WPS4898 P olicy R eseaRch W oRking P aPeR 4898 Including Financial Services in Preferential Trade Agreements Lessons of International Experience for China Constantinos Stephanou The World Bank East Asia and the Pacific Region Poverty Reduction and Economic Management Department Financial and Private Sector Development and Trade Units April 2009 Policy ReseaRch WoRking PaPeR 4898 Abstract The objective of this paper is to address the main used by countries to include financial services in such considerations for China of including financial services agreements, and identifies good practices in preparing in its preferential trade agreements. The paper briefly for financial services negotiations. Particular emphasis is reviews China's financial liberalization process and the placed on lessons from Latin American preferential trade state of its domestic financial system, discusses the main agreements, given their more frequent and extensive considerations of including financial services in China's coverage of financial services compared with other preferential trade agreements, compares and contrasts regions. the different `architectural' approaches that have been This paper--a product of the Financial and Private Sector Development and Trade units, Poverty Reduction and Economic Management Department in East Asia and the Pacific--is part of a larger effort in the department to shed light on the treatment of financial services in preferential trade agreements.. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be contacted at cstephanou@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Including Financial Services in Preferential Trade Agreements: Lessons of International Experience for China by CONSTANTINOS STEPHANOU Keywords: China, trade agreements, financial services, financial liberalization. Constantinos Stephanou (cstephanou@worldbank.org) is a Senior Financial Economist in the World Bank's Financial and Private Sector Development Vice Presidency. The paper is based on a workshop on the same topic that took place in Beijing in early 2008; it reflects solely the views of its author and does not necessarily reflect the views of the World Bank Group or its Executive Directors. The author would like to thank Pierre Sauvé, Sebastian Saez, Mona Haddad, Zijing Niu, Ardo Hanson, the Chinese authorities and financial sector interviewees for their valuable comments and support. 1 Table of Contents I. INTRODUCTION ...............................................................................................3 II. THE STATE OF FINANCIAL LIBERALIZATION IN CHINA ............................4 Liberalization of trade in financial services ................................................................... 4 Financial liberalization and the domestic financial system ........................................... 5 III. INCLUDING FINANCIAL SERVICES IN PTAS ­ KEY CONSIDERATIONS ..7 Approaches to liberalizing trade in financial services ................................................... 7 Including financial services in PTAs ­ international experience ................................... 8 Implications for China .................................................................................................... 9 IV. THE ARCHITECTURE OF FINANCIAL SERVICES IN PTAS......................11 GATS versus NAFTA template...................................................................................... 11 The architecture of financial services in PTAs ­ international experience.................. 12 Implications for China .................................................................................................. 14 V. PREPARING FOR FINANCIAL SERVICES NEGOTIATIONS ......................15 Negotiating financial services in PTAs ­ international experience.............................. 15 Implications for China .................................................................................................. 16 APPENDIX I: LESSONS FROM LATIN AMERICA AND THE CARIBBEAN ....18 PTA developments in LCR ............................................................................................ 18 Coverage of financial services in LCR PTAs................................................................ 18 Country case studies: negotiating team, strategy and consultation process ................ 21 Financial services liberalization commitments compared to the GATS....................... 23 Financial services liberalization commitments compared to the status quo ................ 23 Preferential nature of de novo liberalization commitments ......................................... 24 Comparing `architectural' models for liberalization purposes.................................... 24 REFERENCES ...................................................................................................25 2 I. Introduction The objective of this paper is to address the main considerations for China of including financial services 1 in its preferential trade agreements (PTAs). China has made significant progress in liberalizing its domestic financial system since its WTO accession in 2001, and it is currently engaged in multiple bilateral (Australia, Singapore, Peru, Iceland) and plurilateral (GCC, SACU) PTA negotiations in addition to undertaking feasibility studies for more such agreements. The question arises as to whether ­ and how ­ financial services should be included in such PTAs, as well as the implications in terms of the negotiations preparation process. The paper addresses these issues by reviewing international experience and placing it in the context of China's financial system circumstances and policy priorities. Particular emphasis is placed on lessons from Latin American PTAs given their more frequent and extensive coverage of financial services compared to trade agreements in other regions. The paper is structured as follows: Section II briefly describes China's financial liberalization process (including in financial services trade) and state of the domestic financial system; Section III discusses the main considerations (both positive and negative) of including financial services in China's PTAs; Section IV compares and contrasts the different `architectural' approaches used by countries to include financial services in their PTAs; Section V reviews good practices in preparing for financial services negotiations to facilitate the overall process and achieve the desired outcomes Finally, Appendix I summarizes the experience of including financial services in FTAs for selected countries ­ Chile, Colombia and Costa Rica ­ in the Latin America and the Caribbean (LCR) region 2 . 1 The definition of financial services in this paper covers banking, insurance and insurance-related services, as well as other financial services. 2 The Appendix is based on Goncalves and Stephanou (April 2007). 3 II. The State of Financial Liberalization in China The liberalization of trade in financial services is only a subset of the broader financial liberalization agenda. The purpose of the former is to increase financial market access and remove discriminatory and other access-impeding barriers to foreign competition. By contrast, the chief purpose of the latter is to remove distortions in domestic financial systems ­ for example, interest rate and capital account controls, directed lending policies, restrictions on intra-sectoral activities, preferential treatment of publicly-owned banks, entry barriers for new operators ­ that impede competition and the allocation of capital to its most productive and profitable uses. Financial liberalization can be conceptually divided into domestic financial reform and capital account opening, and there is a broad literature on its appropriate speed and sequencing. In that context, trade liberalization in financial services is part of overall financial liberalization, although in practice there are typically strong overlaps between the two types of policy reforms 3 . Liberalization of trade in financial services China's accession to the WTO in 2001 marked a watershed with regards to its process of liberalization in financial services trade. As others have already pointed out 4 , China's financial services commitments ­ which were phased in over a 5-year period following accession ­ were among the most radical ever negotiated in the context of the WTO. Although the structure of commitments was similar to that made by other countries (e.g. more liberalization in mode 3 and in banking services 5 ), the degree of liberalization ­ both in terms of market access and national treatment ­ went significantly beyond the regulatory status quo as of 2001 and featured far-reaching pre-commitments to future market opening 6 . This reflected both China's relatively low starting point (high entry barriers) and the asymmetric nature of the accession's bargaining process (non- reciprocal concessions). China's commitments in certain banking services are very liberal compared to both developed and developing countries since they allow no discriminatory measures, although their commercial value has been limited thus far for a variety of reasons (see below). By contrast, commitments made in insurance and (particularly) securities services generally lag those made by developed countries in the GATS. 3 Trade in financial services is often linked to capital movements, notably in the context of the establishment of a commercial presence which requires inward direct investment. Certain types of cross- border financial transactions may also involve capital movements and hence require some measure of capital account opening as an inherent part of the service provision. In addition, countries (including China) often seek to promote greater policy coherence by opening up domestic financial markets to foreign competition in the context of broader financial reform efforts. 