World trade

III.Other papers provided by the members

1.The other papers provided by Members give interesting insights regarding the wider, perhaps less direct, channels linking the exchange rate, or exchange rate regimes, and trade. Two theoretical papers, one from Frenkel and Razin (1987), and one from Obstfeld (2001), looked at one of the leading explanatory models of the post-war open macroeconomics literature (the Mundell-Fleming model), to conclude that this model has lost a great deal of relevancy nowadays. The Mundell-Fleming model analyses, inter alia, how the mobility of international capital movements (and related openness to these movements) affects the effectiveness of domestic macroeconomic policies under alternative (fixed and flexible) exchange rate regimes. Obstfeld points to weaknesses in some the model's basic assumptions, for example that flexible prices and highly integrated markets are hardly observable in reality, and that the existence of market segmentation blurs some of the model's conclusions. Frenkel and Razin point to other limitations of the model, in particular that it ignores inter-temporal budget constraints and that it does not include economic agents' expectations.

2.A related paper is that of Chinn and Wei (2008), which examined whether flexible exchange rate regimes have eased - as would have concluded the Mundell-Fleming model - current account adjustments. They analysed data for 170 economies in the period 1971-2005. They found no evidence to support this claim, which was fairly popular during the late 60's and early 70's. According to their findings, the nature of the nominal exchange rate regime has little influence on real exchange rate adjustment, with the parameter exerting the strongest influence on the current account.

3.A number of papers made the point that the trade characteristics, structure and policy regime of a given country influence the way exchange rates impact trade. Kyriacou and Papageorghiou (2010), from the Central Bank of Cyprus, looked at the experience of their country in the run-up to the accession to the euro. According to them, a small open economy with high import content, such as Cyprus, had no interest in letting the exchange rate depreciate. Indeed, given the inelastic structure of imports, exchange rate depreciations would only lead to inflation, and a loss, not a gain, in market shares. For these reasons, the adoption of the euro was not found to be a constraint.

4.Li (2003) analysed the link between trade liberalization and exchange rate movements. He explained that in theory, trade liberalization should lead to a decrease in the relative prices of imports and to an increase in the relative prices of exports in relation to non-tradable goods. If as a result of this process, domestic prices fall below world prices, the real exchange rate would tend to depreciate. He tested this hypothesis for a panel of 45 developing countries and found that the real exchange rates of countries having liberalized trade the most in a short period of time were also those that had depreciated the most. At the same time, he found that countries with multiple liberalization episodes over time tended to see their currencies appreciate in real terms.

5.The wide differences of import and export price elasticities and the high degree of heterogeneity among countries, with varying effects on exchange rate adjustments, is further discussed by Imbs and Méjean (2010). In their paper, they provide calculations of the price elasticities of imports and exports for 30 developed and developing countries accounting for 85 per cent of world trade, considering data for manufactured goods between 1995 and 2004. The authors recognized that export elasticities are fairly large, ranging from 1.43 to 5.21. Developed countries tended to have lower values while developing countries had higher ones, although countries such as China (2.836) and India (2.610) had export elasticities close to those of OECD countries, such as the United States (2.944), France (2.894) and the UK (2.654).

6.Bauman (2010) studied whether regional production networks, in conjunction with preferential regional trade agreements, could generate ‘regional demand multipliers’ by fostering competitiveness and regional economic convergence. They looked at regional experiences in South Asia, East Asia, Central America, Andean Countries and Mercosur, between 1992 and 2008. For each region, the authors defined a ‘hub’ country, whose business cycle seemed to affect the activity of the other neighbouring economies. They chose Argentina, Brazil, China, India, Japan, Korea (Rep. of), and Mexico. According to their 'regional multiplier' hypothesis, due to regional production chains, positive shocks in these 'hubs' provided a demand stimulus to the other regional economies, leading to a virtuous cycle for all the countries in the region, with the hub countries absorbing more imports from the other countries in the region. They found evidence for this hypothesis in the case of East Asia, which has a share of regional trade of over 50 per cent (60 per cent of which denotes intermediate goods) and a high GDP convergence index. Latin America is the second region with the highest regional flows of intermediate goods and inputs, accounting for 20 per cent of its regional trade, although trade with the rest of the world remains far more important than regional trade. GDP convergence, according to the author's definition, is also lower than in Asia. The authors argued that there are marked differences among sub-regions in Latin America, with Mercosur presenting the highest (but still weak) signs of the 'regional multiplier' and Central America the lowest. Given that Latin America has grown significantly in the last few years, due to favourable conditions in the international market for commodities, the authors argue that this region faced a favourable conjuncture to evaluate how regional economic relations could sustainably foster intra-regional competitiveness, in the face of competition from other regions such as East Asia.

