Structural vulnerability and resilience to currency crisis: Foreign currency debt versus export

North American Journal of Economics and Finance 42 (2017) 132–143

Contents lists available at ScienceDirect

North American Journal of Economics and Finance j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / e c o fi n

Structural vulnerability and resilience to currency crisis: Foreign currency debt versus export Ryota Nakatani International Monetary Fund, 700 19th Street, NW, Washington, D.C. 20431, USA

a r t i c l e

i n f o

Article history: Received 6 August 2014 Received in revised form 14 July 2017 Accepted 17 July 2017

Keywords: Currency crisis Exports Foreign currency debt Monetary policy Elasticity Structural vulnerability

a b s t r a c t Is there any factor that is not analyzed in the literature but is important for preventing currency crises? I argue that exports are an important factor to prevent currency crises that has not been frequently analyzed in the existing theoretical literature. Using the third generation model of currency crises, I derive a simple and intuitive formula that captures an economy’s structural vulnerability characterized by the elasticity of exports and repayments for foreign currency denominated debt. I graphically show that the possibility of currency crisis equilibrium depends on this structural vulnerability and also analyze how this vulnerability impacts the effectiveness of monetary policy response. Ó 2017 Elsevier Inc. All rights reserved.

1. Introduction The literature on currency crises has analyzed causes and mechanisms of how the crises occur and what happens when countries experience the crises. Little theoretical literature has focused on factors that prevent currency crises other than policy responses. Is there any factor that is not analyzed in the literature but is important for preventing currency crises? To answer this question, in this paper, I discuss and show that exports are a potentially important factor for the prevention of a currency crisis. I introduce exports into the third generation models of currency crises and derive an intuitive and important formula that captures the structural vulnerability of the economy. I also conduct graphical equilibrium analysis to explore the roles of exports in currency crises and investigate how this structural vulnerability influences the effectiveness of monetary policy response. The organizational structure of this paper includes a review of the literature, in which I briefly summarize and discuss the relevant literature on currency crises models and the specific model that I chose for my analysis and compelling reasons for doing so. I then discuss the importance of exports as addressed in the empirical literature. Furthermore, I develop a theoretical model to analyze the role of exports graphically. 2. Literature and motivation 2.1. Three generations of currency crisis models A currency crisis can be defined as a sudden devaluation of a currency that often ends in a speculative attack in the foreign exchange market. There have been three ‘generations’ of models of currency crises (Glick & Hutchison, 2013). The first E-mail address: [email protected] http://dx.doi.org/10.1016/j.najef.2017.07.009 1062-9408/Ó 2017 Elsevier Inc. All rights reserved.

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

133

generation models focus on inconsistencies between domestic macroeconomic policies, such as a fixed exchange rate regime and a persistent government budget deficit that eventually must be monetized (Flood & Garber, 1984; Krugman, 1979). These models describe a government that attempts to maintain a pegged exchange rate system, but is subject to a constant loss of international reserves, due to the need to monetize government budget deficits. These two characteristics of the policy are inconsistent with each other, and provoke an eventual speculative attack on the reserves of the central bank. Thus, in the first generation models, the key factor is the government activity, and the models predict that the fixed exchange rate regime must collapse. In second generation models, policymakers weigh the cost and benefits of defending the currency and may abandon an exchange rate target. In these models, the government maximizes an explicit objective function, i.e., Obstfeld (1996) developed models in which the central bank minimizes a quadratic loss function that depends on inflation and on the deviation of output from its natural rate. In these models, an interaction between investors’ expectations and actual policy outcomes can lead to self-fulfilling crises. The third generation models focus on how distortions in financial markets and banking systems can lead to currency crises. The basic idea is that banks and firms in emerging countries have currency mismatches on their balance sheets since they borrow in foreign currency and lend or invest in local currency. Aghion, Bacchetta, and Banerjee (2000, 2001) and Nakatani (2016) analyzed the effects of credit constraints on currency crises by focusing on private foreign currency denominated debt. They explored how problems in the financial markets interact with currency crises, and how crises can have real effects on the economy. Another type of third generation model was developed as an application of the Diamond and Dybvig (1983) model by Chang and Velasco (2001), who focused on how distortions in the banking system can lead to currency crises, and Caballero and Krishnamurthy (2001), who developed a model in which firms finance risky long-term projects with short-term domestic and foreign debts and face a liquidity problem caused by uncertainty about future production and limited amounts of internationally accepted collateral. The other type of third generation models is based on the idea that government guarantees to the banking system can generate moral hazard problems that lead to crises (Burnside, Eichenbaum, & Rebelo, 2001; Corsetti, Pesenti, & Roubini, 1999; Dekle & Kletzer, 2002; Dooley, 2000; McKinnon & Pill, 1996; Schneider & Tornell, 2004). 2.2. Aghion-Bacchetta-Banerjee model In this paper, I extend the third generation model focusing on credit constraints of firms that was originally developed by Aghion et al. (2000, 2001). There are several reasons to use this model. As shown in Aghion, Bacchetta, and Banerjee (2001), their model (the ABB model) can include the features of the first and the second generation models. In addition, the possibility of multiple equilibria can also be included. Moreover, by using this model and introducing exports, we can describe the tradeoff between the costs and benefits of large currency depreciation for firms in the open economy. Furthermore, with this type of model, we can have short-run nominal rigidity, which is supported by empirical evidence, and see how financial friction can cause currency crises. Finally, during the recent global financial crisis in 2008-09, central bankers were concerned with the possibility of currency crises in some countries. Those concerns were strong especially for emerging European countries that had high foreign debt in their economies. This ABB model illustrates the situation of those emerging countries accurately. Nakatani (2017) empirically applied the ABB model and found that both productivity shocks and country risk premium shocks affect exchange rates. A recent extension of the ABB model was developed by Bergman and Jellingsø (2010), who examined the medium-term effects of interest rate defense in the ABB model. Their finding was that even though an interest rate hike is successful in preventing a currency crisis in the short-term, it may cause a currency crisis in the medium-term. This occurs because the first-period interest rate hike results in lower inflation in the medium-term, which in turn raises the real interest rate and thus increases the burden of domestic debt. In this paper, I focus on equilibrium in the short-term to obtain clear theoretical and policy implications, because disregarding the medium- and long-term equilibrium effects is not a potential problem, given that it is highly likely that exports do not affect output beyond the short-term. Another extension of the ABB model was developed by Miller, García-Fronti, and Zhang (2006). They introduced demand factors into the ABB model, including exports. This introduction of exports into the model is justified because empirical research found that exports have a key role in preventing and recovering from currency crises, as discussed in greater detail in the next section. However, the problem in their paper is that exports were assumed to be an exogenous constant variable. In other words, exports do not respond to exchange rates in their model. This assumption contradicts the fact that currency depreciation improves price competitiveness in the export sector and leads to an increase in exports. Therefore, in my analysis, I assume that an export is a function of exchange rates, which is a more practically reasonable assumption, and study how this introduction of endogenous exports into the ABB model can change the policy implications. 2.3. Importance of exports Theoretical models of currency crises have focused on the vulnerabilities of external exposures in order to identify causes of the crises. The first and second generation models have focused on the government budget deficit that can be supported by the capital inflows of foreign investors. The third generation models have focused on the foreign currency denominated debt of private firms, external funding of commercial banks and moral hazard problems triggered by the government guarantees.

