經濟學留學essay：優(yōu)化資本結構-思考經濟和其他價值：Optimal Capital Structure: Reflections on economic and other values

1. Introduction 引言

Despite a vast literature on the capital structure of the firm (see Harris and Raviv, 1991, Graham and Harvey, 2001, Brav et al., 2005, for overviews) there still is a big gap between theory and practice (see e.g. Cools, 1993, Tempelaar, 1991, Boot & Cools, 1997). Starting with the seminal work by Modigliani & Miller (1958, 1963), much attention has been paid to the optimality of capital structure from the shareholders’ point of view. 

在過去的幾十年的研究已經生產上的其他利益相關者的利益對資本結構的影響。眾所周知的例子是客戶的利益，誰收到來自該公司的產品或服務的擔保（見如Grinblatt和Titman，2002）。已經收到了相當的關注的另一個領域是管理的激勵機制和資本結構“（同上）之間的關系。此外，企業(yè)control2問題（見詹森及Ruback，1983），有關企業(yè)governance3的問題（見施萊弗和Vishney，1997年），收到一個獅子的資本結構決策更近的學術界的關注.所有這些研究中，有一件事是明確的：資本結構的決定（或更確切地說，隨著時間的推移資本結構的管理）涉及公司的市值僅比最大化的問題。

Over the last few decades studies have been produced on the effect of other stakeholders’ interests on capital structure. Well-known examples are the interests of customers who receive product or service guarantees from the company (see e.g. Grinblatt & Titman, 2002). Another area that has received considerable attention is the relation between managerial incentives and capital structure (Ibid.). Furthermore, the issue of corporate control2 (see Jensen & Ruback, 1983) and, related, the issue of corporate governance3 (see Shleifer & Vishney, 1997), receive a lion’s part of the more recent academic attention for capital structure decisions. 

在本文中，我們給不同的目標，并已在文獻中提出的考慮的概述。我們兩個廣泛的定義的情況之間的區(qū)別。首先是傳統(tǒng)的公司，現有股東的股份價值最大化的努力情況。其他因素比價值最大化每當進入資本結構的決定，這些因素是價值最大化的目標。第二宗個案涉及的堅定，明確地選擇比單靠價值最大化的目標。這可能是因為股東采取多個利益相關者方法，或因支配財政文學的結構比一般的企業(yè)不同的所有制結構。后者的一個例子是合作，一個法人實體，可以發(fā)現在AO許多歐洲國家。為什么企業(yè)都面臨著多重目標的討論，我們是指到Hallerbach和Spronk（2002年a，2002年b）。

From all these studies, one thing is clear: The capital structure decision (or rather, the management of the capital structure over time) involves more issues than the maximization of the firm’s market value alone. In this paper, we give an overview of the different objectives and considerations that have been proposed in the literature. We make a distinction between two broadly defined situations. The first is the traditional case of the firm that strives for the maximization of the value of the shares for the current shareholders. Whenever other considerations than value maximization enter capital structure decisions, these considerations have to be instrumental to the goal of value maximization. The second case concerns the firm that explicitly chooses for more objectives than value maximization alone. This may be because the shareholders adopt a multiple stakeholders approach or because of a different ownership structure than the usual corporate structure dominating finance literature. An example of the latter is the co-operation, a legal entity which can be found in a.o. many European countries. For a discussion on why firms are facing multiple goals, we refer to Hallerbach and Spronk (2002a, 2002b). #p#分頁標題#e#

 

In Section 2 we will describe objectives and considerations that, directly or indirectly, clearly help to create and maintain a capital structure which is 'optimal' for the value maximizing firm. The third section describes other objectives and considerations. Some of these may have a clear negative effect on economic value, others may be neutral and in some cases the effect on economic value is not always completely clear. Section 4 shows how, for both cases, capital structure decisions can be framed as multiple criteria decision problems which can then benefit from multiple criteria decision support tools that are now widely available. 

