ࡱ> Y xbjbjWW ==?YO] X l$P\tt;(L{{{K9M9M9M929 : ;$<>;{{{{;[) [)[)[){  K9 {K9[)[)I.70d@K9pq= 3, as displayed in Table 3. The variables represented in the first three columns of Table 4 aim to proxy the level of the three main types of capital flows. We expect that a lower ("stability-promoting") value of the financial services trade policy indicators developed above is correlated with more FDI and portfolio investment flows, especially relative to "other investment". A higher level of FDI stands for more long-term oriented capital flows, and more portfolio investment reflects more developed financial sectors. A larger share of FDI and portfolio investment relative to other investment is expected to reflect a more balanced financing structure across financial instruments with a smaller share of lending. The next six columns of table 4 report the volatility of capital flows. We expect that higher "modal bias", "instrument", and "restrictiveness" indicators are correlated with more volatile capital flows, both over the 7 year period as a whole (as reflected by the standard deviation), and by the change in flows in the context of the Asian crisis in 1997. We also expect a correlation between trade policy indicators and the volatility of aggregate flows, especially that of portfolio and "other investment" flows. The incidence of financial crisis is represented by a dummy variable which takes the value of one if the country was affected by a financial crisis during 1991-97 and zero otherwise. Financial crisis are reported in Caprio and Klingebiel (1996a and 1996b) for the 1991-95 period. Those sample countries which started experiencing a crisis only in 1996 and 1997 also received a value of one for this variable (see Annex 1 for crisis and non-crisis countries). We expect that the higher (the less stability promoting) the indicators of financial service commitments, the more likely is the incidence of crisis over the observation period. For this estimation, we also anticipate the volatility of capital flows to have significant explanatory power. Overall capital flows may be correlated with the incidence of financial crisis if they reflect large current accounts and economic boom-bust cycles. Independent variables A number of independent variables in addition to the financial service trade commitment indicators has been applied to estimate the influences on capital flows and financial crisis. These variables are also used in previous studies of financial sector stability (e.g., Demirguc-Kunt and Detragiache, 1997). Macroeconomic variables include the logarithm of the average inflation rate for the 1991-97 period, the real interest rate (deposit rate minus inflation), and the exchange regime. High inflation is predicted to be positively correlated with relatively low levels and high volatility of capital flows, and a higher probability of financial crisis, as high inflation reflects macroeconomic instability. High real interest rates are expected to attract non-FDI capital inflows (returns to FDI are reflected in profits rather than in interest rates). A fixed exchange regime is anticipated to be correlated with lower capital inflows or even outflows during 1997 as the Asian crisis unravelled. Demirguc-Kunt and Detragiache (1997) also use the ratio of M2 to reserves as an indicator of external vulnerability. The lower reserves relative to broad money, the more vulnerable are countries to sudden capital outflows causing financial sector difficulties. This variable is, hence, expected to explain the incidence of financial crisis. There are no direct measurements of the quality of financial sector regulation. The "law and order" index as developed by the International Country Risk Guide (Keefer, Knack and Olson, 1995) and the political risk indicator of the Euromoney Magazine serve as proxies for the regulatory environment. We expect high scores on the regulatory quality to be correlated with more balanced and less volatile capital flows, and a lower probability of incurring a financial crisis. Controls on capital flows are also difficult to measure. We use the World Bank WDI indicators for controls on portfolio investment entry as an approximation for capital controls. We expect that such controls have less of an effect on FDI but more on the other two types of capital flows. As the literature typically stresses that capital flows do not cause but exacerbate financial crisis, we do not expect to find a significant coefficient of this variable for explaining the incidence of crisis. Finally, we included two variables estimating existing foreign commercial presence. The first measures the share of foreign-owned banks in the total number of banks, and the other the share of foreign banking assets (for a definition, see Claessens, Demirguc-Kunt and Huizinga, 1998). We expect that foreign presence may be correlated with the less volatile and distorted capital flows and a lower incidence of financial crisis. V. Methodology and Results The results of descriptive statistics and regression analysis, as outlined below, largely confirm the above-made hypotheses that financial services trade policies matter for the "quality" of capital flows and the incidence of financial crisis. Macroeconomic and (to a more limited extent) regulatory variables also contribute to explaining these phenomena. Although the trade policy variables raise the explanatory power of the estimations considerably (and are therefore not marginal), in some estimations the significance of coefficients is not very strong and robust. Therefore, the results illustrate the importance of the claims developed in this paper but individual estimations and numbers should be interpreted with some caution. a. Methodology The following regression analysis will apply OLS to the analysis of capital flows and a binary probit model to the estimation of financial crisis: Capital flows = f (c, macro variables, regulatory variables, capital controls & commitment indicators) Financial crisis = f (macro variables, regulatory and financial variables, capital controls, capital flows & commitment indicators). b. Results from descriptive statistics Before analysing the results of the regression analysis, it is worthwhile discussing some descriptive statistics. Countries which experienced financial crisis during 1991-97 show a combined indicator of financial services trade policies three times as high (= less favourable for financial stability) as countries without a crisis. Crisis countries report an average score of 2.75 (range 0-6) whereas the other countries show an average of 0.9 (range 2 to 2.5). The previous Table 4 also provides some interesting findings. Net foreign direct investment averaged about 2 percent of GDP for all sample countries between 1991 and 1997. As mentioned above, this is similar to the sum of portfolio and "other investment". This finding shows that FDI was the main source of foreign financing during this period, although the opposite is frequently claimed. It