Financing Investment in East Asia: Regional or Global Savings?

Transcription

1 Financing Investment in East Asia: Regional or Global Savings? Soyoung Kim * Sunghyun H. Kim ** Yoonseok Choi *** Seoul National Universy Sungkyunkwan Universy Suffolk Universy and Suffolk Universy Abstract This paper investigates the extent to which domestic investment in East Asian countries is financed by domestic, (East Asian) regional and global savings in order to infer the relative importance of regional vs. global capal markets in East Asia. Panel regression results show that regional saving in East Asia plays a much more important role than global saving in financing investment in the region. The results suggest that global capal flows, despe of s huge volume in East Asia, does not contribute to proper investment financing. The results also show that Japanese saving has significant effects on regional investment but Chinese saving does not. JEL Classification: F4 Key Words: saving, investment, financing, East Asia, capal flows. * Department of Economics, Seoul National Universy, Seoul, Korea. soyoungkim@snu.ac.kr. ** Corresponding Author. Department of Economics, Sungkyunkwan Universy, Jongno-gu, Seoul, Korea and Department of Economics, Suffolk Universy, 8 Ashburton Place, Boston, USA. Tel: , henry.kim@suffolk.edu. *** Department of Economics, Suffolk Universy, Boston, USA. ychoi2@suffolk.edu.

2 1. Introduction International financial market liberalization has substantially raised the degree of capal mobily in East Asian countries since the 1990s. 1 In particular, since the 1997 Asian Crisis, East Asian economies have accelerated regional financial cooperation and integration in part to safeguard the region's financial markets against the spillover of global market instabily, and also to promote financial market development in the region. 2 However, is not yet clear whether the regional financial market integration increased the degree of capal mobily whin East Asia. Various signs suggest that capal mobily whin East Asia may not be as high compared to the degree of capal mobily between East Asian countries and developed countries. 3 Recent buildup of global imbalances may reflect the tendency of "international" mobily of East Asia's capal. For example, while many developing countries in East Asia continue to face significant infrastructure and investment defics, they also accumulate large current account surplus. This paper investigates the evolution of saving and investment relations of East Asian countries, which can provide evidence of regional and global financial integration in East Asia. 4 In particular, we examine the role of regional vs. global capal markets by measuring the extent to which domestic investment is financed by domestic saving, regional saving, and global saving. Based on such relations, we address how regional or global savings have contributed to financing domestic investment of East Asian countries in recent years. Since Feldstein and Horioka (1980)'s seminal contribution, numerous studies have investigated the degree of international capal mobily based on the relation between domestic investment and domestic saving. 5 Under financial autarky, domestic investment and domestic saving should be perfectly correlated as domestic investment is fully constrained by domestic saving. However, if capal is perfectly mobile internationally, domestic investment and domestic saving should exhib a low correlation because capal can move freely to any place wh high returns. We extend the Feldstein-Horioka methodology by adding regional and global savings in 1 See, for example, Kim, Kim and Wang (2007). 2 There have been several important regional iniatives, such as the ASEAN Surveillance Process, the Chiang Mai Iniative, the Asian Bond Markets Iniative, and Asian Bond Fund. 3 Eichengreen and Park (2004) and Kim, Lee, and Shin (2007) document that financial market integration whin East Asian countries has been far slower than financial market integration between East Asia and advanced countries. 4 We examined the same issue wh different data and methodology in Kim et al. (2011). 5 For example, see Murphy (1984), Obstfeld (1986, 1995), Bayoumi (1989), Kim (2001) and Kim, Kim and Wang (2007). 2

