تامین بدهی های خارجی در بازار ناقص سرمایه بین المللی: نظریه و شواهد
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|14611||2010||23 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of International Money and Finance, Volume 29, Issue 2, March 2010, Pages 201–223
We investigate the determinants of foreign borrowing costs in a stochastically growing economy. We find that these increase with the debt-wealth ratio, depending also upon the volatilities of domestic and foreign origin, and the length of debt contract. In addition, the sensitivity of the short-term debt supply to the debt-wealth ratio exceeds that of long-term debt, and the effects of volatility on the borrowing premium, growth of wealth, and its volatility, depend on the relative size of a direct effect and a secondary portfolio-adjustment effect of the initial shock, as well as the length of the debt contract. Panel regressions suggest that the empirical evidence generally support the theoretical predictions.
The basic macrodynamic model of a small open economy typically assumes that the country has unrestricted access to a perfect world capital market, so that it can borrow or lend unlimited amounts at an exogenous constant world interest rate. If, in addition, one invokes the conventional assumption of a fixed constant rate of time preference, these two exogenous constants must in fact be equal in order for an interior equilibrium to prevail. This imposes a stringent “knife-edge” condition on the economy, one that has several important ramifications; see e.g. Turnovsky, 2002 and Schmitt-Grohé and Uribe, 2003. For example, it leads to an extreme form of “consumption-smoothing”, and has the further consequence that a temporary structural change or policy shock may have a permanent effect; see e.g. Sen and Turnovsky (1989). It also has profound implications for the long-run viability of tax policy; see e.g. Frenkel et al. (1991) and Turnovsky (1997, chapter 6). Quite apart from the unpalatable practice of imposing equality upon two seemingly unrelated and independent structural parameters, the assumption that a small economy can borrow or lend unlimited amounts indefinitely in international markets at a constant rate is implausible. At some point the economy will cease to be “small” and its decisions will influence the world capital market. Accordingly, beginning with an early paper by Bardhan (1967), economists have periodically imposed a relationship between the rate at which a country can trade financial assets and its net asset position, thereby breaking the knife-edge constraint noted above.1 The effect of this is to introduce a borrowing premium, which essentially serves as a proxy for the country's default risk. Various specifications of this relationship can be found. Several authors follow Bardhan's original specification and assume that the premium depends upon the level of debt; see e.g. Obstfeld, 1982 and Bhandari et al., 1990, and Fisher (1995). However, as originally argued by Sachs and Cohen (1982) and Sachs (1984), a more appropriate measure reflecting the country's ability to service its debt, is to assess the debt relative to some measure of earning capacity, such as its wealth, capital stock or level of output. Moreover, normalizing in this way becomes necessary if one wishes to incorporate increasing debt costs in an equilibrium of ongoing growth; see e.g. van der Ploeg, 1996 and Turnovsky, 1997, and Turnovsky and Chattopadhyay (2003); see also Mendoza and Uribe (2000), and Sendhadji (2003). Underlying these specifications is the potential for default risk of highly indebted economies. The issue of sovereign default risk and debt repayment stems from the seminal work of Eaton and Gersovitz, 1981 and Eaton and Gersovitz, 1989). Emphasizing the difference between sovereign default and bankruptcy of an individual agent, they present a debt repayment function in which lenders establish a credit ceiling that prevents borrowers from repudiating the debt. Similarly, many researchers focus on the sovereign government's trade off between default and lender-enforced penalties, such as a ban from international credit market (Grossman and van Huyck, 1988), seized assets abroad, or future trade barriers (Bulow and Rogoff, 1989). The upward-sloping supply of debt function has been employed in a variety of contexts. These include: an analysis of the terms of trade shocks (Obstfeld, 1982 and Eicher et al., 2008), international term structure of interest rates (Fisher, 1995), economic growth (van der Ploeg, 1996 and Turnovsky and Chattopadhyay, 2003) and foreign aid (Chatterjee et al., 2003). The idea of inelastic debt supply curve is also applied in the models of financial crisis episodes in less developed countries. Duffie et al. (2003) model the pricing of sovereign debt focusing on the yield spread. Eicher et al. (2001) show that financial market liberalization alone can generate sharp reversals in foreign capital flows if short-term debt supply is more inelastic than is long-term debt. Rodrik and Velasco (2000) provide a theoretical and empirical analysis of the effects of socially excessive short-term capital flows as a cause of more severe crisis. While the specification of borrowing costs as an increasing function of the country's net foreign debt position seems plausible, it is nevertheless ad hoc. Beginning with Bardhan (1967) it has been postulated as an equilibrium reduced form relationship, rather than being derived from the underlying behavior of relevant agents. As a consequence, the existing literature simply postulates an arbitrary (usually convex) relationship between debt and borrowing costs, ignoring the dependence of this relationship upon other relevant aspects, such as the degree and nature of the underlying risk in the economy. This paper has several objectives. First, we wish to derive the equilibrium borrowing premium from the rational behavior of risk-neutral expected profit maximizing international financial intermediaries that allocate funds between borrowers and lenders, while taking account