The Case for Monetary Union in East Asia: From Theory to Empirics

Among others, this paper presents the case for fixed exchange rates and monetary union in East Asia, the underlying motivation behind the rising interest on optimum currency areas (OCAs) in the Asian context. The paper also explicitly proposes the dollar as the monetary anchor. Twenty empirical papers in 1994–2009 are reviewed and the five original members of ASEAN have been commonly indicated to be sufficiently homogenous for monetary unification. Some other common threads can be also observed. First, recent considerations could have played a more important role than the traditional theory in the selection of country samples for analysis. Second,unlike the early intellectual work which had been molded by the debate on adjustment mechanisms, current empirical literature has been driven largely by the enthusiasm to identify homogenous economies. Third, there is a reason to believe that little has been done in light of the endogeneity view of OCA criteria. Finally, an explicit monetary anchor, particularly the dollar anchor, has scarcely been set a priority in empirical analyses.


Introduction
In the interwar years, the difficulties of reinstating the gold standard and the disruptive shock of the Great Depression had prompted a wide-ranging debate on the international monetary system in the West. Today, the birth of euro and the Asian financial crisis were among the drivers which had spurred the proliferation of the intellectual work on monetary arrangement in the context of East Asia.
The theory of optimum currency areas (OCAs) has advanced only minimally since the seminal contributions of Mundell (1961), McKinnon (1963, and Kenen (1969). It remains difficult to move from theory to empirical work and policy analysis. Bayoumi and Eichengreen (1997, p.762) Nevertheless, as the above statement vividly points out, it has remained difficult to operationalize the intellectual work to practical grounds given the complexities found in the real world. In spite of this, numerous empirical studies have attempted to demystify the theory in an effort to identify groups of economies which could possibly come together under one common monetary umbrella.
Against this backdrop, the present paper seeks to present the reasons why the idea of Asian monetary union has gained increasing popularity despite of the obstacle highlighted. The paper also provides the arguments for dollar as the monetary anchor. Finally, it reviews 20 relevant empirical works in 1994-2009 to discover some underlying common threads in them.
Among other findings, the five original ASEAN member countries have been indicated in general to constitute a monetary union. Viewed in this light, policymakers of the original ASEAN members may need to further foster political and economic integration to reap the benefits of a single monetary zone. Meantime, business managers may need to tune their regional decisions based on the fact that macroeconomic circumstances in the region are highly symmetrical.
The remainder of the paper is structured as follows. Section 2 revisits the essence of the classical work on OCA. Section 3 explains the case against flexible exchange rates for East Asia.
Section 4 details the case for monetary union in the region. Section 5 presents the case for US dollar as the monetary anchor. Section 6 provides the review of empirical studies and section 7 discusses relevant interpretations and concludes.

The theory of optimum currency areas
In the classic published in 1961, Robert Mundell proposed that an optimal currency area is characterized by internal factor mobility and external factor immobility. Succinctly, interregional and interindustrial factor mobility can substitute for changes in nominal exchange rates to restore internal and external equilibriums when asymmetric shocks occur between economic regions.
The equilibriums pertain to maintenance of full employment, stable price levels, and balanced international payments.
Based on the presumption that perfect labor mobility hardly prevails, Kenen (1969) provided an alternative to define optimality. He argued that well-diversified economies are more able to cope with asymmetric shocks between members in a monetary union and are therefore more feasible candidates to be part of the union. Another dimension for optimality came from McKinnon (1963) in which he contended that highly open economies are least feasible for flexible exchange rates whereas exchange rate fixation with a putative currency is highly desirable.
In the presence of liberal capital movements, a sustainable monetary bloc entails irrevocably fixed exchange rates; full and irreversible convertibility of currencies; financial market integration; liberalized movements on current transactions; common monetary policy; and harmonization of national financial regulations and structures of institutions (see e.g. Tavlas, 1993).

The case against flexible exchange rate
The fundamental argument raised by Milton Friedman 1 in his 1953 classic for allowing exchange rate to float lies in the ability of floating rate to ease the process of adjustment to external shocks.
