(See handout no.16; reading McKinnon-Schnabl)
The macroeconomic authorities of developing and transition countries are faced with a difficult task.
On the one hand, the global economy is unstable with frequent shocks, crises and volatilities. It is also a place with fierce competition with multinational corporations of EU, US, Japan and Korea, the emergence of China as the factory of the world, dumping and unfair trade practices by some exporters, etc.
On the other hand, the domestic capability of most developing and transition countries is still weak. The market economy is not well developed, and domestic enterprises lack competitiveness. The government is often saddled with inefficiency, corruption, lack of expertise, and political pressure.
The question for the central bank governor and the finance minister is: how do you manage monetary policy, and especially the exchange rate, in order to avoid unnecessary shocks and provide stable environment for economic development? More specifically, in this age of accelerated globalization, what is the appropriate exchange rate system for developing or transition countries?
The question is difficult enough, but to make the matter even more complicated, there are many policy goals but only one exchange rate.
My advice to countries with more than one exchange rate: If a country has the multiple exchange rate system, the IMF will not seriously deal with you and FDI will probably not come. It is a good idea to unify the rates as soon as possible. China unified their rates in 1994 and Vietnam did so in 1989. Unification did not give them any big shock, and they began to receive large amounts of FDI after that, partly because of the improved exchange rate system. The government is often concerned about the social consequences of currency reunification, but people are usually much better off with a unified rate than without. The persistence of the multiple exchange rate practice is mainly a political problem, not economic. It is a huge and hidden subsidy and taxation system, from which some people benefit greatly.
The possible goals for exchange rate management may include the following:
1. Competitiveness
2. Price stability
3. Current account adjustment
4. Domestic financial stability (protection of balance sheets of banks and firms)
5. Public debt management
6. Avoiding speculative attacks
7. Minimizing domestic impact of a large exogenous shock (like a regional conflict or currency crisis)
8. Promoting FDI, growth, or industrialization
Obviously, one exchange rate cannot pursue all these goals simultaneously. You must choose.
The story is a bit more complex, however. These goals are not necessarily equally valid, so they should not be pursued with equal weights. Some of them are very legitimate macroeconomic concerns but others may be impossible or irrelevant. Economists and officials have different views on this matter also.
The first two goals--competitiveness and price stability--are very fundamental and all countries should mind them. The big problem is that these two goals often conflict with each other. To maintain competitiveness (or regain it after domestic inflation), the exchange rate should be flexible and devaluation must be accepted, if necessary. On the other hand, to contain domestic inflation or avoid imported inflation, the exchange rate should be either stable or even moderately overvalued (this is called the use of the exchange rate as a nominal anchor). But clearly, these two requirements are not compatible.
Suppose domestic inflation is rising and the home currency is becoming overvalued. This may be due to the failure of fiscal and monetary policy or a private sector-driven bubble, over-borrowing and over-investment. Either way, the central bank is faced with a dilemma. If it devalues the currency to regain competitiveness, it must accept the inflation that has already occurred (called the accommodation of inflation). This may encourage further inflation and devaluation in the future. If it keeps the exchange rate stable, it will exert a deflationary pressure on the domestic economy so the inflation-devaluation spiral can be avoided. But in this case, competitiveness is lost due to overvaluation, and a recession is likely. The central bank may also choose the half-way solution by devaluing partially.
The third goal--current account adjustment--is popular but controversial. Traditionally, devaluation is recommended to a country with a current account (or trade) deficit. However, whether it really works to reduce the deficit or not must be carefully studied in the context of each individual country. Devaluation is a double-edged sword; as noted above, it may trigger an inflation spiral. It may also affect the macroeconomy in other complicated ways to offset the intended relative-price effect. For surplus countries like Japan and China, external pressure for appreciating the currency is often exerted. Please refer to the discussion of the elasticities approach in lecture 7 and especially lecture 9.
