AEPR series. This is the pre-peer- reviewed version of the following article: The Boom and Bust of the RMB's Internationalization:

Transcription

1 JCER Working Paper AEPR series No This is the pre-peer- reviewed version of the following article: The Boom and Bust of the RMB's Internationalization: A Perspective from Cross-Border Arbitrage, Asian Economic Policy Review, vol. 12, issue 2, which has been published in final form at and DOI: /aepr The Boom and Bust of the RMB's Internationalization: A Perspective from Cross-Border Arbitrage Ming Zhang (Chinese Academy of Social Sciences) Bin Zhang (Chinese Academy of Social Sciences) This paper was prepared for the Twenty fourth Asian Economic Policy Review (AEPR) Conference China s Financial Transformation, October 15, 2016, Tokyo. July 2017 Asian Economic Policy Review Japan Center for Economic Research

2 To authors If you want to introduce the same working paper you wrote and presented at the AEPR conference held in Tokyo on Oct. 15, 2016, in your own/your affiliation s website, please be aware the following requirements. To ensure that all citations and references to your published article are captured by the SSCI (Social Sciences Citation Index), authors are required to amend the cover page of your working paper as soon as practical after publication in AEPR. The amended cover page should include the full article citation, journal name, volume and issue, and DOI, as well as a hyperlink to the published article. The cover page of JCER Working Paper AEPR series has been already amended after publication in AEPR. The face of this working paper is an example of an amended working paper cover page.

3 The Boom and Bust of RMB Internationalization: the perspective from cross-border arbitrage Ming Zhang 1 Bin Zhang 2 September, 2016 Preliminary draft for comments only, please do not cite. Abstract RMB internationalization developed very fast from 2010 to 2015H1, measured by cross-border RMB settlement in trade and investment, the RMB deposit in offshore market, and the sign of bilateral local currency swaps between PBC and foreign central banks. However, the progress of RMB internationalization slowed down significantly since 2015H2, if measured by the above criteria. Cross-border arbitraging activities played a significant role in the boom and bust of RMB internationalization till now. From 2010 to 2015H1, when there was a strong RMB appreciation expectation and a wide onshore-offshore interest rate spread, Chinese and foreign enterprises took this opportunity to make cross-border exchange rate or (and) interest rate arbitrage. Because the arbitraging activities took place under the disguise of RMB settlement in cross-border trade, the rampant cross-border arbitrage resulted in a superficial fast development of RMB internationalization. On the contrary, from 2015H2 up to now, especially after the exchange rate reform on August 11, 2015, when the RMB appreciation expectation turned into depreciation expectation, the onshore-offshore interest rate spread shrank sharply, and the potential risks on China's domestic financial market rose observably, the incentives for cross-border arbitrage weakened significantly. The slowing down of the cross-border arbitrage caused a slower pace of RMB internationalization. A slower pace of RMB internationalization basing more on real demand might become a New Normal in the future. To pursue a more sustainable RMB internationalization in the next decade, Chinese government should maintain a relatively high economic growth, avoid the burst of systemic financial crisis, continue to liberalize the capital account in a gradual and cautious way, accelerate the reform and opening up of domestic financial market, and integrate RMB internationalization with Asian monetary cooperation. Key words: RMB internationalization, Cross-border arbitrage, Systemic financial risk, Capital account liberalization. JEL codes: E41, E58, F31. 1 Senior fellow, Institute of World Economics and Politics, Chinese Academy of Social Science, Beijing, China. Address: zhangming@cass.org.cn. 2 Senior fellow, Institute of World Economics and Politics, Chinese Academy of Social Science, Beijing, China. Address: zbiwep@cass.org.cn.

4 1.Introduction After the burst of U.S. subprime mortgage crisis, the international status of Chinese economy rose markedly, as the result of the sharp contraction of advanced economies. The main motivations for Chinese government to start to promote Renminbi (RMB) internationalization since 2009 included: Firstly, to raise the international status of RMB so as to match RMB's global status with Chinese economy's global status; Secondly, to reduce the reliance on the US dollar (USD) in cross-border trade and investment, which could not only reduce the exchange rate risk and exchange cost assumed by Chinese enterprises, but also lower the necessity for the People's Bank of China (PBC) to accumulate foreign exchange reserve (Zhang, 2015); Thirdly, to facilitate Chinese enterprises' international operations such as outward direct investment or obtaining funds on international financial markets; Fourthly, to stimulate the development of both RMB onshore and offshore financial markets (Kroeber, 2013); Finally, to push forward domestic structural reform by using RMB internationalization as a catalyst or a commitment device (Yu, 2012). Table 1 summarizes the major policies adopted by Chinese government since 2009, which shows that RMB internationalization were pushed forward along two tracks. One track was encouraging the RMB settlement in cross-border trade and investment, and the other track was developing offshore RMB markets in global financial centers such as Hong Kong, Singapore, London and Tokyo, etc. (Subacchi and Huang, 2012). If the above two tracks --- the RMB settlement in foreign trade and investment plus the development of offshore RMB markets --- were used to measure the progress of RMB internationalization, this process developed very fast from 2010 to 2014, however it slowed down notably from 2015 up to now. <Insert Table 1 Here> What had caused the boom and bust of RMB internationalization from 2010 till now? This paper argues that, the cross-border arbitraging activities played a critical role in both the fast development and the slowing down of RMB internationalization. Specifically speaking, when there was RMB appreciation expectation and wide onshore-offshore interest rate spread, the cross-border arbitrage went up, resulting in the seemingly boom of RMB internationalization. On the contrary, when there was RMB depreciation expectation and insignificant onshore-offshore interest rate spread, the cross-border arbitrage receded, leading to the bust of RMB internationalization. The rest of this paper is organized as following: The second section reviews the general development of RMB internationalization from 2010 to the first half of 2016; The third section analyzes the boom of RMB internationalization from the perspective of rising arbitraging activities; The fourth section explains the bust of RMB internationalization from the perspective of fading arbitraging activities; The fifth section provides policy suggestions for Chinese government about how to promote a more sustainable RMB internationalization; The final section concludes. 2.The Progress of RMB Internationalization from 2010 to 2016H1: a Boom and Bust Cycle From July 2009 Chinese government began to push forward RMB settlement in cross-border