Suppose the demand for exports of a country falls, necessitating a fall in relative prices of goods and labor to correct the deficit-it will be easier for the change in terms of trade be accomplished through depreciation or devaluation rather than through some combinations of inflation in the foreign country and unemployment in the home country.
Nonetheless, attempts to increase competitiveness by devaluations would only lead to inflation and retaliations (see McKinnon, 1963;Krugman, 1990). In highly open economies, domestic prices and wages are most likely closely linked to exchange rates of significant trading partners, rendering devaluations or depreciations ineffective in restoring external balance; the net result is more inflation. Also, devaluation is useless and might even be detrimental when a shock comes from the capital account, as when emerging markets are hit by contagion and face sharply higher interest rates; the Latin American and the Indonesian experience had been contractionary irrespective of the degree of devaluation (Calvo, 2002).
In light of the above, for economies which have been integrated in respect of trade and international capital flows, which define most of East Asia today, flexible rates are most probably undesirable. In fact, the sharp fluctuations in the yen-dollar rate, coupled with pseudofixed or soft pegs and incompetent monetary policies were the main culprit behind the Asian crisis. Thus, the foregoing of independent monetary policy and floating rate is likely to be beneficial to developing countries (Calvo and Reinhart, 2002). Even Japan, an advanced economy, is not spared from the devastating effects from floating rates. The Japanese banking system was the casualty of excessive appreciation of the yen between 1985 and 1995 (Mundell, 2003).

The case for monetary union
The primary case in favor of exchange rate fixation against a pivotal currency rests upon the desirability of certainty (Krugman, 1990). By fixing participants' currency values against a hard currency, the resulted system will confer a degree of stability between the participants and the numéraire country, as well as between the participants, as enjoyed by the ever-expanding EMU.
As a matter of fact, the Euro club contains 16 members since Slovakia adopted the Euro on January 1, 2009. The following discussion highlights the case for an Asian monetary union which generally underlies the motivation for extensive studies in the area.

Greater integration
Tighter economic integration in East Asia is ever warranted in the face of rising regional integration elsewhere such as NAFTA, EU, Mercosur, CEMAC 1 , OECS 2 , UEMOA 3 , and CACM 4 . To an extent, these arrangements have brought intra-regional stability but more competition between trade blocs and more volatility between major currencies. Along these lines, East Asia may need to further enhance its intra-regional trade to insulate against disturbances originating from outside the region. Recent free trade deals have encompassed ASEAN 5 , China, Japan, India, Australia, and New Zealand which cover aspects of goods, services, investments, and intellectual property. Trade liberalization and economic integration often require stable exchange rates; otherwise, regional free trade agreements could be undermined by compensatory tariffs demanded by exporters in stable countries against countries that might devalue their currencies (Ngiam and Yuen, 2001).
Moreover, a monetary union which encourages trade and economic integration constitutes a virtuous self-reinforcing circle (Bayoumi and Eichengreen, 1997). Since EMU was established, trade and investment have grown tremendously. The same is true for the case of dollarization, an effective monetary union with the US, which has raised trade, investment, and economic growth significantly (Alesina and Barro, 2001).
Highly open economies would gain much if exchange rates are fixed. When initial trade is large, the size of required price and wage adjustments to accommodate any given external shock will be small (Krugman, 1990). With initial exports of 20 percent of GNP, a one percent deficit (of GNP) would require less fall in prices and wages than if the initial exports were one percent.
Even when initial trade is low, the gains from fixed rates could also be high (Alesina, Barro, and Tenreyro, 2002). Since low initial trade could be due to high trading costs, the trade that did occur must have high marginal values-coupled with lower marginal costs when exchange rates are fixed, higher marginal gains will result.
