The Case Against Currency Boards:
Debunking 10 Myths about the Benefits of Currency Boards
(5가지만 뱔췌)
by Nouriel Roubini
There is currently a broad debate among economists and policy makers about the benefits and costs of currency boards (CB). In recent times, they have been adopted by countries such as Hong Kong (1983), Argentina (1991), Estonia (1992), Lithuania (1994) and Bulgaria (1997). Also, a number of Asian countries (Indonesia in particular, but also Thailand and Malaysia) are considering adopting a currency board system. The talk of currency boards is also widespread in Latin America where countries such as Venezuela, Brazil and Mexico are considering whether to follow the successful experience of Argentina.
I will argue that, in spite of their current popularity, currency boards are overall a very bad policy idea. What matters for an economy is to have stable, sensible and credible economic policies, not a currency board. We will present the arguments against currency boards by considering 10 myths about the alleged benefits of currency boards.
Our analysis analysis will suggest that the case for currency boards is very weak and the experience with them a mixed bag. There are countries in which they seem to work for a while; however, these countries are successful not because of the CB system itself but rather because they follow macroeconomic policies and structural liberalization policies that are consistent with the maintenance of fixed rates. Fixed rates and currency boards without these good policies lead to currency collapse and economic disaster. Conversely, if you do follow the right economic policies you do not need a currency board: you will do as well without one and adopting one may only hurt you when truly exogenous shocks require an adjustment of your nominal exchange rate parity.
There are of course some marginal benefits of CB that one can point to: short-run credibility when you start from an hyperinflation (like in Argentina), stronger incentives not to monetize and run budget deficits under some conditions. But those are all results that a country can achieve without a CB and therefore avoid the other costs of having one. While these days CBs are being proposed as the Holy Grail that will cure every ill and give bliss to a country, their shortcomings are serious and their adoption may actually harm a country rather than improve its economic long-run performance.
Let us then consider the arguments, or myths, about currency boards in more detail.
Myth 1. Speculative attacks against a currency do not occur in a currency board system because you are credibly committed to fixed exchange rates.
One of the main myths about CBs is that speculative attacks against a currency cannot occur in a currency board because you are committed fixed exchange rates. The empirical truth is that CBs do not prevent speculative attacks on a currency. The fact that currency boards do not prevent speculative attacks is evident from the recent experience of Argentina and Hong Kong (HK). As the Mexican peso collapsed in 1994 and the Asian currencies collapsed in 1997, the currencies of Argentina (in 1995) and HK (in 1997) were also subject to speculative attacks. While it is true that these currencies did not collapse (so far), these attacks had very large economic costs. Since the attack implied that investors perceived that the probability of devaluation was positive and high, the monetary authorities of these countries had to cut the money supply (as a capital outflows leads to an automatic cut in the monetary base in a mechanical CB) and increase dramatically the domestic interest rates (as high as 20% in real terms in both countries). This huge increase in interest rates pushed Argentina in a severe recession in 1995 with output collapsing 6% and unemployment rising to 18%.
In the rhetoric of the defenders of a currency board, an attack should never occur and therefore domestic interest rates should remain at the level of the country to which you peg (i.e. 5% as they both peg to the US dollar). That turned out to be false and the expectation of a devaluation forced the monetary authorities to sharply increase interest rates. Since a CB does not mean that a devaluation will not occur (it is just a more strict form of fixed exchange rates), the possibility of a devaluation implies that when an attack occurs, the domestic authorities have to push up interest rates to credibly show that they are committed to the peg parity. These high interest rates can bankrupt domestic banks, domestic firms and lead to a big recession. It happened in Argentina in 1995 and it will happen in HK this year: high interest rates are already destroying growth in HK in 1998. Compare instead the behavior of HK with that of Taiwan and Singapore who did not have a CB an let their fixed parity to the US$ to be abandoned when their currencies were attacked: their currencies devalued somewhat but they avoided a bigger crisis. Growth in Taiwan and Singapore will be better than in HK this year as they did not have to increase interest rates as much and their real exchange rate was allowed to depreciate.