4 See Bhattasali, Li, and Martin eds. (2004) for details. 5 According to the GATS, the four modes of supplying services are via cross-border provision (mode 1), consumption abroad by domestic nationals (mode 2), commercial presence via local establishment by foreign providers (mode 3), and via the temporary entry of foreign individuals for this purpose (mode 4). 6 Commitments were phased in following WTO entry, and culminated in the elimination by December 2006 of several mode 3 restrictions on form of establishment, geographical presence and business scope, particularly for banking services. As is common with the vast majority of trade agreements, China's mode 4 commitments in financial services are horizontal in nature and fairly restrictive. 4 With two special exceptions, China has not gone beyond its WTO financial services commitments in the PTAs that it has completed to-date. In particular, the Closer Economic Partnership Agreements (CEPAs) with Hong Kong Special Administrative Region (SAR) and Macau SAR in 2004 granted service providers owned and controlled by firms (juridical persons) from those territories preferential access to China's market ahead of the WTO liberalization schedule. Most of these preferences were time-bound and expired once China's accession commitments were fully phased-in. The two CEPAs can be considered as special cases since they were intended to promote deeper integration in the broader Chinese market. By contrast, recently-completed bilateral PTAs have either not included commitments on financial services (Chile 2006, Pakistan 2007, ASEAN-TIS 2007) or have merely restated the Parties' respective commitments under the GATS in the relevant Annex (New Zealand 2008). Financial liberalization and the domestic financial system Although China's financial system has been successful at mobilizing savings and is becoming more liberalized, it is not yet fully playing the role of allocating savings efficiently to the economy. China undertook financial liberalization in a slow and gradual approach since the early 1980s, and accelerated such efforts subsequently to its WTO accession 7 . However, there remain significant gaps in this process, including strict controls on the capital account 8 , limited currency convertibility, only partly liberalized interest rates, and insufficient competition from private (domestic and foreign) financial institutions 9 . On the macroeconomic front, this policy has contributes to significant foreign reserves accumulation and a `long debt, short equity' international financial position 10 . The domestic financial system remains heavily bank-dominated and relatively unsophisticated in assessing and pricing credit risk, with large regional disparities in terms of access to credit. Banks continue to earn the vast majority of their revenues from loans, primarily to large state-owned corporate clients. By contrast, firm financing from capital markets is limited; while the domestic banking system's credit portfolio exceeds 150 percent of GDP, the corporate bond market only represents around one percent of GDP. The equity market's importance has increased substantially since 2004, but a large part of the increase has been due to the revaluation of existing assets and the public listing of state-owned firms (including banks) with small amounts of free float 11 . Entry by foreign financial services providers has increased in recent years, albeit at a slow pace. In addition to strategic non-controlling investments in local banks, 7 See Lane and Schmukler (February 2007), Calomiris ed. (2007), and Mattoo et al. eds (2007) for a description of the evolution of the financial liberalization reforms. 8 The Qualified Domestic Institutional Investors (QDII) and Qualified Foreign Institutional Investors (QFII) programs by the Chinese authorities allow domestic/foreign institutional investors to access local and foreign capital markets respectively; however, these programs have been relatively small to-date. 9 See Lardy (September 2008) for an analysis of costs of financial repression in the Chinese banking sector. 10 This means that China's stock of international financial assets (outward investments) is mostly in the form of debt while its stock of liabilities (inward investments) is mostly in the form of equity. 11 See McKinsey Global Institute (May 2006), Allen et al. (2008), WTO (April 2008), and the China-US Symposium discussions (http://www.law.harvard.edu/programs/pifs/chinaus.html) for recent reviews of the Chinese financial system. 5 several foreign banks have established branches and (more recently) subsidiaries in China; however, their share only amounted to around 2.4 percent of system-wide assets in 2007 12 . Insurance penetration by foreign providers, primarily in the form of equal- participation joint ventures with local partners, has increased at a faster pace, with their market share in terms of premiums rising to around 6 percent in 2007. Foreign providers tend to be concentrated in a few large cities where they command a much greater market share than they do at the national level. Foreign participation in domestic securities markets is even lower than in banking and insurance because of greater entry restrictions (minority ownership in joint ventures) and a moratorium on new licenses, while overall cross-border trade in financial services is minimal due to capital account restrictions 13 . As a result, even though domestic financial institutions benefit from access to foreign capital and technical expertise (e.g. risk management, product development etc.), relatively little has been achieved to-date in terms of increased competition or greater product variety outside a few major cities. The slow pace of foreign entry partly arises from policymakers' emphasis on liberalizing in a gradual, orderly manner to ensure financial stability and allow domestic financial institutions to build up expertise and commercial orientation. Low penetration can be partly attributed to the fact that foreign financial institutions were only recently allowed to enter and compete in various parts of the financial system. However, there have also been complaints about China's implementation record of its WTO financial services commitments and of the existence of regulations and administrative procedures (particularly of a prudential nature) that unduly increase the cost and slow down the pace of foreign expansion, effectively acting as indirect and de facto (if not necessarily de jure) discriminatory barriers 14 . Although some of these measures may reflect the focus of the authorities on financial stability or their desire to allow sufficient time for domestic providers to adjust, others are an inevitable consequence of China's momentous `catching up' process in financial sector regulatory policies and the adoption of international prudential standards and codes (`soft laws'). 12 According to Fitch Ratings (May 2007), 29 Chinese banks, accounting for around 53 percent of system- wide assets, have been invested in by foreign banks over the past few years. 13 In spite of significant de novo liberalization as part of the WTO accession, China's cross-border financial services trade commitments remain significantly lower than those made by the majority of developed countries in the GATS. As a result and according to data from SAFE, China's total cross-border trade in financial services (both exports and imports) for 2005 amounted to only around US$300 million. 14 Examples include the inability of foreign banks to acquire more than 20 percent individually ­ or 25 percent in total ­ in domestic banks (differences in interpretation on the right of establishment under WTO commitments) as well as regulatory restrictions or delays on local incorporation, the extent of foreign currency funding/lending, the speed of branch expansion, the introduction of new financial products, the provision of electronic payments processing services by foreign credit card companies, the use of foreign reinsurance, the licensing of new securities firms, and the supply of financial information services by foreign providers. See USTR (December 2007), China's Transitional Review Mechanism correspondence in the WTO Committee on Trade in Financial Services (www.wto.org), the Banking & Securities Working Group section of the European Chamber of Commerce (2007) report, and PricewaterhouseCoopers (September 2007 and June 2008) for more details. 