is also noteworthy that the Asian crisis countries relied much more on "other investment", i.e. (short-term) international lending than, for example, Latin America. The last row of Table 4 shows that the level of capital flows does not depend much on the type of financial service commitments; countries with less favourable and more restrictive commitments do not show capital flows very different from the total sample. Data on the volatility of capital flows is also very revealing. Volatility is highest for "other investment", almost twice as high as for portfolio investment and over three times as high as for FDI. This finding does not confirm the claim that investors buy and sell bonds and equities in rapid succession, and thereby cause much volatility in capital markets. Volatility of "other investment" is much higher, probably as short term lending and depositing allows rapid movements in and out of financial markets. It is also noteworthy that the volatility of "other investment" is above average in Asia and Latin America, and in countries which experienced financial crisis. Volatility of portfolio investment and "other investment" flows is also much above average for countries with high combined indicators of commitments which are less conducive to balanced and stable capital flows (last row). The last 3 columns of Table 4 illustrate some of the events in 1997. FDI and portfolio investment was relatively stable in the total sample and in all sub-groups. The confidence crisis of 1997 largely affected "other investment", as financial institutions were not willing to roll over their loans, and possibly as capital flight set in. The South East Asian 5, for example, experienced a decline in "other investment" flows by 8 percent of GDP, while portfolio investment increased slightly and FDI was almost constant. The table shows that the large decline in "other investment" flows in 1997 also affected other emerging markets which had experienced a financial crisis before and those which have less favourable financial services commitments. Only Latin America experienced an improvement in "other investment" flows. c. Results from regression analysis On balance, the regression analysis fares relatively well in explaining the structure and volatility of capital flows and the incidence of financial crisis, but less well in explaining absolute levels of capital flows. Hypothesis 1 on the importance of mode 3 commitments for balanced capital flows and financial stability The findings for the respective variable "modal bias" confirm the relevance of this hypothesis. Table 5, column 4 illustrates that relatively liberal regulation towards commercial presence is positively correlated with the relative size of portfolio investment flows. Furthermore, more liberal commitments towards commercial presence reduce the volatility of FDI inflows (Table 6, column 1), and raise the stability of portfolio investment flows in 1997 as compared to 1996 (Table 7, column 1). Hypothesis 2 on the importance of commitments on a broad range of instruments This hypothesis as reflected in the variable "lending bias" has been confirmed relatively less by the data. The "lending bias" contributes to explaining the level of FDI flows (Table 5, column 1) and the decline in "other investment" flows in 1997 (not indicated). However, it does not contribute to explaining the volatility of capital flows over the whole 1991-97 period. Table 5: Determinants of the Level of Capital Flows Regression Analysis: OLSDependent variables: level of capital flows (percent of GDP)Foreign direct investmentPortfolio investmentOther investmentOther investment minus portfolio investment(1)(2)(3)(4)Independent variables:Trade policy  "Modal bias"(0.61) (1.64)-0.65 (-1.36)1.67 (2.20)** "Lending bias"-0.47 (-2.73)**-0.12 (-0.27)Macroeconomic: Inflation average 1991-97 (log)-0.03 (-0.12)1.35 (4.49)***-1.35 (-2.23)**-1.74 (-4.18)*** Real interest rate0.11 (2.36)**0.027 (0.22)Regulatory environment & other: Law and order tradition-0.80 (-0.87)0.08 (0.16) Political risk0.05 (2.45)**0.01 (0.47) Capital controls-2.52 (-2.63)**0.93 (0.69)-0.30 (-0.08) Share of foreign banks5.38 (1.24)Number of observations: 25201922 Adj. R2 0.470.630.040.49 *, **, *** = significant at 10, 5, 1 percent level, respectively. Hypothesis 3 on the importance of few restrictive measures for foreign affiliates This hypothesis as represented by the variable "restrictive measures under mode 3" performed very well in explaining the volatility of capital flowsone of the key concerns in this whole debate. All three main types of capital flows and the aggregate of portfolio and "other investment" were significantly more volatile in countries where considerable restrictions on operations, funding, equity and new licenses were present (Table 6, columns 2-4). The composite indicator for all three types of restrictions was significant in the estimation of capital outflows of "other investment", and the sum of portfolio and other investment in 1997. Table 8 illustrates the importance of the total restrictiveness indicator in explaining financial crisis. This indicator also proved very robust over different combinations of variables. In all regressions, the explanatory power declines considerably without the respective trade policy variables (as illustrated for example in the last two lines of Table 6). This supports the claim that trade policy is not just a marginal variable in these estimations. Table 6: Determinants of the Volatility of Capital Flows Regression Analysis: OLSDependent variables: volatility of capital flows (standard deviation) Foreign direct investment (1)Portfolio investment (2)Other investmentPortfolio and (3)other investment (4)Independent variables:Trade policy "Modal bias"0.44 (2.35)** Restrictive measures, mode 3 1.27 (2.93)**0.77 (2.51)**1.04 (3.07)***Macroeconomic: Inflation average 1991-97 (log) 0.16 (1.32)0.96 (2.11)*0.13 (0.40)0.49 (1.37) Real interest rate-0.05 (-2.52)**0.18 (2.80)**Regulatory environment & other: Law and order tradition 0.26 (1.73)0.71 (1.26)0.59 (1.43)0.87 (1.88)* Capital controls -10.5 (-2.62)**-3.75 (-2.01)*-4.56 (-2.22)**Number of observations:21172221 Adj. R20.300.630.230.37 Adj.R2 without trade policy variable0.110.410.080.08 *, **, *** = significant at 10, 5, 1 percent level, respectively. Table 7: Determinants of Changes in Capital Flows 1996-1997 Regression Analysis: OLS Dependent variables - Change in capital flows between 1996 and 1997 (percent of GDP) Portfolio InvestmentOther InvestmentPortfolio and Other Investment(1)(2)(3)Independent variables:Trade policy "Modal bias" -1.21 (-2.82)** Total restrictiveness -1.41 (-2.07)*-1.39 (-2.18)**Macroeconomic: Inflation average 1996-97 -0.01 (-0.40) Exchange regime -1.26 (-1.28)-16.9 (-3.75)***-15.1 (-3.51)***Regulatory environment & other: Political risk 0.06 (2.62)**-0.05 (0.62) Capital controls  Share of foreign bank assets 16.10 (2.43)**9.92 (1.44)Number of observations: 222120 Adj. R2 0.400.200.33 *, **, *** = significant at 10, 5, 1 percent level, respectively. Macroeconomic, regulatory and other variables Inflation was a very good predictor of portfolio and "other investment" flows whereas it does not seem to affect much FDI flows and the volatility of capital flows. Somewhat surprisingly, inflation and all other macro variables were not correlated with the incidence of financial crisis. Portfolio investment flows are correlated with high real interest rates, but "other investment" flows and FDI are not affected. Real interest rates are also significant in explaining the volatility of FDI (negative correlation) and portfolio investment (positive correlation). This means that the higher the real interest rates the more stable FDI flows but the less stable portfolio investment flows. Real interest rates do not explain changes in capital flows in 1997 and financial crises. Table 8: Determinants of Financial Crisis, 1991-1997 Regression Analysis: Dependent variables - Incidence of Financial CrisisBinary probit(1)(2)(3)(4)Independent variables:Trade policy  Total restrictiveness0.44 (1.94)*0.53 (2.02)**0.71 (1.96)**0.50 (2.05)**Macroeconomic: Inflation average 1996-97 0.12 (0.40)0.53 (0.92)Regulatory environment & other: Law and order tradition-0.24 (-1.67)* Political risk-0.04 (-1.87)*-0.04 (-1.88)*-0.05 (-1.35) Reserves (M2/international reserves)0.03 (0.30) Capital flows: Standard deviation (other investment)0.43 (1.69)* Standard deviation (portfolio and other investment)0.45 (1.90)* Average flows, % of GDP (portfolio and other investment)0.67 (1.78)* Average flows, % of GDP (all foreign financing)0.20 (1.96)**Number of observations: 22212121 *, **, *** = significant at 10, 5, 1 percent level, respectively. The exchange regime has only proven to be a good predictor of changes in "other investment" flows between 1996 and 1997. (The significance of the coefficient for the variable reflecting the sum of "other" portfolio investment also seems to be mainly due to "other investment" shifts). "Other investments" were negatively affected by fixed exchange regimes, probably as the exchange rate peg in some Asian country collapsed and confidence in the peg in other countries declined as well. As mentioned above, there is no direct measure of the quality of the regulatory environment. Both the law and order and the political risk indicator were significant predictors of capital flows and financial crisis in a number of estimations. However, more work seems to be needed to develop useful proxies for countries' regulatory environment, and Claessens and Glaessner (1997) have started this work with a number of Asian countries. Regarding the level or structure of capital flows, capital controls only have a significant adverse effect on FDI.  While they do seem to lead to less volatile portfolio and other investment flows in the estimations for the whole 1991-97 period, they do not contribute to explaining changes in capital flows in 1997 nor the incidence of financial crisis. The vulnerability to external shocks as expressed by the ratio of broad money to reserves (as successfully used by Demirguc-Kunt and Detragiache, 1997) was not significant in explaining the incidence of financial crisis. As predicted, selected capital flow variables seem a good predictor of financial crisis. The volatility of "other investment" and the aggregate of "other" and portfolio investment, and the level of "other investment" and all foreign financing had the predicted effect on the incidence of financial crisis. This illustrates the importance of large and volatile short term flows (as imbedded in "other investment"). The significance of the total level of foreign financing (which is strongly correlated with current account deficits) suggests a correlation between large current account deficits over an extended period of time (here 7 years) and financial crisis. Finally, the two variables representing actual foreign presence were mostly not significant in explaining capital flows and financial crisis. Only the share of foreign banking assets was positively correlated with other investment flows in 1997 as compared to 1996. This finding confirms our prediction that in countries with more foreign presence, "herding behaviour" of investors and, therefore, capital outflows of "other investments" are less significant. Somewhat surprisingly, actual foreign presence did not contribute to explaining financial crisis. Furthermore, trade openness variables do not significantly explain foreign presence. These findings suggest that foreign presence in many countries is not (yet) the result of trade openness but due more to historical factors. Historically high levels of foreign presence in otherwise closed financial systems without the possibility of new entry are likely to contribute much less to stability than foreign presence in open systems. As argued above, governments are less likely to introduce useful prudential regulation, adjust macro-policy making and end destabilising domestic financial sector interventions in a closed system, independent of whether it is dominated by foreign or domestic service suppliers. While this may explain the above findings, we would expect foreign presence to become a more important stability-factor in the future as liberalisation rather than history begins dominating these variables. VI. Conclusions and policy implications The purpose of this study was to show the importance of financial services trade policy for capital flows and financial sector stability. The way countries liberalise financial services trade and what type of instruments of protection they apply determine the benefits countries can derive from liberalisation. We have argued and found significant evidence that a balanced liberalisation over the full range of financial instruments, and the liberalisation of foreign commercial presence with minimal strings attached contribute to less distorted and less volatile capital flows and to less crisis-prone financial systems. What are the implications of these findings for countries' financial sector liberalisation strategies? First, liberalisation of capital flows and financial services trade liberalisation should not be confused. Second, countries with a weak financial system which fear that capital flows could exacerbate their problems may nevertheless benefit from the capacity building and efficiency-enhancing effect of certain types of financial services trade liberalisation. GATS commitments allowing the commercial presence of foreign institutions (without undue restrictions on their operations), and liberalising a broad range of instruments should be considered. Such liberalisation requires only limited liberalisation of capital flows in the GATS context. In other words, in countries with weak financial systems, what is sometimes called "modal neutrality", i.e., equal liberalisation commitments as between cross-border supply (mode 1) and supply through commercial presence (mode 3), may not always be