Abstract
In recent years, the IMF has released a growing number of reports and other documents covering economic and financial developments and trends in member countries. Each report, prepared by a staff team after discussions with government officials, is published at the option of the member country.
Outward Financial Spillovers From Global Liquidity Shocks: two Thought Experiments for the UK
The UK is a global hub for cross-border liquidity generation and distribution and is, thus, a natural propagator and potential amplifier of financial shocks. In that context, we study the UK’s role in intermediating two potentially large global financial shocks that could materialize in the future—namely, stress from global shadow banks spilling over to global banks via repo markets in the context of a disorderly renormalization of monetary policy in advanced economies; and China’s anticipated integration into the global financial system. The paper finds a significant potential role for the UK in either case, offers a rough quantification of the spillover impacts (both negative and positive), and points to some policy implications: the need to ensure effective oversight of shadow banking activity and institutions, including through engagement with other regulators and international bodies; and maintain the UK as a reliable and resilient center for international finance.
A. Introduction
1. This paper comprises two distinct thoughts experiments to study outward spillovers that may be generated via the UK’s systemic and globally-interconnected financial system. In section B, which has a nearer-term focus, we study the repo market channel for stress transmission from global shadow banks to UK banks in the aftermath of a world interest-rate “snapback”, while quantifying possible adverse outward credit spillovers, and highlighting some policy implications. Section C, which clearly has a longer-term focus, examines the UK’s potential contribution to the process of internationalization of the renminbi, and to the global distribution of China’s large banking sector cross-border claims in the context of prospective capital account liberalization.
B. Contagion and Spillover from Shadow Banking Stress in World Repo Markets Following Global Interest-Rate Snapback1
2. The current global environment of low interest rates, asset price inflation, and tight bank regulation offers fertile ground for the build-up of risks in the nonbank, or shadow banking, sector. The exceptional monetary stimulus provided by major central banks since the Global Financial Crisis (GFC) has helped stave off full-blown depression, but is also generating new vulnerabilities. For instance, negative real interest rates on safe assets are pushing investors in search of yield into riskier territories. The prevailing record-low implied volatilities in a range of markets (see Figure 1) seem at odds with observable economic vulnerabilities (such as in China and other emerging economies, Japan, and the euro area) and non-trivial geo-political tensions (e.g. crises in Ukraine and in Iraq). In other cases, investors are using cheap, abundant liquidity to leverage up and boost returns on equity, while pushing up asset prices. Against the backdrop of tight bank regulation, a significant amount of this risk-taking is likely occurring in the shadow banking system.2
A Global Under Pricing of Risk
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
3. In the UK, the rising sectoral contribution of the “fund sector” to systemic risk, and to bank contagion, seems to lend preliminary support to this concern.3 Although around two-third of overall systemic risk in the UK financial system can be attributed to major UK banks (left panel, Figure 2a), the contribution to this risk from pension, equity, bond and hedge funds has risen notably since end–2012: for instance, as shown in the right panel of Figure 2a, for equity funds, the contribution has increased by one-fourth, albeit from a relatively small level, i.e. from (from 3.2 to 4 percent). The contribution of the “fund sector” to banking sector distress-dependence has been relatively small compared with insurers over the last year (left panel, Figure 2b plots the evolution of percentage contributions over time). However, it is important to note the disproportionate share within this contribution of bond and hedge funds (right panel Figure 2b). Together, they account for three-fifth of the fund sector’s contribution to banking sector distress dependence, despite being one-fifth of the size of pension funds in the sample.
The Contribution of Nonbanks to UK Systemic Risk
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
The Contribution of Nonbanks to UK Banking Sector Distress Dependence
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
Source: IMF staff estimates4. As a global financial center, the UK is also vulnerable to risks arising from “non-UK” (or global) shadow banks, especially via interactions in world repo markets. For example, several global hedge funds incorporated in tax havens have their fund managers based in the UK. These hedge funds – alongside, inter alia, exchange-traded funds and US-based mortgage realestate investment trusts (mREITs) – borrow cash in world repo markets from US money market funds (MMFs) through G14 dealer banks (see Figure 3). Some of these shadow entities have achieved highly leveraged positions in the past few years, by using the cash obtained initially to buy new assets, posting the latter as collateral to borrow more cash, and so on (mREITs had amassed leverage of 7.3 times). Importantly, many of the G14 dealers intermediating this borrowing are either UK-headquartered G-SIBs (RBS, HSBC and Barclays); or the UK subsidiaries and branches of foreign banks (such as Citibank, UK). The estimated share of UK-resident dealers in this repo chain is c. 30-40 percent.
End-2012 Map of Repo Chain Linking US MMFs to Shadow Banks via G14 Dealers
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
Source: Bank of England staff estimates5. Against this backdrop, we study the possibility of contagion from global shadow banks to UK banks following a world interest-rate snapback, and the resulting outward spillovers.4 Focusing on the US$5 trillion segment of the world repo market described above, we analyze how a snapback can generate stress in leveraged global shadow banks. Specifically, we look at the ability of these shadow banks to meet variation margin calls triggered by the sudden loss in value of “repoed” collateral that results from the interest rate shock. We identify “limit points” at which the shadow banks run out of usable collateral to meet margin calls, necessitating an unwind of their repo positions and possibly fire sales of assets. This opens general and specific channels of contagion to the intermediating G14 dealers, many of which are UK-resident banks, including the possibility of intra-day freezes in the repo market. Given that banks’ banking and trading books would already be under stress due to the snapback,5 this liquidity shock in the secured lending market could easily spill over into the unsecured (LIBOR) market, reminiscent of the developments around the GFC. Banks could end up deleveraging, especially abroad, generating outward financial spillovers.
6. We estimate that variation margin flows following an interest-rate snapback could be sizable and possibly difficult-to-meet for some highly leveraged shadow banks, such as mREITs. As shown in Figure 4, for a 100 basis point rise in the interest rate, the implied variation margin flows for “unchanged” collateral haircuts, would be c. US$550 billion. For 200 basis points, the figure is US$1 trillion. This compares with the G14 dealers’ own collateral “cushion” of ~US$300bn (= Reverse Repo book minus Repo book, in Figure 3), suggesting that inability of leveraged shadow banks (such as exchange traded funds, fixed income hedge funds and mREITs) to meet their margin calls could create stress in the repo market. Figure 4 shows the automatic buildup in shadow bank leverage that will follow from depressed asset prices following the snapback, and the point at which leverage must stabilize or unwind due to the exhaustion of usable assets to post as collateral. For mREITs, this “limit point” is reached with a 200 basis point interest-rate shock.