The benefits are especially important to the highly heterogeneous economic structure of East Asia. All the while, MNCs operating in the region have to diversify their production processes and stages of production across countries to exploit comparative advantages (Ramos, 1994). Examples are the tourism and electronics industries which are highly concentrated in the growth triangles (GTs), subregional economic zones which were set up to foster economic complementation. The leading GTs are the Indonesia-Malaysia-Singapore Triangle (IMS-GT); the East ASEAN Growth Area (EAGA) covering Brunei and parts of Malaysia, Philippines, and Indonesia; a growth zone linking parts of Myanmar, Laos, Thailand, and China; and the southern Chinese Economic Triangle, made up of China, Hong Kong, and Taiwan.
A bolstered economic integration from a firm monetary union will also preclude any undesirable beggar-thy-neighbor policies. Even the implicit dollar peg (or pseudo-exchange rate union) adopted by the Asian economies prior to the Asian crisis had actually insulated each other from harmful depreciations (McKinnon, 2005).

Lower costs
A currency area enhances the role of money as unit of account by setting economies of scale into play and reduces transaction costs, including the costs of information, search, exchange, hedging, and calculation (see Tavlas, 1993). Small economies, including the less developed economies in Indo-China in East Asia, should benefit the most from the unit of account, means of payment, and store of value services provided by a major currency (see Bayoumi and Eichengreen, 1997).
Indeed, the US dollar has been commonly accepted in Vietnam and its neighboring countries since the Vietnam War.
A credible monetary union anchored on a stable currency will also lead to lower cost of capital. Since the uncertainty arisen from currency risk and sudden regime change is removed in this arrangement, the cost of international and hence domestic borrowing becomes lower (see McKinnon 2005;Chang, 2000). In addition, the improved allocational efficiency of financing process in a monetary bloc does provide both borrowers and lenders a broader spectrum of financial instruments, thereby enabling more efficient choices to be made in terms of duration and risk (Robson, 1987).
Lower cost of capital also stems from lower reserve requirement when enlargement of foreign exchange market in a monetary bloc removes volatility of exchange rates and ability of speculators to influence money prices (see Tavlas, 1993). Moreover, if countries are structurally diverse, like those in East Asia, reserves for intra-area transactions too may be substantially reduced because any payments imbalances may be offsetting.

Price stability
A monetary standard based on a credible currency also helps in curbing inflation in several ways (see Tavlas, 1993). First, with respect to inflation targeting, exchange-rate targeting is better than monetary-growth targeting because exchange rates are highly observable. Second, any high inflation country which joins a low inflation monetary bloc can 'import' low inflation reputation without loss of output and employment. The recent past has seen establishments of currency board intended to import monetary policy credibility from a stable developed country (Oomes and Meissner, 2008) 6 . Third, internal monetary policy can be removed from politically dependent domestic authorities and delegate it to a more independent foreign authority.
Countries with currency boards (e.g., Argentina, Estonia, Lithuania, and Bulgaria) have experienced lower inflation and higher growth than those with other regimes (Guide, Kähkönen, and Keller, 2000). Comparable results could be found in dollarized countries too (see e.g., Chang, 2000).
Inflation reduction is also important to less developed countries in East Asia. In general, the degree of monetary authority independence from the executive branch in these countries is far lower than those in the advanced countries, which partly explains why internal monetary policies in small countries are relatively unstable. Based on IMF data, average CPI inflation in Vietnam, Laos, and Indonesia for 2001-2007 is about 4-6 percent higher than the US level while Myanmar's rate is about 24 percent higher. Since there is no permanent Phillips curve trade-off (see Tavlas, 1993), high inflation countries have little to lose in the long run and much to gain by adopting monetary policy of a stable country.

Financial stability
A stable domestic currency against the denominator of liabilities is utmost crucial in times of distress where speculative capital flows could easily deplete foreign reserves even among neighboring countries that are marginally leveraged (Rogoff 2005). Given that many developing countries in East Asia are still substantially indebted in hard currencies especially in dollars (Calvo, 2002;McKinnon, 2005), any steep depreciations would certainly render them insolvent.
The Thai experience during the Asian crisis is a very good instance. This can be precluded if a firm peg against the denominator of debt is adopted in a monetary union.