desirable. Many countries do allow more capital flows than required under GATS. For these countries, the combination of commitments suggested above is likely to promote a balanced instrument and maturity structure of foreign debt, making the financial system less prone to volatility and instability. The cross-border provision of financial services which typically do not involve capital movements, such as the provision of financial information also seems to be unproblematic from this perspective. However, this does not mean that mode 1 liberalisation should be generally discouraged. Countries should only be aware of the stability trade-off arising from the required capital account liberalisation and the possible effect on the structure of capital flows. This makes the advisable policy and institutional prerequisites more stringent for mode 1 than for mode 3 liberalisation. A number of studies mentioned above discuss these prerequisites and some sequencing considerations. Countries with stable financial systems and a sound macroeconomic and regulatory framework have every reason to apply a very broad liberalisation strategy and commit to far-reaching trade liberalisation across all modes of supply, with full integration into global capital markets through capital account liberalisation. Finally, two other concerns need to be addressed. First, some governments may wish to protect their domestic financial service providers to build a "domestic industry". Or they may attempt to use protection of the financial system to subsidise the infant-industrial sector (Henderson, 1998). Such motivations seem to be behind some countries' reluctance to liberalise commercial presence of foreign financial institutions in the domestic financial sector. If countries do not wish to liberalise mode 3 trade, does this mean they should not liberalise at all, given the potential "stability costs" of cross-border liberalisation? The answer to this questions depends on whether the efficiency gains outweigh the "stability costs" from liberalisation. If the efficiency gains are very large, and the costs relatively low (for example, due to relatively well-functioning supervision, transparency etc.), mode 1 liberalisation may be "second-best" to mode 3 but, nevertheless, preferable to completely closed markets. However, a very old lesson from trade policy needs to be reiterated in this context: trade protection is not the first best choice for infant industry support or industrial policies (independent of the desirability of such policy objectives). If domestic financial institutions are protected from foreign competition for infant-industry reasons or if the financial sector is protected to allow quasi-fiscal support of infant-industries in the industrial sector, this can introduce significant distortions and long-term costs. Inefficiency in a variety of guises (unproductive investments, cronyism) often come with protection. The quasi-fiscal costs of directed credits or controlled interest rates are an implicit tax which has to be born by the rest of the economy. The costs in terms of delayed financial sector development are not immediately visible but can be significant as well. Many financial crises have also shown that the costs of inappropriate financial sector intervention must ultimately be borne by the public when the government is forced to bail out the financial system in order to avoid its collapse. Given the high costs of many crises, it is probably less costly to achieve worthwhile public policy objectives through direct budgetary support rather than through intervening in the financial system and delaying valuable financial services trade liberalisation. Second, the GATS framework explicitly allows for measures to be taken for prudential and balance-of-payments reasons which could include restrictions on capital transfers. These provisions are tantamount to a safeguard to prevent or solve a severe crisis. The benefits from recourse to such measures, however, need to be weighed carefully against their costs. Both the suspension of commitments and the re-introduction of capital controls can have considerable long-term costs through a higher risk premium on foreign investment (IMF, ICM, 1998). The above arguments lend strong support for further financial services trade liberalisation in many countries. The GATS provides a useful multilateral framework for doing so, offering sufficient flexibility for countries to pursue an appropriate financial services trade liberalisation strategy, and yet take a more prudent approach towards capital account liberalisation when warranted. Such liberalisation also allows for the necessary degree of prudential regulation and supervision, and provides safeguard mechanisms against financial crisis. Bibliography: Bank of International Settlement (BIS) (1997) Core Principles for Effective Banking Supervision, Basle. Bhagwati, Jagdish (1998) The Capital Myth. The Difference Between Trade in Widgets and Dollars, Foreign Affairs May/June 1998, vol. 77, no.3. Caprio G. and D.Klingebiel (1996a) Bank Insolvency: Bad Luck, Bad Policy, or Bad Banking? in M. Bruno and B.Pleskovic (eds.) Annual World Bank Conference on Development Economics 1996, Washington D.C.: World Bank. Caprio G. and D. Klingebiel (1996b) Bank Insolvencies: Cross Country Experience, World Bank Policy Research Paper 1620. Claessens, Stijm, Asli Demirguc-Kunt, and Harry Huizinga (1998) How Does Foreign Entry Affect the Domestic Banking Market? World Bank Mimeo. Claessens, Stijn and Tom Glaessner (1997) Internationalization of Financial Services in Asia, World Bank Mimeo. Demiguc-Kunt, Asli and Enrica Detragiache (1997) The Determinants of Banking Crises: Evidence from Developing and Developed Countries, IMF Working Paper WP/97/106. Demiguc-Kunt, Asli and Enrica Detragiache (1998) Financial Liberalisation and Financial Fragility, World Bank Working Paper, Washington. Demiguc-Kunt, Asli and Harry Huizinga (1998) Determinants of Commercial Bank Interest Margins and Profitability: Some International Evidence, World Bank Working Paper, Washington. Dooley, Michael P. (1995) A Survey of Academic Literature on Controls over International Capital Transactions. NBER Working Paper 5352. Eichengreen, Barry, Michael Mussa, Giovanni Dell'Ariccia, Enrica Detriagiache, Gian Maria Milesi-Ferretti, and Andrew Tweedie (1998) Capital Account Liberalisation, Theoretical and Practical Aspects, Washington D.C.: International Monetary Fund Occasional Paper 172. Euromoney (1995), London. Francois, Joseph and Ludger Schuknecht (1998) Trade in Financial Services: Long-run pro-competitive effects, mimeo. Garcia-Herrero, Alicia (1997) Monetary Impact of a Banking Crisis and the Conduct of Monetary Policy, IMF WP/97/124. Goldstein, Morris and Philip Turner (1996) Banking Crises in Emerging Economies: Origins and Policy Options, Basle: Bank for International Settlements. Harris, S. and C. Pigott (1997) Regulatory Reform in the Financial Services Industry: Where Have We