5 trade. As Figure 1 shows, the scale of RMB settlement in international trade increased fast from RMB 18 billion in 2010Q1 to RMB 2.09 trillion in 2015Q3. Over the same time, the ratio of RMB settlement to all currency settlement in China's cross-border trade rose from 0.4% to 32.5%. However, Figure 1 also displays that both the scale of RMB settlement and the ratio of RMB settlement to all currency settlement dropped observably from 2015Q3 to 2016Q2. <Insert Figure 1 Here> In January 2011 and October 2011, Chinese government approved RMB settlement in outward direct investment (ODI) and foreign direct investment (FDI) respectively. As exhibited in Figure 2, the scales of RMB settlement in ODI and FDI went up obviously from 2012Q1 to 2015Q3. The scales of RMB settlement in ODI and FDI both peaked in September 2015, and after that RMB settlement in cross-border direct investment declined markedly. In addition, the scale of RMB settlement in ODI was persistently lower than that in FDI, which implied that RMB settlement in cross-border direct investment actually constituted a mechanism under which the offshore RMB flowed back to the onshore market. <Insert Figure 2 Here> As the result of the fast developments of RMB settlement in both international trade and investment, the role of RMB as a global payment currency ascended signally. Figure 3 demonstrates that from December 2011 to August 2015, RMB's share in global payment increased from 0.3% to 2.8%, and RMB's ranking in major global payment currencies rose from the 35th in October 2010 to the 4th in August In November 2015, IMF agreed to adopt RMB into the currency basket of SDR, which marked a milestone achieved by RMB internationalization, especially in the aspect of international payment. Nevertheless, as also shown in Figure 3, both RMB's share in global payment and RMB's ranking in major global payment currencies went down after August In June 2016, RMB's share in global payment was only 1.7%, and RMB's international ranking fell to the 6th. <Insert Figure 3 Here> Hong Kong's commercial banks began to provide RMB deposit service from early As Figure 4 shows, from 2004 to 2009 the scale of Hong Kong's RMB deposit grew very slowly. Only after Chinese government started RMB internationalization, did Hong Kong's RMB market experience a rapid development. From January 2010 to December 2014, the outstanding balance of Hong Kong's RMB deposit rose from 64 billion to 1.0 trillion (the peak), and over the same period, the ratio of RMB deposit to all deposit in Hong Kong went up from 1.0% to 10.0%. However, Hong Kong's RMB deposit began to shrink in 2015 and 2016H1. By the end of June 2016, the scale of Hong Kong's RMB deposit decreased to 712 billion, and the ratio of RMB deposit to all deposit in Hong Kong fell to 6.4%. Figure 5 reveals the same trend like Hong Kong in Taiwan and Singapore's RMB deposit. Taiwan's RMB deposit outstanding increased from 7 billion in January 2012 to 338 billion in June 2015 (the peak), and then decreased to 306 billion in June Singapore's RMB deposit outstanding rose from 60 billion in June 2012 to 234 billion in

6 June 2015 (the peak), and then dropped to 164 billion in March <Insert Figure 4 Here> <Insert Figure 5 Here> Some RMB-denominating financial products gradually emerged on offshore RMB markets, among which the dim sum bond (RMB-denominating bonds issued on offshore market) was obviously the most important one. China Development Bank (CDB) issued the first dim sum bond in Hong Kong in 2007, and from then on dim-sum bond market grew very fast. From 2007 to 2014, the annual scale of new issuance of RMB-denominating bond in Hong Kong increased from 10 billion to 197 billion. Nevertheless, under the persistent RMB appreciation expectation, the majority of the offshore RMB-denominating bond issuers were still the mainland enterprises or financial institutions which had RMB revenues. <Insert Figure 6 Here> In order to provide more RMB liquidity to offshore RMB markets in case of contingency, PBC signed dozens of bilateral local currency swaps with specific foreign central banks. As Figure 7 displays, by the end of June 2016, PBC had signed bilateral local currency swaps with over 30 foreign central banks, and the accumulate scale of these swaps reached RMB 3.5 trillion. However, so far only a few foreign central banks ever activated the bilateral local currency swap to obtain the RMB fund to satisfy the demand from their domestic market, including Bank of Korea and Monetary Authority of Singapore. <Insert Figure 7 Here> To summarize, RMB internationalization developed rather fast from 2010 to 2015H1, but the process slackened apparently since 2015H2. It seems that RMB internationalization has experienced a boom and bust cycle, which will be examined and explained in the rest of this paper. 3. The Boom of RMB Internationalization: the Rise of Arbitrage One of the most important characteristic of RMB internationalization was that Chinese government promoted the process before the full liberalization of both RMB exchange rate and RMB interest rate. Therefore, after the offshore RMB markets were established, two sets of RMB prices would co-exist. Generally speaking, the offshore RMB exchange rate and interest rate would be determined by the market demand and supply, yet the onshore RMB exchange rate and interest rate would be determined by PBC on a larger extent. Then from time to time, there would become a significant spread between onshore and offshore RMB prices, which would create profitable opportunities for both Chinese residents and non-residents to make cross-border arbitrage. More importantly, the international arbitraging activities would be done under the disguise of RMB settlement in cross-border trade, thus the active arbitrages would result in an illusory fast development of RMB internationalization. There were three basic types of cross-border arbitrages: the exchange rate arbitrage, the