Myth 2. Currency boards are good for the stability of the banking and financial system
The reality is opposite: a monetary tightening and interest rates increase when a CB is subject to a speculative attack can bankrupt the domestic financial system and the domestic banks; tight base money mans that, given required reserve ratios, banks are forced to recall loans and firms are going to go bankrupt. This is why it is very dangerous to introduce a CB in a country like Indonesia right now when a major financial crisis is occurring and banks are on the verge of collapse even without a CB. Everything else equal, the fragility of the banking system is an important factor to consider when deciding whether to introduce a CB: the weaker the banks the more dangerous a CB.
The concern with the collapse of the banking system is also the reason why a CB is often implemented with some degree of cheating. When Argentina came under attack in 1995 the rules of a strict CB should have force the country to reduce the monetary base by an amount equal to the large capital outflow. Since this mechanical CB would have led, through the money multiplier in the banking system, to a sharp contraction of bank loans and deposits and banking collapse, the monetary authorities cheated: they cut the monetary base but then they significantly reduced the required reserve ratios of the banks to avoid a sharp fall in the money supply, loans and deposit. This is not what a CB should be and, therefore, during an emergency the role of lender of last resort of the central banks, that should have been eliminated by the CB, was restored. One should certainly approve the Argentinean policy of easing the monetary tightening by changing reserve ratios: the alternative would have been even higher real rates, banking collapse and a even bigger recession than the one that occurred with the monetary squeeze in 1995. However, the Argentinean easing of the reserve ratios proves the point: a true CB does not work because if you let it work as it should, when there is an attack the country and its banking system would go bankrupt. Some role of lender of last resort is always useful to have and a strict CB would eliminate such a role and cause more harm than good when a financial crisis is occurring.
Myth 3. Fixed exchange rate regimes, and CBs in particular, work better than more flexible exchange rate regimes
The experience of the 1990s suggests that rigidly fixed exchange rate are the cause of currency crisis, not flexible exchange rates. Currencies collapse when they are fixed for too long to parities that are not consistent with the equilibrium fundamental value of a currency. The ERM crisis in 1992-93, the Mexican peso crisis and Tequila effect in 1994-95 and the Asian of 1997-98 were caused by the fact that the countries were on a fixed exchange rate system, not because flexible exchange rate systems. People forget that the big currency crises of the 1990s occur with fixed rate regimes (a close cousin of currency boards), not under more flexible exchange rate regimes. The lessons of Latin America (see Chile) is that you should avoid the real appreciation often associated with fixed exchange rates (or CB's): you should rather target the real exchange rate and introduce capital controls on inflows to avoid hot money capital flow leading to nominal and real appreciation of your currency and competitiveness loss. A CB, like most fixed rate regimes, is just a recipe for disaster: in the short-run things look good until the regime collapses with major real economic costs. One should also observe that the historical experience with dozen of CB's experiences suggests almost all CB's eventually collapse or are phased out (this is what Lithuania and Estonia are trying to do now after mistakenly opting for a CB in the early 1990s).
Myth 4. Currency board are good for countries exporting world commodities priced in foreign currency because such countries cannot use the exchange rate to affect their real exchange rate
This argument is made in countries such as Venezuela where some political groups are currently recommending the adoption of a currency board. The argument is that in a country such as Venezuela most exports are oil and other primary commodities whose prices cannot be affected by a devaluation.