Been? Where are We Going? Paris: Organization for Economic Cooperation and Development. Henderson, David (1998) Industrial Policies Revisited: Lessons Old and New from East Asia and Elsewhere, Melbourne: Pelham Paper No. 3. International Monetary Fund (December 1997) World Economic Outlook: Regional and Global Implications of the Financial Crisis in East Asia, Washington D.C. International Monetary Fund (May 1998) World Economic Outlook: Prospects and Policy Issues, Washington D.C. International Monetary Fund (1998) International Capital Markets (ICM), Washington DC. International Monetary Fund (various issues) International Financial Statistics, Washington D.C.. Johnston, Barry, Salim Darbar, and Claudia Echeverria (1997) Sequencing Capital Account Liberalization: Lessons from Experiences in Chile, Indonesia, Korea and Thailand, IMF WP/97/157. Keefer, P., S. Knack and M. Olson (1995) Property and Contract Rights Under Democracy and Dictatorship, Paper presented at CREI Conference on Growth in Barcelona, March 1995. King, Robert G. and Levine, Ross (1993) Financial Intermediation and Economic Development, in Colin Mayer and Xavier Vives (eds.) : Capital Markets and Financial Intermediation, Camebridge: Cambridge University Press, pp. 156-89 Kono, Masamichi, Patrick Low, Mukela Luanga, Aaditya Mattoo, Maika Oshikawa, and Ludger Schuknecht (1997) Opening Markets in Financial Services and the Role of the GATS, Geneva: ϲʹ Special Studies. Krugman, Paul (1998) An Open Letter to Prime Minister Mahathir from Paul Krugman, Fortune Investor, September 1, 1998. Lannoo, Karel and Daniel Gros (1998) Capital Markets and EMU, Report of a CEPS Working Party, Brussels: Centre for European Policy Studies. Levine, Ross (1997) Financial Development and Economic Growth: Views and Agenda, Journal of Economic Literature, 35: 688-726. Low, Patrick and Aaditya Mattoo (1997) Reform in Basic Telecommunications and the ϲʹ Negotiations: The Asian Experience, Geneva: ϲʹ Mimeo. Mattoo, Aaditya (1998) Financial Services and the ϲʹ: Liberalisation in the Developing and Transition Economies, ϲʹ Mimeo, Geneva. Ojeda R.H., R. McCleery and F. DePaolis (1997a) Financial Liberalization, Trade and Regional Macro-Economic Stabilization in the Pacific Rim: A General Equilibrium Analysis, Los Angeles: North American Integration & Development Centre. Ojeda R.H., R. McCleery and F. DePaolis (1997b) The Economy-wide Impact of Financial Liberalization in China and India: A Computable General Equilibrium Simulation, Los Angeles: North American Integration & Development Centre. Rojas-Suarez and Steven R. Weisbrod (1995) Financial Fragilities in Latin America. The 1980s and 1990s, Washington D.C.: IMF Occasional Paper 132. Schuknecht, Ludger (1998) A Simple Trade Policy Perspective on Capital Flows, ϲʹ working paper no. ERAD-98-11. Sorsa, P. (1997) The GATS Agreement on Financial Services A Modest Start to Multilateral Liberalization, Washington D.C.: IMF Working Paper 97/55. World Bank (1997) World Development Indicators (WDI), Washington D.C. World Bank (1998) Global Economic Prospects, Washington D.C.: The World Bank. Annex 1: Sample countries (X = incidence of financial crisis during 1991-97) Argentina Brazil (x) Chile China Costa Rica Czech Republic (x) Egypt Ghana Hong Kong-China Hungary (x) India (x) Indonesia (x) Korea (x) Malaysia (x) Mauritius Mexico (x) Morocco Philippines (x) Poland (x) Romania (x) Senegal Singapore Slovakia (x) South Africa Taiwan Thailand (x) Venezuela (x) Annex 2: Summary of observations on financial services commitments by sample countries in the Uruguay Round (Uruguay Round Schedules, excludes the five sample countries without existing commitments) - Argentina Very distinct difference in commitments, leaving mode 1 unbound and without limitations on mode3, favouring commercial presence. No preference between lending and securities. - Brazil Mode 1 unbound, but mode 3 appears restrictive as well. It is known, however that Brazil has been pursuing gradual liberalisation of foreign commercial presence in banking with the privatisation of state-owned banks. Strong restrictions on operations of foreign commercial presence in banking. - Chile No bias in favour of mode 1 which is unbound for virtually all financial services. An economic needs test (or a national interest test) for mode 3. Measures to discourage short-term capital inflows (eliminated in fall 1998) were included in Chile's schedules despite their essentially prudential nature. No clear preference for either lending or securities. Discriminatory tax on foreign banks' deposits. - Czech Republic Mode 3 appears more liberal than mode 1 with the adoption of the Understanding, and with foreign exchange controls inscribed in the schedule. Although mode 1 is explicitly left unbound in securities, no clear preference either for lending or securities. - Egypt With mode 1 unbound, mode 3 is favoured, with a preference for joint-venture banks. Securities clearly more liberal with no limitations in modes 1 and 3. - Ghana No preference between modes or between banking and securities. - Hong Kong, China With mode 1 unbound, mode 3 is more liberal. However, foreign bank branches or subsidiaries can have offices in only one building. No clear bias either for lending or securities. - Hungary With mode 1 unbound, preference exists for mode 3. Branches are, however, not allowed. Restrictions on commercial presence of securities firms appear more liberal than those on banks. - India With mode 1 unbound, preference exists for mode 3. In mode 3, however, there are numerical restrictions on foreign bank branches, as well as a limitations on the foreign share in total banking assets. Mode 3 slightly more liberal for securities in which 51 per cent ownership by foreigners is allowed. - Indonesia In lending, a clear preference for mode 1, with full mode 1 commitments compared to restricted mode 3 commitments. With mode 1 unbound for securities, preference in favour of loans also exists. As a result of the most recent negotiations, mode 3 has been liberalised significantly for non-banks, but the situation with regard to banks has not changed by very much, except for grandfathering of existing foreign ownership. - Korea Although mode 1 has been kept unbound, preference may have existed in favour of loans compared to securities, due to restrictions on foreign portfolio investment concerning shares (and bonds). Foreign bank branches also have had very limited possibilities for domestic funding which may have led them to rely on imported capital. As a result of 1998 reforms, restrictions on foreign portfolio investment were relaxed substantially, and mode 3 has been significantly liberalised, thereby correcting the preference. - Malaysia With both modes 1 and 3 restricted, difficult to establish which mode is preferred. With emphasis on the establishment of offshore institutions, however, there may be a preference for establishment with a potential for creating large international capital flows. Slight