Variation Margin Flows in Global Repo Market Triggered by Interest Rate Snapback
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
Source: Bank of England staff estimatesLimit Points for Selected Shadow Banks Assuming Unchanged Collateral Haircuts
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
Source: Bank of England staff estimates7. In practice, the lenders to the repo chain could raise collateral haircuts as borrower leverage increases, causing “limit points” to be reached earlier or with smaller interest-rate shocks. The estimates in Figures 4 and 5 are based on the assumption that lenders do not change the required haircuts on the collateral against which they lend. This is a generous assumption as it is likely that haircuts would rise (sharply and unpredictably) with borrower leverage; in some cases lenders may refuse some types of collateral or counterparties altogether: during the GFC, US MMFs suddenly became extremely risk averse and pulled back their funding from major US and EU banks, triggering fire sales of assets, general panic and a surge in counterparty risk. A repeat of such behavior – which forces borrowers in the repo chain to unwind their positions by selling assets in a short period of time – cannot be ruled out.
8. The impact on financial markets and stability would depend critically on the speed with which interest rates rise and on borrowers’ repo maturities. For reasons of tractability, the simulations above assume that interest rates jump instantaneously, but this is too alarmist. Any surge is likely to take some time. This time, coupled with the length of repo maturities, represents the breathing space that vulnerable shadow banks will have to space out their asset sales and meet their repo obligations. Illustratively, at 20 percent, the share of mREITs’ repo borrowing with maturity of less than 30 days, is not trivial. The mere expectation that asset sales may need to occur fast (i.e. anticipation of fire sales) could become self-fulfilling.
9. Stress in any point of the repo funding chain could contaminate the G14 dealer intermediaries, and trigger broader market discontinuities, such as intra-day trading freeze. Such a scenario would be more likely if variation margin flows were large in relation to dealers’ own collateral cushions; asset fire sales had a significant impact on dealers’ trading books and banking books (via higher defaults by debtors); and if dealers’ “own” borrowing from lenders to the repo chain (US MMFs) got cut down or became costlier. The combination of these stresses would typically manifest in higher credit default swap premia for dealers, in turn activating credit valuation adjustments on derivatives exposures. With counterparty risk rising, the cost of borrowing in unsecured markets (measured by the LIBOR-OIS spread) would escalate. Given the complex interconnectedness of financial assets, markets and players, it is impossible to say what the systemic implications would be in such a scenario, except that a major liquidity shock could not be ruled out.
10. If affected dealer banks respond by reducing cross-border lending, outward credit spillovers would result. A shock to the G14 dealer banks, given their size and global interconnectedness, would clearly ripple through world markets, affecting a range of counterparties, and ultimately firms and households. One channel which we quantify is external deleveraging by major UK banks, mirroring their response to the liquidity crunch during the GFC: from end-Sep 2007 to end-Dec 2009, the 3-month UK LIBOR-OIS remained (on average) 80 basis points above its “normal” level of 10 basis points, and was accompanied by an 11 percent decline in UK banks’ foreign claims.6 For illustrative purposes, we assume a shock half that size and duration so that the implied external deleveraging by UK banks is c. 3 percent (US$120 billion, given US$4trn in foreign claims).7
11. The distribution of outward spillovers would depend on how banks allocate the required deleveraging across jurisdictions. In the simplest possible setting, one could assume banks apportion the US$120 billion in line with the share of foreign claims in each jurisdiction. In this case, Hong Kong would face the largest impact, with its domestic credit falling by 2 percent, followed by Mauritius and Bahamas (1.5 percent) and Luxembourg, Ireland, Singapore and South Africa (1 percent). The distribution of spillovers would be quite different if UK banks deleveraged “strategically”, i.e. fully protected the largest (core) jurisdictions, while fully exiting others. In this case, a number of emerging and low-income economies (such as Panama, Djibouti, Pakistan, Mozambique, Philippines, Uruguay) could see credit reducing by 10 percent or more.8 Whether these spillovers would be deemed favorable or unfavorable by the recipient economies would depend, inter alia, on which stage of the economic and credit cycle they were at: i.e. whether the withdrawal of UK bank credit from their jurisdictions eased over-heating pressures, or exacerbated slowdown.
12. Recent Bank of England efforts to strengthen bank balance sheets and liquidity backstops should reduce the likelihood and intensity of a snapback-induced crisis. One important difference from the GFC is that banks, both in the UK and in the US are more resilient, both in terms of capital and funding, and thus less vulnerable to stress in the shadow banking system. Moreover, changes to the Sterling Monetary Framework announced in October 2013 mean that banks deemed healthy ex-ante by the Prudential Regulation Authority (PRA) would be assured access to the Bank’s permanent liquidity facilities at lower cost and longer maturities, against a broader range of collateral, and with less stigma. The decision to extend liquidity supports to central counterparties and major broker-dealers is also timely and appropriate, given the systemic importance of these entities. In both cases, however, stronger backstops will need to be accompanied by commensurately strengthened supervision to ensure beneficiaries do not end up taking more risks.
13. A key policy priority going forward will be ensuring adequate oversight of the nebulous shadow banking system, including global shadow banking activity. To this end, there are already efforts underway at both the Financial Conduct Authority (FCA) and the PRA.9 These efforts are critical given the size and complexity of the shadow banking system (in the UK and globally) and the vulnerabilities and channels of contagion to banks documented in this paper. Any additional budgetary resources, regulatory perimeter adjustments, and international coordination initiatives (in relation to “global” shadow banking), required to support these efforts should be prioritized.
UK Banks’ External Deleveraging-Possible Impact on Recipient Economies
(Percent of recipient economy bank credit)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
C. Spillovers from UK Contribution to the Internationalization of Chinese Banking Liquidity and the Renminbi10
14. China’s capital account is relatively closed, both de jure and de facto, so that the country’s financial integration in the world far undershoots its economic weight. As noted in Hooley (2013), China’s share of world trade and GDP has risen dramatically from 2 percent at the start of the 1990s to over 10 percent today (Figure 7, left panel). The corresponding increase in China’s global financial integration has been modest – from c. 0.5 percent to c. 2 percent, reflecting the country’s (well-enforced) restrictions on capital account convertibility (Figure 7, right panel).