It is true that in the case of a country where a large fraction of exports are set in world markets, the ability to use the exchange rate to lead to a real devaluation that stimulates exports is limited. However, even limited exchange rate flexibility may be useful. In fact, the country also produces and sell non-traditional and non-oil manufactured exports whose demand depends on the nominal and real exchange rate. As Dutch disease phenomenon has already hurt these non-oil exports, you do not need to make things worse by adopting a CB that will lead to even more real appreciation and loss of competitiveness in these goods. Also, since such a country is always going to be subject to terms of trade shocks that negatively affect demand and growth, some exchange rate flexibility is necessary. For example, Texas has a CB (or more precisely a monetary union with the rest of the US). When the price of oil collapsed in the mid 1980s, oil-producing states such as Texas (and Louisiana) went into a deep recession. Suppose that Texas had had its own Texas dollar; then nominal and real depreciation of the Texas dollar relative to the US would have reduced the recessionary effect of a fall in the price of oil. Since your terms of trade are dominated by oil, a fixed exchange rate relative to the US is not optimal: i.e. Venezuela and US do not make an optimal currency union as their economies and the shocks to their productivities are completely different and there is not enough factor mobility across these two countries. There is no good reason to have a fixed parity to the US (either in the form of fixed rates or in the form of a CB or in the form of a monetary union) for a country with large exports of primary commodities subject to large terms of trade shocks. When a shock to the terms of trade occurs, the market will attack this country as the equilibrium nominal parity becomes lower; then the country will either not be able to survive the attack even if they had a CB or will survive the attack paying a larger costs in terms of output loss. Also, in the case of Latin America, there is some likelihood that fixed rate parities in Latin America may collapse in the near future. The Brazilian currency is misaligned in real terms and a major devaluation may occur (especially after the current president is reelected). If the Brazilian currency falls, the Argentinean peso CB will collapse too. At that point, any other country in Latin America considering the adoption (such as Venezuela) and implementing a currency board will be in trouble and be attacked too.
More in general, in any country with large exports of primary goods, a CB is a bad idea. The high volatility of the terms of trade and the high dependence of the economy on primary exports requires a central bank that is able to sterilize capital flows. A currency board would introduce such large fluctuations in the monetary base of the economy that would have a deeply destabilizing influence.
At least in the case of Texas one could make the arguments that Texas and the rest of the US satisfy the conditions for an optimal currency union: while their productivity shocks are not synchronized, at least, the degree of labor of capital mobility between Texas and the rest of the US is very large. As the negative oil shock hit Texas, people and capital left Texas for other states and this dampened the real effect of the shock; high capital and labor mobility helps the adjustment to a terms of trade shocks. The same cannot be said of other countries considering a currency board with
parities pegged to the US dollar. They cannot respond to a large terms of trade shocks by sending their workers to the US.
Debunking 10 Myths about the Benefits of Currency Boards
(5가지만 뱔췌)
by Nouriel Roubini
There is currently a broad debate among economists and policy makers about the benefits and costs of currency boards (CB). In recent times, they have been adopted by countries such as Hong Kong (1983), Argentina (1991), Estonia (1992), Lithuania (1994) and Bulgaria (1997). Also, a number of Asian countries (Indonesia in particular, but also Thailand and Malaysia) are considering adopting a currency board system. The talk of currency boards is also widespread in Latin America where countries such as Venezuela, Brazil and Mexico are considering whether to follow the successful experience of Argentina.
I will argue that, in spite of their current popularity, currency boards are overall a very bad policy idea. What matters for an economy is to have stable, sensible and credible economic policies, not a currency board. We will present the arguments against currency boards by considering 10 myths about the alleged benefits of currency boards.
Our analysis analysis will suggest that the case for currency boards is very weak and the experience with them a mixed bag. There are countries in which they seem to work for a while; however, these countries are successful not because of the CB system itself but rather because they follow macroeconomic policies and structural liberalization policies that are consistent with the maintenance of fixed rates. Fixed rates and currency boards without these good policies lead to currency collapse and economic disaster. Conversely, if you do follow the right economic policies you do not need a currency board: you will do as well without one and adopting one may only hurt you when truly exogenous shocks require an adjustment of your nominal exchange rate parity.
There are of course some marginal benefits of CB that one can point to: short-run credibility when you start from an hyperinflation (like in Argentina), stronger incentives not to monetize and run budget deficits under some conditions. But those are all results that a country can achieve without a CB and therefore avoid the other costs of having one. While these days CBs are being proposed as the Holy Grail that will cure every ill and give bliss to a country, their shortcomings are serious and their adoption may actually harm a country rather than improve its economic long-run performance.
Let us then consider the arguments, or myths, about currency boards in more detail.
Myth 1. Speculative attacks against a currency do not occur in a currency board system because you are credibly committed to fixed exchange rates.