The fourth and fifth goals--protecting the balance sheets of the private and public sectors--are related to the question of exchange exposure and losses. If the currency is devalued, the value of foreign currency-denominated debt will increase in home currency, which creates enormous difficulties for both the private and public sectors. Part of this debt can be hedged, but not all (lecture 4). This concern militates against devaluation (i.e., these goals are inconsistent with the competitiveness concern).
The sixth goal--avoiding speculative attacks--can be achieved by calming the market expectation. How can we do this? Some economists argue that the exchange rate should float. Exchange rate flexibility will remind traders that the currency can go both up as well as down, which discourages them from betting on exchange rate stability or one-way movement. This makes attacks less likely. But exchange rate volatility itself may become a problem. Perhaps the best way to avoid attacks is to avoid overvaluation, and that is attained by frequent reviews and proper adjustments of the exchange rate levels (i.e., fulfilling the competitiveness concern).
The seventh goal--minimizing impact of a large external crisis--essentially is the question of the timing and manner of floating. If the currency of an important neighboring country collapses, your currency becomes suddenly overvalued, relatively speaking. The government must decide whether the home currency should (partly) follow this depreciation or remain stable. There is no easy rule of thumb as to whether you should float earlier, later, or not at all; it depends on individual cases. When you decide to float, it must be accompanied by a good package of other supplementary policies so that the currency will stabilize relatively soon. rather than depreciate endlessly. Kazakhstan's response to the Russian crisis in 1998 gives a nice example of this (lecture 15).
The eighth goal--promoting FDI, growth, or industrialization--is, in my view, a red herring. Such long-term real-sector development goals cannot be pursued by exchange rate management. The only thing that the central bank can do for this purpose is to maintain competitiveness and price stability (namely, doing well in the first two goals). Once, the finance minister of a certain transition country asked the Japanese delegation: "what level of exchange rate promotes industrialization of our country?" We answered, "Your Excellency, the question may not be the right one; we can't give any numbers like that."
To repeat, the first two goals--competitiveness and price stability--are most basic, although mutually contradictory. In the main, exchange rate management should pursue these goals with proper balance and flexibility. The real effective exchange rate (REER) can be used to monitor the current competitiveness.
However, for a developing or transition economy which is small and open, there is an additional problem. Suppose a country which mainly exports (i) primary commodities; or (ii) garments, machines, or consumer electronics assembled from imported parts (contract manufacturing), or (iii) both. The common feature of these two items is that there is very little domestic labor content. Primary commodities are homogeneous in quality and priced in USD in globally integrated markets, so you must accept the international price. The exchange rate does not change your competitiveness but only affects export earnings measured in local currency. In the case of contract manufacturing with imported materials and parts, domestic value-added is normally very small (typically 5-10%) and other (mostly imported) inputs are denominated in USD. Again, the exchange rate does not really matter.
REER is calculated by comparing the prices of domestic and foreign baskets (usually CPI). But REER is largely irrelevant to the competitiveness of such a country, because the exported products hardly contain domestic components represented by CPI.
In such a case, it is not easy to say which exchange rate is consistent with export competitiveness. Any exchange rate will have more or less the same effect, at least for foreign producers. In this case, it is perhaps better to work on real-sector improvements (product quality, technical transfer, efficient logistics, quick response to customers, integration into global value chain and supply chain, OJT, education and training, etc) under price and exchange rate stability, rather than try to manipulate competitiveness artificially by devaluation.
(For details, see handout no.16)
Here is the menu of possible exchange rate systems for developing and transition countries. However, opinions have not converged as to which one is the most desirable.
Managed float
After the Mexican (1994) and the Asian (1997) crises, many economists began to argue that dollar peg was dangerous. They contend that exchange rates should be flexible enough so adjustments are not delayed until too late. Some even argue that IMF should not lend to countries with a dollar peg !
Bipolar view
Some economists--Barry Eichengreen, Stanley Fischer and others--went even further. According to them, the reality of the 21st century with massive financial flows would not allow any country to adopt a "middle" solution such as target zones, adjustable peg, baskets, crawling peg, etc. They recommend that all countries, including developing and transition ones, to converge on either complete fix or 100% free floating. [However, many people, including myself, now regard this as too extreme. After the Argentine crisis where the currency board led to the severe crisis (see below), this view seems to have lost credibility.]