China’s Financial and Economic Weight/Integration in the World
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
Source: Hooley (2013)15. China’s large banking system assets are almost entirely domestically-invested. A simple comparison of the size of the US banking system with China’s reveals just how big the latter is. At about US$ 22 trillion, Chinese banking assets were over 40 percent larger than those of US banks. However, almost all of this was domestically-invested, whereas one-fifth of US banking assets are invested abroad.
Size and External Orientation of China– and US-Resident Banking Systems11
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
16. An opening of the Chinese capital account would have major implications for global finance. In the 13th 5-year plan (2011–15), the Chinese authorities envisaged the start of what is expected to be a long and gradual process of capital account liberalization. Economists have attempted to estimate the implications of this move in various ways. Bayoumi and Ohnsorge (2013) project a net increase of Chinese international assets of US$ 1.6-2.7 trillion12 over the next 15 years or so;13 while Hooley (2013) shows that China’s gross international investment position could increase from 5 percent to 30 percent of world GDP by 2025 under a full liberalization scenario. Should greater capital account openness be accompanied by a similar-to-US external orientation of Chinese banks, gross cross-border claims of almost US$ 5 trillion could be accumulated by Chinese residents.
Cross-border Liquidity Generated in Major Jurisdictions by Subsidiaries and Branches of Foreign Banks Resident in those Jurisdictions
(Average of quarterly stock over 2005–13, US dollars billions)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
Sources: BIS International Locational and Consolidated Banking Statistics; and IMF Staff estimates.17. As a global financial center, the UK could play an important intermediation role in distributing any released Chinese bank claims across jurisdictions. The UK is the world’s leading hub for international banking, where global banks distribute cross-border claims. Over 2005–13, for instance, branches and subsidiaries of foreign-owned banks based in the UK generated over US$ 3.7 trillion in cross-border claims (Figure 9, red column).14 The fact that Chinese banks’ share of these foreign-owned bank assets in the UK is negligible – despite the rapid growth in China’s weight in the world real economy – suggests there is a sizable potential for the UK’s intermediation role.
18. A crude thought experiment provides us with a ballpark of the size of possible “long-term” spillovers from such a UK intermediation role, focusing on gross outflows from China. For this experiment, it is assumed that:
(i) Chinese banks’ cross-border/total asset ratio rises to one-quarter (akin to the average share of US and Japanese banks), implying a release of c. US$ 4.9 trillion in cross-border bank claims of China vis-à-vis the rest of the world.
(ii) 21 percent [US$ 1,029 billion] of these US$ 4.9 trillion cross-border assets are held in the UK, assuming Chinese banks mimic the average of US and Japanese banks’ international asset allocation.
(iii) 43 percent [US$442 billion] of the US$ 1,029 billion is received by UK nonbanks and is thus assumed to “end” in the UK. This mimics the current recipient share of cross-border bank claims of the US and Japan vis-à-vis the UK.
(iv) Of the remaining US$ 587 billion received by UK-resident banking entities, a third is assumed to be “recycled back” to China (broadly consistent with the pattern observed for banking flows into the UK from the US and Japan).
(v) Thus, US$ 391 billion remains as the China-originated gross cross-border bank claims available for global distribution from the UK.15
19. The cross-country allocation of these funds would depend on the final purpose they are intended to serve, as well as the particular competency of the UK. Table 1 shows the cross–country distribution of funds that would obtain if Chinese banks simply followed the pattern of cross-border claim allocation by UK-resident banks today. However, given the Chinese economy’s strong trade focus, it is also possible that parent Chinese banks use the funds as a source of trade finance for importers of Chinese goods in Europe, Middle East and Africa (Hong Kong SAR would likely be the preferred hub for distribution of trade finance to Asia and the Americas). In this case, the US$ 391 billion could conceivably get distributed in proportion to the importing economies’ share in the US$ 580 billion total Chinese exports thereto—see Table 2.
Illustrative Distribution via-UK Chinese Credit In Line with Current Pattern of Cross-Border Claims Allocation by UK-Resident Banks
Illustrative Distribution via-UK Chinese Credit In Line with Current Pattern of Cross-Border Claims Allocation by UK-Resident Banks
| Chinese-originated / UK-intermediated credit flows (In percent of recipient economy GDP avg. 2014–13) |
|||
|---|---|---|---|
| Luxembourg | 13.59 | Hungary | 0.23 |
| The Bahamas | 11.60 | Croatia | 0.27 |
| Liberia | 0.25 | Oman | 0.25 |
| Ireland | 7.16 | Russia | 0.25 |
| Netherlands | 3.45 | Canada | 0.25 |
| Bahrain | 3.23 | Serbia | 0.23 |
| Switzerland | 2.57 | Jamaica | 0.23 |
| Singapore | 2.47 | Albania | 0.23 |
| Hong Kong SAR | 1.88 | Kenya | 0.21 |
| Iceland | 1.56 | Zambia | 0.20 |
| France | 1.48 | New Zealand | 0.19 |
