1

DOMESTIC AND MULTINATIONAL CAPITAL STRUCTURE:

THEORY AND EVIDENCE

J. Edward Graham, Jr.

ABSTRACT

The capital structure of the firm is examined in the domestic environment under successively less restrictive assumptions. Allowances are then made for agency costs and information asymmetries. The domestic capital structure theory is extended to the multinational environment and examined empirically. Results broadly support financial theory. Where the theory is restrictive in its assumptions, it describes multinational capital structure poorly. Subsequent research is proposed.

DOMESTIC AND MULTINATIONAL CAPITAL STRUCTURE:

THEORY AND EVIDENCE

Working Paper

Introduction

What is capital structure and how is capital structure chosen by a corporation? What are the roles of debt and equity in this capital structure? How are these components chosen in a “perfect market” without taxes or bankruptcy costs? How do these roles and proportions change as an allowance is made for taxes and bankruptcy costs? What roles are played by debtholder and equityholder incentives in this structure? What is the impact of agency costs and asymmetric information on the capital structure? How, if at all are these factors and their impact on a firm’s capital structure changed when a provision is made for international variables? Is the capital structure of the multination corporation (MNC) significantly different from the domestic corporations (DC)?

These questions are addressed.

In Section 1, capital structure is defined and considered under the restrictive assumptions of a “perfect market.” The implications of a frictionless capital market are examined. An allowance is made for corporate and personal taxes in Section 2. The following section includes a provision for the impact of uncertainty, leverage and bankruptcy costs. Agency costs, debtholder incentives, and agency cost of debt are considered in Section 4. A brief examination of the impact of asymmetric information on capital structure is presented in Section 5. Selected topics and empirical studies of MNC capital structure are reviewed in Section 6. The paper closes with a summary and encouragement of subsequent research. Each section provides the “domestic” theory relevant to the given topic on capital structure and extends the theory to the multinational setting.

Section 1

The financial structure of a firm is the mix of all the items that appear on the right-hand side of the firm’s balance sheet. The capital structure is the mix of the long-term components of that financial structure. The domestic and multinational corporate capital structure both consist of various elements of liabilities and owner’s equity. Similarities and dissimilarities between MCN and DC capital structure are considered. In the evaluations which follow, the stockholder-owned MNC or DC attempts to select a capital structure that maximizes firm value: this value is composed of the market value of debtholder and equity holder financing.

Does capital structure affect the value of the firm? Are investment decisions affected by the manner in which they are funded? Until the seminal work by Modigliani and Miller (1958), the prevailing view was that firm value was a concave function of its use of financial leverage; that an optimal mix of debt and equity existed (a global maximum) which maximized firm value. Modigliani and Miller (M&M) proposed that firm value was independent of its capital structure, subject to a set of restrictive assumptions. [1]

Central among the provisions of M&M (1958) is an assumption of “perfect” and frictionless capital and financial markets; there is no allowance for taxes, bankruptcy costs or transaction costs. Coupling these features of the M&M modeling with their other restrictions provides support for their proposals: that capital structure is irrelevant as it effects firm value, that investment and financing decisions are independent and that expected return increases linearly with debt, precluding an increase in firm value through the use of leverage. Stiglitz (1969) relaxes the risk class, capital market competitiveness and homogeneous expectations assumptions of M&M (1958) and is able to reach similar conclusions. However, his modeling is not robust to the introduction of bankruptcy costs.[2]

He does find the M&M proof holds, however, under much more general conditions than originally supposed.

An extension of the fairly stringent assumptions of M&M to the multinational setting is difficult. Senbet (1979) questions the extension of the perfectly competitive capital market assumption to the international setting. Yet, given the assumption and the absence of taxes (or an “international tax differential”), he finds the international financing mix is irrelevant. M&M (1958) holds.[3] He still notes that existing financial theory “ has to be amply modified if it is to accommodate” MNCs.

Stiglitz (1969) implicitly acknowledges the potential for an optimal capital structure for the economy as a whole. He merely illustrates irrelevance for the individual firm, given his less-binding restrictions. Nonetheless, he admits the non-robust nature of his M&M examination to bankruptcy costs and the “crucial fallacies of the perfectly competitive capital market:” that lending rates different from borrowing rates, that the nature of a firm’s risky debt changes as its debt ratio changes, and that one firm’s bonds are different from another’s. These three factors are precluded in the perfectly competitive capital market.

