Developing of Emerging Integrated Currency

This research discusses constitution process and forecastingissue for Asian currency units (ACU) in terms of ten Asiancountries including Japan, Indonesia, Malaysia, Hong Kong, SouthKorea, Singapore, Thailand, Philippine, Taiwan and China. Thefirst, this article provides the analysis of forecastingperformance for ACU by employing variant of methods, e.g. BPN, RNN,TDRNN and GARCH. The results showed that ANNs has betterperformance than GARCH and BPN model presents prominent forecastingperformance in most of division conditions. The second part is toidentify the modified components for central value of Asiancurrency units (ACU) via BPN model and compare the results withradial basis function neural network (RBFNN). We constitute eightmodels to evaluate the impacts of macroeconomic variables oncentral value of ACU. The empirical evidence supports that oursample countries need to take foreign direct investment, externaldebt and bank?s claim on private sector into account for expressingthe central value of ACU more effectively. To verify contagioncauses for Asia flu, the experimental model is adaptivenetwork-based fuzzy inference system (ANFIS). The empirical resultwas indicated that the contagion effect would most likely beinfluenced by tight financial linkage and similarly macroeconomiccondition.