| Portugal | 1.01 | Jordan | 0.18 |
| Qatar | 0.95 | Czech Republic | 0.18 |
| Finland | 0.90 | Bulgaria | 0.16 |
| Denmark | 0.87 | India | 0.16 |
| Greece | 0.87 | Lebanon | 0.16 |
| Belgium | 0.83 | Nigeria | 0.15 |
| Germany | 0.83 | Israel | 0.15 |
| Spain | 0.77 | Romania | 0.15 |
| Sweden | 0.73 | Korea | 0.15 |
| United States | 0.58 | Brazil | 0.14 |
| Panama | 0.55 | Saudi Arabia | 0.14 |
| Norway | 0.53 | Kuwait | 0.14 |
| Italy | 0.52 | Poland | 0.13 |
| Australia | 0.49 | Ukraine | 0.11 |
| Turkey | 0.46 | Tanzania | 0.11 |
| South Africa | 0.39 | Malaysia | 0.11 |
| Austria | 0.35 | Slovenia | 0.10 |
Illustrative Distribution via-UK Chinese Credit In Line with Current Pattern of Cross-Border Claims Allocation by UK-Resident Banks
| Chinese-originated / UK-intermediated credit flows (In percent of recipient economy GDP avg. 2014–13) |
|||
|---|---|---|---|
| Luxembourg | 13.59 | Hungary | 0.23 |
| The Bahamas | 11.60 | Croatia | 0.27 |
| Liberia | 0.25 | Oman | 0.25 |
| Ireland | 7.16 | Russia | 0.25 |
| Netherlands | 3.45 | Canada | 0.25 |
| Bahrain | 3.23 | Serbia | 0.23 |
| Switzerland | 2.57 | Jamaica | 0.23 |
| Singapore | 2.47 | Albania | 0.23 |
| Hong Kong SAR | 1.88 | Kenya | 0.21 |
| Iceland | 1.56 | Zambia | 0.20 |
| France | 1.48 | New Zealand | 0.19 |
| Portugal | 1.01 | Jordan | 0.18 |
| Qatar | 0.95 | Czech Republic | 0.18 |
| Finland | 0.90 | Bulgaria | 0.16 |
| Denmark | 0.87 | India | 0.16 |
| Greece | 0.87 | Lebanon | 0.16 |
| Belgium | 0.83 | Nigeria | 0.15 |
| Germany | 0.83 | Israel | 0.15 |
| Spain | 0.77 | Romania | 0.15 |
| Sweden | 0.73 | Korea | 0.15 |
| United States | 0.58 | Brazil | 0.14 |
| Panama | 0.55 | Saudi Arabia | 0.14 |
| Norway | 0.53 | Kuwait | 0.14 |
| Italy | 0.52 | Poland | 0.13 |
| Australia | 0.49 | Ukraine | 0.11 |
| Turkey | 0.46 | Tanzania | 0.11 |
| South Africa | 0.39 | Malaysia | 0.11 |
| Austria | 0.35 | Slovenia | 0.10 |
Illustrative Distribution via-UK Chinese Credit for Trade Finance Purposes
Illustrative Distribution via-UK Chinese Credit for Trade Finance Purposes
| Chinese-originated / UK-intermediated credit | |||||
|---|---|---|---|---|---|
| (In percent of recipient economy nominal GDP avg. over 2014–18) | |||||
| Liberia | 57.7 | Pakistan | 2.7 | Israel | 1.5 |
| Kyrgyz Republic | 39.5 | Cyprus | 2.7 | Oman | 1.5 |
| Djibouti | 36.1 | Hungary | 2.6 | Croatia | 1.4 |
| Togo | 27.2 | Uzbekistan | 2.4 | Poland | 1.4 |
| Benin | 18.0 | Bahrain | 2.3 | Turkey | 1.4 |
| Malta | 14.9 | Latvia | 2.3 | Finland | 1.3 |
| Tajikistan | 11.5 | Sudan | 2.3 | Bulgaria | 1.2 |
| Mongolia | 11.5 | Slovenia | 2.3 | Turkmenistan | 1.2 |
| Mauritania | 7.6 | Estonia | 2.3 | Germany | 1.1 |
| Guinea | 7.4 | IR. of Iran | 2.2 | Dominican Rep. | 1.0 |
| Ghana | 6.0 | Nigeria | 2.2 | Denmark | 1.0 |
| Jordan | 5.5 | Czech Republic | 2.2 | Kuwait | 0.9 |
| United Arab Emirates | 5.0 | Ethiopia | 2.0 | Spain | 0.8 |
| Tanzania | 4.7 | Lithuania | 2.0 | Romania | 0.8 |
| Netherlands | 4.4 | Libya | 1.9 | Greece | 0.8 |
| Mozambique | 3.7 | Angola | 1.8 | Italy | 0.7 |
| Kazakhstan | 3.3 | Belgium | 1.8 | Sweden | 0.7 |
| Lebanon | 3.3 | Slovak Republic | 1.8 | Portugal | 0.7 |
| Senega 1 | 33 | Algeria | 1.7 | Belarus | 0.7 |
| Kenya | 3.2 | Tunisia | 1.7 | Ireland | 0.7 |
| Madagascar | 3.1 | Morocco | 1.6 | Azerbaijan | 0.6 |
| Republic of Congo | 3.1 | Iraq | 1.6 | France | 0.6 |
| Yemen | 3.1 | Zambia | 1.6 | Qatar | 0.5 |
| Georgia | 3.0 | DRC Congo | 1.6 | Norway | 0.3 |
| Mauritius | 3.0 | Côte d’ivoire | 1.6 | Austria | 0.3 |
| South Africa | 2.9 | Saudi Arabia | 1.5 | ||
| Jamaica | 2.8 | Egypt | 1.5 | ||
| Cameroon | 2.8 | Russia | 1.5 | ||
Illustrative Distribution via-UK Chinese Credit for Trade Finance Purposes
| Chinese-originated / UK-intermediated credit | |||||
|---|---|---|---|---|---|
| (In percent of recipient economy nominal GDP avg. over 2014–18) | |||||
| Liberia | 57.7 | Pakistan | 2.7 | Israel | 1.5 |
| Kyrgyz Republic | 39.5 | Cyprus | 2.7 | Oman | 1.5 |
| Djibouti | 36.1 | Hungary | 2.6 | Croatia | 1.4 |
| Togo | 27.2 | Uzbekistan | 2.4 | Poland | 1.4 |
| Benin | 18.0 | Bahrain | 2.3 | Turkey | 1.4 |
| Malta | 14.9 | Latvia | 2.3 | Finland | 1.3 |
| Tajikistan | 11.5 | Sudan | 2.3 | Bulgaria | 1.2 |
| Mongolia | 11.5 | Slovenia | 2.3 | Turkmenistan | 1.2 |
| Mauritania | 7.6 | Estonia | 2.3 | Germany | 1.1 |
| Guinea | 7.4 | IR. of Iran | 2.2 | Dominican Rep. | 1.0 |
| Ghana | 6.0 | Nigeria | 2.2 | Denmark | 1.0 |
| Jordan | 5.5 | Czech Republic | 2.2 | Kuwait | 0.9 |
| United Arab Emirates | 5.0 | Ethiopia | 2.0 | Spain | 0.8 |
| Tanzania | 4.7 | Lithuania | 2.0 | Romania | 0.8 |
| Netherlands | 4.4 | Libya | 1.9 | Greece | 0.8 |
| Mozambique | 3.7 | Angola | 1.8 | Italy | 0.7 |
| Kazakhstan | 3.3 | Belgium | 1.8 | Sweden | 0.7 |
| Lebanon | 3.3 | Slovak Republic | 1.8 | Portugal | 0.7 |
| Senega 1 | 33 | Algeria | 1.7 | Belarus | 0.7 |
| Kenya | 3.2 | Tunisia | 1.7 | Ireland | 0.7 |
| Madagascar | 3.1 | Morocco | 1.6 | Azerbaijan | 0.6 |
| Republic of Congo | 3.1 | Iraq | 1.6 | France | 0.6 |
| Yemen | 3.1 | Zambia | 1.6 | Qatar | 0.5 |
| Georgia | 3.0 | DRC Congo | 1.6 | Norway | 0.3 |
| Mauritius | 3.0 | Côte d’ivoire | 1.6 | Austria | 0.3 |
| South Africa | 2.9 | Saudi Arabia | 1.5 | ||
| Jamaica | 2.8 | Egypt | 1.5 | ||
| Cameroon | 2.8 | Russia | 1.5 | ||
20. Given that Chinese capital account liberalization is likely to be a gradual process, a number of qualifications around the foregoing are warranted when taking a nearer view.
