offer modest support for the modified pecking order theory. However, aspects of the static trade-off theory also seem to play a role in capital structure. Although the multinomial logit provides valuable information as to the determinants of firms’ incremental financing decisions, it does not capture all information potentially present in the data. In particular, it does not test for the presence of a most preferred hierarchy of financing types as suggested by the modified pecking order theory. Hence, ordered logit models are estimated to determine which financing hierarchy fits the data best. At first, 3!=6 different ordered logit estimates and 1/2x6x5=15 bilateral likelihood comparisons are observed. However, every ordering has a twin ordering that yields coefficient estimates of equal magnitude with the opposite sign. This twin ordering has a perfect inverse correlation with the original ordering. Hence, likelihood values are identical. Therefore, only 3 ordered logit estimates and only 1/2x3x2=3 bilateral likelihood comparisons have to be considered. Table VII reports the outcomes of the pair wise LR-tests. The hierarchies are sorted by their likelihood values from lowest to highest. Significance values at the 5% and 1% are 3.84 and 6.64 respectively. The resulting rankings are reported in table VI.
According to table VI, firms have a unique most preferred hierarchy that is identical to the modified pecking order hierarchy. When deciding on the incremental financing type, firms prefer internal finance over external debt financing and external debt financing over external equity financing.
6
24
Table VI Ranking of Financing Hierarchies
This table presents the three possible financing hierarchies and their ranking according to their likelihood. The likelihood and pseudo-R2 are derived from the corresponding ordered logit models estimates.
Hierarchy Internal Finance External Debt Financing External Equity Financing
Likelihood Rank Pseudo-R2
h1 1 2 3 -5,025.41 1 0.164
h2 1 3 2 -5,445.64 3 0.095
h3 2 1 3 -5,441.96 2 0.094
Table VII Likelihood Ratio Test Results
This table presents the outcomes of the pairwise LR-tests. Standard errors * and ** denote significance at 5% and 1% level, respectively.
h1 h3 h2
h1 0.00
h3 833.11 0.00
h2 840.74 7.36 0.00
The pseudo-R2 of the ordered logit model corresponding to h1 is considerably higher as compared to the pseudo-R2 of the ordered logit estimates corresponding to h2 and h3. This implies that h1 fits the data best. However, according to table VII, firms appear to have no ordered preference for financing types. There appears to be no difference between financing hierarchies in a statistical sense.
Again, analyses on the 2004 to 2006/2007 to 2011 subsample7 and the low/high equity subsample8 have been conducted. In addition, ranking hierarchies are analyzed when only modified pecking order variables are used.9 All subsamples yield similar results. In compliance with the modified pecking order theory, h1 always is the most preferred ranking. However, there never is a statistical difference between financing hierarchies. These results offer modest support for the modified pecking order theory.
V. Conclusion
This paper studies the extent to which the modified pecking order theory of capital structure can account for the incremental financing decisions within firms. The empirical evidence of this paper is based on information of United States’ listed companies for the years 2003 to 2011.
25
Recent studies on capital structure have shown contrary and mixed results. Previous studies on capital structure relying on financing deficit regressions have often been criticized as these tend to tell us more about the proportion of debt and equity issues, rather than when and why firms are issuing these two securities. In typical debt ratios, no distinction is made between internal en external equity. This study distinguishes three types of financing: internal financing, external debt financing and external equity financing.
First, a multinomial logit model is estimated to test the modified pecking order theory. Two model specifications are estimated, the first specification contains internal finance variables, debt capacity variables and general control variables. In the second specification static trade-off variables are added.
The results offer modest support for the modified pecking order theory. However, aspects of the static trade-off theory also seem to play a role in capital structure. In line with the modified pecking order theory, results indicate that a substitution exists between internal and external financing. Firstly, more profitable firms prefer to finance investments internally. Second, firms using internal finance are likely to retain their debt capacity. Results for debt capacity appear to be less solid. The modified pecking order theory predicts that firms with limited debt capacity will be more likely to raise additional external equity financing, implying firms with a high debt ratio and a low interest coverage ratio are more likely to issue equity. Our results indicate that firms with a low interest coverage ratio are more likely to issue equity. However, no significant relationship was found between debt ratio and external equity financing. Results for internal finance and debt capacity remain somewhat unchanged when static trade-off variables are added into the equation. However, non-debt tax shields, financial distress and agency costs, three static trade-off variables, do play a significant role in incremental financing decisions. This implies that the static trade-off theory should not be ruled out.
Although the multinomial logit provides valuable information as to the determinants of firms’ incremental financing decisions, it does not capture all information potentially present in the data. Hence, all possible ordered logit models are estimated to determine which financing hierarchy fits the data best. Firms appear to have a unique most preferred hierarchy that is identical to the modified pecking order hierarchy. When deciding on the incremental financing type, firms prefer internal finance over external debt financing and external debt financing over external equity financing. However, there seems to be no difference between financing hierarchies in a statistical sense. Our results offer modest support for the modified pecking order theory.