Currency board
The currency board is an institutional arrangement to tie money supply closely to the amount of international reserves, so the monetary authority has no power to issue money independently (in principle, at least). In the most rigid case, monetary base can be issued or withdrawn only in exchange with foreign assets sold or bought against the monetary authority (however, most currency boards are not so rigid; there are loopholes and lee ways). With a currency board, monetary policy is not needed so the central bank is abolished and a simpler "monetary authority" takes over. The purpose of this system is to regain policy credibility, especially after high inflation and frequent depreciation. Moreover, if the country is very small, it may be a good idea to adopt the currency of a big country to economize the cost of managing money.
However, the demerits of the currency board include the following:
--Inability to conduct monetary policy.
--Lack of the "lender of last resort" (LLR) function; if domestic financial institutions fail and bank runs occur, there is no authority to contain it.
--As overvaluation accumulates, it will eventually lead to a collapse of the system (see below).
Dollarization
As noted in lecture 8, dollarization has two meanings.
Private (or inadvertent) dollarization: people use dollars because they do not trust domestic money or foreign money is more convenient than domestic money.Official dollarization: the government declares USD to be the only official money for the country, and abolishes domestic money.
Dollarization discussed here is the second type. Such officially-declared dollarization is similar to the currency board but even more rigid and extreme. Under a currency board, domestic currency circulates but its quantity is regulated by the amount of international reserves. Under dollarization, domestic money is abandoned and USD notes circulate instead. Again, there is no need for a central bank. The merits and demerits of dollarization are the same as those of the currency board, except that they are further magnified.
Multiple currency basket (proposed for East Asia)
As discussed in the previous lecture, multiple currency baskets consisting of dollar, yen and euro are recommended by some economists to the developing countries in East Asia. These baskets are supposed to automatically smooth the competitiveness shocks arising from the movements of major currencies (but they do not automatically adjust for other shocks--see lecture 13).
Soft dollar zone (for East Asia)
Ronald McKinnon argues that neither floating nor the currency basket is practical in East Asia. He notes that East Asian currencies actually returned to the soft dollar peg after the Asian crisis, the situation which he thinks is desirable and reasonable. The dollar is the key currency in the world economy and monetary stability should be built around it.
Virtual exchange rate stability
This is also proposed by Prof. McKinnon (especially for major countries). There should be a long-term, unchanging nominal exchange rate target based on tradable PPP (for example, $1=110 yen). The two countries (for example, Japan and the US) have the obligation to keep the rate within the narrow band around this target forever. When a large shock hits occasionally (once in a decade?), the rate can deviate temporarily from the target. But after the shock is gone, the rate must return to the original, unchanged band. The purpose of such requirement is to build a strongly regressive expectation around the long-term level. Once such an expectation is formed, maintaining the system will require very little policy intervention. The validity of this system crucially depends on the policy commitment not to change the long-term rate after a crisis.
Double target zones
John Williamson (Institute of International Economics, Washington DC) is the champion of target zone proposals. He has many ideas, and the double target zone is one of them. There should be a "soft" inner band and a "hard" outer band, so the central bank will have three zones where (i) it does not intervene; (ii) it can intervene; and (iii) it must intervene.
Band-basket-crawl (BBC)
This idea, advanced by Rudiger Dornbusch and Y. C. Park, is a variation of the target zone proposal. The central rate should be defined by a multiple currency basket and there should be a band around it. Moreover, there is a built-in inflation sliding of the central rate. This is what I would call a currency basket with inflation slide.
"Eclectic" view
Jeffrey Frankel wrote a paper entitled: "No single currency regime is right for all countries or at all times." The right choice depends on circumstances, and each country should adopt the most suitable system for itself. The attempt to find a one-size-fits-all solution is misguided.
The exit policy problem refers to the question of how to end a fixed exchange rate without crisis. This is a serious issue associated with the regime of more or less fixed exchange rate, such as dollarization, currency board, and adjustable peg with very infrequent adjustments.