134

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

However, there are few studies in the literature that examine the opposite side of the story. What is the factor that benefits from a large depreciation of currency and deters currency crises? Empirical studies have found that exports are a potentially important factor from this perspective.1 The relevance of exports has been considered mainly in empirical analysis characterized by few linkages with theoretical models. Therefore, this paper analyzes the role of exports during currency crises from the theoretical point of view.2 Most empirical papers present evidence of the importance of exports in the context of currency crises.3 Gupta, Mishra, and Sahay (2007) analyzed the behavior of output during currency crises and found that growth of exports and trade openness are statistically positively associated with output growth. Deb (2006) found that a faster export growth rate is a key factor for recovery from currency crises. Eijffinger and Goderis (2008) found evidence that higher exports appreciate the exchange rate. Desai, Foley, and Forbes (2008) used firm-level data to analyze the response of multinational and local firms to currency crises. They found that U.S. multinational affiliates increased sales, assets and investment significantly more than local firms during and after currency depreciations. The results suggest that multinational affiliates expanded economic activity during currency crises when most local firms were financially constrained. Using country-level panel data, Cavallo and Frankel (2008) found robust empirical evidence that economies that trade more with other countries are less vulnerable to sudden stops and currency crises. Bleakley and Cowan (2008) studied over 450 firms in Latin American countries and found that the negative balance sheet effects of a depreciation on firms holding dollar debt are more than offset by higher earnings caused by the competitiveness effect (i.e., increased exports) of depreciation. Next, I provide stylized facts showing the importance of an export sector as a blockade against a currency crisis. If a country has a large share in an export sector, the economy benefits from currency depreciation by increasing competitiveness in that sector. Because foreign investors know this mechanism well, speculative attacks are less likely to succeed in a country with a large export sector. Since the ABB model was developed to explain the Asian currency crisis, I use data from this event. First, I construct an exchange market pressure index (EMPI) to measure the severity of the currency crisis. Using data from the International Financial Statistics published by the International Monetary Fund, the EMPI was calculated as a weighted average of annual percentage of exchange rate depreciation and percentage international reserve losses, with weights such that the two components equal the sample volatility (Kaminsky & Reinhart, 1999). Thus, a high value of EMPI indicates that the country experienced severe pressure on its currency and/or international reserves. Second, I use the ratio of exports to GDP as an indicator of export share. Data on export share were taken from the World Bank’s World Development Indicators. In Chart 1, I show the relationship between the size of the export sector in the economy and the EMPI during the Asian currency crisis in 1997. Because exports can be affected contagiously by the decline in world trade volume during a financial crisis, each country’s export share one year before the Asian crisis, i.e., 1996, is shown in the chart to indicate the exogenous effect of an export sector on a currency crisis. Chart 1 supports the above mentioned story. Specifically, there was a negative correlation between the export share of the economy and the severity of the Asian currency crisis. Countries with a large export sector did not tend to experience large exchange rate depreciation and/or large loss of international reserves. These results remained unchanged in an analysis of this relationship between export share and the severity of the Asian crisis using the sample of Southeast Asian countries (instead of all Asian countries). Despite this empirical evidence and the stylized facts concerning the importance of exports, the export sector is not analyzed frequently in theoretical currency crises models.4 We introduce exports into the third generation model developed by the Aghion et al. (2001), in which firms have foreign currency denominated debt. A compelling reason to use this model, in addition to the advantages described in detail in the previous section, is as follows. Since the main cause of currency crises is the foreign currency debt of firms in the ABB model, we can easily see the tradeoff of a large depreciation of currency for the firms by introducing exports into the model. In other words, currency depreciation has both positive and negative effects on a firm’s retained earnings because it increases sales to overseas countries by stimulating exports, whereas it reduces the cash flow of firms by increasing the burden of foreign currency denominated debt. Thus, the ABB model is the best model for the analysis of exports. 3. Model analysis I introduce exports into the ABB model in this section. Exports are among potentially important factors when the economy faces currency crises. In the original ABB paper, foreign currency denominated debt was the sole key factor in currency 1 Some early studies argued that trade openness reduces vulnerability to financial crises (see Frankel (2005) for a survey of literature). Sachs (1985) argued that Asian countries were less vulnerable to international debt crisis than Latin American countries because they had higher ratios of export to GDP. Guidotti, Sturzenegger, and Villar (2004) suggested that economies with higher trade openness recover fairly quickly from the output contraction of sudden stop. Calvo, Izquierdo, and Talvi (2004) suggested that Argentina’s low ratio of exports to GDP helps explain why the country suffered its worst sudden stop after 1999. Calvo, Izquierdo, and Mejía (2004) and Edwards (2004) found that trade openness is associated with fewer sudden stops and current account reversals. 2 Although depreciation of currency may induce consumers to move from foreign to domestic goods, I do not analyze the effect of imports. This is because the empirical literature has showed that exports are more important than imports for currency crises (Kaminsky, Lizondo, & Reinhart, 1998). 3 Another area of the literature has empirically compared the financial linkages and trade linkages of currency crises and studied the relative importance of those linkages. Eichengreen and Rose (1999), Glick and Rose (1999), Forbes (2002) and Haile and Pozo (2008) have found evidence to support the hypothesis that currency crises spread from one country to another because of trade linkages. In contrast, Kaminsky and Reinhart (2000), Van Rijckeghem and Weder (2001) and Caramazza, Ricci, and Salgado (2004) found that financial linkages play an important role in the propagation of currency crises. However, the purpose of this paper is to analyze theoretically the role of exports in the context of a country’s structural vulnerability to currency crises and not the contagion of crises. 4 One exception is Céspedes, Chang, and Velasco (2004). However, they assumed that the value of home exports is an exogenous constant variable.