 

2. Maximizing shareholder value 實現股東價值最大化

According to the neoclassical view on the role of the firm, the firm has one single objective: maximization of shareholder value. Shareholders possess the property rights of the firm and are thus entitled to decide what the firm should aim for. Since shareholders only have one objective in mind - wealth maximization - the goal of the firm is maximization of the firm's contribution to the financial wealth of its shareholders. The firm can accomplish this by investing in projects with positive net present value4. Part of shareholder value is determined by the corporate financing decision5. Two theories about the capital structure of the firm - the trade-off theory and the pecking order theory - assume shareholder wealth maximization as the one and only corporate objective. We will discuss both theories including several market value related extensions. Based on this discussion we formulate a list of criteria that is relevant for the corporate financing decision in this essentially neoclassical view. 

 

The original proposition I of Miller and Modigliani (1958) states that in a perfect capital market the equilibrium market value of a firm is independent of its capital structure, i.e. the debt-equity ratio6. If proposition I does not hold then arbitrage will take place. Investors will buy shares of the undervalued firm and sell shares of the overvalued shares in such a way that identical income streams are obtained. As investors exploit these arbitrage opportunities, the price of the overvalued shares will fall and that of the undervalued shares will rise, until both prices are equal. 

 

When corporate taxes are introduced, proposition I changes dramatically. Miller and Modigliani (1958, 1963) show that in a world with corporate tax the value of firms is a.o. a function of leverage. When interest payments become tax deductible and payments to shareholders are not, the capital structure that maximizes firm value involves a hundred percent debt financing. By increasing leverage, the payments to the government are reduced with a higher cash flow for the providers of capital as a result. The difference between the present value of the taxes paid by an unlevered firm (Gu) and an identical levered firm (Gl) is the present value of tax shields (PVTS). Figure 1 depicts the total value of an unlevered and a levered firm7. The higher leverage, the lower Gl, the higher Gu - Gl (=PVTS). #p#分頁標題#e#

 

In the traditional trade-off models of optimal capital structure it is assumed that firms balance the marginal present value of interest tax shields8 against marginal direct costs of financial distress or direct bankruptcy costs.9 Additional factors can be included in this trade-off framework. Other costs than direct costs of financial distress are agency costs of debt (Jensen & Meckling, 1976). Often cited examples of agency costs of debt are the underinvestment problem (Myers, 1977)10, the asset substitution problem (Jensen & Meckling, 1976 and Galai & Masulis, 1976), the 'play for time' game by managers, the 'unexpected increase of leverage (combined with an equivalent pay out to stockholders to make to increase the impact)', the 'refusal to contribute equity capital' and the 'cash in and run' game (Brealey, Myers & Allan, 2006). These problems are caused by the difference of interest between equity and debt holders and could be seen as part of the indirect costs of financial distress. Another benefit of debt is the reduction of agency costs between managers and external equity (Jensen and Meckling, 1976, Jensen, 1986, 1989). Jensen en Meckling (1976) argue that debt, by allowing larger managerial residual claims because the need for external equity is reduced by the use of debt, increases managerial effort to work. In addition, Jensen (1986) argues that high leverage reduces free cash with less resources to waste on unprofitable investments as a result.11 The agency costs between management and external equity are often left out the trade-off theory since it assumes managers not acting on behalf of the shareholders (only) which is an assumption of the traditional trade-off theory. 