Though IMF usually advises countries to float their exchange rates in face of domestic crises, emerging middle-income economies are held back by the so-called "fear of floating" dilemma (see e.g., Calvo and Reinhart, 2002). At least two interlocking factors underlie this fear.
First, emerging economies do not have well-developed and diversified financial systems which are able to minimize real sector disruptions resulted from transitory exchange rate variations which made them not able to borrow overseas in their domestic currencies, commonly referred to as "original sin". Second, policymakers in emerging markets suffer from a chronic lack of credibility. As a result, an emerging economy might experience large and frequent shocks to exchange rate expectations or to interest rate risk premiums.

Labor mobility
According to Mundell (1961), the costs of sacrificing the use of exchange rate changes would be minimal if there is mobility or flexibility of the labor markets in geographical and industrial dimensions within a currency area. Alternatively, if labor markets are flexible, real wages can adjust to restore internal and external balances. The following evidence suggests that labor markets in East Asia may be sufficiently mobile or flexible to withstand asymmetric disturbances that may arise in a monetary union.
Asis and Piper (2008) discovered that much of international migration in Asia is intraregional, undocumented, well developed, and well connected. Since early 2000s, the world's largest net labor exporting country is the Philippines. Other main exporters include Indonesia, India, China, Vietnam, Myanmar, Cambodia, and Laos. The common destinations for them are Middle East, Malaysia, and Thailand. Net labor importing countries include Japan, Hong Kong, Taiwan, Korea, Singapore, Brunei, Malaysia, and Thailand, which draw workers from the less developed countries in the region. Notably, China allowed labor export mostly in connection with state contracted projects overseas since the 1978 market reform but international migration has been eclipsed by the much larger internal rural-to-urban migration.
At the same time, Athukorala (2006) found that the number of migrant workers per 1,000 of labor force has increased significantly from 1980s to early 2000s in Japan, Korea, Taiwan, Hong Kong, Singapore, Malaysia, and Thailand. The number has continued to rise in Malaysia and Korea despite the Asian crisis and in Japan in spite of its decade-long recession. Indeed, intra-Asian labor migration had increased approximately from 1 million in the beginning of 1980s to 6.5 million in 2002 (Huang and Guo, 2006). One possible reason would be the establishment of ASEAN Occupational Safety and Health Network in 2001. Eichengreen and Bayoumi (1999) have even indicated that labor mobility in East Asia was higher than that in Western Europe during the 1980s and 1990s. In 1999, the year the euro was adopted ten countries in Western Europe had some kind of minimum wage policy whereas only four East Asian economies had that kind of policy, suggesting that Asian wages could be relatively easily adjusted to clear the labor market (Ngiam and Yuen, 2001). For that year, the unemployment rates in East Asia were also lower than those in Western Europe.

The case for US dollar as the anchor
The big news last week was a speech by Zhou Xiaochuan, the governor of China's central bank, calling for a new "super-sovereign reserve currency"… But they are, apparently, worried about the fact that around 70 percent of those (the China's) assets are dollar-denominated, so any future fall in the dollar would mean a big capital loss for China. (Krugman, 2009) The above excerpt is from Paul Krugman's New York Times column published in April, 2009. Notwithstanding the "flaw" in the US monetary policy and financial sector, the article asserted that the dollar would remain robust in view of the fact that any dollar dumping by China would set downward pressures on the dollar, leading to huge capital loss for the republic. This is clearly reflected in Table 1 which exhibits the currency composition of official currency reserves in the world and in the emerging and developing economies. Despite its declining share, the US dollar is still the most dominant reserve currency till 2008.  Swiss francs Several other factors also make the US dollar the ideal monetary anchor. First, as widely recognized, the dollar is the vehicle currency for transactions across the world, particularly those pertaining to primary products (McKinnon, 2005). In East Asia, the dollar is also the preferred invoice currency even though Japanese trade is as large as the American one (McKinnon, 2005).