6 III. Including Financial Services in PTAs ­ Key Considerations The liberalization of trade in financial services is helpful to, but is not a panacea for, domestic financial system modernization. Liberalization is desirable in order to serve the development needs of the domestic financial system by improving efficiency and the allocation of resources via healthy competition with foreign providers. There already exists a substantial body of literature on the positive relationship between finance and growth 15 and on the spillover effects of foreign (particularly bank) entry 16 , although there are also important sequencing/speed considerations and other preconditions to the process of financial liberalization 17 . More specifically, the size of benefits from trade liberalization in financial services depends critically on the market's attractiveness that will determine the extent to which foreign providers will take advantage of the opportunity (the so-called `if you build it, will they come' argument), as well as on complementary regulatory reforms. The latter include the adoption of international prudential standards, the elimination of financial repression measures, and the strengthening of the enabling `financial infrastructure' environment (i.e. rule of law, credit bureaus, collateral registries, accounting and auditing standards, payment systems etc.). It is important to note that such reforms go well beyond trade policy and cannot therefore be tackled solely in the context of a trade agreement per se. Approaches to liberalizing trade in financial services Countries have achieved the liberalization of trade in financial services in three main ways: (i) unilaterally, by opening their financial systems to international competition in the context of domestic reform efforts; (ii) at the multilateral level under the auspices of the WTO's General Agreement on Trade in Services (GATS); and (iii) on a reciprocal or preferential basis by concluding bilateral or plurilateral PTAs, of which free trade agreements (FTAs) are the most common type. Depending on the country context and circumstances, one or more of these approaches has been used to liberalize specific sub-sectors and/or modes of financial services provision 18 . Autonomous market opening has been the most common form of liberalizing trade in financial services. Most developed countries have adopted such a strategy, progressively liberalizing their financial markets over a relatively long period of time. Many developing countries followed (or are following) the same path, although in some cases ­ such as Mexico ­ the advent of a crisis led to or even accelerated the market opening process. When the time came to negotiate the GATS, these countries typically bound at (or below) the regulatory status quo in their schedules, consolidating the actual degree of openness prevailing at the time of the agreement's entry into force. China and some recent WTO accession countries (e.g. Vietnam), are the clearest examples of 15 See Demirguc-Kunt and Levine (January 2008) for a recent overview of the evidence. 16 See Clarke et al. (Spring 2003) for a review of the literature. 17 See, for example, International Monetary Fund (June 2007). 18 See Saez (December 2006), Arbeláez, Flórez and Salazar (August 2006), and Echandi (September 2006) for the cases of Chile, Colombia and Costa Rica respectively. 7 liberalization of trade in financial services in a multilateral context, while there are few examples of de novo liberalization stemming from PTAs (see Appendix I). Including financial services in PTAs ­ international experience Broadly speaking, there are three main reasons why countries negotiating PTAs include financial services disciplines and commitments in them. The first, and typically the most important reason, is the existence of offensive interests in this sector and of asymmetric bargaining powers between the negotiating counterparts. This should not be surprising: a trade agreement is essentially a mercantilist exercise, so the inclusion of certain sensitive sectors such as financial services reflects the perceived interests and negotiating strengths of the partners. In such cases, a country includes financial services commitments either because it is forced to do so by the negotiating partner (perhaps as a quid pro quo for securing market access in another sector), or because it has offensive interests in financial services and is able to pursue them in the markets of trading partners ­ typically in a North-South PTA context (see Appendix I). The second, less frequently used reason is to use such trade agreements as a vehicle to either `lock in' recent unilateral liberalization or to advance the government's reform agenda in this sector by pre-committing to future market opening, thereby overcoming any lingering resistance from domestic constituencies and providing a positive signal to foreign investors. This seems to have been the case for financial services commitments undertaken by, for example, Argentina in the GATS (Bouzas and Soltz, December 2005) and Costa Rica in the CAFTA-DR-USA trade agreement (Echandi, 2006). Finally, countries can decide to include financial services in PTAs for strategic/political, rather than purely economic, reasons. The clearest examples of such cases involve plurilateral regional trading bloc initiatives, such as the Andean Community and Mercosur in LCR 19 . Arguments in favor of including financial services in PTAs have to be weighed against potential drawbacks. These primarily consist of regulatory sensitivities about further domestic financial markets opening and the related fear of limiting `policy space' in this sector; potential distortions that could arise from giving preferential treatment to specific counterparts (especially in cases where first-mover advantage is important); strategic considerations (commitments made with one PTA counterpart can become a floor in multilateral or other bilateral trade negotiations); and the administrative problems that arise from managing a complex web of financial services liberalization rules with different countries. By far the most important factors until now have been regulatory sensitivities and strategic considerations, which have also been evident in the GATS negotiations and scheduling of commitments. International experience on the inclusion of financial services in PTAs is too recent to allow for an exhaustive evaluation of its effects on domestic financial systems and overall welfare. Although there have been numerous studies on the effects 19 A fourth, and more technical, reason for including financial services in a PTA would be to ensure that the agreement has "substantial sectoral coverage" and therefore constitutes a lawful exemption to the non- discrimination requirement articulated in Art. V of the GATS. This issue can be addressed by not carving out a priori financial services from the scope of the agreement. 8 of foreign entry (particularly banks) from unilateral liberalization, there is no comparable literature on the effects of PTA-induced financial liberalization. The effects do not appear a priori to be very significant since most PTAs seem to be primarily used to consolidate and `lock in' existing unilateral liberalization rather than as means to actively promote further market opening and the process of domestic regulatory reform (see Appendix I). The lack of relevant data and methodologies to assess ex post (as opposed to estimate ex ante) impact remains an important constraint 20 . However, even if a commonly accepted methodology for quantifying impact was established, the short time span since the negotiation or entry into force of PTAs that include financial services (the earliest, and rather atypical, example being the NAFTA in 1994) means that their contribution still cannot be fully assessed. In particular, financial services commitments and disciplines, including dispute settlement mechanisms, have not been put to the test during a market downturn or a significant revision of domestic financial system policy priorities, which is typically when constraints on `policy space' and regulatory sensitivities more fully reveal their binding nature. Implications for China In the near term, the arguments for including financial services in China's PTAs do not appear compelling. The main focus of its trade agreements until now has been to secure market access for goods exports and raw materials imports, as well as to institutionalize economic and political partnerships 21 . China is not a major exporter of financial services and it will likely take some time for this situation to change. While additional liberalization of trade in financial services is necessary and beneficial for China, it is not obvious why it should take place on a bilateral basis as opposed to unilaterally (as is taking place now) or multilaterally (via the WTO). China can successfully resist efforts to liberalize financial services by most PTA counterparts with offensive interests in this sector (the de minimus expectation in such cases would be the lock-in of the regulatory status quo), although it could still be possible to provide such commitments in exchange for export interests in other sectors 22 . This might be a legitimate trade-off in some instances, although it has been strongly resisted by financial sector policymakers in other countries on account of the sensitivities involved (see next section). This concern will likely resonate inside China as well, due to the incomplete financial liberalization process and the emphasis placed by its authorities on gradual, controlled market opening in order to avoid any crisis. Over the longer term, there are likely to be more factors driving the inclusion of financial services in China's PTAs. The authorities are increasingly emphasizing the development of service export capabilities in an attempt to diversify the economy and 20 Unlike trade in goods, assessing the direct impact of including financial services in PTAs goes well beyond trade volumes to include various financial sector outcomes (depth, efficiency, stability etc.). See Hoekman (October 2006) for a recent review of the literature on liberalizing trade in services. 21 See Hufbauer and Wong (October 2005) for a discussion. 