3 the saving-investment regression and investigate the role of regional and global capal markets in financing domestic investment of East Asian countries. 6 Panel regression results show that regional saving plays a much more important role than global saving in financing investment in East Asia. Global capal flows, despe of s huge volume in East Asia, does not contribute to proper investment financing. In particular, the role of Japanese saving is important in driving regional saving: when Japan is included in global saving (instead of regional saving), the coefficient on global saving becomes significant. However, Chinese saving des not play an important role in financing investment in East Asia. This paper adds value to the previous lerature on East Asia's experience of capal account liberalization and financial integration. While past studies have used various measures to document different degrees of regional vs. global capal mobily and financial market integration in East Asia, there have not been any studies that focus on the saving and investment relation. 7 Although some studies have investigated the saving and investment relation in East Asia, they have not examined the issue from the comparative perspective of regional vs. global capal markets of East Asia. 8. The remainder of the paper consists of the following sections. Section 2 provides empirical evidence on capal flows in the Asian region wh portfolio investment data. Section 3 explains the empirical methodology. Section 4 presents the empirical results of savinginvestment regressions. Section 5 concludes the paper. 2. Regional vs. Global Capal Flows Although data on cross-border capal flows by the source and destination countries remain limed, "Coordinated Portfolio Investment Survey (CPIS)" by the International Monetary Fund reveals an interesting trend in this regard. 9 The East Asian region s portfolio investment in regional assets rose from 14.8% of total assets in 2001 to about 27.9% in 2007, while developed economies (US, European Union 15, and Japan) account for about 70.5% of the region s liabilies in 2007, down from 77.5% in See Kim and Kim (2009) for more detailed discussion on methodology and applications to more general samples. 7 For example, see Eichengreen and Park (2004), Kim, Lee, and Shin (2007), Kim and Lee (2008), Kim et al. (2004). 8 See Kim, Kim, and Wang (2007), Kim et al. (2005) and Sinha (2002) 9 Data on cross-border direct investment by the source country are not currently available for most sample countries. 10 See Kim et al. (2011) for details. 3

4 Figure 1 shows relative shares of portfolio investment by the source country---from G6, from East Asia excluding Japan, and from Japan (as a ratio of total portfolio inflows from all countries). CPIS data is available for seven countries in the region; Hong Kong, Indonesia, Korea, Malaysia, the Philippines, Singapore and Thailand (China and Taiwan excluded). The figure shows that in most countries, there is a clear pattern that capal inflows from G6 countries decrease since Since the 2008 global financial crisis hs G6 countries more severely than East Asian countries, portfolio investment by G6 countries decreases during and prior to the crisis period. In Korea and Singapore, we can observe a steady decrease in the share of G6 in portfolio inflows throughout 2000s. In , portfolio inflows from G6 countries increase slightly as these countries recover from the 2008 crisis. Share of portfolio inflows from regional sources increases over time in Korea, Singapore, the Philippines, especially during the crisis period in However, the share of East Asia decreased in some countries such as Thailand and Hong Kong. The relative size of portfolio inflows from regional sources (even after including Japan) is still que small compared to portfolio investment from G6 countries. Exceptions are Indonesia and Malaysia where regional portfolio inflows are more than a half the size of portfolio investment from G6 countries. Singapore is the main reason for a large share of regional capal inflows in these two countries. 3. Empirical Method While the original Feldstein-Horioka saving investment correlation puzzle is based on cross-sectional regression analysis, we start from the following saving-investment regression that has been widely used in past studies in a time-series or panel regression setup. ˆ I = α + βsˆ + ε (1) i where I is domestic investment, S is domestic saving, the subscript i indicates country i, and the subscript t indicates time t. Hat variables denote percentage deviations from the previous period. The coefficient β represents how saving is related to investment, called saving retention coefficient in the previous studies. A high β can be interpreted as an evidence of low degree of international capal mobily. This regression may also be interpreted as showing how investment is financed by domestic saving. A small (or large) β suggests that only a small (large) fraction of domestic 4