26
when factors typically attributed to alternative theories are incorporated. Fama and French (2005) conclude that the least levered non dividend payers make the largest net issues of equity. They interpret this finding as invalidating the modified pecking order theory, as firms with low levels of leverage should be able to raise cheaper additional debt.
Lemmon and Zender (2010) conclude that the modified pecking order theory is a good descriptor of financing behavior. However, their empirical approach was based on the financing deficit regression, their sample covered the period 1971 to 2001.
It seems that none of both theories gives a general explanation of a financing strategy. Hence, these theories could be considered as extremes on a theoretical spectrum of capital structure. Where in reality, truth lies somewhere in the middle, as aspects of both theories are of empirical importance. On the other hand, maybe these theories should not be interpreted as general theories, but as conditional theories of capital structure (Myers, 2001). This may explain why testing these theories on broad heterogeneous samples of firms often leads to contrary results. Hence, in future research it will be useful to test the hypotheses using certain well defined subsamples.
One common problem in capital structure studies is to find the right proxy variables. Tests usually rely on indirect measures or proxies for the unobservable variables that are assumed to drive financing choices. A particular proxy may respond to more than one theory. For example, one can argue that debt ratio is closely related to financial distress, as highly levered firms may have more difficulties in meeting interest and principal payments. If so, a significant coefficient on this particular proxy lacks a clear interpretation as it may be related to both theories. This way a researcher may generate results consistent with one theory even when financing decisions are actually generated by another.
27
REFERENCES
Berger, A. N., Udell, G. F., 1998, The economics of small business finance: The roles of private equity and debt markets in the financial growth cycle, Journal of Banking and Finance, 22, 613-673.
Chirinko, R. S., Singha, A.R., 2000, Testing static tradeoff against pecking order models of Capital Structure: A critical Comment, Journal of Financial Economics, 58, 412-425.
Fama, E. F., French, K. R., 2005, Financing decisions: Who issues stock? Journal of Financial Economics, 76, 549-582.
Frank, M.Z., Goyal, V.K., 2003, Testing the pecking order theory of capital structure, Journal of Financial Economics, 67, 217-248.
Haan, L., Hinloopen, J., 2003, Preference hierarchies for internal finance, bank loans, bond, and share issues: Evidence for Dutch firms, Journal of Economical Finance, 10, 661-681.
Helwege, J., Liang, N., 1996. Is there a pecking order? Evidence from a panel of IPO firms, Journal of Financial Economics 40, 429-458.
Harris, M., Ravis, A., 1988, Corporate control contests and capital structure, Journal of Financial Economics, 20, 55-86.
Harris, M., Raviv, A., 1991, The theory of capital structure, Journal of Finance, 46, 297-356. Hovakimian, A., Opler, T., Titman, S., 2001, The debt–equity choice, Journal of Financial and
Quantitative Analysis, 36, 1–24.
Jensen M., Meckling W., 1976, Theory of the firm: Managerial Behavior, agency costs and ownership structure, Journal of Financial Economics 3, 305-360.
Jensen, M. C., 1986, Agency costs of free cash flow, corporate finance and takeovers, American Economic Review, 76, 323-339.
Leary, M., Roberts, M., 2010, The pecking order, debt capacity and information asymmetry, Journal of Financial Economics, 95, 332-355.
Lemmon, M.L., Zender, J.F., 2010. Debt capacity and tests of capital structure theories, Journal of Financial and Quantitative Analysis, 45, 1161-1187.
Marsh, P., 1982, The choice between equity and debt: an empirical study, Journal of Finance, 37, 121-144.
Modigliani, F., Miller, M., 1958, The cost of capital, corporation finance, and the theory of investment, American Economic Review, 48, 261-297.
28
Myers, S. C., 1977, Determinants of corporate borrowing, Journal of Financial Economics, 5, 147-175.
Myers, S.C., 1984, The capital structure puzzle, Journal of Finance, 39, 575-592.
Myers, S.C., Majluf, N.,1984, Corporate financing and investment decisions when firms have information that investors do not have, Journal of Financial Economics 13, 187-221.
Myers, S.C., 2001, Capital structure, Journal of Economic Perspectives, 15, 81-102.
Sharfman, M., Wolf, G., Chase, R., Tansik, D., 1988, Antecedents of organizational slack. Academy of Management Review, 13, 601–614.
Shyam-Sunder, L., Myers, S. C., 1999, Testing static tradeoff against pecking order models of capital structure, Journal of Financial Economics, 21, 219–244.
