حقوق بستانکار و تخصیص بدهی در شرکت های چند ملیتی

We analyze the optimal debt structure of multinational corporations choosing between centralized or decentralized borrowing. We identify how this choice is affected by creditor rights and bankruptcy costs, taking into account managerial incentives and coinsurance considerations. We find that partially centralized borrowing structures are optimal with either weak or strong creditor rights. For intermediate levels of creditor rights fully decentralized (centralized) borrowing structures are optimal if managers have strong (weak) empire-building tendencies. Decentralized borrowing is more attractive for companies focussing on short-term profitability. Credits are rather taken in countries with better creditor rights and more efficient insolvency systems.

Multinational companies (MNCs) have a wide range of financing options when they set-up a foreign subsidiary. They can rely on capital transferred from the parent company, but they can also raise local credits. How do multinational firms finance their foreign subsidiaries? To what extent do they rely on local financing and why? Empirical evidence suggests that only part of the subsidiaries is financed internally, with capital from the parent company (e.g. Desai et al. (2004)). Furthermore, multinationals seem to choose a different financing strategy depending on where their foreign subsidiary is located. Kang et al. (2004) report that in industrial countries 29 percent of the financing of subsidiaries come from parents and 42 come from host residents, while in developing countries 45 percent of the financing come from US parents and 34 percent come from host country residents. In this paper we focus on one particular aspect of a multinational’s financing decision: the credit financing.1 If (at least) part of the financing has to be done through credits, the question arises whether these should be raised locally in the foreign subsidiary’s host country or via the parent company. The aim of our paper is to determine the optimal debt allocation within a multinational corporation. For this purpose we develop a model of multinational borrowing that explicitly considers agency problems in internal capital markets, the existence of bankruptcy costs and the role of creditor rights. In our model the trade-off between decentralized (local) and centralized (parent) debt financing is driven by two main effects, the incentive and the coinsurance effect. Centralizing the borrowing structure allows the multinational corporation to realize a so-called coinsurance effect.2 In this case the CEO of a MNC can use the net profits of all its subsidiaries to repay debt and avoid costly bankruptcy. Only if the sum of net profits is not sufficient to cover all debt repayments, bankruptcy occurs. Thus, one subsidiary “coinsures” another subsidiary and bankruptcy becomes less likely. This is the positive effect associated with debt centralization. However, debt centralization also entails negative incentive effects. These arise because the coinsurance of the subsidiaries attenuates the disciplining effect of debt. Consider a multinational with two subsidiaries F and H. If, say, the manager of subsidiary F borrows locally, he is directly liable to his debtors. This gives him strong immediate incentives to work hard and avoid the bankruptcy of his subsidiary – at least if he enjoys private benefits of control and does not want to lose his job ( Aghion and Bolton, 1992). 3 Centralizing the borrowing for subsidiary F weakens manager F’s incentives because it reduces the direct link between his success and the liquidation of his subsidiary: Even if he fails, subsidiary F will not be liquidated as long as subsidiary H is successful because he “is coinsured” by subsidiary H. Similarly, centralizing the borrowing for subsidiary H, thus “coinsuring” subsidiary H entails negative incentive effects for the subsidiary manager F as well. Now, internal capital market considerations come into play: If subsidiary H is coinsured and fails but manager F is successful, the profits generated by manager F are used to meet the debt repayments of subsidiary H. As managers are typically interested in having large empires, taking away these funds reduces a manager’s benefits and hence his incentives. This is the downside of reallocating funds within internal capital markets (see for example Brusco and Panunzi (2005)). To summarize, both “being coinsured” by and “coinsuring” the other subsidiary entail adverse incentive effects. These negative incentive effects countervail the positive risk-reducing effect of coinsurance. The trade-off between coinsurance and incentive effects differs for various host countries depending on the strength of creditor rights.4 Stronger creditor rights imply more control rights for the creditor in case of insolvency. As creditors are interested in liquidating insolvent firms, the liquidation of unsuccessful firms becomes more likely when creditor rights are stronger.5 When creditor rights are weak, the threat of liquidation in case of insolvency and hence the disciplining effect of debt is less present than with strong creditor rights. This affects the overall trade-off. We determine the optimal debt structure depending on firm characteristics and the specific legal and institutional settings. In the first part of our analysis we disregard differences in the legal environment of host and home countries. In the second part of the paper we introduce these differences and derive how they affect the optimal borrowing structure. Our main findings are as follows: For MNCs operating in countries with very weak or very strong creditor rights, mixed borrowing structures are optimal. A “mixed borrowing structure” indicates a borrowing structure with centralized borrowing for one subsidiary and decentralized borrowing for the other subsidiary. The optimality of the borrowing structure for intermediate ranges of creditor rights depends on managerial incentives: If managerial empire-building tendencies are weak, a fully centralized borrowing structure is optimal. If empire-building tendencies are strong, a fully decentralized borrowing structure is optimal because it becomes more attractive to provide incentives. Stronger creditor rights increase the attractiveness of substituting parental borrowing with local debt in the foreign affiliate’s country. Furthermore, we find that, due to agency problems, weak creditor rights are associated with higher probability of bankruptcy and higher interest rates for foreign affiliates’ local borrowing. Higher bankruptcy costs increase the attractiveness of centralized borrowing. If the two countries in which the multinational operates differ with respect to bankruptcy costs, the CEO prefers to borrow in the country with a more efficient bankruptcy system. Differences in creditor rights do not have any direct effect on expected profits under any of the borrowing structures. However, as they affect the disciplining effect of debt, they influence managerial incentives and indirectly expected profits. More specifically, weaker creditor rights in the foreign country decrease the attractiveness of a (partially) decentralized borrowing structure. Finally, if the two countries have different growth opportunities and, say, the foreign country exhibits a higher growth potential, centralized borrowing for the foreign affiliate becomes more attractive, whereas it becomes less attractive for the home country affiliate. The remainder of this paper is structured as follows: Section 2 gives an overview of the related literature. Section 3 lays out the set-up and basic mechanisms of our model. In Section 4 we derive the equilibrium outcome and optimality conditions in a national setting. Section 5 analyzes the comparative statics and introduces differences in the legal environment and growth potential between the affiliate’s and the parental country. Section 6 highlights the empirical findings of our model. Section 7 concludes.

