تصمیمات تامین مالی شرکت ها و سیاست های بانک مرکزی

This paper aims to explain why unhedged foreign borrowing by South East Asian corporations rose sharply during the few years prior to the crisis despite little change in fundamentals. We show that decisions of firms and decisions of the central bank are complementary. Consequently, a small shock to fundamentals may have a large and permanent impact on the equilibrium composition of firms’ borrowing.

A few years after the onset of speculative attacks against Asian currencies and the subsequent economic crisis, much remains to be understood about the reasons for the attacks and for the magnitude of the crisis. Arguably, the single most important factor aggravating the crisis has been the enormous amount of foreign currency denominated borrowing by local corporations and financial institutions prior to the crisis. Table 1 shows that foreign bank lending to Asian banks and corporations had been steadily rising prior to the crisis, constituted mostly of short term loans, and frequently exceeded the total level of the country’s reserves of foreign currencies. According to IMF, external financing (including bond financing) for Asian countries has tripled between 1992 and 19961. Furthermore, the bulk of this exposure to foreign currency risk remained unhedged2 Unhedged exposure to currency risk is central to many theories of the Asian crisis3 and has indeed contributed to the amplification of the crisis in at least two ways. First, and most obviously, it led to a much increased debt burden of Asian banks and corporations after the devaluation of local currencies. Second, it was conducive of a Diamond–Dybvig (1983) type of liquidity crisis as short run liabilities exceeded short term assets and lenders refused to roll over the loans. Table 1. International claims held by foreign banks—distributed by maturity and sector Distribution by sector Total Outstanding Short term ≤1 yr Banks Public sector Non-bank private Reserves Short Term/Reserves In billions of US dollars In percentage of total consolidated claims In billions of US dollars Indonesia End 1995 44.5 61.9 20.1 15.1 64.7 14.7 1.9 End 1996 55.5 61.7 21.2 12.5 66.2 19.3 1.8 Mid 1997 58.7 59.0 21.1 11.1 67.7 20.3 1.7 Malaysia End 1995 16.8 47.2 26.4 12.4 60.4 23.9 0.3 End 1996 22.2 50.3 29.3 9.0 61.8 27.1 0.4 Mid 1997 28.8 56.4 36.4 6.4 57.1 26.6 0.6 Thailand End 1995 62.8 69.4 41.0 3.6 55.2 37 1.2 End 1996 70.1 65.2 36.9 3.2 59.6 38.7 1.2 Mid 1997 69.4 65.7 37.6 2.8 59.5 31.4 1.5 Korea End 1995 77.5 70.0 64.4 8.0 27.6 32.7 1.7 End 1996 100.0 67.5 65.9 5.7 28.3 34.1 2.0 Mid 1997 103.4 67.9 65.1 4.2 30.6 34.1 2.1 Sources: Bank for International Settlements and Radelet and Sachs (1998). Table options At this date, two explanations for the huge exposure to currency risk of Asian firms have been suggested. The first explanation simply is that local firms underestimated the risk of devaluation. For instance, the IMF International Capital Markets Report (1998) states that: “large interest rate differentials [between local and foreign borrowing] created a strong incentive for external borrowing, especially when firms regarded the authorities’ ability to sustain their exchange rate arrangements as credible”. However, this argument is not entirely convincing. First, from a theoretical standpoint, the existence of a spread between domestic and foreign interest rates is incompatible with the principle of no arbitrage unless agents actually anticipate that devaluation may take place with positive probability. Second, the spread between domestic and foreign interest rates reached a peak in 1990/1991 while the sharp increase in foreign borrowing took place only in the mid-90’s after the magnitude of the spread had already gone down. Third, most of the financial liberalization in the region took place before 1990 so that the local firms could have already taken advantage of the large spreads in the early 90’s4. Finally, the evolution of ratings of Asian credit risk during the 90’s suggests little change in the way the risk of Asian borrowers was perceived on international financial markets5 In short, there is little evidence of any change in fundamentals large enough to motivate a surge in unhedged foreign borrowing in Asia during the mid-90’s. A second, more appealing, explanation for the exposure to currency risk of Asian firms relies on moral hazard considerations, according to which banks or corporations borrowed externally on the basis of implicit or explicit guarantees by the state. This argument was first brought forward by Krugman (1998). However, at the same time, Radelet and Sachs (1998) pointed out several weaknesses of the moral hazard story: first, one would expect banks to be more often protected by bailout promises than non-bank corporations. Yet, we observe that the bulk of foreign lending in Asia was aimed at non-bank corporations 6 (with the exception of South Korea). Second, even the evidence among non-bank corporations is inconsistent with the moral hazard story as many firms that were too small or unrelated to the government to expect ex-post bail out were able to and did obtain foreign financing prior to the crisis. Finally, from an ex-post point of view, we observe that many of the firms that borrowed abroad have already gone into bankruptcy or are currently facing bankruptcy. We suggest in this paper that the surge in the exposure to currency risk by Asian corporations may have been the equilibrium outcome of strategic interactions between firms and the central bank even in the absence of explicit or implicit promises of bailout. We support this suggestion through a model where firms’ financing choices and the central bank’s decision to defend the currency are jointly