22 The fact that the GATS was completed after the end of the Uruguay Round turned financial services into a single-sector negotiation and may have provided incentives to developing countries with limited export interests in financial services to limit commitments in order to use them as future negotiating chips. 9 move up the export `value chain' 23 . This can also be seen in the financial sector where Chinese financial institutions have begun to cautiously expand abroad by establishing commercial presence in Hong Kong SAR and by taking minority equity stakes in mostly developed country financial institutions 24 . The trend is likely to accelerate as domestic players acquire greater expertise and international experience, and as foreign reserve accumulation, cross-border activities by Chinese corporates and increased domestic competition provide incentives for foreign expansion. Moreover, as the domestic regulatory status quo eventually runs ahead of its WTO accession commitments, China will have more `room to give' in financial services negotiations without generating de novo liberalization ­ an experience that several LCR countries have already gone through (see Appendix I). In such circumstances, including financial services in some of its future PTAs could provide benefits to the domestic industry (preferential treatment and/or unimpeded market access) that could not be secured multilaterally. China would then need to carefully define its offensive interests in financial services and determine how to best promote them via trade agreements. 23 Services (both exports and imports) only accounted for 10 percent of China's foreign trade in 2006. The authorities' plans to accelerate development of the services sector are outlined in the `11th Five-Year Plan for National Economic and Social Development (2007-2012)' and in the `Opinions of the State Council on Accelerating the Development of the Service Sector' (March 2007). 24 See Leung et al. (September 2008) for a report on Chinese banks' cross-border investment activity. 10 IV. The Architecture of Financial Services in PTAs The way in which financial services are covered in PTAs is an important consideration for policymakers. In particular, the lack of a carve-out for financial services does not necessarily entail that all disciplines of the agreement will apply to that activity. Notwithstanding the overall architecture of the trade agreement, there is significant scope to maintain or introduce certain trade- and investment-restrictive measures, where needed, in financial services commitments or reservations 25 . GATS versus NAFTA template PTAs that have developed disciplines on services in general (including financial services) have traditionally followed two distinct architectural models, one based on the GATS and the other on the NAFTA. GATS-type agreements cover the supply of services through four distinct supply modes, whereas NAFTA-type agreements feature separate chapters dealing with cross-border trade in services (modes 1 and 2 of GATS), investment (mode 3) and the temporary entry of business people (mode 4), the latter two being horizontal in character and applying to all subject areas covered by the PTA. The majority of PTAs negotiated by Western Hemisphere countries have followed the NAFTA template, although recent agreements have introduced some changes by mixing in elements of the two templates in the financial services chapter. GATS-type agreements, which have been more popular with European and Asian countries, treat investment in services as a mode of supply (commercial presence) under general disciplines on trade in services, while NAFTA-type agreements distinguish between cross-border trade and investment, the latter governed entirely by disciplines that apply across the board to all economic sectors. There are important nuances in the definition and scope of the two templates that policymakers need to be aware of (see Appendix I). The scheduling of commitments under the two templates is also different. GATS- type agreements are based on a hybrid list approach to scheduling commitments, which combines elements of `positive' or `bottom up' listing (i.e. identifying the sectors and/or modes of supply concerned) and `negative' or `top down' listing (i.e. identifying the limitations and restrictions attached to specific commitments). By contrast, NAFTA-type PTAs use a negative-list or top-down approach in which trade and investment in financial services is assumed to be free from discriminatory treatment except for those non- conforming measures that are explicitly included in annexes containing reservations. This approach obliges countries to list all non-conforming measures prior to an agreement's entry into force (or subject to mutually agreed longer timeframes), otherwise they are deemed to be fully and automatically liberalized ­ a so-called "list it or lose it" approach. Each approach to scheduling commitments has its own benefits and drawbacks. Although both the negative and hybrid list approaches can achieve the same level of liberalization, the former is considered in theory to be more conducive to liberalization as 25 This section is partly based on Sauvé and Molinuevo (May 2008). 11 it introduces a strong element of regulatory transparency and a potentially higher level of commitments to the extent that it typically locks in the regulatory status quo, except in those sectors where Members are explicitly allowed to retain the power to introduce new non-conforming measures. The detailed inventories of non-conforming measures that are appended to PTAs following a negative list approach, while technically more onerous to prepare, allow foreign investors and trade negotiators alike to obtain a comprehensive picture of a country's regulatory landscape. By contrast, only the measures that apply to the sectors, sub-sectors and modes of supply entered in a country's schedule are listed under the hybrid approach, which often differ from the regulatory status quo prevailing at the time that the commitment is scheduled. While neither of the two approaches inhibits the ability of host countries to preserve `policy space', the hybrid approach has the advantage of affording greater latitude in determining the overall level of commitments and related regulatory conditions that might differ from actual practice, allegedly making it more flexible or "development-friendly" than the negative list approach. However, as Fink and Molinuevo (June 2007) note in their empirical analysis of 25 East Asian PTAs, the actual approach to scheduling commitments matters less than is commonly thought because the direction of causality between scheduling approaches and liberalization outcomes often runs both ways. The architecture of financial services in PTAs ­ international experience The architectural approach towards liberalization and regulatory disciplines in financial services is linked to the overall architecture found in PTAs and, in particular, their services and investment chapters. Countries that have followed the GATS approach for their general trade in services chapters have found it natural ­ and more familiar ­ to adopt such an approach towards financial services or to introduce a special chapter expanding on the GATS Annex on Financial Services. Similarly, countries that have previously relied on a NAFTA-like approach may have felt more at ease with a financial services chapter modelled after the latter agreement's separate disciplines on financial services (including for investment). Interestingly, some countries have adopted a range of architectural approaches in their PTA coverage of financial services, which can be attributed to different trading partners (particularly if these are large OECD countries), changes in policy priorities, timing issues, preference for preserving greater policy space in some market segments/modes in financial services etc. There is a broad variety of approaches used by countries to cover financial services in their PTAs. These range from a general chapter on trade in services at one extreme to a self-contained financial services chapter at the other extreme 26 . As Table 2 shows, each of these approaches has its own characteristics, benefits and potential weaknesses. Of course, all competing architectural approaches can theoretically generate 26 The analysis excludes those PTAs that have explicitly carved out financial services. A less radical form of a carve-out is found in the Japan-Mexico FTA where, despite the presence of a chapter dedicated to financial services, the Parties agreed to be merely bound by their respective commitments under the OECD Codes of Liberalisation and Capital Movements and the GATS. See Sauvé and Molinuevo (May 2008) for a discussion of the different approaches. 