7 interest rate arbitrage and the exchange rate plus interest rate arbitrage. (1) the exchange rate arbitrage As mentioned before, after the launching of RMB internationalization, there were two types of RMB exchange rates, i.e. the onshore exchange rate and the offshore exchange rate. The former were determined by PBC according the changes of market situation, while the latter were determined directly by the demand and supply from the offshore market. 3 Therefore, there would become a significant spread between onshore and offshore RMB exchange rates now and then (Figure 8). In a general way, if there was a RMB appreciation expectation, there would be a larger demand for RMB on the offshore market. Because the supply of RMB on the offshore market was much more limited, the offshore RMB exchange rate would be more expensive than the onshore one under RMB appreciation expectation. On the contrary, if there was a RMB depreciation expectation, the offshore RMB exchange rate would be cheaper than the onshore one. <Insert Figure 8 Here> The distinct exchange rate spread between onshore and offshore RMB markets would offer a strong incentive for domestic and foreign enterprises to make cross-border exchange rate arbitrage. The basic model for the international exchange rate arbitrage was: when there was a RMB appreciation expectation, the offshore RMB/USD exchange rate would be higher than the onshore one, therefore it would be profitable for the enterprises to move RMB funds from the onshore market to the offshore market. However, because there were still regulations on cross-border capital flows by PBC, the arbitraging enterprises tended to move the RMB funds abroad through the payment for the unreal import of goods or services. On the contrary, when there was a RMB depreciation expectation, the offshore RMB/USD exchange rate would be lower than the onshore one, hence it would be profitable for the enterprises to move RMB funds from the offshore market back to the onshore market, this time through the revenues from the unreal export of goods or services. Two evidences about the cross-border exchange rate arbitrages could be found. Firstly, according to the above arbitrage model, if there was a RMB appreciation expectation, more RMB payment would take place in the import side (from the mainland's perspective); if there was a RMB depreciation expectation, then more RMB payment would take place in the export side (also from the mainland's perspective). As Figure 9 demonstrates, from 2011Q1 to 2014Q4 (only except for 2012Q3), the quarterly ratio of RMB payment in import to RMB payment in export had been persistently higher than 1.0, which implied the cross-border exchange rate arbitraging activities under the RMB appreciation expectation. However, after RMB appreciation expectation reversed, the quarterly ratio of RMB payment in import to export was persistently lower than 1.0 from 2015Q1 to 2015Q3, which reflected the cross-border exchange rate arbitrage under RMB depreciation expectation. Figure 10 exhibits that the exchange rate spread between CNY and CNH market had a strong and positive correlation with the ratio of RMB payment in import to export, 3 Shu et al. (2014) found that the changes in the exchange rate of RMB/USD in both onshore and offshore markets have a economically and statistically significant impact on the changes in Asian currency rates against USD. Moreover, the CNH's impact on Asian currency rates was distinct from the CNY's impact.