135

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

crises. Therefore, in the original model depreciation of the domestic currency has only negative effects on the economy through deteriorated balance sheets of private firms. By contrast, in my model, depreciation of the exchange rate has both positive and negative effects on the real economy because it increases exports on one hand but reduces retained earnings via increased debt repayments for the foreign debt on the other hand. In this manner, a tradeoff can be observed between exports and foreign currency denominated debt under the circumstance of exchange rate depreciation. I derive a simple formula that states the condition for the occurrence of currency crises. 3.1. Wealth curve Assume that a representative firm produces one type of goods at the domestic price Pt . Those goods are sold to both domestic and overseas consumers. The output is produced using capital and the production function is written as

yt ¼ f ðkt Þ:

ð1Þ 5

We introduce exports into the firm’s profit function. Exports, whose prices are set in foreign currency , are assumed to be the function of real exchange rates and foreign demand.

xt ðEt Pt =Pt ; yt Þ

ð2Þ yt

where Et is the flexible nominal exchange rate (the price of foreign currency in terms of domestic currency) and is the foreign demand. For simplicity, it can be assumed that the price level of foreign countries (Pt ) is normalized to 1. Then the export can be written as a function of the real exchange rates in the following way.

xt ðEt =Pt ; yt Þ

ð3Þ

Note that exports are the increasing function of real exchange rate depreciation and foreign demand, where higher Et =Pt means real exchange rate depreciation. 

dxt =dðEt =P t Þ > 0; dxt =dyt > 0

ð4Þ

In order to concentrate on the role of exports, which is found in the empirical literature, we do not consider the effects of the real exchange rate on domestic consumption here but these effects are worth mention. In the presence of imported consumption goods, a depreciation of domestic currency will induce domestic consumers to move from foreign to domestic goods because of the import price inflation. Thus, a large currency depreciation will also increase the demand for domestic consumption goods. Notably, we don’t include imports in this analysis because the ABB model is a supply-side model, and thus, there is no straightforward way to add imports. The assumption about price rigidity is the same as that in Aghion et al. (2001). Purchasing Power Parity (PPP) is assumed to hold at the beginning of period 1. Following an unanticipated shock, there are deviations from PPP (P1 –E1 ) that are corrected in period 2 (P 2 ¼ E2 ). This shock may be real—such as a change in productivity—or it may be a shift in expectations. The timing of events can be summarized as follows. In period 1, the price P1 is preset and the firm invests. Then, an unanticipated shock occurs, corresponding to a realization of the nominal exchange rate E1 . The shock is accompanied by an adjustment in the monetary policy set by the central bank. Subsequently, output and profits in period 1 are generated and the firm’s debts are repaid. Finally, a fraction of net retained earnings after debt repayment is saved for investment in period 2. Assuming that the working capital kt fully depreciates within one period, the firm maximizes its real profit net of loan repayments