12 the same level of market opening (even as they do not afford the same degree of regulatory transparency) or level of treatment (protection) to investors. Table 2: The Architecture of Financial Services in PTAs Source: Adapted from Sauvé and Molinuevo (May 2008). It is worth noting the preference accorded to a dedicated financial services chapter by an increasing number of countries nowadays. The main reason for its appeal is its flexibility with respect to scheduling financial services-specific commitments and obligations. In NAFTA-type agreements, this approach has led to innovative scheduling of reservations (mix of positive and negative listing depending on sub-sector and mode) and disciplines (differences between cross-border trade and investment in financial services), although these have to be weighed against the increased complexity of managing the agreement's implementation. Dedicated chapters on financial services are also used for purposes of circumscribing the extent to which investors and investments in financial services are subject to various investment provisions, particularly as regards investor-state arbitration. One additional, less often-mentioned reason for its attractiveness is that such an approach allows financial sector policymakers to stay `in the driver's seat' of financial services negotiations on account of its regulatory sensitivities, since they will be subsequently called upon to implement any commitments made. The scheduling of commitments on financial services often shies away from pure negative listing. Most agreements have followed GATS-style hybrid lists in scheduling commitments on financial services, or have introduced a positive listing of sectors for cross-border services, even as they resort to negative lists under services and/or investment chapters. Once again, this reflects countries' sensitivities to certain forms of trade, particularly cross-border commitments that might involve capital flows. 13 Implications for China China's CEPAs with Hong Kong SAR and Macau SAR represent a unique approach covering financial services that is unlikely to be replicated in other PTAs. In particular, the text of these two agreements does not contain any general disciplines or services-specific obligations. The obligations of the Parties stem from the list of specific commitments adopted by China. The lack of general disciplines and of a rules-based dispute settlement mechanism, as well as the need to refer at all times to the schedules of commitments to assess the precise scope of market opening may not make this approach particularly attractive to countries that wish to provide for broad commitments in a greater number of sectors, and signal such undertakings to foreign investors and services suppliers. As for financial services, the lack of generally binding disciplines, such as market access or national treatment, make unnecessary the introduction of a prudential carve-out, or other regulatory provisions. It is debatable whether such a high degree of regulatory discretion would prove attractive to other WTO members, particularly developed country partners, or indeed to Chinese operators in its future PTAs. China may wish to adopt a more sector-specific orientation when including financial services in future PTAs. The existence of different architectural approaches suggests that there is no unique model that all countries should aspire to; specific PTA circumstances will largely determine the approach taken. In addition, the choice of template should be commensurate with a country's level of comfort and/or ambition with market opening in different modes and sub-sectors. However, international experience to- date suggests that coverage of financial services by a dedicated chapter seems to capture the common interests of the Parties in providing for substantial liberalization disciplines, and allows them to tailor such disciplines to the particularities of the financial services sector. To the extent that China wants to `customize' its financial services commitments and obligations or to limit any unforeseen consequences from them, it might be worth adopting a more sector-specific orientation via a self-contained financial services chapter. It is important for the authorities to be cognizant of important nuances between the two main negotiating templates. In particular, negotiators need to be aware of important differences in disciplines and commitments between the GATS and NAFTA templates as they relate to financial services, in order to avoid unintended consequences or limit policy space beyond what is desired 27 . Examples in NAFTA-type agreements include the definition of financial services suppliers as regulated financial institutions; the relationship between financial services and other chapters; the denial of benefits clause; limits to state aid and to preferential arrangements for state-owned financial institutions; restrictions on the imposition of capital controls; existence of ratcheting and/or standstill clauses; and the use of a negative list approach. The dispute resolution mechanisms in financial services are also crucial given the limited degree of jurisprudence generated to trade in this sector within trade agreements, be it at the WTO or in PTAs 28 . 27 See Saez (December 2005) for a discussion. 28 The only relevant experience stems from the NAFTA and brings important lessons on the consequences of the structure of financial services commitments and obligations; see ICSID (July 2006) for details. 14 V. Preparing for Financial Services Negotiations Preparing for financial services negotiations is no different from preparations for negotiations in other sectors. As previously mentioned, trade agreements are essentially mercantilistic exercises in which demandeurs strive to enhance market access opportunities for their services providers rather than to promote mutual recognition or the harmonization of prudential standards. As such, it is important for the authorities to develop a strategy/roadmap for the negotiation process that allows them to classify offensive and defensive interests, identify "red lines", anticipate requests and avoid surprises, as well as ensure the coherence of positions by resolving any internal inconsistencies that could weaken the negotiating stance. Some good practices adopted by other countries to achieve this objective specifically in the case of financial services are described below 29 . Negotiating financial services in PTAs ­ international experience 30 The structure of the financial services negotiations team tends to be similar across countries, although its leadership tends to be country-specific. In particular, the financial services negotiating team typically includes, or relies upon, financial sector officials ­ ministry of finance, central bank, deposit insurance and supervisory agencies ­ in addition to trade specialists. This is an inevitable consequence of the fact that the financial sector is heavily regulated for prudential and market conduct purposes, which requires a high level of technical expertise. Cultural and institutional factors ­ historical involvement, degree of trust and knowledge, relative political powers etc. ­ are important determinants of whether it is financial or trade officials that lead negotiations in the sector 31 . Leadership of the negotiating process will sometimes (though not always) determine the negotiating template, with ministries of finance and central banks typically opting for a dedicated financial services chapter to isolate the negotiations from other chapters and to keep them among financial sector specialists. However, even in such cases, they need to interact with negotiators and stay abreast of developments of negotiations in other chapters that are relevant for the financial sector32 . The consultation process with the private sector and civil society on financial services negotiations differs substantially across countries. As an analysis of a sample 29 See Saez (July 2007) for a discussion on preparing for a bilateral negotiation in financial services. 30 See Saez (December 2006), Arbeláez, Flórez and Salazar (August 2006), and Echandi (September 2006) on the financial services negotiations strategy and process for Chile, Colombia and Costa Rica respectively. 31 For example, it was the Ministry of Finance that led the negotiations process in the Chilean (Chile-USA FTA, Chile-EC AA) and Colombian (Colombia-USA FTA) cases, although it was the Ministry of Trade that led negotiations for Costa Rica (CAFTA-US-DR FTA). In the case of the USA, the USTR handles insurance negotiations and the Treasury handles negotiations in other financial services. 32 Strictly speaking, a financial services chapter does not fully capture all domestic financial system activities, such as foreign investment in domestic non-financial securities, activities of non-regulated financial institutions (in NAFTA-type agreements), provisions on payments and capital movements, and elements of a country's `financial infrastructure' (e.g. accounting services) and of financial institutions' functions (e.g. data processing, telecoms, legal and taxation services). 