آیا دلار سنتی آسیایی می تواند رژیم نرخ ارز ایالات متحده توسعه یافته را برای در بر گرفتن ین ژاپن میخکوب کند ؟

Using daily data for a select set of four Asian exchange rates, namely the Hong Kong dollar, the Singapore dollar, the Taiwan dollar and the Thailand baht, from October 1985 to October 2002, we apply principal components analysis and the O-GARCH model to describe the evolution and persistence in the correlations over time. We also estimate 2-, 3- and 4-variable multivariate GARCH models, without imposing the assumption of constant correlations, to investigate volatility interaction amongst the currencies. To allow for fat tails in the distributions of exchange rate changes, we use the multivariate student-t distribution in maximising our log-likelihood functions. Our results indicate the possibility of designing an Asian exchange rate system involving a number of the region's currencies.

The Asian financial crisis 1997–98 (see AuYong et al., 2004 and Pathan et al., in press) raised many questions about the appropriateness of the region's exchange rate systems. Most regional monetary authorities have traditionally managed their exchange rates relative to the US dollar (see Frankel & Wei, 1994; Kearney & Muckley, 2007a,b; McKinnon, 2000 and McKinnon and Schnabl, 2004). This exchange rate setting practice was motivated by the possibility of achieving price stability alongside the goals of regional financial market and trade integration and the desire to avoid penalising resident investors exposed to foreign exchange rate risk. As the crisis revealed, however, this was not an optimal policy (see Kwan, 2001, McKinnon and Schnabl, 2003 and Mussa et al., 2000). Asian investment and trade linkages are increasingly focused beyond the US dollar area, to include Japan and European countries, and, as a result, a volatile Yen/US$ exchange rate has proven to be harmful for the economies concerned. For example, the US dollar appreciated relative to the Japanese yen by more than 40% from May 1995 through August 1998 and it is alleged that this played a central role in precipitating the Asian financial crisis. It is worthwhile elaborating on this point, many North and Southeast Asian countries, including China and Hong Kong SAR, Indonesia, Malaysia, Singapore and Thailand, already conduct most of their trade within the region, and a significant proportion of this trade occurs with Japan and, moreover, the degree of intra regional trade integration is rising over time (see Huang and Guo, 2006, Kearney and Muckley, 2005 and Shin and Wang, 2004). Furthermore, a characterising feature of the Asia region is the increasing level of foreign direct investment that originates to a significant extent in Japan or the European economies (see Fukao et al., 2003 and Gao, 2005). Moreover, Choudhry, Lin, and Peng (2007) document the growing importance of Japan with respect to long-run relations between the regions' stock markets since the Asian financial crisis. Taken together, these emerging features help to explain why the soft dollar peg maintained by many of the North and Southeast Asian countries rendered these countries increasingly uncompetitive and their exchange rate policies increasingly untenable during the late 1990s. Indeed, these emerging trade and investment linkages imply that a future scenario of a strongly appreciating or depreciating US dollar will have an even more profound influence on the region's economies unless better exchange rate arrangements are implemented. In recognition of this, there is an emerging debate about the need for more cooperative arrangements amongst the region's monetary authorities. Most advocated exchange rate arrangements involve variations on extending the received US dollar pegging regime to a (different) currency specific or a common basket peg arrangement or to an exchange rate system of mutual pegging such as the European Monetary System's Exchange Rate Mechanism. A significant recent development, in this regard, is the extension of the Chiang Mai Initiative in May 2005 from bilateral currency swaps alone to a type of ‘multilateralization’. The extended Initiative provides a pool of reserves valued at 40 Billion US dollars which central banks may draw upon if their currencies come under speculative attack. Researchers at key organisations, such as the Institute for International Monetary Affairs and the Asian Development Bank, have advocated the development of this framework to eventually facilitate the coherence of regional foreign exchange rate and related monetary policies (see Madhur, 2002 and Shinohara, 2006). Moreover, there exists an impressive array of Asian institutions with related responsibilities, including the ASEAN (+ 3) (Association of Southeast Asian Nations plus China, Japan and South Korea), the APEC (Asia Pacific Economic Cooperation), the EMEAP (Executives' Meeting of East Asia-Pacific Central Banks) and the six Markets group. These institutions, with the passing of time, are incessantly extending their remits with respect to supporting and developing regional financial and monetary arrangements. To date, evidence pertaining to informing a more coherent and robust exchange rate arrangement has focused almost exclusively on the determination of the conditional mean of North and Southeast Asian exchange rate returns. Seminal contributions include Bowman (2005), Frankel and Wei (1994), Hernandez and Montiel (2003) and McKinnon (2000). Overall, these articles suggest that while the US dollar continues to be the most influential regional currency, the Japanese yen and the euro exert rising effects on several currencies, especially since the Asian financial crisis. These findings point to the promising possibility of introducing a more co-ordinated basket peg type exchange rate regime in North and Southeast Asia (see Kawai, 2002, Ogawa and Ito, 2002 and Williamson, 2001). While there is a growing literature advocating the basket peg exchange rate regime, as a solution to the Asia problem, there is a paucity of results regarding the compatibility of the Asian currencies, from a time series perspective, with such an exchange rate arrangement. In this article, we provide preliminary evidence with respect to the possibility of extending the existing basket peg exchange rate regime to include the Japanese yen. Japan is the most important regional economy with respect to trade and also, at least prior to the late 1990s, with respect to foreign direct investment (see Kearney and Muckley, 2005 and Urata, 2001). Furthermore, researchers at the Bank of Japan have expressed an interest in deliberately pursuing the possibility of a more co-ordinated regional exchange rate arrangement in the future (seeKamada & Takagawa, 2005). In this article, we build on this possibility by examining three simple and intuitive statistical pre-conditions for the successful extension of the quasi US dollar pegging regime to a common basket peg including the Japanese yen. First, the Japanese yen exchange rates should exhibit approximate stability. Evidently, given the Yen/US$ exchange rate volatility at the crux of the problem, it is expected that the yen rates would not be perfectly stable. Nonetheless, this is an important criterion and should be satisfied to as great an extent as possible. Second, to the extent that the first pre-condition is not satisfied, the Japanese yen exchange rate returns ought to exhibit a high degree of multilateral correlation and, furthermore, the estimated multilateral correlation statistic should be persistent rather than erratic. The presence of this feature in the data would reflect a capacity for the yen to manage its value relative to the candidate set of currencies in the region. Otherwise, it is likely that different currencies may have significantly different optimal management policies with respect to innovations in the Japanese yen exchange rate. For example, Williamson (2001) reports that the Korean won and the Malaysian ringgit appear to have followed a depreciating Japanese yen and the Singapore dollar appears to have followed an appreciating Japanese yen. Finally, periods of relative tranquillity and periods of relative volatility should not coincide across the Japanese yen rate returns. Independent time-varying volatilities are tantamount to an intrinsic aversion – within a candidate bloc – to constraints with respect to co-movements between the exchange rate returns. The satisfaction of this pre-condition would suggest that volatility in the candidate set of currencies is determined by common factors. This is a desirable condition as it indicates that the candidate currencies are sensitive to similar factors, even in periods of relatively intense volatility, thus furthering the possibility of managing these rates together relative to a common currency basket. The remainder of this article is structured as follows. Section 2 summarises the data studied and identifies the set of four yen rates that demonstrates the greatest degree of stability, the first of the aforementioned pre-conditions. Section 3 inspects the same set of currency returns with respect to the criterion of multivariate correlation, relative to all currency sets of that size. Also, the degree of persistence in the multilateral correlations in this set of currencies is calibrated. These perspectives address the second pre-condition. Section 4 estimates the degree of volatility transmission in the set of currencies. This evidence pertains to the third statistical pre-condition. The final section summarises the main findings and draws together the main conclusions.