(i) Chinese banks’ cross-border/total assets ratio may initially only rise to levels seen banks in advanced economies with moderate financial integration. The total increase in cross-border bank claims to the rest of the world in this case would be closer to US$ 1.2 trillion, i.e. less than one-quarter of that assumed above. By extension, the amount remaining as China-originated bank claims available for global distribution from the UK would approach c. US$ 100 billion (and not c. US$ 400 billion).
(ii) Capital account liberalization will have a scale effect by which all types of cross-border assets and liabilities will likely increase; but it will also have a composition effect: currently China’s external assets are predominantly held in low-yielding government bonds. In the future, they are likely to shift towards more FDI and partly portfolio debt and equity investment, as private sector outflows are liberalized. And while currently most of China’s external liabilities are FDI, they can also be expected to shift toward more portfolio investment. This is in line with international experience suggesting that, with capital account liberalization, countries tend to rely relatively less on external FDI liabilities (He et al, 2012).
(iii) Chinese-owned banks still have limited international banking experience and capacity; and, until this changes, Chinese capital account liberalization and the bulk of capital flows will likely largely be facilitated by the world’s big investment banks, many of them operating out of London.
(iv) The UK’s facilitator role over the medium-term will likely center around the following: (a) Chinese institutional investors investing in world capital markets, including via the major investment banks based in London; (b) Chinese corporations issuing debt and equity in London, taking advantage of its deep capital markets in search of lower borrowing costs. This will become especially relevant if future credit conditions tighten in China; (c) taking advantage of London as the world’s leading centre for foreign exchange and derivatives trading, Chinese corporations and banks purchasing derivatives in order to insure against currency and interest rate risks;
(v) While London is likely to play a key role in many of these transactions, the domicile of these transactions may often be located outside of the UK. For example, a lot of the assets that are traded via London-based investment banks are not actually held in London, but are domiciled in jurisdictions such as Luxembourg and Ireland, for tax or regulatory reasons. That said, even though they may remain centers for the listing of other-currency bonds by Chinese entities, the legal, design, sales, marketing and execution work will be carried out in London.
Due to the complexity of the ‘global financial supply chain’ it is difficult to judge what impact policies in the UK will have on other countries, including China. From the perspective of UK regulators, managing and mitigating adverse outward “risk” spillovers from increased Chinese banking activity in the UK will clearly be important.
21. In conjunction with the UK’s role as a cross-border hub for Chinese banking activity, the UK can also facilitate the internationalization of the renminbi. With China’s rising weight in the world economy, the renminbi’s use as a vehicle or reserve currency of the world will inevitably grow. There are already signs of this: with the renminbi now the second most used currency for trade finance after the dollar, overtaking the euro. According to SWIFT, the renminbi is the second most used currency for cross border payments with China and Hong Kong16, and for the Middle East and Central and Latin America, the renminbi is the most used currency for payments with China and Hong Kong. However, while the renminbi has risen to the seventh most used payments currency globally by value, this only puts it at 1.47 percent of global payments. And although the percent of Chinese exports settling in renminbi is projected to rise from below 10 percent in 2011 to about 20 percent in 201417 this is still comparatively small – the share of US exports settled in US dollars is close to 90 percent. Given the large pool of international investors active in London, an increase of renminbi FX and derivatives liquidity there could make a big contribution to the adoption of the currency for trade and financial transactions. This would bring the world closer to a more diversified multi-currency system.
22. The speed of renmbinbi internationalization partly depends on the pace of Chinese capital account liberalization and on how quickly a Euro-renminbi market develops. Unlike the large Eurodollar market – i.e. the market for short-term dollar deposits outside the US – renminbi deposits are scarce outside China and Hong Kong. Such markets bring both benefits and risks: for instance, they help separate currency from country risk, while facilitating off-shore investment banking (which can helpfully support corporate activity). But they also make it easier for financial institutions to arbitrage across regulatory and tax regimes.
23. The UK played a pre-eminent role in the development of the Eurodollar market, which could offer some lessons for the prospective development of a Euro-renminbi market in the future. Box 1 summarizes the various stages of the US dollar’s internationalization, the central role played by the UK in this process, and the lessons that could be drawn from that experience for renminbi internationalization. These are: (i) national policy can play a role in making certain currencies more widely used, but fundamentals matter most; (ii) a multi-currency system is possible; (iii) although the offshore internationalization of a currency is possible even with a relatively closed capital account, it ultimately depends on supply and demand factors; (iv) a large Euro-currency market may make capital controls onshore less effective; and (v) the precise use of Euro-currency markets is highly complex, involving transactions between non-residents; round-tripping transactions by residents (including for potential regulatory arbitrage purposes); and net flows out of and into the home country.