Second, the US is the most important export destination for most East Asian countries.
Third, the importance of dollar can also be recognized by looking at the regional breakdown of FDI inflows into the region (see Kawai and Takagi, 2005 in India (Patnaik and Shah, 2008) in spite of the Asian crisis. Many still opt to maintain their soft dollar pegs because any dollar depreciation will reduce the value of their dollar-denominated assets and increase the value of outstanding debt payments whereas any revaluation would impede their export competitiveness-an impasse duly labeled as "conflicted virtue". Moreover, when a large number of countries are pegging to a currency, it becomes difficult to break out of this pattern (Oomes and Meissner, 2008). 8 Fifth, the dollar is also the 'safe-haven' currency into which nationals in emerging markets fly in the face of a domestic financial crisis (McKinnon, 2005). Even when the Fed had been underperforming during the inflationary 1970s and early 1980s, and in the recent global financial crisis, the dollar-based system proved surprisingly resilient.
Lastly, other candidate currencies may not be suitable enough to serve as the monetary anchor for East Asia. Though Japan's influence in the region is undeniably significant, due to some considerations, the Japanese yen may not be the ideal numéraire.
First, Japan has been facing internal macroeconomic and banking problems and its yen had been very unstable against the dollar in 1985-1995 (Mundell, 2003). Second, because a large part of Japanese trade is invoiced in dollars, any changes in the yen-dollar rate would be passed through to domestic yen prices. This makes the Japanese domestic price levels vulnerable to exchange rate fluctuations. Third, besides the declining dependence on Japan, the economic structures and real business cycles in major Asian economies were significantly different from those in Japan (see Chow and Kim, 2003;Shirono, 2009). Finally, as shown in Table 1, the importance of the Japanese yen in official foreign reserves has actually been diminishing.
What about the Chinese renminbi as the anchor? At present, the Chinese currency is not convertible on capital account, and its financial system is not well developed (Mundell, 2003).
Moreover, the region's power rivalry between Japan and China still makes the dollar the currency of choice in the medium term future (Katada, 2008).
Despite all the above, the most devastating threat to an Asian dollar bloc, however, is the floating yen-dollar rate which may be chaotic when it swings sharply. However, should Japan is also a part of the dollar bloc, this setback virtually disappears. Asian countries were examined. Three country groups displayed significant correlations in exchange rate variability which were comparable to the European ones: (1) Singapore-Malaysia, (2) Singapore-Thailand, and (3) Singapore-Hong Kong-Taiwan. In another study, Loayza, Lopez, and Ubide (2001) utilized 1970-1994 data of 7 Asian economies to present evidence from an error components model. The shock dimensions examined were country-specific, sector-specific, and common shocks. The study discovered significant short-run and long-run co-movements of shocks that were comparable to the European ones in (1) Japan-Korea-Singapore-Taiwan and (2) Indonesia-Malaysia-Thailand.
Using 1978-1999 data and a dynamic factor model on 10 Asian economies, Lee, Park, and Shin (2004) discovered that the Asian region's common shocks in the 1990s were at least comparable to those in Europe. In particular, Indonesia, Korea, Malaysia, Thailand, and the Philippines shared higher degree of regional output co-movements. In 2005, Kawai and Takagi applied a variation of SVAR model to study the impulse response patterns of real GDP and price to exchange rate depreciations among 9 Asian economies. Time period used is 1970-1998.
Symmetry of response pattern in real GDP could be found in (1)  long-run relationships and Engle tests to check for short-run interactions in real outputs. Shortrun common business cycles were found in (1) Singapore-Thailand-Indonesia, and in (2) Hong Kong-Korea-China, as well as between (3) Japan and Taiwan.
Based on fuzzy cluster analysis, Nguyen (2007) 1990-1996, 1990-2000, 1999-2003, and 1990-2003. From 10 economies considered, only the Singapore-Malaysia grouping could weather all the periods. The criteria used are: symmetry of business cycles, volatility of real exchange rate, degree of openness to regional trade, inflation differential from the regional average, and level of export diversification. No reference country is assigned.