Pt ¼ f ðkt Þ  lt ð1 þ it1 ÞPt1 =Pt  lt ð1 þ i ÞEt =Pt

ð5Þ

 lt



where lt is an amount of domestic currency loan, is that of loan denominated in foreign currency, and it and i are interest rates on domestic and foreign currency loans, respectively. We assume that the interest rate on foreign currency loan is constant over time. We assume that whenever profit is positive, the firm retains a proportion ð1  aÞ of the profit and uses it to finance its future investment.6 Thus, the current retained earnings available for capital in the next period are7

W tþ1 ¼ ð1  aÞPt : Owing to the credit constraint, the firm can at most borrow an amount Lt ¼ lt þ

ð6Þ  lt

proportional to its current real wealth

Wt

Lt 6 lW t

5 6 7

ð7Þ

Cook and Devereux (2006) presented evidence of foreign currency pricing of exports in Asian countries. The firm will always save a constant fraction of the profits under the assumption of the logarithmic preference (Aghion, Bacchetta, & Banerjee, 2004). I assume that the marginal product of capital exceeds domestic and foreign interest rates so that constraint (6) is binding.

136

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

Chart 1. Roles of Exports in the Asian Currency Crisis.

where l is a credit multiplier. The rationale for the constant credit multiplier is derived from moral hazard considerations (Aghion, Banerjee, & Piketty, 1999).8 We assume that this credit constraint is binding. Since capital fully depreciates within one period, investment in the current period equals the capital in the next period. Under the credit constraint, this equation of motion for capital can be written as follows:

ktþ1 ¼ ð1 þ lÞð1  aÞPt :

ð8Þ

Then, the output is characterized by the production function

ytþ1 ¼ f ðktþ1 Þ ¼ f ðð1 þ lÞð1  aÞPt Þ:

ð9Þ

The equilibrium condition in the goods market suggests that the sum of domestic (ct ) and export sales equals total production in the economy.9

ct þ xt ðEt =Pt ; yt ÞEt =Pt ¼ f ðkt Þ

ð10Þ

Using this condition, the output in period 2 can be written as



  E1 E1  P0  E1  : y2 ¼ f ð1 þ lÞð1  aÞ c1 þ xt ; y1  ð1 þ i0 Þ l1  ð1 þ i Þ l1 P1 P1 P1 P1

ð11Þ

This is called the ‘‘Wealth curve”; it illustrates the relationship between the exchange rate and future output as originally analyzed by Aghion et al. (2001).10 This curve is characterized by the credit multiplier effect times the wealth of the firm (i.e., savings), where the wealth is calculated as the saving rate times the firm’s profit. The profit is defined as the sum of domestic and export sales minus the sum of domestic and foreign currency denominated debt repayments.11 Taking a derivative of the exchange rate with respect to output allows the slope of the Wealth curve to be obtained.

 dE1  P1 1 n o ¼ R0 dy2 y2 >0 ð1 þ lÞð1  aÞ x1 þ E1 x1E1 =P1  ð1 þ i Þl f k2 1 P1

8

ð12Þ

See Aghion et al. (2001) for the case in which the credit multiplier depends upon real or nominal interest rates. See the second example in Appendix for the model in which the firm determines consumption.   The overall balance of payments account always balances, i.e., ð1 þ i Þl1 ¼ x1 þ l2 . We assume that the country starts out with large foreign liabilities. 11 Although we can think of export credit as another channel that constrains production, this constraint can be eliminated by export credit insurance (Auboin & Engemann, 2013). 9

10

137

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

Using elasticity of export goods with respect to real exchange rate f ¼

E =P 1 1 =P1 1

x1E

x1

, the slope of the Wealth curve can be sim-

plified to

 dE1  P1 1 ¼ R0:  dy2 y2 >0 ð1 þ lÞð1  aÞfð1 þ fÞx1  ð1 þ i Þl1 g f k2

ð13Þ

This equation suggests that the term in the curly brackets in the denominator determines the slope of the Wealth  curve. The condition suggests that if the foreign currency denominated debt (l1 ) is large, the Wealth curve is downward sloping. 



ð1 þ fÞx1 < ð1 þ i Þl1 )

 dE1  <0 dy2 y2 >0

ð14Þ

By contrast, if an export (x1 ) is large or an elasticity of export demand (f) is large, the Wealth curve is upward sloping. 