24. Although Hong Kong SAR is likely to play a dominant role in renminbi internationalization, the UK is emerging as the major renminbi hub outside of China and Hong Kong SAR. The UK already has the strongest financial links with China among all major advanced economies (Figure 12). Moreover, renminbi activity levels in London which are the largest outside of China and Hong Kong SAR, are picking up, albeit from a small base (Table 2), and, by the end of 2013, almost two-third of renminbi trading outside China and Hong Kong SAR was taking place in London. This corresponds with London’s position as the world’s leading financial center for FX trading (40 percent share) and OTC derivatives trading (50 percent share), which is vital if agents are to take positions and conduct transactions in renminbi outside of China.
UK Banks have Relatively Outsized Claims on China
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
Source: Hooley (2013)Renminbi Activity Levels in UK are Rising
Renminbi Activity Levels in UK are Rising
| Renminbi activity in London | ||||
|---|---|---|---|---|
| RMB billions | ||||
| 2011 | 2012 | 2013 | ||
| Foreign exchange trading | 10.6 | 16.8 | 22.3 | |
| o/w deliverable | 2.5 | 7.7 | 15.6 | |
| o/w non-deliverable | 8.1 | 9.1 | 6.8 | |
| Bond issuance | 7.0 | 12.4 | N/A | |
| Deposits | 15.1 | 11.9 | 14.6 | |
| o/w retail | 0.3 | 0.2 | 0.1 | |
| o/w private banking | 3.6 | 2.8 | 2.0 | |
| o/w corporate | 2.9 | 2.1 | 1.1 | |
| o/w interbank | 8.3 | 6.8 | 11.4 | |
| Trade finance | 8.6 | 24.9 | 27.9 | |
| o/w letters of credit | 0.2 | 1.0 | 3.3 | |
| o/w import financing | 6.8 | 19.2 | 20.3 | |
| o/w export financing | 1.7 | 4.7 | 4.3 | |
Renminbi Activity Levels in UK are Rising
| Renminbi activity in London | ||||
|---|---|---|---|---|
| RMB billions | ||||
| 2011 | 2012 | 2013 | ||
| Foreign exchange trading | 10.6 | 16.8 | 22.3 | |
| o/w deliverable | 2.5 | 7.7 | 15.6 | |
| o/w non-deliverable | 8.1 | 9.1 | 6.8 | |
| Bond issuance | 7.0 | 12.4 | N/A | |
| Deposits | 15.1 | 11.9 | 14.6 | |
| o/w retail | 0.3 | 0.2 | 0.1 | |
| o/w private banking | 3.6 | 2.8 | 2.0 | |
| o/w corporate | 2.9 | 2.1 | 1.1 | |
| o/w interbank | 8.3 | 6.8 | 11.4 | |
| Trade finance | 8.6 | 24.9 | 27.9 | |
| o/w letters of credit | 0.2 | 1.0 | 3.3 | |
| o/w import financing | 6.8 | 19.2 | 20.3 | |
| o/w export financing | 1.7 | 4.7 | 4.3 | |
25. Recent policy developments have been supportive of a more prominent role for the UK, both as a cross-border hub for Chinese banking activity and for renminbi internationalization. In June 2013, the Bank of England and the People’s Bank of China established a sterling-renminbi swap line to better backstop offshore currency needs. In October 2013, the PRA outlined its position in relation to non-EEA wholesale branches in the UK, paving the way for Chinese wholesale branches to be set up in the UK. China has allocated a quota of RMB 80 billion for UK-based asset managers to invest directly into Chinese onshore securities. And, in June 2014, the People’s Bank of China appointed China Construction Bank as clearing bank for offshore renminbi in London, allowing onshore settlement of offshore transactions.
26. In sum, there is a long way to go with regard to China’s capital account liberalization and the renminbi’s internationalization, and the UK’s contribution thereto, together with associated outward spillovers, will evolve accordingly. In the near term, Chinese residents will increase their foreign direct investment (FDI) while the renminbi will become more important for trade invoicing. Most activity via the UK will thus be centered on FDI and trade finance. The major spillover will be the signaling effect of renminbi usage in London. In the next stage, Chinese institutional investors may start to buy and sell global cross-border assets on a large scale via the London capital market, traded by major banks; and Chinese corporations will likely start issuing offshore bonds on a large scale, facilitated by the UK financial infrastructure. The spillover effects of these developments could be a major increase in global equity and bond market assets and liabilities, requiring a smooth functioning of UK capital, FX and derivatives markets.18 Only in the final stage will Chinese-owned banks – including through their operations in the UK – likely become major global players and facilitators of China’s capital account liberalization. Effective supervision by UK and Chinese authorities will be needed to safeguard their expansion into global markets, ensuring global financial stability.
The Development of the Eurodollar Market
1870–1925: The rise of the dollar a currency for invoicing international trade
The US dollar was initially a dominant currency for trade invoicing and settlement and only became used in global finance much later. By 1870 the US had overtaken the UK as the world’s largest economy and by 1912 it had become the world’s largest exporter. Thirteen years later, in 1925, the dollar overtook sterling as the major currency for foreign trade acceptances. Just as then, even though China is currently the world’s largest exporter only a small fraction of world trade is actually invoiced in renminbi.
There was some role of policy in internationalizing the dollar at the initial stage. Although the US dominated global trade (as China does today), the market for dollar trade and trade credits was still illiquid, partly due to financial institutions being unfamiliar with the asset class. In 1913 the Fed was established, it allowed branches abroad and became the market maker in the nascent market for US$-denominated trade credit. In 1920 the Fed started purchased trade acceptances, stabilizing and lowering the interest rate on trade acceptance credit; by the second half of the 1920s half of all existing dollar trade credits were on the Fed’s balance sheet (Eichengreen & Flandreau, 2010). In addition, (i) WWI brought about a scarcity of trade credit in London; (ii) UK trade fell relative to the US and (iii) sterling had become increasingly volatile from 1915 onwards. This all made using the dollar more attractive. This may be the phase most resembling the stage of renminbi internationalization at the moment, as China is trying to promote the liquidity of markets for trade finance and other renminbi financial products. For example, the PBoC has set up renminbi swap lines with various other central banks and established clearing banks. Thus far, many of these markets are still relatively illiquid compared to those in other currencies.