15 of LCR FTAs shows (see Appendix I), the scope and frequency of consultations depend on the perceived importance and sensitivity of financial services in the overall PTA negotiations. The consultation process with the financial industry is typically extensive, and consists of measures ranging from information sessions with financial institutions and relevant associations to "side room" participation by those actors during the negotiations and their feedback/commissioning of reports on negotiating positions. The process often includes, at least in the early stages, training on trade law concepts and principles to enhance the understanding of the private sector on the full scope and business implications of the negotiations. The consultation process with civil society varies more widely across countries, and tends to also reflect the country's state-civil society traditions 33 . Use of a trade-related regulatory audit in financial services can facilitate the negotiations and potentially generate positive spillovers in the consultation process and in domestic regulatory conduct and design 34 . Use of this audit was pioneered in the context of preparing the negative lists of non-conforming measures under the NAFTA 35 . Conducting such an audit is a useful means of preparing for services negotiations (to master the sectoral intricacies), ensuring that key regulatory objectives are met in the most efficient manner, identifying antiquated or inefficient regulations (which can then yield useful negotiating currency), encouraging the adoption of pro- competitive regulation (where feasible), and deepening the dialogue and building trust within the government (trade negotiators, line ministries and sectoral regulators) as well as with key stakeholders. Therefore, the process of preparing an inventory of discriminatory and trade- and investment-impairing measures may help not only to anticipate partner country negotiating requests, but serve as an exercise in regulatory transparency and generate positive spillovers in the way that domestic regulation is developed and implemented. However, it is worth noting that the preparation of the list of non-conforming measures is a useful starting point in financial services but it is usually not sufficient given the presence of non-discriminatory measures, including measures of a prudential nature, which may nonetheless be unduly burdensome or possible disguised restrictions to trade. Implications for China Securing the active collaboration of financial sector officials in financial services negotiations would be of great benefit to Chinese trade negotiators. As previously mentioned, negotiators need to be aware of important nuances in financial services disciplines and commitments in order to avoid unintended consequences or limit policy space beyond what is desired. Given prevailing regulatory complexities and policy 33 For example, in the case of the Chile, there were no financial services-specific consultations due to perceived lack of interest and the technical nature of the discussions, while in the case of Costa Rica they included public presentations following each negotiating round as well as a broad communications strategy. 34 See Sauvé (May 2008) for details. 35 This corresponds to a negative list of measures that, absent their inscription in reservation lists, would be found in breach of the key liberalizing provisions found in trade agreements ­ national treatment, market access (quantitative restrictions), local presence requirements, and most favored nation (MFN) treatment. 16 sensitivities, the input of financial sector officials ­ who will, after all, be responsible for implementing any commitments that are made ­ is critical. This is a challenging process in most countries because of the existence of institutional silos and attendant bureaucratic turf battles, differing interests and policy `languages' used by financial and trade officials. The creation of institutionalized collaboration mechanisms (as opposed to relying solely on ad hoc or personal contacts) and the active participation of financial sector officials in the negotiations (as opposed to being passive recipients of information) would greatly facilitate this process. Developing a trade-related regulatory audit approach in financial services can serve as a conduit for discussions between financial sector and trade officials, and it can facilitate such a `whole-of-government' cooperation process. Developing a robust consultation process with non-governmental stakeholders will also enhance the effectiveness of financial services negotiators. To be in a position to make informed and meaningful negotiating offers and requests, a country requires knowledge of the export potential or priorities of its domestic service suppliers, and needs to feel confident of its ability to manage the regulatory, sectoral and economy- wide implications of its own commitments. This cannot be achieved merely by consultations within government: as financial services get covered in future PTAs and as Chinese financial institutions develop the expertise and incentives to expand abroad, strengthening domestic processes of policy dialogue and consultations with key stakeholders in the private sector and civil society will assume heightened importance. Although the scope and frequency of such consultations vary widely across countries, they have nevertheless represented an important exercise in transparency and in enhancing the legitimacy of negotiated outcomes. 17 Appendix I: Lessons from Latin America and the Caribbean PTA developments in LCR Since the 1990s, the world economy in general ­ and LCR in particular ­ has witnessed an unprecedented proliferation of PTAs. Almost every WTO Member is party to at least one PTA, and LCR has been particularly active in this regard with trade agreements ranging from FTAs to custom unions and common markets. Two LCR countries in particular ­ Mexico and Chile ­ have been the initial driving forces behind the proliferation of PTAs in the region over the last decade, although other countries have also recently joined the fray. The entry into force of the NAFTA between Mexico, the US and Canada in 1994 is widely considered as a defining event in developing countries' attempts to engage in PTAs in order to secure greater access to key markets and consolidate recent domestic reforms. The proliferation of PTAs in recent years can be explained by political, strategic and economic reasons that may differ depending on the level of development of participating countries. In general, these agreements are considered relatively easier and faster to negotiate, particularly in markets where geographic or cultural proximity matters. Such agreements may also allow for progress in areas where multilateral reforms are less advanced. North-South agreements take the form of FTAs, while some South-South agreements aim for a higher level of economic/political integration at a sub-regional level via customs unions and common markets. The four main South-South PTAs in LCR are the Southern Cone Common Market (MERCOSUR, established in 1991), the Andean Community (CAN, established in 1969), the Caribbean Community and Common Market (CARICOM, established in 1973) and the Central America Common Market (CACM, originally established in 1960 and reinstated in 1991). Coverage of financial services in LCR PTAs Financial services are typically covered via provisions included in a separate, self-contained chapter. This chapter is, for analytical purposes, the principal vehicle for influencing the operations of the domestic financial system although, strictly speaking, it does not fully capture all such activities 36 . The treatment of financial services by PTAs in LCR follows 3 main approaches: no coverage because of the exclusion of services in general, or of financial services in particular, from the scope of a particular trade agreement 37 36 Areas relevant to the domestic financial system that are not captured by the financial services chapter include foreign investments in domestic non-financial securities, activities of non-regulated financial institutions (only in NAFTA-type agreements), provisions on payments and capital movements, as well as accounting, data processing, telecoms, legal and taxation services. 37 Some of these PTAs include an obligation to negotiate a financial services chapter some time after the entry into force of the agreement. 18 direct coverage via the inclusion of dedicated provisions in a separate chapter or annex dealing exclusively with financial services indirect coverage via `horizontal' provisions of a more generic character covering services and/or investment that partly apply to financial services. The vast majority of PTAs with financial services chapters that were negotiated by LCR countries have followed the NAFTA model. This is partly explained by the role played by Mexico in using the NAFTA template in its own subsequent FTAs. Because of the experience acquired in negotiating this type of FTA and the influence of the US, other LCR countries have developed templates similar to NAFTA in their own PTAs. Recent agreements (e.g. EU-Mexico and EU-Chile FTAs) have introduced some changes to the traditional template, by mixing in elements of the GATS, NAFTA, and of the GATS's Understanding on Commitments in Financial Services. In the case of FTAs negotiated with the US, the NAFTA template still serves as the basis for the treatment of financial services, but provisions from the GATS have also been incorporated 38 . Figure 1: Map of Financial Services Trade Commitments in LCR (end-2007) Note: Lines indicate the existence of a financial services chapter/Annex in a FTA between the relevant countries (dashed lines indicate that the agreement has not yet been ratified or implemented), while the ovals indicate the presence of a trade agreement for the creation of a common market or customs union. 38 Examples include the adoption of a positive-list approach for the cross-border supply of financial services, the treatment of potentially sensitive information, as well as the introduction of a binding ­ as opposed to "best endeavors" as in NAFTA ­ market access provision listing the types of restrictive measures that Parties cannot adopt or maintain with regard to investors or providers of another Party. 