ð1 þ fÞx1 > ð1 þ i Þl1 !

 dE1  >0 dy2 y2 >0

ð15Þ

The intuition behind this result is straightforward. When the economy has a large foreign currency denominated debt, the real exchange rate depreciation increases the debt burden for the firm and this reduces investment, resulting in lower output in the second period. In contrast, if the economy has a large export industry, a currency depreciation increases the sales and profits of exporting firms and their investments. Therefore, output in the next period increases. In Appendix, two examples of an export function are analyzed. 3.2. Interest-Parity-LM (IPLM) curve The other curve that is analyzed in the ABB model is the IPLM curve, which is mainly determined by the policy of the central bank. The IPLM curve is the same as in the original model, and it is derived on the basis of the assumption about PPP, the interest parity condition and the money market equilibrium. The IPLM curve provides the relationship between E1 and y2 : 

E1 ¼

1þi M S2 D 1 þ i1 m ðy2 ; i2 Þ

ð16Þ

where M S2 is the nominal money supply in period 2 and a real money demand mD ðy2 ; i2 Þ is a standard function of output and the interest rate. Bergman and Jellingsø (2010) showed that the first-order derivative of the IPLM curve is 

dE1 1þi M S2 ¼ mD < 0: dy2 1 þ i1 ½mD ðy2 ; i2 Þ2 y2

ð17Þ

Thus, the slope of the IPLM curve is negative. An increase in future output raises the future demand for domestic real money balances, thus resulting in a future appreciation of the domestic currency. This anticipation of a future appreciation increases the attractiveness of holding the currency today, leading to an appreciation of the exchange rate. We can easily see that the slope of the IPLM curve is steep when the money supply is large and the domestic interest rate is low.

E1

E1 Wealth Curve

Wealth Curve

A IPLM Curve

y2

IPLM Curve

C

y2

Fig. 1. Role of Exports in the Context of Currency Crises: Before the Shock. 1.1: Large Export. 1.2: Large Foreign Debt.

138

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

Fig. 2. Role of Exports in the Context of Currency Crises: After a Negative Productivity Shock or a Tightening of the Credit Market. 2.1: Large Export. 2.2: Large Foreign Debt.

3.3. Graphical equilibrium analysis The equilibrium of the model is defined by the intersection of the IPLM and the Wealth curves. The first three figures show the relationship between exports and the occurrence of currency crises. Here, I use an arctan type of export function (see Example 1 in Appendix) to draw the Wealth curve only to determine the shape of the Wealth curve, although the results do not depend on the type of export function. In other words, the only necessary condition to derive the same result here is that a depreciation of exchange rate increases the revenue from exports measured in domestic currency.   Fig. 1.1 shows the case where an effect from export is greater than the repayments for foreign debt, ð1 þ fÞx1 > ð1 þ i Þl1 . When a negative productivity shock (a shift in the f ðÞ function) or a tightening of the credit market (a shift in l) occurs, the Wealth curve shifts to the left (Fig. 2.1). In this case, the exchange rate depreciates, but the country can avoid a crisis equilibrium if it starts from a good equilibrium ‘‘A” and the size of the negative productivity shock is not significant, thus resulting in a new equilibrium ‘‘B”. This occurs because earnings from the export sector are so large that they can offset the negative effect that comes from the credit constraint of the foreign currency denominated debt. However, if the negative shock is so large that the firms decide not to produce goods in the second period, because they cannot earn positive profits as a result of a large negative productivity shock or a tremendous tightening of the credit constraint, then there may be a crisis equilibrium shown as ‘‘B” in Fig. 2.1. Note that in the case of a positive slope of the Wealth curve, there is no possibility of multiple equilibria. By contrast, Fig. 1.2 illustrates the economy in which the effects from the foreign currency denominated debt are greater   than those from the export, ð1 þ fÞx1 < ð1 þ i Þl1 ; hence, the slope of the Wealth curve is negative. Note that the Wealth curve includes an upward segment of the vertical axis because the firm produces nothing when profit is negative because of a huge foreign debt repayment caused by the large currency depreciation. In this case, multiple equilibria are possible under the shock (Fig. 2.2). Thus, we can see the tradeoff between the benefits of currency depreciation and detriments from foreign debt using a simple formula characterized by the elasticity of exports and repayments for foreign debt. Next, we analyze the case in which the economy is hit by an unanticipated expectational shock in the financial markets (Fig. 3). In this case, the IPLM curve can be written as 

E1 ¼

1þi M S2 þg D 1 þ i1 m ðy2 ; i2 Þ

ð18Þ

where g is the foreign exchange risk premium after the shock. This increase in risk shifts the IPLM curve upward. Starting from a good equilibrium ‘‘A” in Fig. 3.1, currency depreciation will increase the output via boosted exports to the new equilibrium ‘‘G” when the economy has a large export sector. In contrast, as shown in Fig. 3.2, if the effects from foreign currency debt dominate, starting from a good equilibrium ‘‘C”, this upward shift in the IPLM curve again leads to a multiple equilibria situation that contains currency crisis equilibrium ‘‘J”. Note that this possibility of a currency crisis is reinforced by the fact that an increase in the foreign exchange premium raises the interest rate on foreign borrowing, which in turn will move the Wealth curve downward. Next, I analyze the appropriate monetary policy to avoid a currency crisis. The monetary policy, i.e., the interest rate policy, affects both the IPLM and Wealth curves. The effect of interest rate it on the IPLM curve is straightforward. As Eq. (16) suggests, an increase in the policy interest rate is associated with appreciation of domestic currency by shifting the IPLM curve downward, owing to the interest parity condition.

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

139

Fig. 3. Role of Exports in the Context of Currency Crises: After an Expectational Shock. 3.1: Large Export. 3.2: Large Foreign Debt.

E1

Wealth Curve

IPLM Curve

y2 Fig. 4. Monetary Policy Response for the Wealth Curve with Positive Slope.