Titman S., 1984, The effect of capital structure on a firm's liquidation decision, Journal of Financial Economics, 13,137-151.
29
APPENDIX A
Table AI Correlation Matrix
This table reports the correlations for the independent and control variables.
Variable (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Internal finance
Profitability (1) 1.00
Liquidity (2) -0.06 1.00
Pay out ratio (3) -0.08 0.13 1.00
Debt capacity
Debt ratio (4) -0.03 -0.16 -0.03 1.00
Interest coverage ratio (5) 0.11 -0.01 -0.09 -0.08 1.00
Static trade-off
Debt tax shields (6) -0.51 0.01 0.07 0.07 -0.03 1.00
Non debt tax shields (7) 0.14 0.10 -0.04 -0.13 0.02 -0.10 1.00
Financial distress (8) -0.01 -0.24 0.05 0.06 -0.03 0.03 0.16 1.00
Agency costs (9) 0.05 -0.35 -0.17 0.23 0.00 -0.05 -0.37 -0.39 1.00
General control
Size (10) 0.36 -0.23 -0.32 0.07 0.15 -0.29 0.13 0.02 0.25 1.00
30
APPENDIX B
Table BI Descriptive Statistics
This table presents the descriptive statistics for all variables for the period 2004 to 2010.
Variables Observations Mean Median St. Dev Min Max Internal finance
Profitability ratio 6,646 -0.29 0.06 2.00 -56.44 0.55
Liquidity ratio 6,646 0.16 0.09 0.17 0.00 1.00
Pay out ratio 6,646 -0.01 0.00 0.01 -0.12 0.00
Debt capacity
Debt ratio 6,646 0.21 0.15 0.28 0.00 3.29
Interest coverage
ratio 6,646 0.05 0.00 0.16 -1.31 1.47
Static trade-off
Debt tax shields 6,646 0.05 0.01 0.29 0.00 8.50
Non debt tax shields 6,646 -0.04 -0.03 0.03 -0.39 0.00
Financial distress 6,646 0.21 0.15 0.21 0.00 0.92
Agency costs 6,646 0.25 0.17 0.23 0.00 0.96
General control
31
APPENDIX C
Table CI Multinomial Regression Results
This table presents the results of the multinomial logit model explaining financing choices. The figures reported are the marginal effects, being the partial derivatives of the probabilities with respect to the explanatory variables evaluated at their respective means. The first specification covers the period 2004 to 2006. The second specification covers the period 2007 to 2011. Standard errors are robust to heteroscedasticity and clustered at the firm level. Both specifications include year and industry dummies. Standard errors * and ** denote significance at 5% and 1% level, respectively.
Variable (1) Internal finance Debt financing Equity financing (2) Internal finance Debt financing Equity financing
Internal finance
Profitability 0.51* -0.44* -0.07* 0.42** -0.37** -0.05**
Liquidity 0.20* -0.24** 0.04 0.20** -0.21** 0.01
Pay out ratio -0.01 0.15 -0.13 0.12 -0.11 -0.01
Debt capacity
Debt ratio -0.17** 0.16** 0.01 -0.09* 0.10* -0.01
Interest coverage ratio 0.38* -0.29* -0.08 0.36** -0.29** -0.07*
Static trade-off
Debt tax shields 0.68 -0.60 -0.07 -0.45 0.38 0.06
Non debt tax shields -1.79** 1.58** 0.22 -0.78** 0.73** 0.04
Financial distress -0.37** 0.38** -0.01 -0.41** 0.42** -0.01
Agency costs
General control
Size 0.02* 0.001 -0.02** 0.02 0.01* -0.01**
Previous debt financing -0.02 0.03 -0.01 -0.04* 0.05** -0.01
Previous equity financing -0.10 -0.05 0.15** -0.14** 0.01 0.12**
N Obs. 2283 4636
Prob. 0.00 0.00
32
APPENDIX D
Table DI Multinomial Regression Results
This table presents the results of the multinomial logit model explaining financing choices. The figures reported are the marginal effects, being the partial derivatives of the probabilities with respect to the explanatory variables evaluated at their respective means. The first specification presents results of the low equity sector. The second specification presents the results of the high equity sector. Standard errors are robust to heteroscedasticity and clustered at the firm level. Both specifications include year and industry dummies. Standard errors * and ** denote significance at 5% and 1% level, respectively.
