National Sun Yat-sen University Institutional Repository:Item 987654321/39389

行政院國家科學委員會補助專題研究計畫成果報告

※※※※※※※※※※※※※※※※※※※※※※※※※

※

The Causes of International Capital Flows: ※

※ The Review of Developing Countries

※

※ 國際資本流通之成因：開發中國家之探討

※

※※※※※※※※※※※※※※※※※※※※※※※※※

計畫類別：▓個別型計畫 □整合型計畫

計畫編號：NSC 90 – 2415 – H – 110 – 006

執行期間：90 年 8 月 1 日至 91 年 07 月 31 日

計畫主持人：印永翔

本成果報告包括以下應繳交之附件：

□赴國外出差或研習心得報告一份

□赴大陸地區出差或研習心得報告一份

□出席國際學術會議心得報告及發表之論文各一份

□國際合作研究計畫國外研究報告書一份

執行單位：國立中山大學中山學術研究所

中 華 民 國 91 年 10 月 31 日

行政院國家科學委員會專題研究計畫成果報告

計畫編號： NSC 91 – 2415 – H – 110 – 006

執行期限： 90 年 8 月 1 日至 91 年 7 月 31 日

主持人：

印永翔，國立中山大學中山學術研究所

計劃參予人：岳俊豪，國立中山大學經研所。李佳倩，銘傳大學經研所。陳瑞芬，文化

大學國企所

１

．Abstract 中文摘要 自從布林頓森林體系在1973年瓦解後，歐美主 要工業國家陸續解除資本管制措施，投資人為了尋 找最佳投資地區與而分散風險，國際資本跨越國界 開始形成。資本在1990年初大量流入亞洲地區， Calvo et al. (1996)指出，工業國家90年代的低利率 與開發中國家國內穩定經濟成長，為主要原因。 本計劃針對為何國際資本流入亞洲國家：印 尼、韓國、馬來西亞、菲律賓、台灣、新加坡、泰 國為本計畫研究對象，特別是影響資本跨國移動之 外部及內部因素之相對重要性

。

。 關鍵詞：資本移動、結構性向量自我回歸。 Abstract

This project is designing to understand the cause of capital inflows in Asia-7 countries, Indonesia, Korea, Malaysia, Philippines, Singapore, Taiwan, and Thailand. The causes of capital flows can be attributed to external and internal factors. In general, economists believe that internal factors are categorized as sounding policies and stronger economic performances. External factors are fluctuation of world output shock and the fluctuation of world interest rates. We found that the influence of external and internal factors play relatively less important in explaining capital movements in Asia. SVAR shows that capital its own shocks have become more important since Asian currency crisis.

Keywords: Capital Flows, SVAR

２．Background

Asian countries received substantial amounts of foreign capitals in the late 1980s and early 1990s, until the year before the Asian currency crisis. For example, there are more than $73 billion capital inflows into Korea, Indonesia, Malaysia, Philippines, and Thailand in 1996. In this project, I am specifically to investigate why substantial amounts of capital are attracted to this area. There are seven countries, Indonesia, Korea, Malaysia, Philippines, Singapore, Taiwan, and Thailand included in this project. The inclusion of these seven developing countries has following particular reasons. Firstly, there are $190 billions of capital inflows to developing

countries in the year, 1999. However, 75% of these massive capitals flew into 12 developing countries (Alejandro, 1999). Thailand was the one receiving the most share of international capital among other countries in Asia. Secondly, Asian currency crisis substantially cooled down the heat of capital inflows into Asian economies. Indonesia, Korea, Malaysia, Philippines, and Thailand were the five countries experienced severe economic downturn1. At the same time, Taiwan and Singapore took relatively little impact of Asian crisis. Thus, it is of interest to investigate the fundamental difference how capitals were motivated into these countries. To the extent that we decompose the magnitudes of those underlying factors to motivate capital inflows into these 7 Asian economies.

A large number of literatures has provided the explanations that capital inflows into developing countries resulted from internal and external factors. The debt crisis of developing countries in the 1980s had alleviated because governments implemented a series of sounding monetary policies. In terms of internal reasons, Chuhan et al. (1998) indicate that country-specific factors are particularly important to explain the recipient of capital inflows for Asian economies. Those factors are rates of return on stock markets, credit rating and secondary market prices of sovereign debt. Calvo et al. (1993 & 1996) point out that the historical low interest rate in the United States and stagnant growth rate of industrial production of industrialized countries are external reasons to cause capital inflows to developing countries. For the Asian countries, Chuhan et al. (1998) conclude that country-specific factors are nearly three to four times more important than external factors (or global factors) in motivating these flows.