 

In Myers' (1984) and Myers and Majluf's (1984) pecking order model12 there is no optimal capital structure. Instead, because of asymmetric information and signalling problems associated with external financing13, firm's financing policies follow a hierarchy, with a preference for internal over external finance, and for debt over equity. A strict interpretation of this model suggests that firms do not aim at a target debt ratio. Instead, the debt ratio is just the cumulative result of hierarchical financing over time. (See Shyum-Sunder & Myers, 1999.) Original examples of signalling models are the models of Ross (1977) and Leland and Pyle (1977). Ross (1977) suggests that higher financial leverage can be used by managers to signal an optimistic future for the firm and that these signals cannot be mimicked by unsuccessful firms14. Leland and Pyle (1977) focus on owners instead of managers. They assume that entrepreneurs have better information on the expected cash flows than outsiders have. The inside information held by an entrepreneur can be transferred to suppliers of capital because it is in the owner's interest to invest a greater fraction of his wealth in successful projects. Thus the owner's willingness to invest in his own projects can serve as a signal of project quality. The value of the firm increases with the percentage of equity held by the entrepreneur relative to the percentage he would have held in case of a lower quality project. (Copeland, Weston & Shastri, 2005.) #p#分頁標題#e#

 

The stakeholder theory formulated by Grinblatt & Titman (2002)15 suggests that the way in which a firm and its non-financial stakeholders interact is an important determinant of the firm's optimal capital structure. Non-financial stakeholders are those parties other than the debt and equity holders. Non-financial stakeholders include firm's customers, employees, suppliers and the overall community in which the firm operates. These stakeholders can be hurt by a firm's financial difficulties. For example customers may receive inferior products that are difficult to service, suppliers may lose business, employees may lose jobs and the economy can be disrupted. Because of the costs they potentially bear in the event of a firm's financial distress, non-financial stakeholders will be less interested ceteris paribus in doing business with a firm having a high(er) potential for financial difficulties. This understandable reluctance to do business with a distressed firm creates a cost that can deter a firm from undertaking excessive debt financing even when lenders are willing to provide it on favorable terms (Ibid., p. 598). These considerations by non-financial stakeholders are the cause of their importance as determinant for the capital structure. This stakeholder theory could be seen as part of the trade-off theory (see Brealey, Myers and Allen, 2006, p.481, although the term 'stakeholder theory' is not mentioned) since these stakeholders influence the indirect costs of financial distress.16 

 

As the trade-off theory (excluding agency costs between managers and shareholders) and the pecking order theory, the stakeholder theory of Grinblatt and Titman (2002) assumes shareholder wealth maximization as the single corporate objective.17 

Based on these theories, a huge number of empirical studies have been produced. See e.g. Harris & Raviv (1991) for a systematic overview of this literature18. More recent studies are e.g. Shyum-Sunder & Myers (1999), testing the trade-off theory against the pecking order theory, Kemsley & Nissim (2002) estimating the present value of tax shield, Andrade & Kaplan (1998) estimating the costs of financial distress and Rajan & Zingales (1995) investigating the determinants of capital structure in the G-7 countries. Rajan & Zingales (1995)19 explain differences in leverage of individual firms with firm characteristics. In their study leverage is a function of tangibility of assets, market to book ratio, firm size and profitability. Barclay & Smith (1995) provide an empirical examination of the determinants of corporate debt maturity. Graham & Harvey (2001) survey 392 CFOs about a.o. capital structure. We come back to this Graham & Harvey study in Section 3.20 

Cross sectional studies as by Titman and Wessels (1988), Rajan & Zingales (1995) and Barclay & Smith (1995) and Wald (1999) model capital structure mainly in terms of leverage and then leverage as a function of different firm (and market) characteristics as suggested by capital structure theory21. We do the opposite. We do not analyze the effect of several firm characteristics on capital structure (c.q. leverage), but we analyze the effect of capital structure on variables that co-determine shareholder value. In several decisions, including capital structure decisions, these variables may get the role of decision criteria. Criteria which are related to the trade-off and pecking order theory are listed in Table 1. We will discuss these criteria in more detail in section 4. Figure 2 illustrates the basic idea of our approach. #p#分頁標題#e#

 

3. Other objectives and considerations 其他目的和注意事項

A lot of evidence suggests that managers act not only in the interest of the shareholders (see Myers, 2001). Neither the static trade-off theory nor the pecking order theory can fully explain differences in capital structure. Myers (2001, p.82) states that 'Yet even 40 years after the Modigliani and Miller research, our understanding of these firms22 financing choices is limited.' Results of several surveys (see Cools 1993, Graham & Harvey, 2001, Brounen et al., 2004) reveal that CFOs do not pay a lot of attention to variables relevant in these shareholder wealth maximizing theories. Given the results of empirical research, this does not come as a surprise. 