Variable (1) Internal finance Debt financing Equity financing (2) Internal finance Debt financing Equity financing
Internal finance Profitability 0.94** -0.87** -0.08** -0.14** 0.41** -0.07** Liquidity 0.39** -0.36** -0.03 0.18** 0.22** 0.03 Payout ratio -0.4 0.48 -0.08 0.6 0.78 0.18 Debt capacity Debt ratio 0.06 -0.07 0.00 -0.14** -0.14** 0.00
Interest coverage ratio 0.75 -0.66 -0.09 0.33** 0.28** -0.05
Static trade-off
Debt tax shields -1.04 0.95 0.1 0.16 -0.14 -0.01
Non debt tax shields -1.34* 1.62* -0.29* -1.20** 0.96** 0.24*
Financial distress -0.30** 0.35** -0.05 -0.53** 0.50** 0.03
Agency costs
General control
Size -0.01 0.01 -0.01** 0.01* 0.01 -0.02**
Previous debt financing -0.07* 0.06* 0.01 -0.02 0.04* -0.02
Previous equity finanicng 0.008 -0.06 0.05** -0.18** 0.02 0.16**
N Obs. 1,493 3,793
Prob. 0 0
33
APPENDIX E
Table EI Ranking of Financing Hierarchies Subsample 2004-2006
This table presents the outcomes of the pairwise LR-tests covering the period 2004 to 2006. This subsample consists of 2,208 financing events. Standard errors * and ** denote significance at 5% and 1% level, respectively. Hierarchy Internal Finance External Debt Financing External Equity Financing
Likelihood Rank Pseudo-R2
h1 1 2 3 -1752.82 1 0.152
h2 1 3 2 -1874.95 2 0.093
h3 2 1 3 -1889.09 3 0.087
Table EII Likelihood Ratio Test Results 2004 to 2006
This table presents the three possible financing hierarchies and their ranking according to their likelihood. Rankings are based on a data covering the period 2004 to 2006. This subsample consists of 2,208 financing events. The likelihood and pseudo-R2 are derived from the corresponding ordered logit models estimates.
h1 h2 h3
h1 0.00
h2 244.25 0.00
h3 272.54 28.29 0.00
Table EIII Ranking of Financing Hierarchies Subsample 2007-2011
This table presents the outcomes of the pairwise LR-tests covering the period 2007 to 2011. This subsample consists of 3,793 financing events. Standard errors * and ** denote significance at 5% and 1% level, respectively. Hierarchy Internal Finance External Debt Financing External Equity Financing
Likelihood Rank Pseudo-R2
h1 1 2 3 -3250.48 1 0.175
h2 1 3 2 -4363.00 3 0.099
h3 2 1 3 -3540.46 2 0.102
Table EIV Likelihood Ratio Test Results 2007-2011
This table presents the three possible financing hierarchies and their ranking according to their likelihood. Rankings are based on a data covering the period 2007 to 2011. This subsample consists of 3,793 financing events. The likelihood and pseudo-R2 are derived from the corresponding ordered logit models estimates.
h1 h3 h2
h1 0.00
h3 579.95 0.00
34
APPENDIX F
Table FI Ranking of Financing Hierarchies Low Equity Subsample
This table presents the outcomes of the pairwise LR-tests for the low equity subsample. The low equity subsample comprises 1,493 financing events and consists of the following sectors: wholesale, primary, education, hotels, transport and
construction.
Standard errors * and ** denote significance at 5% and 1% level, respectively. Hierarchy Internal Finance External Debt Financing External Equity FinancingLikelihood Rank Pseudo-R2
h1 1 2 3 -1066.01 1 0.115
h2 1 3 2 -1106.77 2 0.081
h3 2 1 3 -1120.11 3 0.070
Table FII Likelihood Ratio Test Results Low Equity Subsample
This table presents the three possible financing hierarchies and their ranking according to their likelihood based on the low equity subsample. The low equity subsample comprises 1,493 financing events and consists of the following sectors: wholesale, primary, education, hotels, transport and construction. The likelihood and pseudo-R2 are derived from the corresponding ordered logit models estimates.
h1 h2 h3
h1 0.00
h2 81.52 0.00
h3 108.20 26.68 0.00
Table FIII Ranking of Financing Hierarchies High Equity Subsample
This table presents the outcomes of the pairwise LR-tests for the high equity subsample. The high equity subsample comprises 3,793 financing events and consists of the following sectors; other services, machinery, publishing & printing and post & telecommunications Standard errors * and ** denote significance at 5% and 1% level, respectively.
Hierarchy Internal Finance External Debt Financing External Equity Financing
Likelihood Rank Pseudo-R2
h1 1 2 3 -2942.61 1 0.175
h2 1 3 2 -3214.02 3 0.099
35
Table FIV Likelihood Ratio Test Results High Equity Subsample
This table presents the three possible financing hierarchies and their ranking according to their likelihood based on the high equity subsample. The high equity subsample comprises 3,793 financing events and consists of the following sectors; other services, machinery, publishing & printing and post & telecommunications. The likelihood and pseudo-R2 are derived from the corresponding ordered logit models estimates.
h1 h3 h2
h1 0.00
h3 496.60 0.00