of the potential for default risk. Domestic borrowers are subject to two sources of risk, (i) an internal source, due to domestic production risk, and (ii) an external source, due to the stochastic pricing of foreign bonds, as a result of which they risk defaulting on their loans. We show how this leads to an equilibrium in which the borrowing cost increases with the economy's debt-wealth ratio, with the relationship depending upon the domestic/foreign sources of risk and the length of debt contract. This equilibrium relationship turns out to be a complex one and we therefore analyze it numerically, by calibrating a plausible replication of a representative small open debtor economy under various conditions of risk and borrowing costs. Any changes in these factors will shift the supply curve of debt, causing subsequent reallocation of resources. Along the supply schedule, the agent's resource reallocation further adjusts the borrowing cost. We also find that the curvature of the debt supply curve, a critical determinant of the borrowing premium, is highly sensitive to the length of the debt contract. Although the short-term debt supply curve begins at a lower interest rate with flatter slope than does the one with a longer maturity, its shape dramatically changes with the increase in the relative size of debt. This result is consistent with the findings in the debt crisis literature (for example, Rodrik and Velasco, 2000). After developing the debt cost function in detail, we then introduce it into the stochastic dynamic general equilibrium growth model of Turnovsky and Chattopadhyay (2003). The resulting equilibrium is a balanced growth path along which all variables grow at a common stochastic rate, determined by the parameters of preference, production, and risks. In doing so, we focus on comparing the consequences of replacing the arbitrarily specified debt-cost function, employed by Turnovsky and Chattopadhyay with the form derived in this paper. We find that the simplified specification of the debt-cost function generally leads to an understatement of the borrowing premium in the case of short-term debt and overstatement in the case of long-term debt. The final stage of our analysis is to supplement our theoretical and numerical analysis with empirical evidence. Panel regressions of 20 developing debtor economies over 42 quarters between 1993 and 2004 provide an empirical support to our findings. In particular, we find that the effect of debt on the borrowing cost is steeper for short-term debt, consistent with our numerical simulations. We also find that the borrowing premium for both types of debt increases with the terms of trade volatility, while output volatility has a strong impact on the borrowing premium for long-term debt, but only a marginal effect on the premium for short-term debt. This empirical evidence confirms the importance of deriving the cost of debt function from underling rational behavior. The rest of the paper is organized as follows. Section 2 develops a formal analytical model of the interest premium for risky foreign debt in the context of a dynamic general equilibrium. Section 3 presents results from the calibration to analyze the effects of volatility on the balanced growth path equilibrium. In Section 4, we briefly explain the theoretical background of our regressions and develop a basic regression model, followed by the description of the data and regression results. Section 5 summarizes our main conclusions and an Appendix provides some of the technical details.
نتیجه گیری انگلیسی
The fact that developing economies are subject to default-risk and hence have limited access to the international credit market is an issue of great concern to both researchers and policymakers. Macrodynamic models have tended to incorporate this fact by assuming that such economies face an upward sloping supply of debt, whereby their borrowing costs increase with their debt position. While this formulation has appeal as a plausible reduced form equilibrium relationship, it is essentially ad hoc. In this paper we have developed a model of foreign debt supply, which incorporates the rational behavior of participants in the foreign debt market, in light of the risks they face. Using this framework we derive the foreign-debt supply curve as an upward sloping function of the relative size of debt, measured by debt-wealth ratio. In addition, the foreign-debt supply curve is characterized by the volatility of domestic output growth (internal volatility) and the volatility of bond prices (external volatility), together with their interaction with the debt-wealth ratio. Second, the borrowing premium depends upon the length of the debt contract, with the curvature of the foreign debt supply curve varying inversely with its length. Third, the external source of volatility exerts a larger effect on the agent's portfolio/consumption decision and his welfare than does internal volatility. The growth rate of wealth and its volatility can either increase or decrease, depending on the direct effect of a shock and the indirect portfolio adjustment effect. Using an unbalanced panel of 20 countries over 40 quarters, we find evidence that both the short-term and the long-term interest premia are positively related with the relative size of debt, measured by debt-GDP ratio. Furthermore, we find that the short-term debt supply curve is more inelastic than the long-term curve in the reasonable range of debt size. Also, we examine the effects of a change in domestic and external risk on the short-term and the long-term interest premia. We find that the real exchange rate volatility has a strong, positive effect on the interest premia in all specifications in our regressions, while the effect of the domestic output volatility is weaker. We also find that an increasing volatility may affect the slope of the debt supply curve as well as its position by interacting with the relative debt size. Overall, we find that this empirical evidence offers encouraging support for the underlying theoretical model.