We then separate those underlying factors into two primary categories; external (push) factors and internal (pull) factors in our analysis. Push factors are global output and global real interest rate. Pull factors are domestic output and domestic real interest rate. Noticeably, the 7 countries of this project are considered as small open economies. Business cycles fluctuations are easily transmitted through trade channels. Moreover, empirical studies show that capital mobility of developing countries has increased over time (Hussein and De-Mello, 1999). The ____________________________________

1

In the end of year 1997, the crashes of stock markets for these five economies are as follows, -44.6% for Indonesia, 49.5% for Korea, -44.8% for Malaysia, -33.5% for

integration of capital markets of developing countries with the world financial markets offers the opportunity of international capitals perusing higher rate of return. The impressive economic growth in Asian regions in the early 1990s provided alternatives for investors in industrial countries, when sluggish growth of western economies. More specifically, the influence of capital inflows is decomposed into goods and assets channels.

Since we are interested in the underlying factors to motivate capital flows rather than forecasting errors of the model. Thus, we applied Structural Vector Autoregressive (SVAR) rather than unrestricted Vector Autoregressive framework (VAR). To observe those unobservable shocks relies on placing long-run restrictions2.

The flows of the project are as follows. Section I and II gives brief introduction the purposes of this study. We offer the discussions of theoretical models and empirical works in Section III. Last section concludes remarks.

３．Context

In this section we develop an empirically tractable, small, intertemporal open-economy model. The model is similar to those proposed in Clarida and Gali (1994), Ahmed . et al (1993), Lee and Chinn (1998), and Ying and Kim (2001). We consider an open economy model, which incorporates both global and country-specific productivity and financial shocks. Therefore, the underlying theoretical framework of the empirical model is similar to the two-country

Mundell-Fleming (MF) model, along with a variety of others. As with the MF model, we allow foreign shocks from Asian economies’ major trading partners to influence capital inflows through one of two

channels, trade flows or financial flows (capital flows). Consider the similar model proposed in Chang et al. (2002): (1)

y

_t*



y

*_t1





_ty* (2)

i

_t*



i

_t*





y

_t* (3)

i

_t*



i

_t*_1





_t* (4)

y

_t



y

_t



ri

_t* (5)

y

_t



y

_t1





_ty ____________________________________ 2

The long-run restrictions tend to be less controversial and more readily accepted than other assumptions, they are not without criticisms. Faust and Leeper (1997) indicate that structural inferences under the long-run scheme may not be reliable as the long-run effects of shocks are imprecisely estimated in finite samples and the long-run identification scheme transfers this imprecision of the estimates to other parameters of the model. Another criticism, which also applies when the underlying model has more sources of shocks than does the estimated model - is that the estimated disturbances will commingle the underlying disturbances. I would like thank the anonymous referees in concern the criticisms about SVAR in terms of restrictions when they reviewed the proposal of this project in year 2000.

(6)

i

_t



y

_t





i

_t*



i

t

(7)

i

t



i

t1





_ti

(8)

k

_t



y

_t





i



k

t

(9)

k

t



k

t1





k

where (*) indicates that variable is the foreign level and a

(

)

suggests that the variable level is

determined exogenously. All variables in the system are in logs, where y, i, k, represent output, the interest rate, and capital movement, respectively. The importance of Asian economies as trading partners to 7 countries in this study has increased in recent periods. We applied weighted average industrial output index as world output (y*) and weighted average interest rate (i*) as foreign variables rather than US variables served as y*and i*. For domestic variables, industrial productions and market interest rates represent domestic output (y) and domestic interest rates (i), respectively. Capital inflows are extracted from the balance of payments using definition of difference between change in foreign reserves and trade balance for these seven countries. The inclusions of variables above are to incorporate push factors, foreign variables and pull factors, domestic variables to motivate capital inflows. All data are available in International Financial Statistics published by International Monetary Funds.