After the initial rise of the dollar a de facto multi-currency system existed. Though data is sparse, denomination of trade was split mainly between the dollar and sterling until after World War II (Eichengreen et al. 2014). But the rise of the dollar was not linear. The Great Depression in the US was deeper and longer than in the UK. This caused the role of the dollar to decline temporarily. But after WWII it had become the main international currency again. For today, this means that we are likely to see a multi-polar currency system in the medium-term, with a gradual rise of the renminbi, absent a major crisis in either China or the US.
1955–65: The development of the Eurodollar market
The Eurodollar market developed in London in the mid–1950s–1960s due to market forces in a context of highly constrained international capital flows. There were several demand factors encouraging the use of offshore dollars: (i) the dollar was already the leading world trade currency, making it an important means of invoicing; (ii) the dollar was seen as a store of value, even though there were concerns about credibility of its peg to gold throughout the 1960s; (iii) higher relative interest rates on offshore dollar deposits and bonds–due to regulatory arbitrage of US banks avoiding deposit caps at home – made Eurodollar instruments an attractive investment opportunity (Schenk, 1998; see Figure 1, first row). On the supply side, loopholes in capital controls permitted further offshore accumulation of dollars once current account restrictions were lifted from the mid–1950s onwards (Eichengreen, 2008). At the same time – while restricting convertibility of sterling for non-trade use – the UK authorities chose not to impose restrictions on cross-border banking in dollars. Other countries did restrict Eurodollar banking which encouraged the growth of London as the major centre. And the US tolerated the existence of the Eurodollar market by allowing onshore settlement of offshore transactions (He & McCauley, 2010).
Once the US started to run a current account deficit, the US authorities imposed stricter capital controls, but they never attempted to restrain Eurodollar banking via restrictions on settlement. This episode shows that offshore use of a currency is possible in the presence of significant capital controls. But there still needed to be channels allowing onshore settlement and at least some degree of convertibility for purely financial transactions. China seems to be moving to fulfilling these supply side conditions. But the demand-side factors may remain unachieved for several years to come. Last, the episode also shows that London’s singular role in the internationalization of the dollar was due to its more open regulatory framework – a comparative advantage that it does not have to the same extent today.
1965–80: Increasing pressure on the Bretton Woods System, but continuation of Eurodollar use
Once the US began running current account deficits, it started to impose stricter capital controls which increased the use of the Eurodollar market. By the early 1960s, both the US and the UK were running current account deficits, which led to the so–called Triffin Dilemma: once the offshore stock of dollars was larger than US gold reserves, the credibility of the peg to gold declined. This prompted the US to impose stricter capital controls, eg the introduction of a tax on offshore capital gains. This, in turn, further increased the incentives for conducting cross-border investments via the Eurodollar market, where the tax did not apply (Duffy & Giddy, 1994). During this period, the Euro-bond market in dollars was established: the offshore issuance of dollar-denominated debt. And the Eurodollar market became heavily used for transactions between non-US-residents (Figure 1, second row). But attempts to prevent fears of devaluation proved futile and, eventually, Britain was forced to devalue in 1967 and the US devalued in 1968. This finally led to the break-down of the Bretton Woods system of fixed exchange rates in 1971. Subsequently capital controls were lifted and financial systems were liberalized. The arbitrage advantages of using dollar offshore disappeared. But the Eurodollar market did not decline – in fact it more than doubled in size over the next 20 years. While the role of the Eurodollar market in the breakdown of Bretton Woods was certainly not central, it may have been a contributing factor by weakening the impact of capital controls that were aimed at preventing it.
1980-today: Evolution of the Eurodollar market
Even though many arbitrage opportunities disappeared the Eurodollar market continued to grow from about 10 percent of total dollar intermediation in the 1970s to about one-quarter in 2011. The UK has the largest share of Eurodollar banking at 20% of the total Eurodollar market (He & McCauley, 2010). And according to some estimates, 70% of all Eurobonds are traded in London. While US-residents were at first net creditors on the Eurodollar market, they became net debtors by the mid–1990s. The use of Eurodollars is manifold: (i) 60 percent are claims of non-US residents on other non-US residents (pure offshore); (ii) one-third are claims of US residents vis-à-vis non-US residents that are offset by claims of non-US residents against US-residents (round-tripping) and (iii) roughly ten percent are net flows from non-US to US-residents (Box Figure 1, third row) (He & McCauley, 2011). This highlights the complex and manifold uses of a global currency: If the renminbi is to become a dominant currency it will need to be used for transactions between non-residents; it will be used by residents for round-tripping transactions, including for potential regulatory arbitrage purposes and it will also comprise net renminbi flows out of and into mainland China.
The Development of the Eurodollar Market
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A004
D. Concluding Remarks
27. The foregoing analysis shows that the UK’s systemic and globally inter-connected financial system can be a potential source of outward spillovers, negative and positive. The two thought experiments discussed in this paper—repo market stress spilling over from global shadow banks to UK banks (and then to the rest of the world), following an interest-rate snapback; and the UK’s medium-to-long-term role in the internationalization of Chinese banking activity and the renmbinbi—lay out possible transmission mechanisms and quantifications for these spillovers. They also point to the appropriate policy implications for the UK authorities, namely: ensuring effective oversight of shadow banking activity and institutions, including through engagement with other regulators and international bodies; and positioning the UK financial system for China’s rising global financial integration, including through maintaining the UK as a reliable and resilient center for international finance.
References
Benelli, R, 2011. “Potential Impact on Global Bond Markets of Reallocating Reserves,” China: Spillover Report for the 2011 Article IV Consultation and Selected Issues, (Washington: International Monetary Fund).
Chitu, Eichengreen & Mehl, 2014. “Network effects, homogenous good and international currency choice—new evidence on oil markets from an older area,” ECB Working Paper Series, No 1651; March.
Dufey, G., and I. Giddy, 1994, The International Money Market, Prentice Hall International.
Eichengreen, B, 2001. “Capital Account Liberalization: What Do Cross-Country Studies Tell Us?” World Bank Economic Review, Vol. 15, No. 3, pp. 341–65, (Washington).
Eichengreen, 2008. “Globalizing capital—a history of the international monetary system,” Princeton University Press.
Eichengreen, B., M. Flandrea, 2010. “The Federal Reserve, the Bank of England, and the Rise of the Dollar as an International Currency, 1914–1939,” IHEID Working Papers 16–2010, Economics Section, The Graduate Institute of International Studies.
