سلسله مراتب های اولویت برای تأمین منابع مالی داخلی، وام های بانکی، اوراق قرضه، و مسائل مربوط به سهم: شواهد برای شرکت های هلندی

We estimate the incremental financing decision for a sample of some 150 Dutch companies for the years 1984 through 1997, thereby distinguishing internal finance and three types of external finance: bank borrowing, bond issues, and share issues. First, we estimate a multinomial logit model, which confirms several predictions of both the static trade-off theory and the pecking order theory as to the determinants of financing choices. Next, we estimate all possible ordered probit models to determine which financing hierarchy fits the data best. The results suggest that Dutch firms have a unique most preferred financing hierarchy: (i) internal finance, (ii) bank loans, (iii) share issues, and (iv) bond issues.

Ever since the seminal contribution of Modigliani and Miller (1958), who show that under special circumstances (including the absence of frictions and taxes, and perfectly working capital markets) there is no optimal capital structure, one of the more intriguing challenges in corporate finance is to provide a satisfactory explanation as to why in practice 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. This view holds that a firm chooses that debt–equity mixture that optimizes its value. The resulting ‘optimal capital structure’ is determined by trading off the costs and benefits of equity and debt, including tax shields, financial distress, and agency costs of debt and equity. The second line of reasoning is that there is a pecking order as to the type of financing preferred by managers. When making their incremental financing decision, Donaldson (1961) observed that firms appear not to target specific capital structures. Rather, they choose a type of capital according to a preference order: (i) internal finance, (ii) debt, and (iii) share issues. Myers and Majluf (1984) explain Donaldson's observation by referring to the inherent asymmetry of information associated with acquiring external finance. Insiders (owners and/or managers) know more about the firm's value than outsiders (investors) do. The former avoids issuing equity when they believe that shares are undervalued. The latter, realizing the former's reluctance to issue undervalued shares, would thus interpret a share issue as conveying unfavourable information as to the value of the firm. As a result, share issues are typically followed by a decrease in valuation of the issuing firm's assets. Insiders are therefore reluctant to raise equity capital and prefer to accumulate retained earnings in order to fund incremental investments. 2 Which of the two hypotheses regarding the financing behaviour of firms is more relevant remains an empirical question. Most tests of the static trade-off theory consist of estimating models that relate firms' characteristics to their capital structure, measured by the ratio of debt over assets (or some transformation of this ratio, such as debt over equity). Depending on which exogenous variables are found to be statistically significant, the potential determinants of firms' target debt ratios are to be accepted or rejected (for an overview of this literature see, e.g. Harris and Raviv, 1991). There are however at least two drawbacks to this approach. First, debt ratios represent the proportion that debt takes in all accumulated liabilities since the firm's birth. In some sense, it is a snapshot of a firm's complete history of financing choices at a particular point in time. Information on the timing of acquiring debt or issuing equity is ignored. Second, internal equity (e.g. retained earnings) and external equity (obtained with share issues) are not distinguished in typical debt ratios. This distinction is essential however when considering the effect of asymmetry of information on incremental financing choices. The second route of investigation, especially as to the validity of the static trade-off theory, involves estimating discrete choice models of firms' incremental financing decisions. The focus here is on establishing the relevant determinants for the choice of financing type. A pioneering example of this approach is the study of Marsh (1982). He finds that UK companies are heavily influenced by market conditions and the past history of security prices when choosing between issuing bonds and shares. He also finds that companies appear to make their choice of financing instrument as if they have target debt levels in mind. The study of Marsh (1982) implicitly reveals the possibility of direct tests of the pecking order theory. Helwege and Liang (1996) for instance, examining the financing choices of US firms that did an initial public offering, estimate first a binary logit model regarding the choice between internal and external finance, followed by a multinomial logit analysis for the choice of external finance type. Their results are inconsistent with pecking order behaviour in that the probability of obtaining external funds is unrelated to the shortfall in internally generated funds. De Jong and Veld (2001) estimate a binary logit model on a sample of Dutch quoted firms regarding the choice between bond and share issues. They find no confirmation of pecking order financing, notably as the influence of ‘slack’ (basically cash and liquid assets) on the probability of bond and share issuance is found to be insignificant. Other empirical studies of the pecking order theory focus on the reaction of stock prices to the announcements of share and/or bond issues (see, e.g. Bayless and Chaplinsky, 1990 and De Jong and Veld, 2001). Empirical findings thus obtained suggest the existence of a lemon premium, particularly in case of new share issues, hence confirming the existence of asymmetric information between insiders and outsiders. Shyam-Sunder and Myers (1999) explicitly compare the static trade-off theory with the pecking order theory using a panel of US firms. Their results suggest greater confidence in the pecking order theory than in the static trade-off theory. Although companies in their sample did have well-defined optimal debt ratios, it appears that their managers were not trying to obtain these. However, as shown by Chirinko and Singha (2000), this test implicitly presumes that equity issues cover only a relatively small portion of firms' external financing needs. Accordingly, it cannot be applied under all circumstances. In this paper, we examine the incremental financing decision for a sample of some 150 Dutch firms for the years 1984 through 1997. To that end, we first estimate a multinomial logit model that is quite standard in the recent literature. Our analysis stands out in the sense that we distinguish four financing types: internal finance, bank loans, bond issues, and share issues. Including internal finance among the financing choices is highly relevant. Indeed, it is well documented that for firms in the Netherlands, internal finance is important, as it is for firms in other countries.3 Our motivation for distinguishing bank debt from bonds is that the Dutch economy is bank-based in the sense that bank credit is the major source of external finance, while the corporate bond market is still relatively underdeveloped (see, e.g. Saunders and Schmeits, 2002). This distinction also does justice to the differences in the asymmetry of information between private and public debt (Leland and Pyle, 1977). Indeed, the pecking order theory predicts that firms prefer less risky, negotiable bank debt over nonnegotiable public debt.4 The relevance of including internal finance and bank loans as separate options into the set of possible financing types is confirmed by our estimates. The estimated bank loan equation and the internal financing equation are relatively strong in a statistical sense. The estimation results suggest that the static trade-off theory and the pecking order theory are both of empirical importance in explaining the financial choices of our sample of companies, which is in line with, e.g. Bontempi (2002). Our second contribution to the literature is to examine which ordering of financing types, so-called financing hierarchies, suits the data best. Note that the pecking order theory predicts one particular hierarchy to come out as most preferred by firms: financing internally is preferred over bank borrowing, bank borrowing is preferred over issuing bonds, and issuing bonds is preferred over issuing shares. We propose a new testing methodology using ordered probit analyses. In particular, we estimate separate ordered probit models for each possible financing hierarchy. We then test which one of these hierarchies suits best the data. According to this methodology, Dutch firms prefer internal finance over bank loans, bank loans over share issues, and share issues over bond issues. This hierarchy is very close to the pecking order hierarchy except for the reversal of the last two financing types. The finding that in the Netherlands, public debt is the least preferred form of finance could well be due to the fact that the Dutch corporate bond market is relatively underdeveloped (see, e.g. Saunders and Schmeits, 2002). Indeed, as documented by Esho et al. (2001), firms that reside in a country with a well-developed bond market have a higher probability of tapping this market for their incremental financing needs. We proceed as follows. In the next section, our data set is described, followed by Section 3 in which we introduce the explanatory variables that are used in the discrete choice models throughout the paper. The estimation results for the multinomial logit model, explaining financing choices, are discussed in Section 4, while Section 5 presents the analysis of the ordered probit estimates and the resulting most preferred finance hierarchy. Section 6 concludes.