There are five structural shocks in the system:(i) a foreign productivity shock

(



_ty*

)

, (ii) a foreign financial shock

(



ti*

)

, (iii) a domestic productivity

shock

(



_ty

)

, (iv) a domestic financial shock

(



_ti

)

, and (v) a capital own shocks

(



tk

)

. The reduced from

of (1) – (9) is illustrated, in first differences, as follows: (1.1)



y

*_t





_ty* (1.2)



i

_t*





_ti*





_ty* (1.3)



y

_t



r

(



_ty*





_ti*

)





_ty (1.4)



i

_t



(

r





)(



_ty*





_ti*

)





_ty





_ti (1.5)

(







)



_ty*



(







)



_ti*





_ty





_ti





_tk

The system (1.1) – (1.5) suggests that Asian-7 countries capital inflows can be modeled as: (10)

K

_t



f

(



_ty*

,



_ti*

,



_ty

,



i

,



k

)

A. Identifying Restrictions

The structural shocks in Equation (10) are

unobservable. Because of the unobservable nature of the structural shocks, additional identifying

assumptions are necessary to recover the underlying structural shocks from data. To extract the 4 structural shocks, we consider a 4-variable reduced form model, such as:

(11)



  





0 i t i t i t

A

A(L)U

Y

U

where Yt = (



yt*,



it*,



yt,



it,



kt)’, Ut =

(



_ty*

,



_ti*

,



_ty

,



_ti

,



_tk

)'

,



 



0

,

)

(

i i i

L

A

L

A

and Ai is the matrix of impulse responses of endogenous variables to structural shocks.

The model is estimated separately for each country with the quarterly data from 1978:1 to 2001:2. Equation (11) assumes that all endogenous variables are integrated of order 1, I(1). In addition, the five structural shocks,

(

*

,

*

,

,

,

tk

)'

i t y t i t y t











have unit

variance and are mutually orthogonal.

To efficiently estimate the dynamics of the system in (11), we need to make some identifying restrictions on the long-run impacts of the structural shocks. These effects can be presented in matrix form as: (12)

(

1

)

{

(

1

)

}

0



 





i

A

i

A

a

ij

The identifying restrictions we make are as follows: 1. Foreign variables are affected in the long run by

foreign shocks only. This assumption is commonly used for a small open-economy analysis for both short run and long run effects. Thus this assumption implies that a13(1) = a14(1)

= a23(1) = a24(1) = 0.

2. A foreign financial shock has no long-run effect on foreign output. This assumption is

incorporated as a12 (1) = 0.

3. The similar assumption as above applied in domestic interest rate has no long-run influence on real variable, output, thus, a34 (1) = 0.

4. we further assume that shocks to capital inflows are transitory and have no long-run effects on variables in the system. Thus, this assumption leads, a15 (1), a25 (1), a35 (1), and a45(1) = 0.

The long-run restrictions can be summarized in a matrix form. Due to these identifying restrictions, the A(1) matrix is a lower triangular matrix. The restrictions can be shown as following matrix form. (13)













































































































k t i t y t i t y t t t t t t

k

i

y

i

y











* * * *

*

*

*

*

*

0

*

*

*

*

0

0

*

*

*

0

0

0

*

*

0

0

0

0

*

B. . Estimation Strategy

The empirical method to recover the structural shocks from the observed variables is based on structural VAR analysis pioneered by Shapiro and Watson (1988) and Blanchard and Quah (1989). Ahmed, et al (1993), Clarida and Gali (1994), Rogers (1998), Ying and Kim (2001) have applied the framework to the analysis of the open economy.

In the system of 5 variables, the identification of the structural shocks, from reduced-form shocks, Vt,

requires 10 restrictions (=5*4/2). The following

paragraphs describe how the restrictions can be used to uniquely recover the structural model. The discussion follows Rogers (1998). See also Ahmed and Murthy (1994) for an alternative but equivalent discussion.

Since the techniques to recover the unobservable structural shocks have become standard. The detail discussion can be referred Anders (1995) and Ying and Kim (2001).

C. Empirical Results

Before running the SVAR for Asian-7 countries, the following graphs show the relationship between the current and capital account.