In order to stop hyperinflation or monetization of fiscal deficits, suppose the government adopts one of the fixed exchange rate systems. It may even succeed in stopping inflation (as many Latin American countries did in the 1990s) and win domestic political support. Everything goes well at first. The popularity of the government is closely associated with this exchange rate policy. Over time, it becomes politically more and more difficult to get out of the fixed exchange rate, because that will be judged as policy failure leading to the collapse of the current government.
But economically, the long-term fixing of the exchange rate will often cause gradual overvaluation and the loss of competitiveness. Devaluation is necessary, but the government is politically locked into the fixed exchange rate system. The economy starts to stagnate, and speculators prepare to attack. It becomes too late to avoid the crisis.
This is exactly what happened in Argentina.
The Argentine peso was fixed one-to-one against USD and the currency board was installed in 1991. This system gained credibility as hyperinflation gradually subsided. Everything went smoothly until 1999, when Brazil, a big neighboring economy, devalued. The Argentine peso became seriously overvalued. During 2000-01, the loss of competitiveness and a tight budget (to reduce the fiscal deficit) caused social instability. In December 2001, popular discontent led to political crisis and the change of government. In January 2002, after a few failed governments, President Duarte announced the "New Economic Policy" including the devaluation (and floating) of the peso and the introduction of dual exchange rates. But social and economic instabilities continued even after the floating began.
For a relatively big country like Argentina, a permanently fixed exchange rate may look fine at the beginning, but it will eventually lead to a crisis. If the fixity of the exchange rate is written into law (sometimes even into the constitution) and policy credibility is built around it, the government cannot exit from it. The lesson from this episode is as follows: it is necessary to choose a system where exchange rate adjustment is regarded as normal and does not surprise people.
My own view is close to Jeffrey Frankel's eclecticism. More specifically, it can be stated in two major recommendations.
FIRST, for exchange rate management of developing and transition economies under normal circumstances (i.e., no crisis), aim at short-term stability and long-term flexibility against USD. For this purpose, flexibly adjust the level as well as the system of the exchange rate. Don't get stuck with one rate or one system. There are many systems that can achieve this goal, including:
--Adjustable peg against the dollar, with frequent reviews
--Crawling peg (a sort of inflation slide)
--Basket peg, with frequent reviews
--Managed float
--Band-basket-crawl (BBC)
--Any other system that can combine short-term stability with long-term flexibility
It is not the name of the system, but how it is actually managed, that is important. The central bank should monitor the domestic and world economy as various shocks occur (not just the movements of major currencies !) using many indicators including REER and act swiftly when the need arises.
For developing countries, both free float and permanent fix are to be strongly avoided, because neither offers the flexible choice between stability and flexibility.
My suggestion is essentially to filter out short-term noise but permit long-term trends in the movement of the exchange rate. Short-term fluctuations should be smoothed because they contain too much noise unrelated to fundamentals. This practice is also good for pre-empting bubbles and speculations.
SECOND, when a crisis occurs at home or in the neighboring countries, select the right timing to switch from normal management to crisis management. Under the crisis mode, the normal operation suggested above must be temporarily suspended and the monetary authority must concentrate on minimizing damage to the national economy. How exactly to do this may depend on each country's situation, so there is no general rule. We can at least say that, in a crisis,
--The government must decide whether and when to devalue (float). Because of many policy goals (see above) and the multiplicity of equilibria (lecture 12), there is no simple answer.
--When you float at the time of crisis, devaluation of 20-30% is common, so don't be afraid of such a fall (but 50% is fairly large; 75% is disastrous.) The most important thing is how quickly the devaluation ends and the exchange rate stabilizes, so the country can go back to the "normal" mode. A continuing depreciation for several months or more should be regarded as a failure.
--When you float, a package of measures to avoid excessive or prolonged depreciation and to protect people and firms from unnecessarily large shocks, should be implemented simultaneously. These may well include temporary non-market measures.
<References>
Frankel, Jeffrey...