8 which meant that when the offshore RMB exchange rate was more expensive than the onshore one, more RMB settlement would happen in the import side, and when the offshore RMB exchange rate was cheaper than the onshore one, then more RMB settlement would happen in the export side. <Insert Figure 9 Here> <Insert Figure 10 Here> Secondly, according to the above arbitrage model, under the RMB appreciation expectation, more RMB funds would flow from the onshore market to the offshore market, causing the expansion of the offshore RMB deposit; oppositely, under the RMB depreciation expectation, more RMB funds would flow from the offshore market back to the onshore market, thus resulting in the contraction of offshore RMB deposit. As shown in Figure 4 and 5, the scales of RMB deposit in Hong Kong, Taiwan and Singapore all rose from 2010 to 2015H1 when RMB appreciation expectation was strong, however they all dropped from 2015H2 till now. To provide an additional evidence, Figure 11 displays that the change of Hong Kong's RMB deposit was strongly and positively correlated with the exchange rate spread between CNY and CNH market, which again emphasized the relationship between the direction of RMB flows and the expectation about RMB exchange rate. <Insert Figure 11 Here> (2) the interest rate arbitrage Chinese government has not fully liberalized its capital account up to now, which means that PBC still has the capacity to maintain an independent domestic interest rate. As Figure 12 shows, Chinese domestic deposit interest rate has been persistently higher than Hong Kong's RMB deposit interest rate since early 2009, and this interest rate spread would create a strong incentive for the enterprises to make cross-border interest rate arbitrage. However, Figure 12 also displays that the onshore-offshore interest rate spread has narrowed significantly since 2014Q4. Chinese economic growth kept slowing down in the recent years, so PBC began to cut benchmark interest rates since 2014Q4, which led to the shrinking of onshore-offshore interest rate spread. <Insert Figure 12 Here> If onshore RMB interest rate was persistently higher than the offshore one, the way of cross-border interest rate arbitrage would be that the enterprises borrowed RMB funds from the offshore market and then moved RMB funds back to the onshore market to obtain higher yields. Because there were still regulation on cross-border capital flows by PBC, the arbitraging enterprises tended to make the arbitrage under the disguise of RMB settlement in international trade. Figure 13 summarizes the basic model for onshore-offshore interest rate arbitrage, which includes four major steps. Firstly, the domestic enterprise asked a domestic commercial bank to issue a long term letter of credit (L/C); Secondly, the domestic enterprise sent the L/C to its subsidiary enterprise in Hong Kong, as the payment for an ostensible import from the latter; Thirdly, the Hong Kong subsidiary used the L/C as a collateral to apply for a RMB loan from a

9 Hong Kong commercial bank; Fourthly, the mainland enterprises obtained the RMB funds from its Hong Kong's subsidiary, as the revenue from an ostensible export to the latter. There would be two results of the above basic arbitrage model (which was called the internal guarantee and external borrowing model): on the one hand, the Chinese domestic enterprise obtained a low interest rate loan from the Hong Kong commercial bank; On the other hand, to facilitate the cross-border fund movement, two faked RMB trade settlement took place, the first settlement in import and the second settlement in export, from the mainland's perspective. <Insert Figure 13 Here> Figure 14 provides some evidence about the cross-border interest rate arbitrage. From the beginning of 2009 to the mid of 2014, Hong Kong commercial banks' claims on both Chinese mainland commercial banks and non-bank customers rose fast, which meant that Chinese entities borrowed heavily from Hong Kong's commercial banks during that period. However, over the same period, Hong Kong commercial banks' liabilities to both Chinese mainland commercial banks and non-bank customers only increased very moderately. This significant mismatch about cross-border borrowing and lending testified the dynamic onshore-offshore interest rate arbitrage. Nevertheless, figure 14 also demonstrates that, from the end of 2014 to 2016H1, Chinese entities' borrowing from Hong Kong's commercial banks fell significantly, as the result of the shrinking of onshore-offshore interest rate spread. <Insert Figure 14 Here> (3) the exchange rate plus interest rate arbitrage In the past several years, a more profitable arbitraging strategy was to combine the cross-border exchange rate arbitrage with the interest rate arbitrage. For example, under the RMB appreciation expectation, the Hong Kong subsidiary in Figure 13 could apply for a USD loan, not RMB loan, from the Hong Kong commercial bank. After that, the Hong Kong subsidiary could exchange the USD funds into RMB funds on Hong Kong's market, then it would sent the RMB funds to the mainland through the channel of RMB trade settlement. By doing that, the enterprise could not only make profit from the onshore-offshore interest rate spread, but also enjoy the benefit from RMB's appreciation against USD. However, the risk of this mixed arbitraging strategy was that, if RMB began to depreciate again USD, the potential profit of this arbitrage would decline fast, or even turn into loss. 4. The Bust of RMB Internationalization: the Declining of Arbitrage As demonstrated by Section 2, RMB internationalization achieved a fast progress from 2010 to 2015H1, however it decelerated notably from 2015H2 up to now. This section tries to prove that the declining of cross-border arbitrage played a crucial role in the bust of RMB internationalization recently. The US Fed began the normalization of its monetary policy since In 2014 the Fed exited from the quantitative easing (QE) policy, then Fed increase the federal rate firstly in the end of The normalization of US monetary policy made USD the strongest currency during the period of 2013 to Because RMB was roughly pegged to USD during that period, this peg