7.In this vein, the authors argue that a higher volume of intra-regional trade and convergence could be achieved in Latin America by setting a single currency, and that dollarizing could be a first step towards that end. The Andean Community of Nations (2001) compiled a book of seven interventions made in the context of a Seminar on “Dollarization in Ecuador: Effects on the Andean Trade”, held in Lima on August 25th 2000. Some speakers argued that dollarization promoted macroeconomic stability and convergence, and reduced inflation, an old problem in the region. The Director General of the Andean Community of Nations' (CAN) Secretariat made the point that the de facto dollarization of Bolivia and Peru made sense for the region. Other speakers underlined the challenges of dollarization. First, while a common monetary zone could have positive macroeconomic effects on the region, it would require a very high level of coordination and convergence of economic policies. Second, the domestic banking system had to be upgraded to sustain the capital inflows and outflows inherent to the adoption of the dollar as a currency. Moreover, it was argued that dollarization imposed a significant upgrade of competitiveness in some fragile countries, resulting in a real appreciation of the exchange rate. This view is consistent with the findings of the paper by Astroga et al. (2003), who argued that as a result of dollarization in Ecuador, exchange rate appreciation contributed to the deficit in the trade balance to a level of 10 per cent of GDP. The authors proposed in return a contingent mechanism that would subject selected imports to safeguards to offset the depreciations of trading partners.

8.The link between capital flows and the trade balance was mentioned, in particular, in Lane and Milesi-Ferretti (2002). Their paper provide empirical evidence of a positive long-run impact of a country’s net foreign assets position on real exchange rate appreciation, taking into account OECD countries between 1970-1998. They decomposed this relationship in two sub-components: the relation between the net foreign asset position and the trade balance; and the relationship between the trade balance and the real exchange rate. With regards to the first component, the authors found that the relation between the net foreign assets position and trade balance was driven by the rate of return differential between external assets and liabilities, and the growth rate of output of each particular country. With regards to the second component, the authors found that in the long-run, large trade deficits were associated with real exchange rates that had appreciated. They concluded by saying that the country size tended to increase the magnitude of the trade balance effect on the real exchange rate.

9.Some papers submitted by Members dealt with the relationship between exchange rates, domestic macroeconomic policies and global imbalances. In a paper analysing previous episodes of global imbalances, Krugman (1987) argued that narrowing global imbalances would require not only a shift in demand from deficit to surplus countries, but also a real exchange rate adjustment. In particular, if world spending moved away from the United States, there would be an excess supply of goods produced in that country, which required a relative price adjustment. Currency depreciation in the United States would solve the coordination problem of changing all the relative prices at once, by lowering all domestic wages relative to foreign wages. In a more recent paper, Rajan (2011) argued that global imbalances are the outcome of wrong domestic policies, overconsumption in developed countries, and over-exportation and savings in the largest emerging economies. "Fixing" currency policies would not be enough, and hence preventing a currency war should not be the main focus of international macroeconomic debates. According to this author, the only sustainable solution to global imbalances was a demand shift from developed to emerging economies. To achieve this, and avoid an accompanying credit boom-and-bust cycle in the emerging countries, tougher regulation would be needed in the banking sector of all countries, particularly systemically important ones. In particular, emerging countries ought to discourage short-term foreign currency-denominated loans to their banking system. Moreover, Rajan claims that a more stable and less aggressive US monetary policy is needed to achieve a more sustainable US growth rate, and hence reduce the volatility of capital flows to emerging countries. Birdsall and Fukuyama (2011) argued that emerging economies should adopt an more inward-looking policy agenda, based on increased reliance on domestic demand, continued accumulation of foreign reserves, strengthened regulation of the local banking sector, and lower reliance on the free flow of capital, asin the end such policies should help reduce global imbalances.

10.Improved macroeconomic policy coordination is often seen as a means of ensuring that the exchange rate does not hinder the expansion of international trade. Kasa and Huh (2001) proposed a model according to which central banks have incentives to coordinate their exchange rates in order to avoid competitive devaluations against one another, such as those that arose in the aftermath of the Great Depression of the 1930s. In this respect, Eichengreen and Irwin (2009) analysed the policy reactions that originated in the 1930s in the face of the global downturn, and found that in the 1930s, the world economy suffered from different, inconsistent and uncooperative economic policies, with some countries abandoning the gold standard while other kept it , leading the way for beggar-thy-neighbour reactions. In the same vein, Ahamed (2011) argued that the global economy was unlikely to see a spiral of competitive devaluations similar to the one that occurred during the Great Depression. His main arguments was that increased trade protection and competitive devaluations were not in countries' best interests in a world of global supply chains and production networks, where countries are far more dependent on imported inputs than any time in the past. In this respect, Busch and Levy (2010) provided a legal analysis of WTO provisions if a currency case was brought to dispute settlement. They predicted that such a case would not be successful for the claimant although they argued that the WTO would constitute an important forum for international dialogue and discussion regarding these issues.

11.A few papers have discussed the choice of particular currencies as transaction currencies in international trade. Using a theoretical model, Rey (2001) explained why there seems to be some inertia in the use of a specific international currency. Part of this inertia is linked to the fact that if multiple currencies are being used, higher transaction costs would pass through to export prices. There is an incentive to use only one invoicing currency to maintain lower international prices and competitiveness. The currency of reference is chosen according to the "thick market externality" principle, whereby the transaction costs of using a particular currency in the market are reduced with market size. Therefore, the currencies of countries with large trading power, high levels of openness and substantial bilateral trade flows are more likely to be chosen. Chandrasekhar (2010) argued that the resilience of the dollar as an international reserve and transaction currency was based on overall US strength in international relations. He acknowledged that using alternative reserve currencies, such as IMF Special Drawing Rights (SDRs), presented problems on its own. For example, SDRs can only be used by governments and not by private entities in regular transactions. The author calls for a reform of the current financial architecture.