19 The proliferation of PTAs in recent years has contributed to greater financial services liberalization commitments for many LCR countries and led to an increasingly complex regional `commitments map' (or financial services `spaghetti bowl'). Most progress in financial services rule-making and market opening has been achieved via FTAs; by contrast, LCR countries that have relied on the multilateral framework and on sub-regional customs unions have not made much progress to-date (see Figure 1). Figure 2: Foreign Bank Penetration and Ownership in Selected LCR Countries 100% Costa Rica Nicaragua Honduras 90% El Salvador South-owned Banks as % of Foreign Bank Assets 80% 70% 60% 50% Bolivia 40% Dom. Rep. Jamaica 30% Panama Uruguay Paraguay 20% Ecuador Barbados 10% Colombia Argentina Peru Chile Mexico Antigua Trinidad Brazil Venezuela 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Foreign Bank Assets as % of Total Bank Assets Note: A foreign bank is defined to have at least 50 percent foreign ownership. Figures reported represent averages over 2000-2004 in each country. South-owned banks are foreign banks headquartered in a developing country. The inclusion of financial services in LCR PTAs depends greatly on whether it is a North-South or South-South agreement. Unsurprisingly, developed countries in North-South agreements have been the main proponents of including financial services. Most LCR countries are net importers of financial services and have few ­ if any ­ perceived `offensive' interests linked to a "demandeur" or domestic constituency, both of which lessen the scope for striking reciprocal bargains within the sector. Conversely, the inclusion of financial services in most North-South agreements likely reflects the fact that the majority of foreign financial institutions in LCR countries are headquartered in developed countries (see Figure 2), as well as the relative (and asymmetric) bargaining powers between the negotiating counterparts. In fact, only two LCR countries have tended to include financial services chapters in South-South agreements ­ Mexico 20 (primarily in the immediate post-NAFTA period 39 ) and Panama (which is an offshore financial center and a net exporter of financial services). Country case studies: negotiating team, strategy and consultation process An analysis of financial services chapters has been undertaken for a sample of LCR countries that have recently participated in FTAs (Chile, Colombia and Costa Rica) 40 . These comprise the Chile-US and Chile-EU FTAs, the (yet to be ratified) Colombia-US FTA and the DR-CAFTA as it applies to Costa Rica. The results strongly indicate that each country's financial services-related strategy and negotiating structure/process were shaped by initial conditions and historical experience. Chile: Chile had unilaterally liberalized its domestic financial system following the 1982 crisis, and enjoyed strong macroeconomic and financial stability at the time of FTA negotiations. Its only previous financial services commitments were made in the GATS, and the FTA with the US was seen as an opportunity to set a precedent with other countries. It was decided to negotiate a separate, self-contained financial services chapter independently of other areas (whenever possible) to comfort financial supervisors and market players, with the Ministry of Finance undertaking full responsibility for it. The negotiating stance was aimed at locking-in the status quo, both because the authorities were comfortable with the existing (very high) level of market openness, and because they perceived FTAs as a "third-best alternative" for trade liberalization, behind unilateral and multilateral trade negotiations. The right to introduce capital controls and limitations on Balance of Payments transfers was the most politically sensitive issue, and was negotiated separately by a small team of specialists from the Ministry and the Central Bank. While the private sector was involved via meetings with financial industry associations and the commissioning of a study on the FTA effects, there were no broad consultations with civil society or other stakeholders in the area of financial services. Colombia: At the time of FTA negotiations with the US, the Colombian financial system was still recovering from the effects of the 1998 crisis, which had also disrupted the progressive market opening process of the 1990s. Prior to the FTA, Colombia had made (limited) financial services commitments in the GATS and the G3 (with Mexico and Venezuela). The negotiating process for the financial services chapter was led by the Ministry of Finance, with the collaboration of the financial supervisors. Since the negotiations with the US were initially scheduled to take place at the same time for several Andean countries, Colombia coordinated with its peers to reach common negotiating positions whenever possible (e.g. on social security and collective investments). In particular, there was an exclusive emphasis on defensive interests and an attempt to prevent a non-level playing field that was perceived to arise from various US 39 The NAFTA was the first FTA by a LCR country that included trade in financial services within its scope and devoted a specific self-contained chapter to such trade. Subsequently, Mexico has been a key player in incorporating financial services in its preferential trade agreements; of the twelve PTAs (all are FTAs) entered into by Mexico, seven contain a financial services chapter. 40 See Saez (December 2006), Arbeláez., Flórez and Salazar (August 2006), and Echandi (September 2006) for the cases of Chile, Colombia and Costa Rica respectively. 21 demands (e.g. insurance/bank branching and certain cross-border trade activities). The private sector was actively involved via `side room' participation and preparation of relevant studies by financial industry associations; public presentations also followed each negotiating round. Costa Rica: Costa Rica enjoyed a high degree of market openness but significant state ownership (particularly in the insurance sector) at the time of DR-CAFTA negotiations. The only previous financial services commitments that it had made were in the GATS. Unlike the other two countries, it was the Ministry of Commerce ­ with support from financial supervisors ­ that took responsibility for preparing the negotiations `roadmap' in financial services. The 5 Central American countries had created a joint negotiating team vis-à-vis the US, and a coordination protocol and common objectives for financial services were agreed. These were sufficiently flexible to accommodate offensive interests, although Costa Rica (which acted as the secretariat of the negotiation process) only had defensive interests in this sector. Given Costa Rica's institutional and cultural experience, extensive consultation mechanisms with the private sector and with civil society were utilized. Figure 3: Market Access Commitments ­ Proportion of Financial Services Committed by Chile, Colombia and Costa Rica in the GATS and Subsequent FTAs 100% 90% 80% % of Financial Services Sub-Sectors 70% 60% 50% 40% 30% 20% 10% 0% Mode 1 Mode 2 Mode 3 GATS - Chile US-Chile EC-Chile GATS - Colombia US-Colombia GATS - Costa Rica DR-CAFTA Note: Three levels of market access commitments are applied in the above analysis by financial services sub-sector and mode: none (value of zero), partial (value of 0.5) and full commitment (value of 1). Mode 4 commitments, as well as all national treatment commitments and horizontal restrictions, are excluded. All pre-commitments to liberalize additional sub-sectors and/or modes in future years are included. 22 Financial services liberalization commitments compared to the GATS An analysis of market access and national treatment commitments scheduled in the financial services chapters for the aforementioned sample of LCR FTAs provides evidence of significant additional liberalization commitments when compared to the GATS. This is not unusual given the time elapsed and the (unilateral) market opening undertaken by these countries since the mid-1990s. Additional commitments tend to span all financial sub-sectors, including those that were not well covered in the GATS, such as insurance, securities-related and other financial services. The same is true in modal terms, with significant new commitments particularly in mode 2. Commitments are in general more extensive across all modes for FTAs involving the US, particularly for mode 2. By contrast, mode 1 commitments, while better than what has been harvested to date under the GATS, remain relatively more timid (see Figure 3). Financial services liberalization commitments compared to the status quo De novo liberalization ­ which has chiefly taken the form of pre-commitments to future market opening ­ is relatively rare for the sample of LCR countries under review. An analysis of individual LCR country experiences is much more difficult to undertake because of insufficient information on the regulatory status quo prior to, during and after the implementation of such trade agreements. Apart from Costa Rica's insurance sector that was opened for the first time as a result of its FTA with the United States, real liberalization appears to have mostly taken place in the cross-border provision of some insurance services, as well as in asset management and auxiliary financial services. Although there is limited available data on the actual market size of these sub- sectors and modes, anecdotal evidence suggests that they are relatively less important than `core' banking services. However, the abolition of numerical quotas (e.g. economic needs test) and certain juridical restrictions on forms of entry (e.g. insurance branching 41 ) might also contribute to further liberalization in other sub-sectors under mode 3. The above findings offer strong indications that, with few exceptions, PTAs are primarily used to consolidate and `lock in' existing unilateral liberalization rather than as means to actively promote further market opening and the process of domestic regulatory reform. The fact that the LCR countries under review appear to have already largely liberalized their domestic financial systems on a unilateral basis prior to their engagement in PTA negotiations has also contributed to this outcome. Of course, consolidation of the regulatory status quo and the application of certain disciplines in trade agreements remain important because they can limit the arbitrary use (and abuse) of `policy space' by host country authorities. New disciplines such as those on regulatory transparency, as well as the lock-in of the current policy regime via commitments, standstill and ratcheting clauses, enhance predictability, prevent potentially costly policy reversals, and can thus benefit both domestic and foreign 41 However, it should be noted that domestic authorities retain the right to regulate such branches as they deem necessary for prudential purposes, including via the establishment of local capital requirements. 