10 made RMB's effective exchange rate appreciate dramatically. However, considering that Chinese economy growth kept declining over the same time, thus the fast appreciation of effective exchange rate finally caused the over-value of RMB. As shown in Figure 15, from 2014Q2 to August 2015, the daily opening price of RMB/USD exchange rate was persistently higher than the closing price, implying that there was a strong depreciation expectation about onshore RMB/USD exchange rate, and PBC tried to prevent the large depreciation of RMB against USD through intervening (pushing up) the daily opening price. As the result of PBC's effort, although there had been a heavy depreciation pressure since 2014Q2, the RMB/USD exchange rate was maintained at the level of 6.1 to 6.2 from 2014Q2 to the exchange rate reform on August 11, <Insert Figure 15 Here> To facilitate the inclusion of RMB into the SDR's currency basket, PBC started a new round of RMB exchange rate reform on August 11, 2015 (8/11 reform). The key point of the 8/11 reform was that PBC gave up the authority to manage the daily opening price of RMB/USD exchange rate, and then let the opening price equal to the closing price of the last trading day. Because there was still a heavy RMB depreciation pressure on the market during the reform, the direct result of 8/11 reform was the dramatic depreciation of RMB/USD exchange rate. From August 10 to August 13, the RMB/USD exchange rate dropped from to 6.401, depreciating about 4.7% within 4 days. The dramatic RMB depreciation gave rise to large market fluctuations on both domestic and international financial markets. As the result, just a few days after 8/11 reform, PBC had to step in to resume intervention so as to stabilize RMB/USD exchange rate. Considering PBC had already given up the authority to manage daily opening price of RMB exchange rate, after 8/11 reform Chinese central bank had to rely on the open-market operations to stabilize the exchange rate, i.e. by selling USD and purchasing RMB persistently and in large scale. This open market operation would naturally lead to the loss of foreign exchange reserve. From July 2015 to July 2016, the scale of China's foreign exchange reserve shrank from USD 3.65 trillion to USD 3.20 trillion. In order to strike a balance between stabilizing RMB's exchange rate and saving foreign exchange reserve, PBC allowed the RMB/USD exchange rate depreciated gradually and mildly after 8/11. From the end of July 2015 to the end of August 2016, the RMB/USD exchange rate felled from to 6.691, depreciating about 9.4%. In other word, PBC tried to further liberalize RMB exchange rate formation mechanism in 8/11 reform, however it was forced to intervene the exchange rate against shortly after the reform. Therefore as the result of this incomplete reform was, the RMB depreciation expectation did not vanish, but persisted, and even exacerbated. Under the persistent RMB depreciation expectation, Chinese economy began to face a continuous capital outflow. As Figure 16 displays, from 2014Q2 to 2016Q2, China had been facing persistent capital account deficit for nine quarters. More importantly, the capital account deficits were larger than the current account surpluses from 2015Q3 to 2016Q2, which meant that the demand for USD would be much larger than the supply of USD on China's foreign exchange market, further pushing down RMB/USD exchange rate. To put another way, the capital outflow and the RMB depreciation expectation could reinforce each other, and even form a vicious cycle. <Insert Figure 16 Here>

11 After the analysis on the breakdown of China's capital outflow, we can find that the outflow of cross-border borrowing and lending (the other investment) dominated the net capital outflow up to now, and both residents and non-residents played a significant role in the outflow of other investment. For Chinese residents, along with the economic growth slowing down and capital account opening up, both the households and the enterprises had a stronger incentive to invest abroad to diversify their portfolios. For the non-residents, in the past foreign financial institutions provided a large amount of loans to Chinese entities (as mentioned before in the part of cross-border interest rate arbitrage). After the RMB appreciation expectation turned into depreciation expectation, Chinese entities began to pay back the foreign borrowing in order to avoid further loss on the exchange rate. Both the residents' asset diversifying and the non-residents' withdrawing of loans made significant contributions to China's capital outflow so far. Another important incentive for both residents and non-residents to move funds out of China was that, along with the declining economic growth, the investment returns in China decreased significantly and the potential financial risks rose quickly. One of the most unique characteristic of Chinese economic growth model was the high investment in both manufacture and infrastructure sectors. In the past, Chinese manufacturing enterprises borrowed heavily from domestic commercial banks to make investment. When both internal and external demands were strong, the high production capacity of Chinese enterprises could be utilized fully, finally bringing in high profits. However, after the burst of US subprime crisis, both external and internal demands shrank notably, which led to a serious over-capacity problem to Chinese manufacture sector. Not only the traditional sectors such as the coal, steel, cement and PVC faced serious over-capacity, but also some new emerging sectors such as solar energy, wind energy and ship building also faced the same problem. The direct influence of over-capacity in manufacture sector was the declining of the profitability of industrial enterprises. As shown in Figure 17, in the past the average growth rate yoy of China's Industrial value added was about 14% to 15%, yet this growth rate dropped to only 5% to 6% recently, meaning that the profit margin of Chinese manufacturing enterprises deteriorated seriously. <Insert Figure 17 Here> Moreover, because the high capacity in Chinese manufacture sector had been supported by the high leverage, the exacerbation of over-capacity would unavoidably result in the deleveraging process of Chinese corporate sector, causing big problem to domestic commercial banking sector. As Figure 18 demonstrates, Chinese commercial banks began to face a negative impact since The non-performing loan ratio kept rising and the ratio of provision to loan kept decreasing from then on. Furthermore, considering that the commercial banks had a strong incentive to hide the real NPL numbers by moving the bad assets out of the balance sheets and by rolling over the bad loans, the real problem faced by Chinese commercial banks would be much more serious than the official statistics. <Insert Figure 18 Here>