5 investment is financed by domestic saving. If domestic investment is not fully financed by domestic saving, a fraction of domestic investment is likely to be financed by foreign saving, which implies a non-zero degree of international capal mobily. In the following, we extend this interpretation explicly, in order to evaluate the relative role of regional vs. global capal markets (or saving) in financing domestic investment. We add regional and global savings as explanatory variables to the equation (1): ˆ I = α + βsˆ + γsˆ + δsˆ + ε (2) i R G where S R is East Asian regional aggregate saving (excluding own country s saving), and S G is global aggregate saving (excluding East Asian countries). The regression shows how domestic investment is related wh domestic, regional, and global savings. We can interpret β as the usual saving retention coefficient. Further, γ and δ can be interpreted as how much domestic investment is financed by Asian regional aggregate and global saving. If we extend the interpretation that a low β implies a high degree of international capal mobily, a high γ (a high δ) can be interpreted as a high degree of regional (global) international capal mobily, because domestic investment is likely to be less related wh domestic saving but more related wh foreign saving when the degree of international capal mobily is high. One potential problem in interpreting the estimated β as the (inverse of) degree of international capal mobily is that changes in saving are not exogenous to changes in investment. That is, changes in investment may affect changes in saving. However, we can still interpret the coefficients β, γ, and δ as a simple relationship between saving and investment. The estimated β shows how domestic saving and investment are correlated and can provide some information on the low bound of the degree of international capal mobily. Nine East Asian countries are analyzed in this paper: China, Hong Kong, Indonesia, Korea, Malaysia, the Philippines, Singapore, Thailand, and Taiwan. We use annual data from 1980 to Saving and investment data are constructed from the Penn World Table, ver Saving is defined as the GDP minus government consumption and private consumption. We use percentage deviations of the level data for panel regressions. 12 Regional saving is constructed by 11 For national GDP, we use constant price GDP per capa data (rgdpl) multiplied by population. Consumption, government spending and investment data are calculated by multiplying relative share data series (kc, kg, ki) by GDP. For China, we use the China version 2 data. 12 We do not use saving and investment ratios over GDP in the regression because the ratio data are subject to business cycle fluctuations and may produce spurious regression results. 5

6 adding saving data of the nine East Asian countries in the sample excluding own country. We use two different versions for regional aggregate, one whout Japan and the other wh Japan. For global aggregate saving, we use the sum of G7 countries savings (or G6 countries when Japan is included in the regional aggregate saving). Since we use the data series denominated in international dollar at the Penn World Table, we do not have a un of account problem in aggregating national savings. For panel data regression, we used fixed effects panel regression. Standard errors are corrected for heteroscedasticy by panel-corrected standard errors (PCSE). 4. Empirical Results 4.1. Basic Statistical Properties of Saving and Investment Table 1 reports main statistical properties of saving and investment (as a ratio of GDP) in East Asia. The first column reports the statistics of the whole time period ( ). Average savings and investment rates of the ten East Asian countries are 35.0% and 31.3%, respectively. Singapore shows the highest savings rate (about 54%) among the eleven countries, while China shows the highest investment rate of 39.7%. The Philippines show the lowest savings and investment rates, 16.2% and 19.6%, respectively. Average savings rate is higher than the average investment rate in all countries except for the Philippines, Korea and Thailand. In subsequent columns, we divide the sample period into , , and Most countries experienced an increase in both saving and investment rates from 1980s to 1990s, especially Thailand, Malaysia and Korea. Some countries exhib a decrease in saving rates in these periods, including Hong Kong, Indonesia and the Philippines. Another striking pattern is that the investment rate significantly decreased from 1990s to 2000s in all sample countries, partly due to global financial crisis in We also report average statistics of sub-group countries: emerging Asia (nine countries, excluding Japan), big three (China, Japan and Korea), ASEAN (Indonesia, Thailand, the Philippines, Malaysia, and Singapore), and Greater China (Hong Kong, Taiwan and China). Both savings and investment rates of the big three and Greater China are larger than the average of full ten countries. The average savings rate of emerging Asia tends to increase over time. The average investment rate of emerging Asia increased from the 1980s to the 1990s but decreased from the 1990s to the 2000s. A number of past studies examined cross-sectional saving and investment relation to infer the degree of capal mobily. Table 2 reports the estimated coefficient (β) in the cross-sectional 6