 

The survey by Graham and Harvey finds only moderate evidence for the trade-off theory. Around 70% have a flexible target or a somewhat tight target or range. Only 10% have a strict target ratio. Around 20% of the firms declare not to have an optimal or target debt-equity ratio at all. 

 

In general, the corporate tax advantage seems only moderately important in capital structure decisions. The tax advantage of debt is most important for large regulated and dividend paying firms. Further, favorable foreign tax treatment relative to the US is fairly important in issuing foreign debt decisions23. Little evidence is found that personal taxes influence the capital structure24. In general potential costs of financial distress seem not very important although credit ratings are. According to Graham and Harvey this last finding could be viewed as (an indirect) indication of concern with distress. Earnings volatility also seems to be a determinant of leverage, which is consistent with the prediction that firms reduce leverage when the probability of bankruptcy is high. Firms do not declare directly that (the present value of the expected) costs of financial distress are an important determinant of capital structure, although indirect evidence seems to exist. Graham and Harvey find little evidence that firms discipline managers by increasing leverage. Graham and Harvey explicitly note that ‘1) managers might be unwilling to admit to using debt in this manner, or 2) perhaps a low rating on this question reflects an unwillingness of firms to adopt Jensen’s solution more than a weakness in Jensen’s argument'. 

 

The most important issue affecting corporate debt decisions is management’s desire for financial flexibility (excess cash or preservation of debt capacity). Furthermore, managers are reluctant to issue common stock when they perceive the market is undervalued (most CFOs think their shares are undervalued). Because asymmetric information variables have no power to predict the issue of new debt or equity, Harvey and Graham conclude that the pecking order model is not the true model of the security choice25. #p#分頁標題#e#

 

The fact that neoclassical models do not (fully) explain financial behavior could be explained in several ways. First, it could be that managers do strive for creating shareholder value but at the same time also pay attention to variables other than the variables listed in Table 1. Variables of which managers think that they are (justifiably or not) relevant for creating shareholder value. Second, it could be that managers do not (only) serve the interest of the shareholders but of other stakeholders as well26. As a result, managers integrate variables that are relevant for them and or other stakeholders in the process of managing the firm's capital structure. The impact of these variables on the financing decision is not per definition negative for shareholder value. For example if ‘value of financial rewards for managers’ is one the goals that is maximized by managers – which may not be excluded – and if the rewards of managers consists of a large fraction of call options, managers could decide to increase leverage (and pay out an excess amount of cash, if any) to lever the volatility of the shares with an increase in the value of the options as a result. The increase of leverage could have a positive effect on shareholder wealth (e.g. the agency costs between equity and management could be lower) but the criterion 'value of financial rewards' could (but does not have to) be leading. Third, shareholders themselves do possibly have other goals than shareholder wealth creation alone. Fourth, managers rely on certain (different) rules of thumb or heuristics that do not harm shareholder value but can not be explained by neoclassical models either27. Fifth, the neoclassical models are not complete or not tested correctly (see e.g. Shyum-Sunder & Myers, 1999). 

 

Either way, we do expect variables other than those founded in the neoclassical property rights view are or should be included explicitly in the financing decision framework. To determine which variables should be included we probably need other views or theories of the firm than the neoclassical alone. Zingales (2000) argues that ‘…corporate finance theory, empirical research, practical implications, and policy recommendations are deeply rooted in an underlying theory of the firm.’ (Ibid., p. 1623.) Examples of attempts of new theories are 'the stakeholder theory of the firm' (see e.g. Donaldson and Preston, 1995), 'the enlightened stakeholder theory' as a response (see Jensen, 2001), 'the organizational theory' (see Myers, 1993, 2000, 2001) and the stakeholder equity model (see Soppe, 2006). 