12 To make matters worse, Chinese property market has been under a obvious bubble and thus has been facing a downward adjustment pressure. There are two prominent problems at present. Firstly, there is a huge property inventory accumulated in Chinese 3-tier and 4-tier cities, which would be very difficult to digest. Secondly, the property price in Chinese 1-tier cities is so high that Shenzhen and Shanghai's property prices could dwarf any other international financial centers such as London, New York or Tokyo, if measured by any traditional indicator such as the ratio of house price to rent, or the ratio of house price to family income. For example, as the data from Wind database shows, by the end of 2015, the ratio of property price to family income in Beijing and Shanghai reached 20 years, and the ratio about Shenzhen reached 25 years! Considering that Chinese property bubble has been driven mainly by banking credit, so any significant downward adjustment of the property market would result in a severe loss in Chinese banking sector. The darkest scenario would be that the cooperate sector deleveraging and the downward adjustment of property market happen together. If this scenario comes true, Chinese commercial banking sector would be destined to suffer a systemic crisis. Along with the continuous declining of the economic growth, PBC started a new round of monetary policy easing since 2014Q4. From 2014Q4 up to now, there has been 6 times of benchmark interest rates cutting and 5 times of required reserve ratio cutting. The frequent lowering of benchmark interest rates caused two consequences: First, the onshore-offshore interest rate spread shrank markedly, weakening the incentive for the enterprises to make cross-border interest rate arbitrage. As shown in Figure 12, the onshore-offshore RMB deposit interest rate spread fell from 3% in 2011 to less than 1% in 2016H1; Second, the shrinking of onshore-offshore interest rate spread would further pushing down RMB/USD exchange rate. As a matter of fact, the reason why PBC slowed down the pace of monetary policy easing in 2016H1 was exactly to avoid that the further cutting of either benchmark interest rate or required reserve ratio would cause a stronger RMB depreciation expectation and thus trigger a larger capital outflow. In other word, PBC began to face a traditional international trillemma: when domestic capital kept flowing out, to maintain the stability of RMB/USD exchange rate, PBC had to give up the monetary policy independence in some sense. Therefore, in an effort not only to increase the monetary policy effectiveness but also to alleviate the RMB depreciation pressure, PBC strengthened capital account regulations significantly after 8/11 reform. Now it is much more difficult for both Chinese households and enterprises to move money abroad, compared to the period before August As the result of the strengthening capital account controls and the postpone of Fed's interest rate hiking, Chinese economy faced a weaker pressure in capital outflow and RMB depreciation in 2016H1. However, the strengthening of capital account control had two other results: First, it reversed PBC's effort to accelerate capital account liberalization since 2012, impairing PBC's policy reputation; Second, it had a negative impact on the process of RMB internationalization. Finally, when RMB exchange rate faced great depreciation pressure after 8/11 reform, PBC made interventions not only on the onshore but also on the offshore market. One example is that, to limit the potential ability of the international speculators to short RMB on the Hong Kong market, PBC required the Hong Kong branch of Bank of China (who was the major supplier of RMB in Hong Kong) to stop lending RMB to foreign investors. Another example is that, there were market rumors that PBC required some Hong Kong branches of Chinese major commercial banks to sell USD for RMB on the Hong Kong market, which aimed to prevent the further depreciation

13 of RMB's CNH price and to control the further expansion of CNH/CNY exchange rate spread. Undoubtedly, PBC's intervention on the offshore market did achieve some effects in the short term, however these measures also impaired the independence of offshore RMB market, thus weakening the confidence of international investors about offshore RMB markets in the future. To sum up, several important factors caused the declining of cross-border arbitraging activities, leading to the bust of RMB internationalization: The first was the change from persistent RMB appreciation expectation to persistent depreciation expectation, as the result of both the over-appreciation of RMB's effective exchange rate and PBC's actions during and after 8/11 reform; The second was the shrinking onshore-offshore interest rate spread, resulting from PBC's easing of monetary policy to revive Chinese economic growth; The third was the rising of domestic systemic financial risks, especially the risks inside Chinese banking sector, as the result of corporate sector deleveraging and downward adjustment of the property market; The fourth was PBC's new measures to strengthen capital account controls, which aimed to increase monetary policy effectiveness and to mitigate RMB depreciation pressure; The fifth was PBC's interventions on the offshore RMB market to relieve RMB depreciation pressure. As the incentives for cross-border arbitrage weakened, the pace of RMB internationalization slowed down accordingly. 5. How to Pursue a More Sustainable RMB Internationalization in the Future? RMB internationalization developed fast during 2010 to 2015H1, but it slowed down observably since 2015H2. As proved before, cross-border arbitrage played an important role in the boom and bust of RMB internationalization. Just as Yu (2014) pointed out, the profits earned from cross-border arbitrage were the main drivers of, but could not constitute a sustainable basis for, RMB internationalization. Therefore, the deceleration of RMB internationalization may not be terrible. After the bubble of cross-border arbitrage receded, the RMB internationalization might be driven more by real needs, thus the future path might be more solid and sustainable. Here we put forward some policy suggestions to promote the more sustainable RMB internationalization in the future. First, Chinese government should try to maintain a relatively high and efficient economic growth in the next decade, by adopting both structural reforms and counter-cyclical macro-economic policies. In the long run, the future of RMB internationalization would be determined by the trajectory of Chinese economic growth. Moreover, the relatively rapid growth of Chinese economy could also increase the opportunity for the international community to use RMB (Huang et al., 2014). To deal with the decline of China's potential economic growth rate caused by demographic aging, the diminish yield of capital investment and the slowing down of total factor productivity growth, Chinese government should push forward the structural reforms such as breaking the monopoly of SOEs in service sector and opening up this sector to private enterprises, liberalizing domestic factor prices, and accelerating the income redistribution reforms both across different sectors and inside household sector. To deal with the current negative output gap, Chinese government should further loose both fiscal and monetary policy. Second, Chinese government should try to avoid the burst of systemic financial crisis in the next decade. As mentioned before, Chinese commercial banking sector tend to face a significant and negative impact in the near future, as a result of the deleveraging of corporate sector and the downward adjustment of property market. To avoid the breaking out of a systemic banking