23 financial services providers and local consumers. It is therefore conceivable that a PTA could exert positive impact on the business environment (including for financial services) even if real liberalization commitments remain limited to the status quo. However, the issue of policy space is a double-edged sword, and policymakers need to decide on the level of policy flexibility and regulatory discretion that properly balances policy considerations that go beyond trade liberalization objectives per se. Linked to this issue is the need for policymakers negotiating the financial services provisions of PTAs to be cognizant of the important nuances in disciplines and commitments that might create unintended consequences or limit policy space beyond what was envisaged. The short time span since the entry into force of most PTAs means that their contribution ­ whether anticipated or unanticipated ­ still cannot be fully assessed. Preferential nature of de novo liberalization commitments An interesting additional finding of the analysis is that many de novo liberalization commitments are actually not preferential in nature. While some commitments (e.g. the abolition of an economic needs test for Chile) are country-specific and benefit the financial services providers of the PTA counterpart, others (e.g. permitting branching or opening up the insurance industry to private providers for Costa Rica) require new `horizontal' regulations or laws that would apply to the entire industry and could actually benefit financial service providers from third countries. Therefore, unlike trade in goods, the extent to which financial services commitments were preferential was not primarily determined by rules of origin 42 . This result would seem to suggest that there might not always be important first-mover advantages or serious economic distortions created by using PTAs to promote market opening in financial services, although much depends on whether the relevant liberalization commitments are actually preferential in nature. Comparing `architectural' models for liberalization purposes Based on the experience of the 3 countries that were analyzed, it is unclear whether GATS- or NAFTA-type `architectural' models actually lead to greater liberalization in financial services. A simple review of the FTAs mentioned above would seem to favor the negative list approach and broader rules and disciplines embedded in NAFTA-type agreements, primarily on grounds of heightened regulatory transparency, but this can be largely attributed to the involvement of the US in many such agreements. In addition, it has been argued 43 that the direction of causality between scheduling approaches and the level of liberalization commitments can run both ways. Moreover, even NAFTA-type agreements have tended to use a hybrid list for financial service commitments ­ in fact, both models have introduced new features in recent years that borrow from each other, revealing signs of convergence around a hybrid approach. 42 Provisions on denial of benefits ­ also known as rules of origin ­ allow parties to a PTA to deny the benefits of the agreement to a financial services supplier that is not owned or controlled by nationals of the other party. These provisions can be important because their restrictiveness helps determine the extent of preferential treatment entailed in market opening commitments undertaken by the parties. 43 See, for example, Fink and Molinuevo (June 2007). 24 References Allen F., Quian J. and Qian M. (2008), "China's Financial System: Past, Present and Future", found in China's Great Economic Transformation book edited by T.G. Rawski and L. Brandt, Cambridge University Press. Arbeláez, M. A., Flórez A. and Salazar N. (August 2006), "Financial Services in the Colombia-U.S. Free Trade Agreement", World Bank mimeo. Bhattasali D., Li S. and Martin W. eds. (2004), China and the WTO: Accession, Policy Reform, and Poverty Reduction Strategies, World Bank and Oxford University Press. Bouzas R. and Soltz H. (December 2005), "Argentina and the GATS: A Study on the Domestic Determinants of GATS Commitments" in Gallagher P., Low P. and Stoler A. (eds.) Managing the Challenges of WTO Participation, WTO. Calomiris C.W. ed. (2007), China's Financial System at a Crossroads, Columbia University Press. Clarke G., Cull R., Martinez Peria M.S., and Sanchez S. (Spring 2003), "Foreign Bank Entry: Experience, Implications for Developing Economies, and Agenda for Further Research", The World Bank Research Observer, Vol. 18, No 1. Demirguc-Kunt A. and Levine R. (January 2008), "Finance, Financial-Sector Policies, and Long-Run Growth", World Bank Policy Research Working Paper 4469. Dobson W. and Kashyap A.K. (September 2006), "The contradiction in China's gradualist banking reforms", paper prepared for the Brookings Panel on Economic Activity. Echandi R. (September 2006), "The DR-CAFTA-US FTA Negotiations in Financial Services: The Experience of Costa Rica", World Bank mimeo. European Chamber of Commerce (2007), "European Business in China ­ Position Paper 2007/2008", report. Fink C. and Molinuevo M. (June 2007), "East Asian FTAs in Services: Roaring Tigers or Timid Pandas?", World Bank Policy Research Report, Trade Issues In East Asia ­ China Focus. Fitch Ratings (May 2007), "Foreign Banks' Expansion into China, with a Focus on Hong Kong's Banks", Special Report. Goncalves M.P. and Stephanou C. (April 2007), "Financial Services and Trade Agreements in Latin America and the Caribbean: An Overview", World Bank Policy Research Working Paper 4181. Hoekman B. (October 2006), "Liberalizing Trade in Services: A Survey", World Bank Policy Research Working Paper 4030. Hufbauer G.C. and Wong Y. (October 2005), "Prospects for Regional Free Trade in Asia", Institute for International Economics Working Paper 05-12. 25 International Center for Settlement of Investment Disputes (July 2006), "Fireman's Fund Insurance Company vs. the United Mexican States", Case No. ARB(AF)/02/01. International Monetary Fund (June 2007), "Reaping the Benefits of Financial Globalization", Research Department discussion paper. Kose M.A., Prasad E., Rogoff K., and Wei S.J. (August 2006), "Financial Globalization: A Reappraisal", IMF Working Paper 06/189. Lane P.R. and Schmukler S.L. (February 2007), "The International Financial Integration of China and India", World Bank Policy Research Working Paper 4132. Lardy N.R. (September 2008), "Financial Repression in China", Peterson Institute for International Economics, Policy Brief Number PB08-8. Leung F., Michaelis H., and Tang T. (September 2008), "Venturing Abroad: Chinese Banks and Cross-Border M&A", Boston Consulting Group report. Liu L. (September 2005), "The Impact of Financial Services Trade Liberalization in China", RIETI Discussion Paper Series 05-E-024. Mattoo A., Stern R.M., and Zannini G., eds. (2007), A Handbook of International Trade in Services, Oxford University Press. McKinsey & Company (May 2006), "Putting China's Capital to Work: The Value of Financial System Reform", McKinsey Global Institute Report. McKinsey & Company (June 2008), "Global investment strategies for China's financial institutions", The McKinsey Quarterly. PricewaterhouseCoopers (September 2007), "Foreign insurance companies in China ­ September 2007", report. PricewaterhouseCoopers (June 2008), "Foreign banks in China ­ June 2008", report. Saez S. (December 2005), "Trade in services negotiations: A review of the experience of the United States and the European Union in Latin America", CEPAL SERIE Comercio Internacional No. 61. Saez R. (December 2006), "Trade in Financial Services: The Case of Chile", World Bank mimeo. Saez R. (July 2007), "Preparing for a Bilateral Negotiation in Financial Services", World Bank mimeo. Sauvé P. (May 2008), "Conducting a trade-related regulatory audit in financial services", World Bank background note. Sauvé P. and Molinuevo M. (May 2008), "Architectural approaches to the treatment of trade and investment in financial services in selected preferential trade agreements", World Bank background note. United States Trade Representative (December 2007), "2007 Report to Congress on China's WTO Compliance", report. World Trade Organization (April 2008), "Trade Policy Review: China", Report by the Secretariat. 26