The suitability of a monetary union in East Asia: What does the cointegration approach tell?

Abstract

The issue of whether or not to form a monetary union in East Asia remains a hot issue in the study of the East Asian economies. Most of the existing studies apply a framework focusing on the symmetric issue of the fundamental shocks and the extent of correlations by applying the Blanchard and Quah [O.J. Blanchard, D. Quah, The dynamic effects of aggregate demand and supply disturbances, American Economic Review 79 (1989) 655–673] structural vector autoregression (VAR) technique, which includes first-differenced variables in the model and examines only the bilateral relationships. However, the shock symmetry does not necessarily require the co-movements of the real output variables between the countries concerned. The present paper employs the Johansen [S. Johansen, Statistical analysis of cointegration vectors, Journal of Economic Dynamics and Control 12 (1988) 231–254] cointegration approach to check the long-run co-movements of real outputs among the East Asian countries, Japan and the United States to draw some implications about forming a monetary union in the region. The results suggest that some groups of Asian NIEs plus the United States will be potential candidates to form a monetary union. Mainland China is not suggested as a member country of a monetary union with any of the grouped economies. More interestingly, the ASEAN countries alone are not a feasible group to form a monetary union unless Japan is included, which has important implications for the role of Japan towards the formation of a regional monetary union.

Introduction

In recent years, the feasibility of forming a monetary union and establishing a regional (common) monetary unit in East Asia has been a subject of lively debate in industrial, governmental and academic arenas. The post-crisis experience has provoked calls among politicians for further monetary integration and alternative regional exchange rate arrangements in East Asia to enhance the stability and credibility of the exchange rate system. In 2000 ASEAN plus China, Japan and Korea (known as ASEAN + 3) agreed to establish a currency-swap network in order to help avert any future crisis,¹ and again in 2003 decided to establish an ASEAN Economic Community (AEC) by 2020. At the ADB 39th annual meeting in 2006, ASEAN + 3 finance ministers agreed to take over the ADB's controversial plan to launch an Asian Currency Unit (ACU) with the hope of making it the region's legal currency in the future.

According to the seminal work of Mundell [19] and McKinnon [18], the incentive for two economies to peg their bilateral exchange rates rises with the bilateral intensity of trade, flexibility of factor markets, and symmetry of underlying shocks. By doing so, both will be able to forsake nominal exchange rate changes as an instrument of adjustment and to reap the reduction in transactions costs associated with a common currency. In general, the identified preconditions for forming a monetary union include: (i) the openness and degree of goods market integration; (ii) factor market integration; (iii) similarity in economic structure and symmetry in (real) shocks; (iv) financial market integration; and (v) policy coordination.² Most of the existing studies focus on the symmetric nature of the fundamental shocks and use the Blanchard and Quah [5] structural vector autoregression (VAR) technique to identify the fundamental shocks. Specifically, the Blanchard–Quah approach typically imposes long-run restrictions so that only the supply (real) shock affects the real output growth in the long-run and both the supply and demand (nominal) shocks have a long-run influence on the domestic price inflation. Given that real output fluctuation is subject to supply shocks in the long-run, the symmetry in the supply shock is an important precondition for forming a monetary union. If the supply shocks to respective economies are symmetric, the cost of relinquishing the discretionary monetary policy is likely to be outweighed by the benefits of establishing a common currency. In contrast, if the shocks are asymmetric, it will be more costly to give up the benefits of autonomous monetary policy and, hence, to establish a monetary union.

The shocks correlation approach³ has several weaknesses. First, a correlation analysis of shocks identified by the structural VAR is inherently a bivariate method, whereas it is obvious that an analysis of OCA must be based on a multi-country framework. More specifically, what a bivariate approach reveals is just the country-to-country correlations without taking into account the relationship with other possible partner countries. A multi-country approach will thus be more appropriate to examine the monetary arrangement within a particular group of countries.

Second, it is important to distinguish between the short- and the long-run dynamics in consideration of a monetary union.⁴ If real output variables are not cointegrated amongst the countries concerned, each output variable wonders randomly over time, which may lead to a different growth path for each country. Since the nominal exchange rate changes as well as other macroeconomic stabilization policies have only transitory effects on the economy, the long-run economic divergence among the economies can be an obstacle to forming a monetary union. The Blanchard–Quah approach generally employs a bivariate VAR model including first-differenced variables, which as a result will be likely to remove any information about the long-run equilibrium relationship.

In contrast to the previous studies, the novelties of this paper are twofold. First, we apply the multivariate cointegration technique to the issue of an East Asian monetary union. Recently Cheung and Yuen [6] and Sato and Zhang [22] employ the cointegration technique to assess the integration issue, but Cheung and Yuen [6] focus only on the three Greater China economies (the Mainland, Hong Kong and Taiwan). Even though Sato and Zhang [22] investigate the whole East Asian economies, they employ a bivariate VAR of real output series for possible pairs of countries within a two-country framework. In contrast, the present paper features an investigation of whether a group of East Asian countries share a common stochastic trend of real outputs by using a multivariate VAR framework.⁵ Second, we attempt to investigate 60 groups of countries to detect possible regional currency areas, which is far more comprehensive than the previous literature.⁶ We include in this study Japan and the United States in addition to nine East Asian economies including three Asian NIEs (Korea, Taiwan and Hong Kong), ASEAN5 (Singapore, Malaysia, Indonesia, Thailand and the Philippines), and Mainland China to investigate the co-movements of the real output variables spanning a period from 1978Q1 to 2006Q4. Based on a multivariate cointegration approach, this study will provide important implications for cost effectiveness in establishing a regional monetary union.

The rest of the paper is structured as follows. In Section 2 we discuss the analytical framework for this study. Section 3 describes the data and presents the results of the empirical examination. Finally, Section 4 concludes the paper.