14 crisis, Chinese government must adopt a sound counter-cyclical macro-prudential policy framework, make preparations for the exacerbating of commercial banks' balance sheets, and use more market-oriented measures to deal with the non-performing loans in the future. Third, Chinese government should follow a more appropriate sequence for the financial reforms, which means that the capital account liberalization in the future should still be gradual and cautious. We argue that there are at least three preconditions for the full liberalization of China's capital account: The first is the full liberalization of RMB's interest rate; The second is the full liberalization of RMB's exchange rate (Xu and He, 2016); The third is that the existing financial vulnerabilities on domestic financial market should be handled firstly. Eichengreen (2015) warned that Chinese government should not emulate European's experiences which removes capital account controls before prudential regulation framework had been adequately established and strengthened. Qu et al. (2013) also pointed out that, before the full opening up of its capital account, Chinese government must put its domestic financial house in order, by developing a stronger institutional and regulatory framework to not only improve market efficiency but also control financial risks. If Chinese government accelerate the capital account liberalization before the satisfying of the pre-conditions, there might be a massive and persistent capital outflow, hurting both the economic growth and financial market stability, or even triggering a systemic financial crisis. Fourth, Chinese government should continue to reform the financial markets, making them more liquid, deep and diversified. Ultimately, the competitiveness of one currency is determined by the development and openness of the financial markets of the issuing country (Ito and Chinn, 2014). As pointed out by Frankel (2012), by the criteria of openness, breadth and liquidity of financial markets, RMB still had a long way to go to catch up with other major international currencies, especially USD. Therefore, Chinese government should speed up the reforming and improving of the financial markets, especially the stock market and the bond market. As for the stock market, an exit mechanism for the listed companies should be introduced, the current approval based system of stock offering should be reformed into a registered based one (Cruz et al., 2014), and more domestic and international institutional investors should be welcomed. As for the bond market, more government bonds should be issued by Chinese central and local governments (Ma and Yao, 2015), the fragmented corporate bond markets should be unified and regulated by a singular authority, and more defaulting events should be allowed to release the potential credit risks. Fifth, Chinese government should try to integrate the RMB internationalization with the Asian monetary cooperation. In the future, more real demand for RMB would come from Asian countries, not remote European or American countries, considering the current distribution of global production networks. Gao and Yu (2012) asserted that the regionalization of RMB would be an inevitable step toward the internationalization of RMB. Subacchi (2014) found that the experience of Japanese Yen's internationalization highlighted the importance of regional integration as a method to overcome market inertia and network externalities. Fratzscher and Mehl (2013) approved that since the mid 2000s RMB had become a key driver of currency movements in Asia. Therefore, Chinese government should continue to encourage other Asian countries to use RMB in international payments, financing and investment. For example, Chinese government should encourage the foreign governments, financial institutions and companies to issue RMB-denominating bond on Chinese domestic market (Panda Bonds). Chinese government

15 should also try to establish a regional financial safety network with other Asian countries, on the basis of CMIM and AMRO. Actually, only after RMB become a regional currency, does RMB have the capacity to grow into a real global currency. 6. Conclusion RMB internationalization developed very fast from 2010 to 2015H1, measured by cross-border RMB settlement in trade and investment, the RMB deposit in offshore market, and the sign of bilateral local currency swaps between PBC and foreign central banks. However, the progress of RMB internationalization slowed down significantly since 2015H2, if measured by the above criteria. Cross-border arbitraging activities played a significant role in the boom and bust of RMB internationalization till now. From 2010 to 2015H1, when there was a strong RMB appreciation expectation and a wide onshore-offshore interest rate spread, Chinese and foreign enterprises took this opportunity to make cross-border exchange rate or (and) interest rate arbitrage. Because the arbitraging activities took place under the disguise of RMB settlement in cross-border trade, the rampant cross-border arbitrage resulted in a superficial fast development of RMB internationalization. On the contrary, from 2015H2 up to now, especially after the exchange rate reform on August 11, 2015, when the RMB appreciation expectation turned into depreciation expectation, the onshore-offshore interest rate spread shrank sharply, and the potential risks on China's domestic financial market rose observably, the incentives for cross-border arbitrage weakened significantly. The slowing down of the cross-border arbitrage caused a slower pace of RMB internationalization. A slower pace of RMB internationalization basing more on real demand might become a New Normal in the future. To pursue a more sustainable RMB internationalization in the next decade, Chinese government should maintain a relatively high economic growth, avoid the burst of systemic financial crisis, continue to liberalize the capital account in a gradual and cautious way, accelerate the reform and opening up of domestic financial market, and integrate RMB internationalization with Asian monetary cooperation. Reference Cruz, Prince Christian & Yuning Gao & Leilei Song. (2014). The People's Republic of China's Financial Markets: Are They Deep and Liquid Enough for Renminbi Internationalization? ADBI Working Paper Series, No.477. Eichengreen, Barry. (2015). Sequencing RMB Internationalization, CIGI Papers, No.69. Frankel, Jeffrey. (2012). Internationalization of the RMB and Historical Precedents, Journal of Economic Integration, 27(3), Fratzscher, Marcel and Arnaud Mehl. (2013). China's Dominance Hypothesis and the Emergence of a Tri-polar Global Currency System, The Economic Journal, 124(12), Gao, Haihong and Yongding Yu. (2012). Internationalization of the Renminbi, BIS Papers, No.61. Huang, Yi Ping & Daili Wang & Gang Fan. (2014). Paths to a Reserve Currency: Internationalization of the Renminbi and its Implications. ADBI Working Paper Series, No.482.