7 regressions of investment on saving for 9 emerging East Asian countries using OLS estimation method. Cross-sectional data are constructed by taking averages of saving and investment rates over different time periods. From the regression of the whole data period, the estimated β is 0.46, while in the 1980s is 0.79, decreases to 0.48 in the 1990s, and further decreases to 0.41 in the 2000s. We also perform a rolling regression wh 15 year window. The results show that the estimated β consistently decreases over time; from 0.68 in the period to around 0.40 in the period and after. All these results confirm that the estimated β decreased significantly over time in East Asia, which may indicate that long run capal mobily has increased over time in this region. Before running panel regression, we use panel un root and panel cointegration tests to find the appropriate data form for regressions. Table 3 reports the results for nine emerging Asian countries. The panel un root test results, based on Levin et al. (2002) and Im et al. (2003), show that we can reject the null hypothesis of un root for the first differenced data of saving and investment rates, but not for the level data. Panel cointegration test based on Pedroni (1999) show that the null hypothesis of no cointegration cannot be rejected. This implies that we should use the first differenced data and considering a cointegrating relation is not necessary Investment and Foreign Savings Table 4 reports the panel regression results for the regression of domestic investment on domestic, regional, and global aggregate savings (in equation (2)). We run four types of regressions over the whole sample period: only domestic saving as a regressor (model 1), domestic and East Asian regional aggregate savings as regressors (model 2), domestic and global savings as regressors (model 3), and finally domestic, regional and global savings as regressors (model 4). For all regressions, we include Japan in Asian regional aggregate (first column) and in global aggregate saving (second column) to see how Japanese savings can affect the relative importance of regional and global savings in determining domestic investment. First, the estimated coefficient on domestic savings rate, β, ranging from , is significant in all panel regressions. If we interpret β as the (inverse of) degree of international capal mobily, the estimated coefficient implies that the degree of international capal mobily of East Asia is far from perfect, which is consistent wh the findings in past studies, i.e., Kim, Kim, and Wang (2007). The coefficient on regional aggregate saving is significant in all regressions wh 1% significance level, 0.66 (model 2) and 0.64 (model 4). This result suggests that a large fraction of 7

8 domestic investment of emerging East Asian countries has been financed by Asian regional saving, which implies a strong role of Asian regional capal market. It is also interesting to note that the estimated coefficient on regional aggregate saving is larger than that on domestic saving in all regression models. However, the coefficient on global saving is not significant in most cases; is significant at 5% level only in model 3 when Japan is included in global aggregate. In addion, the coefficient is relatively small, ranging from 0.03 to This result suggests that the role of global saving in financing domestic investment in East Asia has been relatively small. On the other hand, comparison of the results wh Japan in the regional aggregate and those wh Japan in the global aggregate suggests that Japanese saving has been playing a significant role in financing Asian investment. When Japan is included in Asian regional aggregate, the coefficient on regional aggregate is much larger than the case when Japan is included in global saving (0.66 vs in model 2, 0.64 vs in model 4). In addion, the coefficient on global aggregate is smaller when Japan is included in regional aggregate (0.21 vs in model 3, 0.03 vs in model 4). Table 5 shows the regression results of model 4 using sub-period samples wh fifteen year window. The sub-period panel regressions show that the estimated coefficient on Asian regional aggregate saving was small and insignificant during earlier sub-periods (from to ), but for later sub-periods (from onwards), the estimated coefficient is posive around and significant. When Japan is included in global aggregate saving (right side panel), the coefficient on regional aggregate becomes smaller than the case when Japan is included in regional aggregate. However, the coefficients are still significant wh 1% significance level. When Japan is included in global aggregate, the coefficient on global aggregate saving becomes significant in some sub-sample periods, especially in sub-samples including the Asian Crisis period. In sum, the sub-period estimation results confirm that the Asian regional saving (not global aggregate saving) plays an important role in financing investment in emerging Asia and Japan is likely to be the key country Role of Japanese and Chinese Savings To confirm this conjecture, we run a regression that includes Japanese saving as a separate regressor as follows (model 5). 8