In contrast, the effects of interest rate policy on the Wealth curve are ambiguous, as Eq. (11) suggests. There are two channels of interest rate policy on the Wealth curve.12 The first channel is the export channel. A real depreciation of domestic currency, induced by lowering the policy interest rate, improves the competitiveness of the export sector and boosts exports, thus shifting the Wealth curve to the right. The second channel is the valuation channel, i.e., the changes in foreign currency debt due to exchange rate changes. If the exchange rate depreciates as a result of monetary loosening, the amount of foreign currency denominated debt increases when evaluated in domestic currency, thereby shifting the Wealth curve to the right. These two channels offset each other’s effects on the Wealth curve due to interest rate policy. In the following graphical analysis, I compare an economy with a large export sector to an economy with a large foreign currency debt. The first case is an economy with a large export sector compared with the foreign currency debt (Fig. 4). In this case, an appropriate monetary policy response is to lower the policy interest rate, because a depreciation of the real exchange rate will boost exports and increase output. Although the nominal exchange rate will depreciate, the monetary authority can avoid a crisis equilibrium with zero output by shifting the IPLM curve upward to attain a new equilibrium with positive output.

12 In general, there is another channel called the debt service cost channel. An increase in the interest rate has adverse macroeconomic effects by raising the interest burden on a firm’s debt, by shifting the Wealth curve to the left. This is not the case in my model, because the interest burden on domestic debt is predetermined at the beginning of the first period, and it cannot be affected by the monetary policy during a currency crisis (i.e., in the first period).

140

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

Fig. 5. Role of Structural Vulnerability on the Effectiveness of Monetary Policy under a Negative Productivity Shock or a Tightening of the Credit Market. 5.1: Steep Wealth Curve. 5.2: Gradual Wealth Curve.

Then, we analyze the second case, in which the effects of foreign currency debt are larger than those of exports. Fig. 5.1 is       the case when ð1 þ fÞx1 < ð1 þ i Þl1 and ð1 þ i Þl1  ð1 þ fÞx1 is large. Fig. 5.2 is the case when ð1 þ fÞx1 < ð1 þ i Þl1 and   ð1 þ i Þl1  ð1 þ fÞx1 is small. Given the IPLM curve, the probability of facing currency crises is higher in the case of Fig. 5.1 than in the case of Fig. 5.2. Put differently, the monetary authority must raise the interest rate to a large extent so as to avoid a crisis.13 Similar arguments hold for the cases when expectational shock occurs in the financial market (Fig. 6). Fig. 6.1 shows that, compared with the gradual Wealth curve depicted in Fig. 6.2, a small shift in risk perception can increase the possibility of a currency crisis when the economy has large foreign currency debt and the Wealth curve is steep. Thus, the greater the foreign currency debt, the more aggressively the monetary authority must raise the interest rate. 4. Conclusion In this paper, I surveyed the three generations of currency crisis models and argued that exports are important factors during currency crises that have not been analyzed frequently in the literature. Thus, I introduced exports into the third generation model that was developed by Aghion et al. (2001). In the original ABB paper, foreign currency denominated debt was the sole key factor in the occurrence of currency crises. Namely, in the original model, depreciation of the domestic currency induced by an unanticipated shock has only negative effects on the economy through deteriorated balance sheets of private firms. The introduction of exports into the model suggests that depreciation of the exchange rate has both positive and negative effects on the real economy because it increases exports, on the one hand, but reduces retained earnings via increased debt repayments of the foreign debt, on the other hand. In this manner, the tradeoff between exports and foreign currency denominated debt can be analyzed under the circumstance of exchange rate depreciation in my model. I showed that graphical explanations with the Wealth curve and the IPLM curve are helpful to see this tradeoff. I derived a simple and intuitive formula that determines the slope of the Wealth curve when firms are exporting to foreign countries. In that formula, I found that the elasticity of exports plays a crucial role in the context of currency crises. I also showed that structural vulnerability is important for the prevention of currency crises and the effectiveness of a monetary policy response. Acknowledgements I am grateful for helpful discussions and comments from the two anonymous reviewers, Philippe Aghion, Michael M. Hutchison, Carl E. Walsh, Kenneth M. Kletzer, Michael P. Dooley, Nirvikar Singh, Johanna L. Francis, Alan C. Spearot and seminar participants at the Development Economics Seminar at the University of California, Santa Cruz. I particularly appreciate very much the support that I received from the following people who carefully read my draft and provided helpful comments. I very much appreciate my advisor, Michael M. Hutchison, for guidance on my research and his sound advice. I also 13

Nakatani (2014) has performed some empirical analyses based on this argument.

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

141

Fig. 6. Role of Structural Vulnerability on the Effectiveness of Monetary Policy under an Expectational Shock. 6.1: Steep Wealth Curve. 6.2: Gradual Wealth Curve.

express special thanks to Carl E. Walsh for careful reading of my draft and providing detailed comments on this theoretical model. I express my heartfelt appreciation to Kenneth M. Kletzer for commenting on and proofreading this research as well as providing advice for publication. Lastly, I express my sincere appreciation to Johanna L. Francis for providing comments on an earlier version of this paper. The views expressed in this article are those of the author and do not reflect those of the institutions he belongs to. Appendix A. Export and consumption functions A.1. Example 1: An arctan type of export function If we assume a constant level of foreign demand and an arctan type of export function (Fig. A1), which has lower and upper limits on the volume of exports, and also assume that domestic consumption is not affected by the real exchange rates and the amount of foreign debt is negligible, then the Wealth curve would be similar to the curve depicted in Fig. A2. A.2. Example 2: Export and consumption functions based on microeconomic foundation To derive an export function explicitly, the firm’s behavior must be modeled on the basis of the microeconomic foundation. The firm’s profit maximization problem can be set as

Pt ¼ ct þ



 Et Pt1 Et   Et  xt  ð1 þ it1 Þ lt  ð1 þ i Þ lt  U ct þ xt ; yt Pt Pt Pt Pt

ðA:1Þ

x1

x Export function

x E1 P1 Fig. A1. An Arctan Type of Export Function.