The OCA criteria used are volatility in real GDP, volatility in real exchange rate, volatility in interest rate, trade openness, and convergence of inflation. The adjusted Maastricht criteria are budget deficit/GDP, external debt/GDP, exchange rate volatility, inflation differential, and annual prime lending rate. Results for pre-crisis period indicated groupings of Indonesia- Another support came from Kawai (2008) who discussed how regional integration has been proceeding in trade, FDI, and other activities; exchange rate arrangements in Asia; possible unwinding of global payments imbalances and surges in capital inflows; and challenges for monetary coordination. Period studied is 1989-2003. Comparisons to post-euro EMU were made.
More recently, Sato, Zhang, and Allen (2009) managed to identify two prospective groups: (1) US-Taiwan-Hong Kong-Singapore and (2) Thailand-Malaysia-Singapore-Philippines-Indonesia-Japan. The study employed Johansen cointegration to check for long-run comovements of real outputs. Data series were 1978-2006 and were seasonally adjusted using Census X-12 method. Ten Asian countries were sampled. Interestingly, the ASEAN countries were prospective only when Japan was included.
Quah (2009) compared the values of the OCA dimensions, namely inflation convergence, export diversification, labor market flexibility, and external indebtedness of 17 Asian economies to those in EMU and dollarized countries. The anchor currency used is the US dollar. Dataset was segmented into 1980-1996, 1997-2000, and 2001-2007, which contain post-euroization and post-dollarization periods. Results suggested that inflation rates and levels of export diversification in Asia were comparable to those in dollarized economies; labor markets in the region were at least as flexible as those in EMU; external debt levels in Asia have fallen considerably in comparison to the dollarized countries, indicating reduced incentive to fix exchange rates to the dollar; and the most prospective countries for a dollar bloc were India, Thailand, and Malaysia.
7. Discussion and conclusion Some common threads can be observed from the review. First, though the empirical papers have used different methods, some common 'groupings' can still be found in many of the results.
Based on pre-1997-dataset results, two general groupings can be recognized: the Northeast Japan-Korea-Taiwan group and the Southeast Thailand-Malaysia-Singapore-Indonesia group.
The pre-crisis (growth period) data appear to have generated groupings by level of economic development, that is, the more developed Northeast group and the less developed Southeast group.
For those using pre-2000 datasets, the "Asian Tigers" Taiwan-Hong Kong-Singapore group, the 'crisis' Korea-Thailand group, and the Southeast Malaysia-Philippines-Indonesia group can be detected. Obviously, the dataset which encompasses the pre-crisis and the crisis period has produced the Asian Tigers group which has been robust during the crisis period, the crisis group which has been severely distressed, and the Southeast group which has been relatively less affected. When pre-2008 datasets are utilized, an 'extended Southeast' Korea-Philippines-Thailand-Malaysia-Singapore-Indonesia grouping can be commonly found. It is apparent that this group represents the countries which have been substantially impacted during the crisis but have since rebounded significantly. highly open small economy with total trade more than its GDP (see Kawai and Takagi, 2005). Of course Macau is also a highly open small economy. If Hong Kong, a China's territory, can be treated as a separate entity in the studies, so does Macau.
Else, the authors could have been adhering to another facet of the traditional theory when selecting the countries. Consider this: If a prevailing exchange rate regime, fixed or flexible, can maintain external balance without causing unemployment (or demand-induced wage inflation), that regime is optimal. (Kenen, 1969, p. 41).
The excerpt above is an interpretation given by Peter Kenen on the definition of optimality implied by Mundell (1961). Hence, if an Asian country has already achieved those objectives with existing exchange regime, moving into a monetary union is not necessary. But then again, the reasons for including certain countries and not including the others are rarely made clear.
Along these lines, it is not totally unfounded to conjecture that early theoretical views might have been overshadowed by current considerations 11 , such as the case for monetary union.