9 ˆ I = α + βsˆ + γsˆ + δsˆ + φsˆ + ε i R G Japan (3) where neher regional nor global aggregate saving includes Japanese saving. Table 6 confirms the conjecture. The coefficient on Japanese saving is posive (0.36) and significant at 5% level, while the coefficient on regional aggregate saving is still posive (0.46) but this value is lower than the case when Japan is included in regional aggregate (0.66). Coefficient on global aggregate saving is minus and insignificant. The sub-period estimation shows that the coefficient on Japanese saving is near zero in early 1980s but become posive and significant once the Asian Crisis period ( ) is included in the sample (subsamples from to ). During these periods, the coefficient on Japanese saving is as large as that on regional aggregate saving. However, the role of Japan becomes insignificant again in more recent periods. As the role of China is increasing in regional financing these days, we examine the role of Chinese saving in financing investment in emerging Asia. In Table 7, we include Chinese saving as a separate regressor and repeat the same exercise. In this case, since China is part of emerging Asia, the sample size is now 8 instead of 9 countries. Japan is included in global aggregate saving, not regional saving. The results show that the coefficient on Chinese saving is near zero, as is the coefficient on global aggregate saving. The role of regional aggregate is high and significant (0.81). China may use s excess domestic saving to invest in many emerging Asian countries, but the regression based on saving and investment does not yet reflect this phenomenon. The sub-period analysis shows that the coefficient on Chinese saving does not change much over time and remain insignificant. 5. Conclusion Following capal account liberalization and financial market deregulation in the early 1990s, East Asian countries experienced a surge in international capal flows. Theoretically, improved international capal mobily can provide a benef by efficiently allocating world-wide saving into the place wh appropriate investment opportunies. This paper investigates the extent to which domestic investment of East Asian countries is financed by domestic, regional, and global savings, in order to infer the role of regional and global capal markets in funding domestic investment. The empirical results are as follows. First, global saving (as proxied by G6 saving, excluding Japan) has not played any role in financing domestic investment of emerging East 9

10 Asian countries. Since the 1990s, there has been large capal flows between East Asian countries and industrial countries. These large capal flows have often been blamed as a source of the currency crisis and instabily of asset prices and financial systems in East Asian countries. Despe of such drawbacks of high international capal mobily, theories suggest that global capal markets can also provide a benef. However, our empirical result shows that huge international capal flows between major industrial countries and East Asian countries have not really been a meaningful source of domestic investment in East Asia. Second, East Asian capal market including Japan has been successful in financing domestic investment. It appears that Japanese saving has played the most important role in financing domestic investment of emerging East Asian countries. China has not been a major investor in the region. This result is interesting as past studies often suggested that East Asian regional capal markets are not much integrated. However, when we extend the coverage of regional capal market to include Japan, is rather well integrated and found effective in funding investment of individual countries in the region. East Asian countries took various steps to improve regional financial cooperation and integration since the 1997 Asian financial crisis. Improved regional financial cooperation together wh various measures to facilate the market liberalization and deregulation process has likely contributed to this increased role of regional financial markets. 10

