Should China reduce its foreign reserves?

Before the Global Financial Crisis (GFC) China ran large trade surpluses and faced pressure to let the renminbi (RMB) appreciate. Having abandoned the peg to the U.S. dollar in 2005, the RMB has appreciated by 36% on a real effective basis.

Until 2009 China ran surpluses in both of its major customs regimes: processing trade and ordinary trade. Processed exports are final goods (such as tablet computers) that are produced using parts and components (such as microprocessors) that come primarily from east Asia. Ordinary exports are goods such as clothing and toys that are produced primarily using Chinese factors.  Since the GFC, China’s surplus in ordinary trade has disappeared while its surplus in processing trade has soared (see Figure 1).

Figure 1. China’s ordinary, processing, and total trade balances

Source: China customs statistics.

One possible explanation for these divergent responses is that the RMB appreciation since 2005 has affected the price competitiveness of ordinary exports more than the price competitiveness of processed exports. For ordinary exports, most of the value added is produced in China, while for processed exports, much of the value added comes from imported parts and components. Cheung et al. (2012) – using dynamic ordinary least squares (DOLS) estimation and quarterly data over the 1994-2010 period – find statistically significant price elasticities for China’s ordinary exports. They report that a 10% RMB appreciation would reduce ordinary exports by between 13% and 19%. Their findings imply that the RMB appreciation since 2005 has reduced steady state ordinary exports by 50% or more. On the other hand, processed exports are produced within global value chains. It is thus necessary – as Bayoumi et al. (2013) highlight – to take account of foreign value added when calculating the response of exports to exchange rates.

New evidence on exports

In Thorbecke (2014) I present data on value-added exchange rates for processed exports, and investigate their impact on trade. I conduct two types of tests over the 1993-2013 period.

• First I employ panel DOLS estimation with value-added exchange rates across the supply chain relative to individual countries importing processed exports.
• Second, I use Johansen maximum likelihood techniques with value-added real effective exchange rates and aggregate data on processed exports.

The first model employs annual data and the second model uses quarterly data. With these two very different data sets and methodologies, the results in every specification indicate that there is a strong and statistically significant relationship between exchange rates throughout the supply chain and processed exports. The findings from the panel data estimation imply that a 10% appreciation across the supply chain would reduce processed exports by between 13% and 19%. The findings from the time series estimation indicate that a 10% appreciation would reduce processed exports by between 22% and 29%.

While the RMB has appreciated by 36% since 2005, exchange rates in supply chain countries have depreciated.  This is clear in Figures 2 and 3. Figure 2 presents data from the panel estimation and Figure 3 from the time series estimation. Figures 2 and 3 use very different data sets, but tell a similar story. According to the measure in Figure 2, although the RMB has appreciated by 45% between 2005 and 2013, the value-added exchange rate for processed exports (the integrated exchange rate) has only appreciated by 7%. The reason for this is that weighted exchange rates in supply chain countries have depreciated. Figure 3 shows that although the RMB has appreciated by 36% on a real effective basis between 2005Q1 and 2013Q4, the integrated real effective exchange rate has only appreciated by 14%. The reason for this is that, according to this measure also, exchange rates in key supply chain countries have depreciated. Thus China’s surplus in processing trade has soared partly because depreciations in supply chain countries have offset the appreciation of the RMB.

Figure 2. Weighted averages of the bilateral integrated exchange rate, the bilateral exchange rate in supply-chain countries, and the bilateral RMB exchange rate with 24 importing countries

Source: The CEPII-CHELEM database, China Customs Statistics, and calculations by the author. Note: Weights are determined by the share of processed exports going to each of the 24 countries.

Figure 3. The integrated real effective exchange rate (IREER), the RMB REER, and the weighted REER in supply-chain countries

Source: The Bank for International Settlements, China Customs Statistics, the International Monetary Fund International Financial Statistics, and calculations by the author.

The two largest suppliers of parts and components to China – Taiwan and South Korea – ran global current account surpluses between 2005 and 2013 that averaged almost 9% and almost 3% of gross domestic product , respectively. Nevertheless, Taiwan’s real effective exchange rate depreciated during these nine years and Korea’s real effective exchange rate appreciated by less than 5%. Taiwan, Korea, and China accumulated foreign exchange reserves to slow exchange rate appreciations. For instance, China’s external reserves increased by $508 billion in 2013 and by $125 billion in the first quarter of 2014 (Troutman 2014).

Many Asian economists argue that further reserve accumulation is not beneficial for the region. Yu (2014) observes that resources are misallocated when central banks sterilise the impact of reserve accumulation on domestic liquidity, because small and medium-sized enterprises are denied access to funds. Yoshitomi (2007, p. 32) notes that “the same income that is being salted away in the form of dollar reserves could instead be used to underwrite investments in housing, water supply, roads, and infrastructure generally, not to mention education and health care for Asian populations.”  Fang et al. (2012) find that an additional year of education in China produced returns of 20% per year, far exceeding the return available from U.S. dollar reserves.

Conclusion

If central banks in east Asian surplus economies together reduced their rates of reserve accumulation and gave greater play to market forces, the surpluses that they run in processing trade and in their overall current account balances would generate pressure for a concerted appreciation of East Asian currencies against importers’ currencies. The evidence reported here indicates that such an appreciation would help to rebalance processing trade.

A joint appreciation would increase the purchasing power of Asian citizens and allow them to import more medicines, foods, and other goods from the rest of the world.  It would also have an attenuated effect on export competition between East Asian economies in third countries because their currencies would be appreciating together. Finally, it would help to maintain intra-regional exchange rate stability and thus facilitate the flow of parts and components within production networks (see Tang 2014).

Published in collaboration with VoxEU

Author: Willem Thorbecke is a Senior Fellow at Japan’s Research Institute of Economy, Trade, and Industry.

Image: Chinese 100 yuan banknotes are seen in this picture illustration taken in Beijing July 11, 2013. REUTERS/Jason Lee