Secondly, while the early intellectual debate has been spurred by the fact that homogenous regions or countries hardly prevail and hence adjusting mechanisms such as factor mobility, product diversification, flexible exchange rates, etc., are needed to achieve the objectives of price stability, full employment, and external balance, the empirical literature appears to have always been on the search for homogeneous countries. There seems to be a consensus among the empirics that symmetrical countries are extremely important and can be identified in Asia so that adjustment mechanisms may be less needed. Nonetheless, as pointed out in Mongelli (2002) Frankel and Rose, 1998). Suppose OCA criteria are indeed endogenous, that is, achieved only after monetary union is formed, using post-euro benchmarks would be much more appropriate than using pre-euro benchmarks. In this respect, studies done after euroization, those which used post-1999 Asian data (e.g. Huang and Guo, 2006;Sato and Zhang, 2006) have obvious advantage of using the post-euro data. Among the papers reviewed, however, only Rana (2007), Kawai (2008), and Quah (2009) have demonstrated this in a limited manner. It is not unfounded to say that little has been done in view of the endogeneity criticism.
Lastly, notwithstanding the obvious reasons for the dollar as the monetary anchor for East Asia, the dollar has nevertheless gained little attention among the empirics. Perhaps due to the ambiguity of a center economy in Asia (unlike in EU where Germany is commonly accepted as the center economy), a monetary anchor is rarely set a priori in the empirical studies.
In conclusion, policymakers of the original ASEAN members may really need to look into further political and economic integration in light of the extensive indication that ASEAN could feasibly constitute an optimal currency area. Specifically, the US dollar may be a very good numéraire currency. To regional business managers, the finding here may confirm the decisions which have been based on the homogeneity assumption of the ASEAN region. In another aspect, decision-makers may consider locate some of their operations outside of ASEAN to reap the advantages of diversification. Nonetheless, one must be aware of the "shortcomings" implied in the empirical literature that might undermine the validity of the result.
Unquestionably, the conclusions made here are limited in the sense that only 20 empirical papers dated 1994-2009 have been reviewed and hence may not necessarily reflect the studies in the field at large. Meanwhile, Singapore, Japan, Thailand, Korea, Indonesia, and the Philippines had about two-thirds of their currency basket weights on the dollar. 8 Since the days before the Asian crisis, most Asian economies had informally soft-pegged their currencies to the dollar, a move which made them vulnerable to the depreciating yen. However, dollar pegs were entirely rational from the perspective of Asian economies-to facilitate hedging by merchants and banks against exchange risks, and to help central banks anchor their domestic price levels. Nevertheless, since their dollar pegs were 'soft', the obvious Achilles heel was the vulnerability to one-way speculation which struck during the crisis. In contrary, if their exchange rates were securely locked to the dollar with credible regional arrangements, the system as a whole would definitely be durable. 9 Demand shocks were not examined in this paper as they were thought to be unlikely to be invariant to demand management policies and currency regimes. 10 Long-run purchasing power parity (PPP) implies that real exchange rates are stationary. A vast literature has, however, shown that they are nonstationary. This is because fundamental macroeconomic variables that determine real exchange rates are nonstationary. A system of nonstationary real exchange rates may have a long-run equilibrium path in common since the individual nations will experience a set of common real macroeconomic shocks. This is termed as GPPP hypothesis. 11 For instance, Font-Vilalta and Costa-Font (2006) selected the set of countries according to availability of data and affiliation to the Japanese economy. Clearly it is not based on whether the existing arrangement is not optimal or otherwise. 12 First, residents of a country can hold claims to output in other countries. Dividends, interests, and rental revenues from these claims will insure income as long as output is imperfectly correlated. Such ex-ante inter-regional insurance allows the smoothing of both temporary and permanent shocks. Second, a country's residents can adjust their wealth portfolio in response to income fluctuations by buying and selling assets and by borrowing and lending on inter-regional, or international, credit markets. Such ex-post adjustment of asset portfolios allows for the smoothing of transitory shocks. This argument which was put forward by Robert Mundell can be found in Mongelli (2002).