11 References Bayoumi, T., Saving-Investment Correlations: Immobile Capal, Government Policy, or Endogenous Behavior? IMF Working Paper 66. Eichengreen, B., Park, Y.C., Why has there been less Financial Integration in Asia than in Europe? MAS Staff Papers, Monetary Authory of Singapore, Feldstein, M., Horioka, C., Domestic saving and international capal flows. Economic Journal 90, Im, K, Pesaran, M.H., Shin, Y., Testing for un roots in heterogeneous panels. Journal of Econometrics 115, Kim, S.H., The Saving-Investment Correlation Puzzle is Still a Puzzle. Journal of International Money and Finance 20, Kim, H., Oh, K.Y., and Jeong, C.W., Panel Cointegration Results on International Capal Mobily in Asian Economies. Journal of International Money and Finance 24, Kim, S., Kim, S.H., International Capal Mobily: Global vs. Regional Perspective, Working Paper. Seoul National Universy. Kim, S., Kim, S.H., Park, C.Y., International Capal Mobily of East Asian Countries: Global vs. Regional Perspective. In: Devereux, M., Lane P., Park, C.Y., Wei, S.J., (Eds.). The Dynamics of Asian Financial Integration: Facts and Analytics, Routledge. Kim, S., Kim, S.H., Wang, Y., Global vs. Regional Risksharing in East Asia. Asian Economic Papers 3, Kim, S., Kim, S.H., Wang, Y., Saving, Investment and International Capal Mobily in East Asia. Japan and the World Economy 19,

12 Kim, S., Lee, J.W., Shin, K.H., Regional and Global Financial Integration in East Asia. In: Eichengreen, B., Wyplosz, C., Park, Y.C., (Eds.). China, Asia and the New World Economy. Oxford Universy Press. Kim, S., Lee, J.W., Real and Financial Integration in East Asia. Working Paper Series on Regional Economic Integration No. 17, Asian Development Bank. Levin, A., Lin, C., Chu, C., Un root tests in panel data: asymptotic and fine-sample properties. Journal of Econometrics 108, Murphy, R., Capal Mobily and the Relationship between Saving and Investment in OECD countries. Journal of International Money and Finance 3, Obstfeld, M., Capal Mobily in the World Economy: Theory and Measurement. Carnegie-Rochester Conference Series in Public Policy 24, Obstfeld, M., International Capal Mobily in the 1990s. In: Kenen, P. (Ed.), Understanding Interdependence: The Macroeconomics of the Open Economy. Princeton Universy Press, Princeton. Pedroni, P., Crical Values for Cointegration Tests in Heterogeneous Panels wh Multiple Regressors. Oxford Bulletin of Economics and Statistics 61, Sinha, D., Saving-Investment Relationship for Japan and other Asian Countries. Japan and the World Economy 14,

13 Table 1. Properties of Saving and Investment (Percentage over GDP, Period Average) S/Y I/Y S/Y I/Y S/Y I/Y S/Y I/Y China Hong Kong Indonesia Korea Malaysia Philippines Singapore Taiwan Thailand Japan Average Big ASEAN Greater China Emerging Asia Note: Big 3 (Korea, Japan, China), ASEAN (Indonesia, Thailand, Malaysia, Singapore and the Philippines), Greater China (Hong Kong, Singapore, China), Emerging Asia (9 sample countries excluding Japan). 13

14 Table 2. Cross-sectional Saving-Investment Correlation (I/Y)i=α+β(S/Y)i period β whole period 0.46* ** ** * ** ** * ** ** ** ** * * * * * * * * * Note: Data include 9 emerging East Asian countries (excluding Japan). ** and * denote 1%, and 5% significance levels, respectively. 14

15 Table 3. Panel Un Root and Cointegration Tests Panel un root test (Levin, Lin, Chu test) S 1.17 ΔS -6.82** I 1.83 ΔI -9.78** (Im, Pesaran, Shin test) S 3.22 (1.00) ΔS -8.84** I 1.91 (0.97) ΔI -9.62** Panel cointegration test (Pedroni residual test wh max lag length at 2) (S, I) v-statistics 0.30 rho-statistics 0.69 PP-Statistics ADF-Statistics Note: Cointegration test statistics are for the null hypothesis that none of the variables are cointegrated. Panel un root and cointegration test results use the data of 9 emerging East Asian countries. ** and * denote 1%, and 5% significance levels, respectively. 15