One of the main myths about CBs is that speculative attacks against a currency cannot occur in a currency board because you are committed fixed exchange rates. The empirical truth is that CBs do not prevent speculative attacks on a currency. The fact that currency boards do not prevent speculative attacks is evident from the recent experience of Argentina and Hong Kong (HK). As the Mexican peso collapsed in 1994 and the Asian currencies collapsed in 1997, the currencies of Argentina (in 1995) and HK (in 1997) were also subject to speculative attacks. While it is true that these currencies did not collapse (so far), these attacks had very large economic costs. Since the attack implied that investors perceived that the probability of devaluation was positive and high, the monetary authorities of these countries had to cut the money supply (as a capital outflows leads to an automatic cut in the monetary base in a mechanical CB) and increase dramatically the domestic interest rates (as high as 20% in real terms in both countries). This huge increase in interest rates pushed Argentina in a severe recession in 1995 with output collapsing 6% and unemployment rising to 18%.
In the rhetoric of the defenders of a currency board, an attack should never occur and therefore domestic interest rates should remain at the level of the country to which you peg (i.e. 5% as they both peg to the US dollar). That turned out to be false and the expectation of a devaluation forced the monetary authorities to sharply increase interest rates. Since a CB does not mean that a devaluation will not occur (it is just a more strict form of fixed exchange rates), the possibility of a devaluation implies that when an attack occurs, the domestic authorities have to push up interest rates to credibly show that they are committed to the peg parity. These high interest rates can bankrupt domestic banks, domestic firms and lead to a big recession. It happened in Argentina in 1995 and it will happen in HK this year: high interest rates are already destroying growth in HK in 1998. Compare instead the behavior of HK with that of Taiwan and Singapore who did not have a CB an let their fixed parity to the US$ to be abandoned when their currencies were attacked: their currencies devalued somewhat but they avoided a bigger crisis. Growth in Taiwan and Singapore will be better than in HK this year as they did not have to increase interest rates as much and their real exchange rate was allowed to depreciate.
Myth 2. Currency boards are good for the stability of the banking and financial system
The reality is opposite: a monetary tightening and interest rates increase when a CB is subject to a speculative attack can bankrupt the domestic financial system and the domestic banks; tight base money mans that, given required reserve ratios, banks are forced to recall loans and firms are going to go bankrupt. This is why it is very dangerous to introduce a CB in a country like Indonesia right now when a major financial crisis is occurring and banks are on the verge of collapse even without a CB. Everything else equal, the fragility of the banking system is an important factor to consider when deciding whether to introduce a CB: the weaker the banks the more dangerous a CB.
The concern with the collapse of the banking system is also the reason why a CB is often implemented with some degree of cheating. When Argentina came under attack in 1995 the rules of a strict CB should have force the country to reduce the monetary base by an amount equal to the large capital outflow. Since this mechanical CB would have led, through the money multiplier in the banking system, to a sharp contraction of bank loans and deposits and banking collapse, the monetary authorities cheated: they cut the monetary base but then they significantly reduced the required reserve ratios of the banks to avoid a sharp fall in the money supply, loans and deposit. This is not what a CB should be and, therefore, during an emergency the role of lender of last resort of the central banks, that should have been eliminated by the CB, was restored. One should certainly approve the Argentinean policy of easing the monetary tightening by changing reserve ratios: the alternative would have been even higher real rates, banking collapse and a even bigger recession than the one that occurred with the monetary squeeze in 1995. However, the Argentinean easing of the reserve ratios proves the point: a true CB does not work because if you let it work as it should, when there is an attack the country and its banking system would go bankrupt. Some role of lender of last resort is always useful to have and a strict CB would eliminate such a role and cause more harm than good when a financial crisis is occurring.
Myth 3. Fixed exchange rate regimes, and CBs in particular, work better than more flexible exchange rate regimes
The experience of the 1990s suggests that rigidly fixed exchange rate are the cause of currency crisis, not flexible exchange rates. Currencies collapse when they are fixed for too long to parities that are not consistent with the equilibrium fundamental value of a currency. The ERM crisis in 1992-93, the Mexican peso crisis and Tequila effect in 1994-95 and the Asian of 1997-98 were caused by the fact that the countries were on a fixed exchange rate system, not because flexible exchange rate systems. People forget that the big currency crises of the 1990s occur with fixed rate regimes (a close cousin of currency boards), not under more flexible exchange rate regimes. The lessons of Latin America (see Chile) is that you should avoid the real appreciation often associated with fixed exchange rates (or CB's): you should rather target the real exchange rate and introduce capital controls on inflows to avoid hot money capital flow leading to nominal and real appreciation of your currency and competitiveness loss. A CB, like most fixed rate regimes, is just a recipe for disaster: in the short-run things look good until the regime collapses with major real economic costs. One should also observe that the historical experience with dozen of CB's experiences suggests almost all CB's eventually collapse or are phased out (this is what Lithuania and Estonia are trying to do now after mistakenly opting for a CB in the early 1990s).