We analyse the determinants of the incremental financing choice for some 150 Dutch nonfinancial companies that are quoted on the Amsterdam Exchanges for the years 1984 through 1997. We thereby distinguish a broader range of financing types than is usual in the literature: internal finance, bank borrowing, bond and share issues. The inclusion of both internal finance and bank borrowing is especially relevant for the case of the Netherlands where bank loans are the most important source of external finance and where internal finance is the most important source of finance in general. This is confirmed by our estimation results in the sense that the internal finance and bank loan equations perform quite well from an econometric point of view. The estimation results indicate that the static trade-off theory and the pecking order theory are both of empirical importance in explaining the financial choices of our sample of companies, although not all aspects of both theories are confirmed. We then introduce a new methodology to specifically test whether a preferred hierarchy as to the type of incremental financing exists. In particular, we estimate ordered probit models for all possible financing hierarchies and compare these by means of a likelihood ratio test. This procedure yields a ranking of financing hierarchies in terms of the best fit to the data. The resulting ranking, which is also in concordance with the individual parameter estimates, shows that the most preferred hierarchy is very close to that predicted by the pecking order theory: Dutch firms prefer internal finance over external finance, and among the external financing types, they prefer bank loans over shares and shares over bonds. Indeed, the only difference with the pecking order hierarchy is that the preference order for shares and bonds is reversed. This low preference of Dutch firms for bonds could well be attributable to the relatively low level of development of the Dutch corporate bond market.