16 Table 4. The Role of Domestic, Regional, and Global Savings: whole-period analysis ˆ I = α + βsˆ + γsˆ + δsˆ + ε i R G Model Japan in regional aggregate Japan in global aggregate Model 1 own saving (β) 0.51** 0.51** (0.07) (0.07) Model 2 own saving (β) 0.44** 0.45** (0.07) (0.07) regional aggregate (γ) 0.66** 0.55** (0.18) (0.16) Model 3 own saving (β) 0.48** 0.47** (0.07) (0.07) global aggregate (δ) * (0.12) (0.13) Model 4 own saving (β) 0.44** 0.43** (0.07) (0.07) regional aggregate (γ) 0.64** 0.48** (0.20) (0.17) global aggregate (δ) (0.13) (0.14) Note: Hat variables denote percentage changes. Numbers in parenthesis are standard errors. ** and * denote 1%, and 5% significance levels, respectively. 16

17 Table 5. The Role of Domestic, Regional, and Global Saving: sub-period analysis ˆ I = α + βsˆ + γsˆ + δsˆ + ε i R G Japan in regional aggregate Japan in global aggregate Period β γ δ β γ δ ** ** ** ** ** ** ** ** ** 0.93** ** 0.46* ** 1.20** ** 0.69** 0.84* ** 1.49** ** 0.85** 0.94* ** 1.50** ** 0.86** ** 1.38** ** 0.81** 0.62* ** 1.22** ** 0.78** ** 1.24** ** 0.99** ** 0.89** ** 0.83** ** 0.80** ** 0.81** ** 0.79** ** 0.78** ** 0.88** ** 0.80** ** 0.92** ** 0.80** 0.04 Note: Regression results are based on regression model 4 in Table 4. ** and * denote 1%, and 5% significance levels, respectively. 17

18 Table 6. Role of Japanese Saving ˆ ˆ ˆR ˆG ˆJapan I = α + βs + γs + δs + φs + ε i Whole-period Model 5 own saving (β) 0.42** (0.07) regional aggregate (γ) 0.46** (0.16) global aggregate (δ) (0.15) Japanese saving (φ) 0.36* (0.15) Sub-period Period β γ δ φ ** ** ** ** ** 0.51* ** 0.69** ** ** 0.77** ** ** 0.79** ** ** 0.75** ** ** 0.66** ** ** 0.82** * ** 0.72** ** 0.74** ** 0.70** ** 0.72** ** Note: Regional aggregate includes 9 East Asian emerging countries (whout Japan) and global aggregate includes G6 countries. The numbers in parenthesis are standard errors. ** and * denote 1%, and 5% significance levels, respectively. 18

19 Table 7. Role of Chinese Saving ˆ ˆ ˆR ˆG ˆChina I = α + βs + γs + δs + φs + ε i Whole-period Model 5 own saving (β) 0.39** (0.07) regional aggregate (γ) 0.81** (0.16) global aggregate (δ) 0.01 (0.16) Chinese saving (φ) (0.16) Sub-period Period β γ δ φ ** ** ** ** ** 0.56** ** 0.66** 0.97* ** 0.88** 1.05* ** 0.97** ** 0.94** ** 0.90** ** 0.92** ** 0.94** ** 1.01** ** 0.99** ** 0.90** ** 0.90** Note: Regional aggregate includes 8 Asian emerging countries (whout Japan or China) and global aggregate includes G7 countries (wh Japan). The numbers in parenthesis are standard errors. ** and * denote 1%, and 5% significance levels, respectively. 19

20 Figure 1. Sources of Capal Inflows in East Asia: CPIS Data (percentage share of total portfolio investment inflows) Hong Kong Korea Indonesia Malaysia Philippines Thailand 20 Singapore Share of G6 Share of East Asia (excluding Japan) Share of Japan