An extension of the traditional North and Southeast Asian exchange rate regime, i.e., an exclusive US dollar peg, to a common currency basket including the Japanese yen, is motivated by the emerging intra-regional and international nature of regional investment and trade linkages. These emerging linkages underscore the deficiencies of the traditional US dollar peg exchange rate regime and exacerbate the corresponding intrinsic susceptibility of that regime to financial crises. In the light of this motivation, this article has specified three simple and intuitive statistical pre-conditions for the successful extension of the traditional exchange rate regime to include the Japanese yen. These pre-conditions relate to the stability, multivariate correlations and volatility of yen exchange rate returns. Overall, the evidence gathered is consistent with the contention that the Hong Kong dollar, the Singapore dollar, the Taiwan dollar and the Thailand baht comprise an eligible nucleus to a more co-ordinated regional exchange rate arrangement, based on the extension of the traditional US dollar peg, to a basket of currencies including both the Japanese yen and the US dollar. Specifically, this article examined nine North and Southeast Asian currencies, observed at a daily frequency, over the period 1st October 1985 through 1st October 2002. The unconditional statistical properties motivated the adoption of the multivariate student t distribution in our volatility modelling and the aforementioned set of yen rates was identified as exhibiting the greatest degree of stability. This evidence concerns the first statistical pre-condition. This article then adopted a principal components framework alongside the orthogonal GARCH model and found that the same set of currencies exhibits the highest degree of multivariate correlation, of all sets of that size. The correlations were demonstrated to be markedly time varying, however, it was established that there was also a high degree of persistence in the level of correlation in this set of currencies. The persistent co-movements lend considerable support to the prospect of managing these currencies as a homogeneous set relative to the Japanese yen as well as the US dollar. This constitutes evidence that the selected set of currencies satisfies, to a convincing extent, the second pre-condition. Finally, the article calibrated the degree of volatility transmission in the set of currencies using a suite of 2- 3- and 4-variable, BEKK specified, t-distributed multivariate GARCH models. This multi-model approach, in conjunction with a carefully implemented estimation procedure, yields results that are relatively robust to idiosyncracies in the likelihood space. Evidence consistent with common factors determining volatility was revealed. This evidence pertains to the third statistical pre-condition. As the three statistical pre-conditions are approximately satisfied it is apparent that, from a time series perspective, these bilateral yen rates might act as the nucleus of a more integrative exchange rate arrangement in North and Southeast Asia.