determined. Firms decide upon the optimal amount of foreign financing for a fixed size investment project depending on their expectations of the risk of devaluation. We assume that the cost of financing born by local firms enters negatively into the central bank’s objective function. The central bank decides whether it should devalue by comparing the cost of raising domestic interest rates to defend the currency with the cost of devaluation. If firms expect that the central bank has strong incentives to defend the currency, they use mostly foreign borrowing, thus raising the cost of devaluation to the central bank and the incentives for other firms to use foreign borrowing. An implication of this strategic interaction is the possibility of multiple equilibria, characterized by vastly different financing decisions of firms. A jump from a low to a high foreign borrowing equilibrium may be caused by even a small change in fundamentals. The impact on foreign borrowing will be especially large when strategic firm managers realize the role their financing decisions play in the central bank’s decision-making and collude to influence central bank policy to their advantage. Hence, high foreign borrowing may have arisen in South East Asia because of a slightly different moral hazard problem from the one emphasized in the existing literature, i.e. because of firms trying to change the ex-ante incentives of the central bank to devalue rather than expecting “ex-post” bail-out in case of a devaluation. Note that our explanation of the high level of foreign currency denominated debt incurred by the South East Asian firms does not rely on firms systematically under-estimating the risk of devaluation. On the contrary, the decisions of the firms are motivated by a full understanding of the trade-offs faced by the central bank. It is widely recognized that, in the recent Asian crisis as well as in other crises, central banks were confronted with a difficult policy dilemma (see for instance Corsetti et al., 1998 or Dornbusch, 1998). On the one hand, the weak economy and the collapse of the real estate and asset prices, which weakened the banking sector’s portfolio, made raising interest rates very costly. On the other hand, keeping interest rates low at the cost of a devaluation of the local currency would cause the bankruptcy of the many firms that had incurred significant foreign debt. In the case of Thailand, the central bank committed substantial reserves to the defense of the currency and raised the interest rate before finally giving up the peg. In this paper, we postulate, first, that central banks were, as far as possible, trying to minimize the losses incurred by local firms borrowing on foreign capital markets and, second, that these agents perfectly understood their importance for the local authorities. We are not alone in viewing the high level of foreign borrowing in Asia as an equilibrium phenomenon. Many authors, such as Burnside et al. (2001) or Gande et al. (2000) argue, as we do, that Asian institutions voluntarily exposed themselves to currency risk as an optimal response to incentives provided by the central bank. However, our analysis differs from theirs in two dimensions. First and most importantly, they suggest that the incentives provided by the central bank took the form of ex-post bailout guarantees, which is subject to the Radelet and Sachs (1998) critiques mentioned earlier. Second, while their model can explain a high level of exposure of Asian firms to currency risk prior to the crisis, it cannot explain why that level changed so dramatically between 1992 and 1996 despite little change in the economic, financial and political environment in Asia during that period. Two other much related papers are Bris and Koskinen, 2002 and Bris et al., 2001. Their work and ours share the same motivation in that we both study the interaction between central bank policy and corporate financial policies. However, they focus on corporate leverage and risk-taking behavior of exporting firms while we focus on the composition of firms borrowing and on multiple equilibria, which do not arise in their set-up. The rest of the paper is organized as follows: Section 2 describes the basic model. Section 3 solves for the equilibrium with competitive firms. Section 4 analyzes the impact of the presence of strategic agents on the equilibrium. Section 5 concludes.

In this paper, we develop a simple model of the choice by a firm between foreign and domestic financing in the presence of a risk of devaluation. We showed the existence of a strategic complementarity between financing decisions of firms and central bank policy. We showed that, as a consequence, a small temporary shock to fundamentals may have a large permanent impact on the equilibrium composition of firms borrowing. Finally, we argued that strategic behavior and collusion on the part of firms’ managers could exacerbate even further the impact of small changes in fundamentals on foreign borrowing. We conclude from these results that the large amount of foreign borrowing observed in South East Asia may be due to an unusual moral hazard problem different from the one emphasized in the existing literature, i.e. to firms trying to change the ex-ante incentives of the central bank to devalue rather than to firms expecting ex-post bail-out in case of a devaluation. A final remark though is that the two types of moral hazard problems are not mutually exclusive. Even if guarantees of ex-post bail-out cannot explain the bulk of the surge in foreign borrowing in Asia in the mid 90’s, they certainly played a role at least as far as the banking sector is concerned. Hence, we view the argument of this paper as complementary to those of the existing literature.