preference for securities, with the banking sector unbound for new licenses and with many restrictions on branching and operations. - Mexico With mode 1 unbound, preference for mode 3 seems to exist. Slight preference for lending compared to securities, as underwriting appears unbound. - Morocco Preference for mode 3, as mode 1 is unbound for deposit-taking while fully bound for lending. Mode 3 is fully bound for banking and securities except for trading of foreign exchange and trading of securities for own account. A preference may exist for lending, as lending is fully bound both in modes 1 and 3, while mode 1 is not fully bound for securities. - Philippines With the entry "commercial presence required" in mode 1, preference for mode 3 evident. Securities more liberal than lending, as no foreign equity limitation exists for securities dealers, while for banks the limit is 40 per cent. - Poland With mode 1 unbound, preference exists for mode 3. Preference seems to exist for lending compared to securities, as trading in securities is entirely unbound. Foreign exchange controls retained in horizontal section of schedule. - Romania An apparent preference for mode 1 exists, as there are no limitations in mode 1 for lending, while an authorisation requirement and other limitations exist for commercial presence of banks. Trading in foreign exchange and securities completely unbound, resulting in a preference for lending compared to securities. - Singapore With mode 1 unbound for the most part, a slight preference seems to exist for mode 3, despite the fact that no new commercial banks are allowed, and no commitment is made for allowing new merchant banks. Only one office permitted for foreign banks, and many restrictions apply on their operations. Preference for securities apparently exists, as no limitations apply to securities trading and underwriting in mode 3, although new membership on the stock exchange is unbound. - Slovak Republic Mode 3 appears more liberal than mode 1 with the adoption of the Understanding, and with foreign exchange controls inscribed in the schedule. Although mode 1 is explicitly left unbound in securities, no clear preference either for lending or securities, except that a citizenship requirement exists for banks' board of directors. - South Africa With mode 1 unbound, preference for mode 3 exists. Lending is preferred over securities, as trading of securities and underwriting are both kept unbound. - Thailand With mode 1 unbound, a preference for mode 3 may have existed. Preferences appear to have existed in allowing different forms of commercial presence of foreign financial institutions, by allowing branches with IBF (International Banking Facilities) licenses priority in new establishment; since those entities do not have a domestic commercial base or any adequate means of funding, they may have acted as vehicles for excessive borrowing from abroad in the form of short-term loans. As a result of the 1997/98 reforms, mode 3 has been liberalised to a certain extent, thereby correcting somewhat the preference. - Venezuela With mode 1 unbound, a slight preference for mode 3 exists, although restrictions appear tight on mode 3 as well. Securities more liberal than banking.  Arranging a loan with a foreign bank abroad via telephone would fall under mode 1 whereas the same loan arranged through the domestic subsidiary or branch of a foreign bank would fall under mode 3.  Financial services which typically give rise to current transfers and payments without necessarily involving cross-border capital movement are services such as insurance intermediation, stock brokerage, provision and transfer of financial information and advisory services. If foreign institutions are chosen for their advance technology or expertise, trade liberalisation can be quite valuable regardless of whether capital movement is allowed.  It is also recognized, however, that restrictions on capital outflows, such as the repatriation of a portion of the invested capital, would discourage inward direct investment. The discussion here only refers to liberalization required by the commitments made by Members under the GATS.  The extent to which transfers of capital related to the commercial presence need to be liberalized under a commitment in mode 3 is not defined in the GATS (Article XVI footnote 8). See the description below of the relevant GATS provisions.  See GATS Article XVI, Footnote 8. What constitutes an "essential part of the service "for mode 1 trade and an "inflow related to the supply of the service" under mode 3 trade is not further specified.  This provision is also likely to constrain inflows: if the repayment of a loan from abroad arranged through a foreign affiliate can not be made due to controls on capital outflows, this is likely to discourage such loans, regardless of whether a generous or narrow interpretation of GATS provisions regarding inflows is applied.  This literature, in turn, extends the more generic theoretical and empirical debate on the importance of financial services for economies (see Levine, 1997, for a survey on the economic role of the financial sector, and King & Levine, 1993, for the importance of skill and knowledge transfers).  See also IMF WEO, May 1998;and ADB, 1998, for the use of such arguments in explaining the Asian crisis, and Demirguc-Kunt and Detriagache (1998) for an empirical study of financial instability. The debate on capital controls contrasts quantitative restrictions/prohibitions and tariff-like protection such as reserve requirements and transaction taxes (see Schuknecht, 1998 for a trade policy perspective on capital controls).  See, for example, the arguments by Bhagwati (in Foreign Affairs, May-June 1998) and Krugman in his open letter to Prime Minister Mahathir of Malaysia (Fortune Investor, September 1998).  Johnston, Darbar and Echeverria (1997) have developed a blue-print for the phasing of domestic policy reform and capital account liberalisation.  For the theoretical underpinnings of this discussion, see, e.g., Levine (1997) and King and Levine (1993). The literature on the international trade dimension is very limited (see, e.g., Francois and Schuknecht (1998).  Infrastructure includes trading facilities, payment systems, trading personnel, and communication facilities. Market development refers mainly to the depth of markets and the financial instruments available. The Basle Core Principles for Effective Banking Supervision (BIS, 1997) discuss the key elements of effective regulation and supervision, and transparency.  Perverse incentives could arise in the absence of an orderly exit policy (IMF, ICM, p.76).  For a theoretical analysis see Francois & Schuknecht, 1998; for a CGE analysis see Ojeda, McCleery and DePaolis, 1997a; and for empirical studies see Demirguc-Kunt & Huizinga, 1998, Ojeda, McCleery and DePaolis, 1997b (India and China), and for what can be expected in the Euro area, see Lannoo and Gros, 1998.  