142

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

E1

Wealth Curve

y

y2

Fig. A2. Wealth Curve.

where Uðct þ xt Þ is a cost function that includes various costs such as operating and/or fixed set-up costs.14 The firm maximizes its profit subject to an export demand that is a function of real exchange rates and foreign demand xt ¼ xd ðEt =Pt ; yt Þ.15 The first order conditions for the firm’s profit maximization problem yield the following supply conditions for domestic sales and exports, which in turn determine domestic consumption and export volumes.

1¼

   @ U ct þ xd PEtt ; yt @ct

   d Et  Et @ U ct þ x Pt ; yt ¼ @xt Pt

ðA:2Þ

ðA:3Þ

At an optimum, the firm equates the marginal revenue from the domestic consumption, unity, to the marginal cost of producing the product. The same optimum condition holds for exports. In general, note that the function U captures the complementarity (or substitutability) of production between exports and domestic consumption. If we assume negative cost complementarities between export and domestic sales, @c@t U@xr < 0, a depreciation of real exchange rates increases exports, which in turn reduces marginal costs of production for domestic sales and hence also results in an increase in production for domestic consumption. References Aghion, P., Bacchetta, P., & Banerjee, A. (2000). A simple model of monetary policy and currency crises. European Economic Review, 44, 728–738. Aghion, P., Bacchetta, P., & Banerjee, A. (2001). Currency crises and monetary policy in an economy with credit constraints. European Economic Review, 45, 1121–1150. Aghion, P., Bacchetta, P., & Banerjee, A. (2004). A corporate balance-sheet approach to currency crises. Journal of Economic Theory, 119, 6–30. Aghion, P., Banerjee, P., & Piketty, T. (1999). Dualism and macroeconomic volatility. Quarterly Journal of Economics, 114, 1359–1397. Auboin, M., & Engemann, M. (2013). Trade finance in periods of crisis: What have we learned in recent years? WTO Staff Working Paper ERSD-2013-01. Bergman, U. M., & Jellingsø, M. (2010). Monetary policy during speculative attacks: Are there adverse medium term effects? The North American Journal of Economics and Finance, 21, 5–18. Bleakley, H., & Cowan, K. (2008). Corporate dollar debt and depreciations: Much ado about nothing? Review of Economics and Statistics, 90, 612–626. Burnside, C., Eichenbaum, M., & Rebelo, S. (2001). Hedging and financial fragility in fixed exchange rate regimes. European Economic Review, 45, 1151–1193. Caballero, R. J., & Krishnamurthy, A. (2001). International and domestic collateral constraints in a model of emerging market crises. Journal of Monetary Economics, 48, 513–548. Calvo, G. A., Izquierdo, A., & Mejía, L. (2004). On the empirics of sudden stops: The relevance of balance-sheet effects. Proceedings, Federal Reserve Bank of San Francisco, issue Jun. Calvo, G. A., Izquierdo, A., & Talvi, E. (2004). Sudden stops, the real exchange rate, and fiscal sustainability: Argentina’s lessons. In V. Alexander, J. Mélitz, & G. M. von Furstenberg (Eds.), Monetary unions and hard pegs: Effects on trade, financial development and stability (pp. 151–182). Oxford: Oxford University Press. Caramazza, F., Ricci, L., & Salgado, R. (2004). International financial contagion in currency crises. Journal of International Money and Finance, 23, 51–70. Cavallo, E. A., & Frankel, J. A. (2008). Does openness to trade make countries more vulnerable to sudden stops, or less? Using gravity to establish causality. Journal of International Money and Finance, 27, 1430–1452. Céspedes, L. F., Chang, R., & Velasco, A. (2004). Balance sheets and exchange rate policy. American Economic Review, 94, 1183–1193. Chang, R., & Velasco, A. (2001). A model of financial crises in emerging markets. Quarterly Journal of Economics, 116, 489–517. Cook, D., & Devereux, M. B. (2006). External currency pricing and the East Asian crisis. Journal of International Economics, 69, 37–63. Corsetti, G., Pesenti, P., & Roubini, N. (1999). Paper tigers? A model of the Asian crisis. European Economic Review, 43, 1211–1236. Deb, S. (2006). Trade first and trade fast: A duration analysis of recovery from currency crises. Departmental Working Papers 2006–07, Department of Economics, Rutgers University. 14

For example, Nguyen and Schaur (2010) analyzed the case in which the quadratic cost function can be written as

Uðct þ xt Þ ¼ H þ K þ bðct þ xt Þ þ 12 ðct þ xt Þ2 where H is the fixed cost of production, K is the sunk entry cost of exporting and b is the firm’s idiosyncratic marginal cost of production. 15 While it has not been analyzed in this article, this model should be analyzed further by incorporating imports into the model.