Myth 4. Currency board are good for countries exporting world commodities priced in foreign currency because such countries cannot use the exchange rate to affect their real exchange rate
This argument is made in countries such as Venezuela where some political groups are currently recommending the adoption of a currency board. The argument is that in a country such as Venezuela most exports are oil and other primary commodities whose prices cannot be affected by a devaluation.
It is true that in the case of a country where a large fraction of exports are set in world markets, the ability to use the exchange rate to lead to a real devaluation that stimulates exports is limited. However, even limited exchange rate flexibility may be useful. In fact, the country also produces and sell non-traditional and non-oil manufactured exports whose demand depends on the nominal and real exchange rate. As Dutch disease phenomenon has already hurt these non-oil exports, you do not need to make things worse by adopting a CB that will lead to even more real appreciation and loss of competitiveness in these goods. Also, since such a country is always going to be subject to terms of trade shocks that negatively affect demand and growth, some exchange rate flexibility is necessary. For example, Texas has a CB (or more precisely a monetary union with the rest of the US). When the price of oil collapsed in the mid 1980s, oil-producing states such as Texas (and Louisiana) went into a deep recession. Suppose that Texas had had its own Texas dollar; then nominal and real depreciation of the Texas dollar relative to the US would have reduced the recessionary effect of a fall in the price of oil. Since your terms of trade are dominated by oil, a fixed exchange rate relative to the US is not optimal: i.e. Venezuela and US do not make an optimal currency union as their economies and the shocks to their productivities are completely different and there is not enough factor mobility across these two countries. There is no good reason to have a fixed parity to the US (either in the form of fixed rates or in the form of a CB or in the form of a monetary union) for a country with large exports of primary commodities subject to large terms of trade shocks. When a shock to the terms of trade occurs, the market will attack this country as the equilibrium nominal parity becomes lower; then the country will either not be able to survive the attack even if they had a CB or will survive the attack paying a larger costs in terms of output loss. Also, in the case of Latin America, there is some likelihood that fixed rate parities in Latin America may collapse in the near future. The Brazilian currency is misaligned in real terms and a major devaluation may occur (especially after the current president is reelected). If the Brazilian currency falls, the Argentinean peso CB will collapse too. At that point, any other country in Latin America considering the adoption (such as Venezuela) and implementing a currency board will be in trouble and be attacked too.
More in general, in any country with large exports of primary goods, a CB is a bad idea. The high volatility of the terms of trade and the high dependence of the economy on primary exports requires a central bank that is able to sterilize capital flows. A currency board would introduce such large fluctuations in the monetary base of the economy that would have a deeply destabilizing influence.
At least in the case of Texas one could make the arguments that Texas and the rest of the US satisfy the conditions for an optimal currency union: while their productivity shocks are not synchronized, at least, the degree of labor of capital mobility between Texas and the rest of the US is very large. As the negative oil shock hit Texas, people and capital left Texas for other states and this dampened the real effect of the shock; high capital and labor mobility helps the adjustment to a terms of trade shocks. The same cannot be said of other countries considering a currency board with
parities pegged to the US dollar. They cannot respond to a large terms of trade shocks by sending their workers to the US.
'형설지공 > 경제경영' 카테고리의 다른 글
| 공적자금은 공짜가 아니다 (0) | 2001.03.03 |
|---|---|
| 예금보험제도와 예금보험공사 (LG 주간경제) (0) | 2001.03.03 |
| 통안증권 증발과 본원통화 및 국가채무 영향 (0) | 2001.03.03 |
| "Cash Poor": E-Cash 정착이 어려운 이유 (0) | 2001.03.03 |
| Dollarization의 의미와 의의 (0) | 2001.03.03 |