Errunza and Senbet (1981) note that any international investment by a firm competing with “the locals” is a departure from “Market perfection.” Given perfectly competitive markets, these locals could acquire the necessary technologies and capital on their own. Discovering a systematic positive relationship between “current degree of international involvement and excess market value,” they illustrate the abbreviation of market perfection in the international arena.[4] They discover that the factors leading to these financial market imperfections are priced. They provide an indication of the contributions of international barriers to foreign direct investment. This impacts MNC capital structure.

Many of the arguments against the restrictive assumptions of M&M (1958) apply in both the domestic and multinational setting. Differential tax, lending and borrowing treatment exist for

Corporations with and between nations. The limitations mentioned above are addressed and successively less restrictive assumptions are provided; a number of insights concerning the capital structures of the DC and MNC are gleaned. Although capital structure theory is still inadequate, studies since 1958 have encouraged certain tenets of this theory and discouraged others.[5]

Section 2

What is the impact on capital structure if an allowance is made for taxes? Revisiting their earlier work, M&M (1963) reveal that taxes and leverage (an effective government subsidy of debt) do give the firm an advantage. The implication is that an optimal cattail structure exists for the firm. Extending this argument, if the only imperfection in the financial world is corporate taxes, the firm uses all debt financing.[6] Treating the tax shield as perpetuity, they find firm value is maximized through the use of debt.

Specifying the value of the firm without a provision for personal taxation, M&M (1963) main point is that required returns are lowered using debt. The value of the firm is an increasing function of debt. Since firm value is more easily maximized using debt and the firm undertakes investment when the value of the firm is increased, the probability of investment and magnitude of investment are also increased.

Perhaps some of the liveliest examinations of DC and MNC capital structure have had taxes and international taxation as their fulcrum. Differential international taxation (a potential barrier to the smooth operation of international financial markets) has invited extensive study. The differential government subsidy of business has invited and discouraged the flow of capital across borders.[7]

Black (1974) addresses capital market equilibrium, given explicit barriers to international investment, as a function of international tax differentials. Broadly defining “tax” as any kind of barrier to foreign investment, he posits that where these barriers are effective, a world CAPM does not hold. [8] Financial and capital markets are imperfect and a firm will structure its mix of debt and equity to take advantage of these imperfections. An optimal multinational capital structure, long in domestic assets and short in foreign ones, may exist. Where these barriers are ineffective, he seems to imply capital striker irrelevance as it relates to international capital market imperfections. His work does not specifically model the impact of personal taxes. His basic premise is that models of international investment assume “specified or unspecified barriers” to multinational investment and that these barriers cause changes in the pricing of international securities. Tax-free investors are subsidized by these barriers, to which they are not subject. Senbet (1979) also notes that international tax differentials affect a firm’s financing mix and investment.

Miller (1977) extends the theoretical consideration of capital structure to personal taxes. Proposing that bankruptcy costs and agency costs are insufficient to offset the tax advantage of debt, he posits that the value of the firm, in equilibrium, is independent of capital structure. He acknowledges (as did Stiglitz (1969)) an aggregate macroeconomics capital structure, but denies an optimal structure for the individual firm. [9] He proposes that, in equilibrium, the marginal tax rate for a buyer of corporate debt equals the marginal tax rate for the issuing corporation; implying a macro-level capital structure optimum but a micro-level irrelevance.