He, D., L. Cheung, W. Zhang and T. Wu, 2012. “How Would Capital Account Liberalization Affect China’s Capital Flows and the Renminbi Real Exchange Rates?” China and the World Economy, Vol. 20, No. 6, pp. 29–54.
He, D. and R. McCauley, 2012, “Eurodollar banking and currency internationalisation,” BIS Quarterly Review, June.
He, D. and R. McCauley, 2010. “Offshore Markets for the Domestic Currency: Monetary and Financial Stability Issues,” BIS Working Paper No. 320, September.
Hills, R., J. Hooley, Y. Korniyenko and T. Wieladek (forthcoming). “The International and Domestic Transmission of Bank Liquidity Shocks: Evidence from the UK,” mimeo, Bank of England.
Hooley, J, 2013. “Bringing Down the Great Wall? Global Implications of Capital Account Liberalization in China,” Quarterly Bulletin Q4 2013, Bank of England.
Schenk, C., 1998, “The Origins of the Eurodollar Market in London: 1955–1963”; Explorations in Economic History, vol. 35, 221–238
Segoviano, M, 2006. “Consistent Information Multivariate Density Optimizing Methodology,” London School of Economics. Financial Markets Group, Discussion Paper 557.
Segoviano, M. and C. Goodhart, 2009. “Banking Stability Measures,” IMF Working Paper 09/4
Segoviano, M., S. Malik, A. Bouveret and P. Lindner, 2014. “Financial System Stability Measures: Systemic Risk and Interconnectedness including Banks and non-Banks,” IMF WP forthcoming.
Tamim, B. and F. Ohnsorge, 2013. “Do Inflows or Outflows Dominate? Global Implications of Capital Account Liberalization in China,” IMF Working Paper 13/189, Washington, DC.
Tarashev, N., C. Borio, K. Tsatsaronis, 2010. “Attributing Systemic Risk to Individual Institutions,” BIS Working Paper, No. 308.
This section was prepared by S. Ali Abbas (EUR), with contributions from Miguel Segoviano and Shehryar Malik (MCM), and Mathieu Vital and Robert Hills (Bank of England).
In concept, shadow banks are all nonbank institutions capable of liquidity and maturity transformation and the creation of leverage. For the purposes of this section, the focus is primarily on the “fund sector”, notably bond and equity funds, hedge funds and mortgage real-estate investment trusts.
This part of the analysis considers the financial system to be a high-dimensional portfolio of bank and nonbank institutions (Segoviano et al (2014)). Under this assumption a joint distribution of portfolio losses can be constructed from which, in turn, a measure of systemic (tail-) risk is derived. For the purpose of this analysis the system’s “expected short fall” (ES) is chosen as the measure of systemic risk in the financial system. The systemic ES takes into account the size of each institution in the system and interconnectedness among institutions. The marginal contribution to systemic risk (MCSR) by any institution or sector can be computed as the percentage contributed to the systemic ES. The MCSR is based on the (shapley-value based) risk attribution methodology proposed by Tarashev, Borio and Tsatsaronis (BIS, 2010).
The snapback is modeled as a symmetric instantaneous upward shift in the yield curves for the dollar, euro and sterling. A steepening of the yield curve (rise only in long rates alone) is perhaps more reasonable for the euro area, but in the particular context of the question we pose, the benefits to assuming this are inconsequential. Also, we abstract from the reasons for the snapback. We abstract from the “cause” of the snapback, noting only that even if the cause were a good one (say, stronger growth expectations in advanced economies), a sudden rise in rates could be disruptive.
UK banks could be affected by a sudden rise in world interest rates due to a number of reasons, e.g.: banks’ funding costs rise, directly from the increase in the risk-free rate, and indirectly from stress in the banks’ counterparties; default rates rise by firms and households vulnerable to the interest-rate shock; and mark-to-market losses depress banks’ own asset values (in both banking and trading book), so that market valuation of equity falls (leverage rises) with knock on effects on banks’ CDS spreads and funding costs.
See paper by Hills et al (forthcoming) for more granular evidence.
End-December–2013 consolidated banking claims on ultimate risk-basis (Bank of International Settlements).
Another possible scenario could be generated by taking the actual cross-country distribution of UK banks’ deleveraging during the GFC and imposing it on UK banks’ cross-border assets today. A caveat with this approach would be, however, that UK banks’ cross-country exposures have substantially changed since the GFC (for some banks, jurisdictions exposures have been unwound to almost zero), so that earlier deleveraging may only offer limited insights into how UK banks might behave today.
These include (i) joint work with the FSB and IOSCO to identify global systemically important finance companies, insurers, investment funds, hedge funds; (ii) Bank of England work on monitoring systemic linkages between US money market funds and UK banks, including through funding channels; (iii) initiatives to enhance the transparency of repo and securities lending markets, with a view to reducing their pro-cyclicality; and (iv) regular reviews by the FPC of the adequacy of the regulatory perimeter for UK financial institutions and activities. For further details, the June 2014 Financial Stability Report (Box 9).
This section was prepared by S. Ali Abbas (EUR), John Hooley (AFR) and Carsten Jung (Bank of England), with contributions from Matthew Cowie (HM Treasury) and Glenn Hoggarth (Bank of England).
Consists of local operations of domestically-owned banks, foreign subsidiaries and foreign branches, end–2012.
This is imputed using the percent of GDP range estimated by the authors.
This could have the effect of reducing interest rates for a number of recipient economies. For instance, Benelli (2011) argues that, ceteris paribus, every US$ 0.5 trillion diverted from advanced markets into EM sovereign bonds lowers the yields on those bonds by 240 bps.
The UK’s contribution to cross-border liquidity provision was obtained by subtracting the consolidated cross-border claims of UK-headquartered banks from the locational cross-border claims of all banks and branches located in the UK. This difference provides a proxy for the global claims originated from the UK by the subsidiaries and branches of banks headquartered outside the UK.
Somewhat higher numbers obtain with top-down calculations, i.e. assuming China’s share of the red bar in Figure 9 approaches China’s long-term share in world GDP.
Renminbi tracker, SWIFT, June 2014.
Deutsche Bank, 2014.
Bayoumi & Ohnesorge (2013) estimate that the increase in Chinese residents’ cross-border portfolio assets could account for up to 3 percent of global bond and equity markets