In this paper we develop a framework for understanding the debt allocation process within multinational corporations. In our analysis we show that the debt structure within multinationals matters beyond tax issues – a fact that had almost been completely neglected in the literature so far. In particular, we highlight that the legal environment is key in determining the degree of debt centralization within a MNC. However, as our analysis suggests, different aspects of the legal environment have differing effects on the borrowing structure. Although very stylized, our model and results do reflect the existing empirical findings related to multinational finance and creditor rights. While we provide a rationale for mixed borrowing structures, we demonstrate how the trade-off between incentive problems in internal capital markets and coinsurance determines the optimal borrowing structure. Our analysis highlights the relevance of creditor rights for a multi-entity firm’s capital structure in general and for multinational corporations in particular. Differences in the legal environment induce a bias of the debt allocation towards the country with a better legal environment, i.e. stronger creditor rights and lower bankruptcy costs. A major contribution of our paper is to highlight the importance of a comprehensive view on multinationals’ borrowing decision due to feedback effects on internal capital markets – an aspect that current research on MNC finance did not focus on yet. A more comprehensive model would endogenize the incentive problems of the home subsidiary as well. The basic trade-off of our model would not be affected but there may be room for reinforcing incentive effects between the subsidiaries. This must be left to future research. Further questions to be addressed in future research relate to the effect of creditor rights on several aspects of multinational finance. An extension of our work could incorporate the choice between equity and debt into a model of multinational finance. A particularly interesting question is how the legal environment affects the multinational’s choice between internal debt, i.e. parental borrowing for the subsidiaries, and internal equity. Another interesting aspect which needs further empirical investigation is a differentiated analysis of the effect of creditor rights on the different aspects of the insolvency regime, for example also the actual liquidation of insolvent firms.