On a related note, observers have sometimes blamed cultural factors for the relatively underdeveloped securities markets in developing countries. Trade restrictions and lack of incentives to adopt modern financing techniques could provide an alternative explanation for such observations.  The policy commitments are listed in the ϲʹ Members' Schedules of Specific Commitments made at the end of the Uruguay Round in December 1993. Those commitments entered into force under the GATS agreement on 1 January 1995. For many countries, the commitments were revised as a result of the financial services negotiations in 1995 (entry into force on 1 September 1996) and those in 1997 (entry into force on 1 March 1999). Such changes, which are overwhelmingly liberalising commitments, are not reflected in the analysis, as they have in large part entered into force too recently to affect fluctuations in capital movement appearing in the data. Regarding the commitments of ϲʹ Members, account has been taken of the fact that in many countries, there are "horizontal" limitations, or restrictions applying to all service sectors inscribed in the Schedules which affect capital movement.  This does not mean that mode 3 is "good" and mode 1 is "bad", but that "ceteris paribus" these modes give rise to different impacts on capital flows and financial stability. The later estimations will also control for the effect of sound macro- and regulatory policies through appropriate variables.  The underlying assumption is that a bias in commitments towards cross border trade and related capital flows can be detrimental to financial stability. The indicator does not incorporate the positive effect on efficiency as emanating from cross-border liberalisation.  Complete data on bank flows are only available for a more limited number of countries in this sample which does not permit proper econometric regressions. Time-series data distinguishing short term and long term flows and bank lending as a fraction of "other investment" is only available for the period after 1995.  Inflation, deposit rates, M2, GDP and exchange regime data are from IMF IFS.  Claessens and Glaessner (1997) have started to develop such indicators for a number of Asian countries.  The compilation of capital controls by the IMF is less useful in this context, as it does not distinguish qualitative differences between countries' restrictions.  Fourteen of the 27 sample countries experienced financial crisis over the 1991-97 period.  The latter variable was a good predictor of financial crisis in Demirguc-Kunt and Detragiache (1997).  Possibly, controls on entry for portfolio investment (as captured by this variable) are not effective or they do not capture the type of controls which really "bite" and restrict capital flows.  (..continued)  PAGE 3  DATE \@ "dd/MM/yy" 02/03/99,  TIME \@ "HH:mm" 16:38,  FILENAME ERAD-98-12.doc Capacity - Transparency and information - Regulation & supervision - Infrastructure & market development, risk management Efficiency - Competition - Technology transfer - Skill transfer & development Financial services trade Financial sector stability Capital flows - Quantity - Structure (term, instrument) - Volatility -.6789:<T~5:@ͺ͵}oeoWoeO6@OJQJ5<CJOJQJmHnH5<CJOJQJj5<CJOJQJU"jQ5@CJOJQJU5@CJOJQJj5@CJOJQJU5@CJOJQJOJQJ5CJOJQJjUhmHnH @OJQJj@OJQJU@OJQJj@OJQJU5@OJQJ9;<U~ $*$ 0 $*$ 0U"p#"$&*$ND $d %d &d 'd /.+Dт 0$*$ 0$*$ 0 $*$dx 0 $*$d 0$*$ B#9;<U~;A  6Lp ' ( ) * 3 mnL]/W!\#]#^###?%J'[*\****+8+9+]++   a$*$ ) `0p@  $*$ 0$*$ 0 $*$ 0!p# $*$ 0;A  6Lp$*$d$*$$*$d ) `0p@  $*$ ) `0p@  $*$ ) `0p@  @AY[    ' 3 L\\##?%%[***8+9+]+++Q,j.k...00011@1112K3L3M3333x47ŹѶCJj0JCJU5B*B*j0J@U5@@H*@j5CJU5j5U56CJCJH*5>*jUhmHnH@H*OJQJ @OJQJ @OJQJ: ' ( ) * 3 mnL]/W!\# $*$d$*$$*$$*$d$*$d\#]#^###?%J'[*\****+8+9+]++++++,0,,0$$l  # $*$dh$ $*$d$*$$*$0+++++,0,O,P,Q,.00000011@111112222x4759<<@E1MSSTTaUW|Z}ZZZZZZZZZZZZZZZ[[[[[[[[[ [ [ [ [ [[[[[[[_``:a;acff5g6g7g8g9g:g;g=g>g?g@g    `0,O,P,Q,.00000011@1111<*$$4F#dh$dh $*$d$*$d$$l  # $*$dh$112222x4759<<@E1MSSTTaU$*$$*$0 $*$d$*$d  & Fddh*$$4F#dh$77778849 <<<<<@@EEKKK/L0L.M/MST]U^UOVPV|Z}ZZZZZZZZZZZZ[ __`Oafff4gCgDgIgJgggggggggggghhehfhhhhhhӻ >*B*CJ B*OJQJB*CJOJQJB*CJB*CJ5B* jUmH50J@5@ B*OJQJ@j0J@UJaUW|Z}ZZZZZZZZZZZZZZZ[[[[[[[[[ [ J ] $*$d [ [ [ [ [[[[[[[_``:a;acff5g6g7g8g9g:g$$$$*$d$*$ $*$d J ]:g;g=g>g?g@gAgBgCgDgEgFgGg o:$$B` s  #$$$N$$Bֈ s  # @gAgBgCgDgEgFgGgHgIgJgKgLg`gag{gggggggggggggggggggggggggggggggghh hhhhhLhMhRhYh^hehfhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhiiiii?i@iGi  bGgHgIgJgKgLg`gag{ggg0$$$N$$ֈ s  #$$ ggggggggggg[N$$Bֈ s  #$$$N$$Bֈ s  # gggggggggggggggggg`N$$Bֈ s  #$$$$B s  #$ggghh hhhhhLhMhRhYh^hehfhhhhhhh &$$B8 s  #$$$B s  #$$hhhhhhhhhhhhhhhhhhhhhhh٨P$$B s  #$$$&$$B8 s  #hhhhhiiaibiiiiiiiiii/j0jcjdjjjkjjjjj3k4ksktkkkPlQlllmmWrdrww3x4xPz{{/{&1+„8ÆĆՆ߆ſŵ>*B*CJOJQJ>*B*CJOJQJ B*OJQJB*CJOJQJB*CJOJQJ5B*5B*OJQJj0J@U5@@B*CJ >*B*CJB*CJChhhiiiii?i@iGiLi\iaibiwixiiiiiiiiiii$$B s  #$$$GiLi\iaibiwixiiiiiiiiiiiiiiiiiiiiiiiiiiijjj#j(j/j0jJjKjPjWj\jcjdjejfjgjhjijjjkjljzj{j|j}j~jjjjjjjjjj.k/k0k1k2k3k4knkokpkqkrksktkkkkkkkkKlLlMlNlOlPlQlll  biiiiiiiiiiiiiiiiijjj#j(j/j0jJjKjPj$$$B s  #$$PjWj\jcjdjejfjgjhjijjjkjljzj{j|j}j~jTN$$Bֈ s  #$$$$B s  #$~jjjjjjjjjj.k/k0k1k2k3kN$$Bֈ s  #$$$B s  #$$3k4knkokpkqkrksktkkkkkkk0$$BL s  #$$$:$$B` s  #kkKlLlMlNlOlPlQllllllllmmmm`$$$N$$Bֈ s  #llllllmmmmmmmmpWrswPzQzRz{{%+,„78†ÆĆņƆ͆89@GNUVopwz}ÇćԇՇ܇   bmmmmmpWrswPzQzRz{{%+,$*$ $*$d$d$*$dD$$Bt s  #$$,„78N$$Bֈz l #$$$$*$d$*$†ÆĆņYN$$Bֈz l #$$N$$ֈz l #$$ ņƆ͆N$$Bֈz l #$$$ ߆UVÇć#$WX~܈݈Ήωqrŗrդ &(ξ CJOJQJ5CJOJQJ 5OJQJ j0JUj0J@U@5@ 5@CJ B*OJQJ>*B*CJOJQJB*CJOJQJB*CJB*CJOJQJB89@GNUVop`$$Bz l #$$$N$$Bֈz l #pwz}ÇćԇՇ܇߇ &$$B8z l #$$$$Bz l #$܇߇#$FGKOSWXmnrvz~LjΈՈ܈݈pωInjĐėŗ}rsդ֤ !)]jw|   b#$FGKOSWP$$$Bz l #$$WXmnrvz~ٜ$$Bz l #$$$&$$B8z l #LjΈՈ܈݈pωInjĐ $*$d$*$d$$Bz l #$$$ėŗ}rsդ֤ !)]jw|$$ R$  $*$d$*$d¥ĥʥ˥Υϥ"$-.0:?DPkpuæȦЦ%;@EUbgl~§ӧۧݧߧ #4GLQcpuzŨҨ>*CJOJQJ5CJOJQJ CJOJQJ\ĥ˥ϥӥ#$%&'()*+,-./0:<?BDGJMPRcdkmpsuy}æƦȦʦ̦ΦЦҦԦզ  b,$$$c$$64֞ F"&``````  ĥ˥ϥӥxupppppppppp$$$$$64  Fy="&    #$%&'()*+,-g0d$$$64  Fy="&  $$ -./0:<?BDGJMPli$$$64  Fy="&  $$$ PRcdkmpsuy}n$$64  Fy="&  $$$ qn$$$64  Fy="&  $$ æƦȦʦ̦ΦЦqtn$$$64  Fy="&  $$ ЦҦԦզqn$$$64  Fy="&  $$ q0n$$$64  Fy="&  $$ !#%'+,;=@CEHMPUW[\bdgjlpty~§ŧɧΧӧէ٧ڧۧܧݧާߧ   b!#%qn$$$64  Fy="&  $$ %'+,;=@CEHMPUqn$$$64  Fy="&  $$ UW[\bdgjlpty~qn$$$64  Fy="&  $$ ~qn$$$64  Fy="&  $$ §ŧɧΧӧqn$$$64  Fy="&  $$ ӧէ٧ڧۧܧݧާߧq0n$$$64  Fy="&  $$  qn$$$64  Fy="&  $$ !#'+/4qn$$$64  Fy="&  $$ !#'+/48<=GILOQVZ_ceijpruxz~¨èĨŨƨǨȨҨԨרڨܨݨިߨ %(,01=?BE  b48<=GILOQVZ_cqn$$$64  Fy="&  $$ ceijpruxz~qn$$$64  Fy="&  $$ qn$$$64  Fy="&  $$ ¨èĨŨq0n$$$64  Fy="&  $$ ŨƨǨȨҨԨרڨܨݨިߨqln$$$64  Fy="&  $$ Ҩרܨ(=BGYglq~©ǩ̩ԩ!1EJO[kpuܭ`bST*tu),-D:Z5>*CJOJQJjCJOJQJUmHOJQJ 5OJQJj0J@U5@@5CJOJQJ CJOJQJKqn$$$64  Fy="&  $$  %(qn$$$64  Fy="&  $$ (,01=?BEGKOTYqn$$$64  Fy="&  $$ EGKOTY[_`giloqtx{~©ĩǩʩ̩ΩЩҩԩ֩ة٩!$),1378EGJMORUX[]abkmpsuz  bY[_`giloqtx{~qn$$$64  Fy="&  $$ ~q0n$$$64  Fy="&  $$ qn$$$64  Fy="&  $$ ©ĩǩʩ̩ΩЩҩԩq|n$$$64  Fy="&  $$ ԩ֩ة٩qn$$$64  Fy="&  $$ !$),1qn$$$64  Fy="&  $$ 1378EGJMORUX[qn$$$64  Fy="&  $$ []abkmpsuz~qn$$$64  Fy="&  $$ z~ѪҪIDܭݭlabU)*tuvwxyz{|}~()-;<=>?