We introduce an organizational balance sheet which is based on the organizational theory of Myers (1993). The intention is to offer a framework to enhance a discussion about criteria that could be relevant for the different stakeholders of the firm. In Myers' organizational theory employees (including managers) are included as stakeholders; we integrate other stakeholders as suppliers, customers and the community as well. Figure 3 presents the adjusted organizational balance sheet. #p#分頁標題#e#

 

Pre-tax value is the maximum value of the firm including the maximum value of the present value of all stakeholders' surplus. The present value of the stakeholders' surplus (ES plus OTS) is the present value of future costs of perks, overstaffing, above market prices for inputs (including above market wages), above market services provided to customers and the community etc.28 Depending on the theory of the firm, the pre-tax value can be distributed among the different stakeholders following certain 'rules'. Note that what we call 'surplus' in this framework is still based on the 'property rights' principle of the firm. Second, only distributions in market values are reflected in this balance sheet. Neutral mutations are not29. 

 

Based on the results of Graham and Harvey (2001) and common sense we formulate a list of criteria or heuristics that could be integrated into the financing decision framework. Some criteria lead to neutral mutations others do not. We call these criteria 'quasi non-economic criteria'. Non-economic, because the criteria are not based on the neoclassical view. Quasi, because the relations with economic value are not always clear cut. We include criteria that lead to neutral mutations as well, because managers might have good reasons that we overlook or are relevant for other reasons than financial wealth. 

 

The broadest decision framework we propose in this paper is the one that includes both the economic and quasi non-economic variables. Figure 4 illustrates the idea. The additional quasi non-economic variables are listed in Table 2. This list is far from complete. 
 

flexibility could be relevant for at least employees and the suppliers of resources needed for these projects. As long as managers only would invest in zero net present value projects this variable would have no value effect in the organizational balance sheet. But if it influences the value of the sum of the projects undertaken this will be reflected in this balance sheet. Of course, financial flexibility is also valued for economic reasons, see Section 2 and 4. 
 

The probability of bankruptcy influences job security for employees and the duration of a 'profitable' relationship with the firm for suppliers, customers and possibly the community. For managers (and other stakeholders without diversified portfolios) the probability of default could be important. The cost of bankruptcy is for them possibly much higher than for shareholders with diversified portfolios. As with financial flexibility, the probability of default influences shareholder value as well. In Section 2 and 4 we discuss this variable in relation to shareholder value. Here the variable is relevant, because it has an effect on the wealth or other 'valued' variables of stakeholders other than equity (and debt) holders. We assume owner-managers dislike sharing control of their firms with others. For that reason, debt financing could possibly have non-economic advantages for these managers. After all, common stock carries voting rights while debt does not. Owner-managers might prefer debt over new equity to keep control over the firm. Control is relevant in the economic framework as well, see Section 2 and 4. #p#分頁標題#e#

 

In practice, earnings dilution is an important variable effecting the financing decision. Whether it is a neutral mutations variable or not30, the effect of the financing decision on the earnings per share is often of some importance. If a reduction in the earnings per share (EPS) is considered to be a bad signal, managers try to prevent such a reduction. Thus the effect on EPS becomes an economic variable. As long as it is a neutral mutation variable, or if it is relevant for other reasons we treat EPS as a quasi non-economic variable. 

The reward package could be relevant for employees. If the financing decision influences the value of this package this variable will be one of the relevant criteria for the manager. If it is possible to increase the value of this package, the influence on shareholder value is ceteris paribus negative. If the reward package motivates the manager to create extra shareholder value compared with the situation without the package, this would possibly more than offset this negative financing effect.