Lee And Zechner (1984) extend the Miller (1977) hypothesis to the international setting and find that the “Miller equilibria do not hold” in the presence of different tax subsidies between countries. They analyze capital structure in a two country setting and find, on a theoretical basis, that firms tend to be all equity or all debt depending on location. They likewise identify “differential tax subsidies of debt if inflation rates differ across countries.” Differential rates of inflation and national corporate tax rates can change the equilibrium structure of international capital markets. In the presence of barriers to international investment (capital markets are not perfectly integrated), the Miller irrelevance theorem holds; in the sense that “leverage is indeterminate for the individual firm while it is determinate for the aggregate economy.” Given integrated international capital markets, high-tax country firms tend to be all debt-financed and low-tax, all equity. [10]

Hodder and Senbet (1990) also generalize the Miller analysis to an international setting. Given differential international taxation, they note implications for national differences in capital structure, the international Fisher effect and yield differentials. They propose that corporate tax arbitrage plays a key role in generating an international capital structure equilibrium. Contrary to Lee and Zechner (1984) they show that, “ if corporations engage in international tax arbitrage on an equal footing, no optimal capital structure exists for the individual firms.” Miller (1977) holds. Thus, differences in international tax rates alone are not sufficient to dictate a firm’s capital structure. Assuming otherwise “perfect” capital markets, their international analogy of Miller is robust to different tax rates. [11] Apparent international differences in capital structure are not due solely to differences in personal and corporate tax rates. The international Fisher effect holds despite differences in taxes and inflation. Additionally, there exists no “induced preference for corporate borrowing in a particular currency.” Much of the difference in the findings between Lee and Zechner (1984) and Hodder and Senbet (1990) can be attributed to their adopted

Restrictions. Nonetheless, the contrasts of those two studies are noteworthy.

Section 3

What are the implications for the firm’s cattail structure when an allowance is made for uncertainty and bankruptcy costs? Kim’s (1978) modeling provides support for the traditional view; that a firm’s value is a concave function of its debt financing and that an interior optimum exists where the slope of these function equals zero. He proposes that the value of the levered firm is equal to the value of the unlevered firm plus the present value of the tax-shield provided by debt less tax credits lost and costs incurred in bankruptcy. This traditional argument flows from the M&M (1958 and 1963) argument by allowing for taxes and bankruptcy costs. He shows an optimal level of debt financing (optimal capital structure) at less than the firm’s debt capacity.

Haugen and Senbet (1978) counter that the expectations of bankruptcy costs and other costs of financial distress do not impact capital structure. They also find Kim’s one-period model deficient. They propose that the choice of liquidation and/or reorganization exists as an effective alternative to bankruptcy. They argue that the costs of bankruptcy are bounded by the costs of informal reorganization. They posit, given rationality and large financial markets, that the costs of bankruptcy “must be small” and that the liquidation decision is made independently of capital structure. They note that all claimants have “ an incentive to avoid bankruptcy.” Given the widespread use of bankruptcy in firm reorganization, their arguments seem to lack intuitive appeal.

Titman (1984) notes that traditional approaches to the consideration of capital structure have not considered all claimants. He introduces a third claimant – workers, customers, community, and suppliers. Noting the shallow nature of the claims of Haugen and Senbet (1978) that these third parties should join in reorganization coalitions to protect their interests, he proposes the use of preferred stock to protect these third parties. His arguments do not extend easily into the multinational arena.

Bradley, Jarrel, and Kim (1984) (BJK) acknowledge that the general view (though not a consensus) is that optimal capital structure involves “balancing the tax advantage of debt against the present value of expected bankruptcy costs.” The “upshot” of extensions of Miller’s (1977) modeling is the recognition that the existence of an optimal capital structure is essentially an empirical issue as to whether or not various leverage-related costs are significant enough to influence the costs of corporate borrowing.[12] Their examinations reveal that firm leverage ratios are negatively related to the volatility of earnings (given non-trivial costs of financial distress) and that “permanent ratios” are related to the firm industrial classification. They generate strong findings of intra-industry similarities in capital structure and a persistence of inter-industry differences.

These theoretical and empirical considerations of capital structure, leverage, and bankruptcy costs have been extended to the international setting. Shaked (1986) examines the failure probabilities of a sample of MNCs. His findings suggest that the probability of bankruptcy is greater for DCs than MNCs. He observes that MNCs are significantly more capitalized with lover returns volatility than DCs. His study seems to lack an adequate explanation for these findings, however. MNCs might be expected to have higher debt ratios, given their less volatile and internationally diversified operations, than their domestic counterparts. Given also the hedging of local political risk with higher debt ratios by foreign affiliates and the reduction of bankruptcy probability (expected bankruptcy costs) through international diversification, this seems counter-intuitive.