@ABCDEQ_jv#./01234567   bqn$$$64  Fy="&  $$ ѪҪIDܭqmmiggggg7$$64  Fy="&  $$ ܭݭlabU)*tuvwxyz{|} $$64&<$$ $*$dh $*$d$*$d}~()-Zl=$$64\"&  $$$_$$6֞d n5&&  -;<=>?@ABCDEQ_s$$6  bB""&  $$$ _jvvs$$$6  bB""&  $$ #./0123456v,sssssss$$$6  bB""&  $$ 6789:NZ_dinsx}<||||||||$$z$$6  bB""&  $789:NZ_dinsx} "#:RW\afkpuz8nop(/:01Z.b+,R.O   b}Dzz$|$$6/  bB""&  $$  "$zzzzzzzzz$$$|$$6   bB""&  #R78lbc:01W*+,S/OIKsxe VXY[jyCJH*CJ5CJ56@j0J@U5@@CJ CJOJQJ5CJOJQJP"#:RW\afkpuzxzzzzzzzzz$$$|$$6[  bB""&  zzzzzzzzz$$$|$$67  bB""&  8nop(/|||||rjjjj$*$d $*$d|$$6  bB""&  :01Z.b+,R.Od$ $*$d$*$ $*$d$*$d IJKOSW[\stuvwx +19>HNX^ij~ !"#$6<FKRX   b IJKOøgHd_$$$V$$l4r 4% d$4$$lR40 %  $d$ OSW[\stuvwpz d$ LC d$ C d$  d$ Cd$W$$lr 4% $$ wxL d$ LC d$ C d$  d$ Cd$Q$$lr 4%  $ LC$ C$ $ C$ =S$$l+r 4%  $ LC$ C$ $ C$ =S$$lr 4%   P$ LC$ C$ $ C$ =S$$l!r 4%   +19>HNX^i$ LC$ C$ $ C$ =S$$l)r 4%  ij~$ LC$ C$ $ C$ =S$$l9r 4%  $ LC$ C$ $ C d$ =Q$$lr 4%  $ LC$ C$ $ C$ =S$$lr 4%   !"#$ LC$ C$ $ C$ =S$$l4r 4%  #$6<FKRX`a$ LC$ C$ $ C$ =S$$lr 4%  X`abz{| VWXYZ[uy <=>CLQZ_ijyz{|}~   babz{|$ LC$ C$ $ C$ =S$$lr 4%  |$ LC$ C$ $ C$ =S$$l/r 4%   V$d$d$ $*$dQ$$lr 4%  VWXYZ[uy `4$$l640 %  $$$2$$l40 %`  x$$$V$$l4r %`` p?HQ$$lr %  $ C$ C$ C$V$$l4r %    {r d$ C d$ C d$ C d$ =Q$$lr %  $ C$ C$ C$ <=>CLQZ_i0$ C$ C$ =S$$lr %  ijyz{|}P d$ C d$ C d$ C d$ =Q$$lr %  }~P$ C$ C$ =S$$lr %  '()*+,EFKRW^cjowx',16;<=345J[oz{|   b$ C$ C$ C$ =S$$lr %  '()*+ dd$ C dd$ C dd$ C dd$ =S$$lr %  '35z|8Qbdu34IJ~"1m4CXq 4g+x+++i8N9!;;@@ABB CEFF?]@]j0JCJU@CJ6@0J@j0J@U@5@CJH*CJ5CJ5CJH*CJ5CJL+,EFKRW^cjow0$ C$ C$ C$ =Q$$lr %  wx$ C$ C$ C$ =S$$lr %  $ C$ C$ C$ =S$$lr %  x d$ C d$ C d$ C d$ =S$$lr %  ',16; d$ C d$ C d$ C d$ =S$$l r %  ;<=3$$$S$$l r %  345J[oz{|y8J$$l4\ }>#  $$$2$$l40 #  lVDF$$l\ }>#  $ tC$ C$ C$J$$l4\ }>#  UH$$l\ }>#  $ C$ C$ =F$$l\ }>#  $ tC  !"#>?EMNOPabhpv"+078PQTWZ[dejotuv 34J~   b !"LH$$l'\ }>#  $ tC$ C$ C$ = "#>?EMNOPaU F$$l\ }>#  $ tC$ C$ C$ =H$$l\ }>#  abhpvH$$l^\ }>#  $ tC$ C$ C$ = U\F$$l\ }>#  $ tC$ C$ C$ =H$$l \ }>#  "+078PQTWZF$$l\ }>#  $ tC$ C$ C$ =Z[dejotuvhWUUF$$l\ }>#  $ tC$ C$ C$ =F$$l\ }>#  34J~L4$$lR40 %  $d$d$ $*$d$*$dHGpW$$lr 4% $$$X$$l4r 4% LQ$$lr 4%  d$ LC d$ C d$  d$ Cd$  !"123456RSX_dklm#(01234CDEFGH\pu}~ -FGHINW  b !<$ LC$ C$ $ C$ =S$$l+r 4%  !"12345P$ LC$ C$ $ C$ =S$$l!r 4%  56RSX_dkl$ LC$ C$ $ C$ =S$$l)r 4%  lm$ LC$ C$ $ C d$ =S$$l9r 4%  $ LC$ C$ $ C$ =S$$lr 4%  $ LC$ C$ $ C$ =S$$l4r 4%  #(0123$ LC$ C$ $ C$ =S$$lr 4%  34CDEFGP$ LC$ C$ $ C$ =S$$l/r 4%  GH\pu}~$ LC$ C$ $ C$ =S$$l9r 4%   0$ LC$ C$ $ C$ =S$$l/r 4%  -FGHINWXpqtwz$ C$ $ C$ =S$$l/r 4%  $ LCWXpqtwz}~a > p  4!'C)g+h+i+x+++q,r,J-K---S.T...k/l///008191G2H2b2c222P3Q33344354555>6?6666677h8i8O9P9:::: ;!;;;1<2<<<==>>#?$?   bz}~a > p  4$*$d $*$dS$$l/r 4%  $ LC!'C)g+h+i+x+++q,r,J-K---S.T...k/l///0$*$ $*$d$*$d008191G2H2b2c222P3Q33344354555>6?6666677h8i8$*$i8O9P9:::: ;!;;;1<2<<<==>>#?$???.@/@u@v@AA$*$*$$???.@/@u@v@AAA)A/A5A@ASAYA_AoA{AAAAAAAAAAAABBB$B1B?B@BBB C CCCCCCCDDEEFFFFGGGGZH[HcHdHHHHHoIpIzI{I6J7J?J@JuKvKKK-M.M6M7M=O>OIOJOPPPP~QQQQRRSSSS dAA)A/A5A@ASAYA_AoA{AAAAAAAAAAAABBB$B1B?B@BB$*$BB C CCCCCCCDDEEFFFFGGGGZH[HcHdHHHH*$$*$HHoIpIzI{I6J7J?J@JuKvKKK-M.M6M7M=O>OIOJOPPPP~QQQQ*$QRRSSSSSSTTTT(V)V5V6VXX'X(XtYuYYY!Z"Z-Z.Z\*$SSSTTTT(V)V5V6VXX'X(XtYuYYY!Z"Z-Z.Z\\\\>]?]^_``abc"efghhljjl'mopqr)txttuuLvwwww"w#w$w%w'w(w*w+w3w4w6w7wwwwww x x xx'x>x^x_xxxyxxxxxxxxx#   ]\\\\?]^_``abc"efghhljjl'mopqr)txttd$*$$*$@]^ ^___``aaabbbccc"e#efffffggh hhh jjljmjnjjjjjllll&m'm(m)mppqqrrr(t)t*txtytttuuuuuuuKvLvMvwwww!w#w$w%wCJ jCJU@CJ6CJ@@CJj0JCJUCJ j0JUj0JCJUCJNtuuLvwwww"w#w$w%w'w(w*w+w6w7wwwwww x x xx #"$1$d$*$%w&w(w)w+w,w2w3w4w5w7w8wLwMwUwVwXwYwjwkwpwqwswtw~wwwwww x xx^x_xxxyxxxxxxx@CJ>*CJ0J$<mHnH0J$<j0J$<U <mHnH< j<U 0J$mHnH0J$ j0J$UCJ*x'x>x^x_xxxyxxxxxxxxxx$*$ 000&P . A!"#$%7,...()()())))() 000&P . A!"#$%7,...()()())))()QD phoenixSDphoenixQD phoenix4 [<@<Normal1$CJOJQJhmH nH \"\ Heading 1!$$ & F 1$0@&5;OJQJhnHb2b Heading 2+$$ & F 1$0@& h:OJQJhnHbBb Heading 3+$$ & F 1$0@& h5OJQJhnH`R` Heading 4+$$ & F 1$0@& hOJQJhnHTT Heading 5$ & F 1$0@&6OJQJhnH<A@<Default Paragraph FontRB@R Body Text$ & F 1$ hOJQJhnH0+0 Endnote TextCJ6*@6Endnote ReferenceH*2@"2 Footnote TextCJ8&@18Footnote ReferenceH*>>TOC 1*$0 $ mH ::TOC 2*$0 $ mH ::TOC 3*$p0 $ mH ::TOC 4*$@ 0 $ mH ::TOC 5*$0 $ mH 66TOC 6*$0 $mH ..TOC 7 *$0mH 66TOC 8*$0 $mH 66TOC 9*$0 $ mH > >Index 1*$` $ mH > >Index 2*$0 $ mH :.: TOA Heading *$ $mH &"&Caption CJ2O2_Equation Caption,@",Header " 9r , @2,Footer # 9r &)@A& Page NumberTPRT Body Text 2%$ & F 1$ OJQJhnHXQbX Body Text 3&$ & F 1$p OJQJhnH`r` Body Text 4('$ & F 1$p0 8pOJQJhnHTT Body Text 5($ & F 1$ OJQJhnHH0H List Bullet)$ & F 1$OJQJhnH^6^ List Bullet 2!*$ & F1$0 OJQJhnH^7^ List Bullet 3!+$ & F1$ @OJQJhnH`8` List Bullet 4$,$ & F1$p0 OJQJhnH`9` List Bullet 5$-$ & F1$ OJQJhnHR1R List Number.$ & F1$ hOJQJhnH^:^ List Number 2!/$ & F1$0 OJQJhnH^;^ List Number 3!0$ & F1$ OJQJhnH^<^ List Number 4!1$ & F 1$p0 OJQJhnH^="^ List Number 5!2$ & F 1$ OJQJhnHd2dParagr. Num. - WTO3$ & F1$ hOJQJhnHj*K//33<G/H.I]QOR [3tqSb*t n L - l9I }Nt}NQ t  &&~~~~~~~~~@7h߆Ҩ@]%wx;gi\#0,1aU [:gGgggghhiPj~j3kkm,ņpW-PЦ%U~ӧ 4cŨ(Y~ԩ1[ܭ}-_6"Ow i#aVi}+w;3"aZ!5l3Gz0i8ABHQ\txx     !"#$%&'()*+,./012345679:<=>?@ABCDFGHIJKLMNPQRSTUVWXYZ\]^_abcdfhj+@gGil܇Ez7XW$?Sx -8EO[`et !#&;DGY_bm|!tt tt8@,( 2 Z  s *{? Z  s *{? Z  s *{? (2  (2  H  #   N  3   NB   S DNB   S D NB   S D NB   S D (   (  H  #  (  H  #  NB  S DNB  S DNB  S DN  3  HB  C DHB  C DHB  C DB S  ?9~VVVVVVVVVVVVVVVVV)*+tB#!p@B#!p@B#!p@  et0t0 t0t 20t rt <  t P\ \ t \ @\ t@< 0 t< "t Pt@ $ta At t-it Yt@&tpt0t $*+5"1 + 1 2 : @-B-----..u4x44444@5J5q7777&84888<#<(<1<=<F<H<U<Z<b<<<<<<<<<<<<<^=b=c=h=??aGlGmGrG>JGJ_JhJmJvJ0K9K/L8L=LFLYYYY>bBbCbHbRb[b`bib3d=dggiiii:m>m@mEmkztzV`w{#-+  9FKV''''r(x())))))))))9-D-e-q-s-y-z--------Q/Z/}44445%5P5T5V5_5n5t5u5{5}5555555555556 6P6X6c6j6666666!7'777772888:8A8888888999999::::::$;.;0;6;;;>?ZZ[[\\ʱ````aŽaabb%,鳦ܳddddfff$&.3;Dzgggggggggg&38"#$%'(*+57ղݲssstҽ>?bװ99::;;%9p)2T%&'.... OJQJo( OJQJo( OJQJo( OJQJo(hh. hhOJQJo(0.0.0.0()h.0()p0p()()p@ ()ho(0o(()p0po(()p0po(-0o(()0o(()0o(-p0po(()@ 0@ o(()ho(. 0OJQJo(-.h.h.h()()h.0()0()8h8-~}|247 >%<'z@\\MARCOPOLO2\2067-C1Ne00:winspoolHP LaserJet 4\\MARCOPOLO2\2067-C1pc XXA4 7''''\\MARCOPOLO2\2067-C1pc XXA4 7''''s t@GTimes New Roman5Symbol3& Arial7"Univers9CG Times")&3F&3Fy2f1_!)@d]LCommitments Towards Financial Services Trade Liberalization, Capital Flows, SchuknechtHornOh+'00 @L h t  MCommitments Towards Financial Services Trade Liberalization, Capital Flows, eomm Schuknecht chu Normal.dot Hornl.d2rnMicrosoft Word 8.0 @@v*_@d@d1՜.+,D՜.+,< hp   OMC - ϲʹt_] MCommitments Towards Financial Services Trade Liberalization, Capital Flows, Title 6> _PID_GUIDAN{FD0988B5-E329-11D1-BD96-000629B067E1}  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklnopqrstvwxyz{|}~Root EntryingxPb4ssB_z F dtdtData.logz b% gudj j  m1Table@^hj'TDETECT.COMhu?^WordDocument' xybddSummaryInformationpagefile.(@>b?zdZZDocumentSummaryInformationiles8Pd?\tdCompObjRECYCLERh@ bdjObjectPoolxZb zdD$D$tdtdTillf  FMicrosoft Word Document MSWordDocWord.Document.89q