R. Nakatani / North American Journal of Economics and Finance 42 (2017) 132–143

143

Dekle, R., & Kletzer, K. M. (2002). Domestic bank regulation and financial crises: Theory and empirical evidence from East Asia. In S. Edwards & J. A. Frankel (Eds.), Preventing currency crises in emerging markets (pp. 507–552). Chicago: University of Chicago Press. Desai, M. A., Foley, C. F., & Forbes, K. J. (2008). Financial constraints and growth: Multinational and local firm responses to currency depreciations. Review of Financial Studies, 21, 2857–2888. Diamond, D. W., & Dybvig, P. H. (1983). Bank runs, deposit insurance, and liquidity. Journal of Political Economy, 91, 401–419. Dooley, M. P. (2000). A model of crises in emerging markets. Economic Journal, 110, 256–272. Edwards, S. (2004). Thirty years of current account imbalances, current account reversals, and sudden stops. IMF Staff Papers, 51, 1–49. Eichengreen, B., & Rose, A. K. (1999). Contagious currency crisis: Channels of conveyance. In T. Ito & A. O. Krueger (Eds.), Changes in exchange rates in rapidly development countries: Theory, practice, and policy issues (pp. 29–51). Chicago: University of Chicago Press. Eijffinger, S. C. W., & Goderis, B. (2008). The effect of monetary policy on exchange rates during currency crises: The role of debt, institutions, and financial openness. Review of International Economics, 16, 559–575. Flood, R. P., & Garber, P. M. (1984). Collapsing exchange-rate regimes: Some linear examples. Journal of International Economics, 17, 1–13. Forbes, K. J. (2002). Are trade linkages important determinants of country vulnerability to crises? In S. Edwards & J. A. Frankel (Eds.), Preventing currency crises in emerging markets (pp. 77–124). Chicago: University of Chicago Press. Frankel, J. A. (2005). Mundell-Fleming lecture: Contractionary currency crashes in developing countries. IMF Staff Papers, 52, 149–192. Glick, R., & Hutchison, M. M. (2013). Models of currency crises. In G. Caprio (Ed.), The evidence and impact of financial globalization (pp. 485–497). Oxford: Elsevier Inc. Glick, R., & Rose, A. K. (1999). Contagion and trade. Why are currency crises regional? Journal of International Money and Finance, 18, 603–617. Guidotti, P., Sturzenegger, F., & Villar, A. (2004). On the consequences of sudden stops. Economia, 4, 171–214. Gupta, P., Mishra, D., & Sahay, R. (2007). Behavior of output during currency crises. Journal of International Economics, 72, 428–450. Haile, F., & Pozo, S. (2008). Currency crisis contagion and the identification of transmission channels. International Review of Economics and Finance, 17, 572–588. Kaminsky, G. L., Lizondo, S., & Reinhart, C. M. (1998). Leading indicators of currency crises. IMF Staff Papers, 45, 1–48. Kaminsky, G. L., & Reinhart, C. M. (1999). The twin crises: The causes of banking and balance-of-payments problems. American Economic Review, 89, 473–500. Kaminsky, G. L., & Reinhart, C. M. (2000). On crises, contagion, and confusion. Journal of International Economics, 51, 145–168. Krugman, P. (1979). A model of balance-of-payments crises. Journal of Money, Credit and Banking, 11, 311–325. McKinnon, R. I., & Pill, H. (1996). Credible liberalizations and international capital flows: The ‘overborrowing syndrome’. In T. Ito & A. O. Krueger (Eds.), Financial deregulation and integration in East Asia (pp. 7–42). Chicago: University of Chicago Press. Miller, M., García-Fronti, J., & Zhang, L. (2006). Supply shocks and currency crises: The policy dilemma reconsidered. CEPR Discussion Paper No.5905. Nakatani, R. (2014). The effects of financial and real shocks, structural vulnerability and monetary policy on exchange rates from the perspective of currency crises models. UTokyo Price Project Working Paper No.61, Graduate School of Economics, University of Tokyo. Nakatani, R. (2016). Twin banking and currency crises and monetary policy. Open Economies Review, 27, 747–767. Nakatani, R. (2017). The effects of productivity shocks, financial shocks and monetary policy on exchange rates: An application of the currency crisis model and implications for emerging market crises. Emerging Markets Finance and Trade. http://dx.doi.org/10.1080/1540496X.2016.1216836. In press. Nguyen, D. X., & Schaur, G. (2010). Cost linkages transmit volatility across markets. EPRU Working Paper Series 2010–03, Economic Policy Research Unit (EPRU), Department of Economics, University of Copenhagen. Obstfeld, M. (1996). Models of currency crises with self-fulfilling features. European Economic Review, 40, 1037–1047. Sachs, J. (1985). External debt and macroeconomic performance in Latin America and East Asia. Brookings Papers on Economic Activity, 2, 523–564. Schneider, M., & Tornell, A. (2004). Balance sheet effects, bailout gurantees and financial crises. Review of Economic Studies, 71, 883–913. Van Rijckeghem, C., & Weder, B. (2001). Sources of contagion: Is it finance or trade? Journal of International Economics, 54, 293–308.