16 Ito, Hiro and Menzie Chinn. (2014). The Rise of the "Redback" and the People's Republic of China's Capital Account Liberalization: An Empirical Analysis of the Determinants of Invoicing Currencies, ADBI Working Paper Series, No.473. Kroeber, Arthur. (2013). China's Global Currency: Lever for Financial Reform, Brookings-Tsinghua Center for Public Policy, Monograph Series, No.3. Ma, Guonan and Wang Yao. (2015). Can the Chinese Bond Market Facilitate A Globalizing Renminbi? Fung Global Institute Working Paper, May People's Bank of China. (2015). RMB Internationalization Report 2015, June. Qu, Hongbin & Junwei Sun & Mackel Paul & Ju Wang. (2013). The Rise of the Redback II: An Updated Guide to the Internationalization of the Renminbi, HSBC Global Research, March Shu, Chang & Dong He & Xiaoqiang Chen. (2014). One Currency, Two Markets: The Renminbi's Growing Influence in Asia-Pacific, BIS Working Paper, No.446. Subacchi, Paola. (2014). Expanding Beyond Borders: The Yen and the Yuan, BIS Working Paper, No.450. Subacchi, Paola and Helena Huang. (2012). The Connecting Dots of China's Renminbi Strategy: London and Hong Kong, Chatham House Briefing Paper, No.IE BP 2012/02. Xu, Qiyuan & Fan He. (2016). Influence of RMB Cross-Border Settlement on the Chinese Economy, China & World Economy, 24(1), Yu, Yongding. (2012). Revisiting the Internationalization of the Yuan, ADBI Working Paper Series, No.366. Yu, Yongding. (2014). How Far Can Renminbi Internationalization Go? ADBI Working Paper Series, No.461. Zhang, Ming. (2015). Internationalization of the Renminbi: Developments, Problems and Influences, CIGI New Thinking and the New G20 Paper Series, No.2. Time July 2009 August 2010 January 2011 August 2011 October 2011 October 2011 December 2011 March 2013 September 2013 Table 1 The milestones of RMB Internationalization Events The pilot scheme of RMB settlement in cross-border trade in Shanghai and four cities of Guangdong Province (Guangzhou, Shenzhen, Zhuhai, Dongguan). Open domestic inter-bank bond market to three kind of foreign institutions (foreign central banks, foreign RMB clearing banks, foreign RMB settlement participant banks) RMB settlement in outward direct investment permitted. The pilot scheme of RMB settlement in cross-border trade expanded to the whole country, and the scope of RMB settlement expanded to all transactions under current account. RMB settlement in foreign direct investment permitted. Domestic commercial banks could provide RMB loans to overseas projects. RQFII scheme started. RQFII scheme expanded in both experimental institutions and proportion limit of investment. Foreign Investors could use RMB to establish, merge and invest financial

17 Billion RMB institutions in China. December 2013 The RMB purchase & sale business changed from quota management to macro-prudential management. September 2014 Qualified foreign non-financial companies could issue RMB bonds in domestic interbank bond market. November 2014 The two-way cross-border RMB fund pooling service for qualified transnational companies permitted. November 2014 RQDII scheme started. November 2014 Shanghai Hong Kong connection in stock markets started. May 2015 The approved foreign RMB clearing and participant banks could do repo transactions in China's inter-bank bond market. Source: People's Bank of China (2015). Figure 1 The development of RMB Settlement in cross-border trade 2,500 2,000 1,500 1, % 30% 25% 20% 15% 10% 5% - 0% RMB Settlement in Cross-border Trade The ratio of RMB settlement to all settlement in Cross-border Trade (RHS) Source: CEIC. Figure 2 The development of RMB settlement in cross-border direct investment

18 Billion RMB RMB settlement in ODI RMB settlement in FDI Source: CEIC. Figure 3 The global share and ranking of RMB settlement in global payments 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% RMB's Share in Global Payments RMB's Ranking in Global Payments (RHS) Source: Wind. Figure 4 The change of HK's RMB deposit outstanding

19 Billion RMB Billion RMB 1,200 1, % 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% RMB Deposits in HK The ratio of RMB deposit to HK's total deposit Source: CEIC. Figure 5 The change of Taiwan and Singapore's RMB deposit outstanding RMB Deposits in Taiwan RMB Deposit in Singapore Source: Wind. Figure 6 The development of Panda bond in HK