C a p i t a l a n d C u r r e n t A c c o u n t o f I n d o n e s i a 1 9 8 1 1 9 8 3 1 9 8 5 1 9 8 7 1 9 8 9 1 9 9 1 1 9 9 3 1 9 9 5 1 9 9 7 1 9 9 9 2 0 0 1 - 7 .5 - 5 .0 - 2 .5 0 . 0 2 . 5 5 . 0 7 . 5 I N D C A A I N D K A A C a p it a l a n d C u r r e n t A c c o u n t o f K o r e a 1 9 6 0 1 9 6 4 1 9 6 8 1 9 7 2 1 9 7 6 1 9 8 0 1 9 8 4 1 9 8 8 1 9 9 2 1 9 9 6 2 0 0 0 - 2 0 - 1 5 - 1 0 - 5 0 5 1 0 1 5 2 0 K O R C A A K O R K A A C a p i t a l a n d C u r r e n t A c c o u n t o f M a l a y s i a 1 9 6 0 1 9 6 4 1 9 6 8 1 9 7 2 1 9 7 6 1 9 8 0 1 9 8 4 1 9 8 8 1 9 9 2 1 9 9 6 2 0 0 0 - 2 8 - 2 1 - 1 4 -7 0 7 1 4 2 1 M A C A P M A K A P C a p i t a l a n d C u r r e n t A c c o u n t o f P h il i p p in e s 1 9 8 5 1 9 8 7 1 9 8 9 1 9 9 1 1 9 9 3 1 9 9 5 1 9 9 7 1 9 9 9 2 0 0 1 -4 -3 -2 -1 0 1 2 3 P H I C A A P H I K A A

C a p ital a n d C u rren t A cco u n t o f S in ga p ore 1 97 8 1 9 80 1 98 2 1 9 84 1 9 8 6 1 9 88 1 9 90 1 9 9 2 19 9 4 1 9 9 6 19 9 8 2 00 0 -8 -6 -4 -2 0 2 4 6 8 S IN CA S IN KA C a p i t a l a n d C u r r e n t A c c o u n t o f T a i w a n 1 9 8 1 1 9 8 3 1 9 8 5 1 9 8 7 1 9 8 9 1 9 9 1 1 9 9 3 1 9 9 5 1 9 9 7 1 9 9 9 2 0 0 1 - 7 .5 - 5 .0 - 2 .5 0 .0 2 .5 5 .0 7 .5 TA I C A TA I K A C a p i t a l a n d C u r re n t A c c o u n t o f T h a ila n d 1 9 7 6 1 9 7 9 1 9 8 2 1 9 8 5 1 9 8 8 1 9 9 1 1 9 9 4 1 9 9 7 2 0 0 0 - 6 - 4 - 2 0 2 4 6 8 1 0 T H C A A T H K A A

Figures 1. Current and capital account for Asian-7.

The current account and capital account in the graphs are denoted as thin and thick-dash lines, respectively. Taiwan is the only exceptional case, which current account is surplus over 2 decades in contrast to the current account deficits for other Asian developing countries. Mirror images of current account and capital account in the graphs above denotes consistent current account deficits require the foreign capital financing in the spending for domestic investment and consumption. It is clear to see that reversal of capital flows in all Asian-7 during the Asian currency crisis from 1997-1998 from the graphs.

We then investigate the stationarity of variables in the system. According to popular Augmented Dicky Fuller (ADF) and Phillips and Perron (PP) test, the higher power in finding unit roots tests as KPSS and Leybourne and McCabe (LMC) are applied. According to those unit roots tests, we define those variables needed first difference to be stationary3.

C.1. Variance Decompositions

The variance decompositions provide the relative importance for structural shocks to capital inflows in Asian-7.

Table 1. The percentage of structural shocks to capital inflows

____________________________________

3

The detail of unit roots tests are available if request.

Country y* t



*

,



ti y t



i t



k t



Indonesia 24.3 5.1 7.1 14.0 50.0 Korea 8.1 8.8 4.1 5.8 73.1 Malaysia 10.3 7.3 17.0 3.0 62.4 Philippines 11.3 3.2 8.5 25.8 51.1 Singapore 6.0 3.6 4.1 17.6 68.6 Taiwan 7.9 11.0 9.2 2.9 69.0 Thailand 19.7 21.5 19.3 14.5 25.0 The foreign productivity and interest shocks to Indonesia and Thailand are particularly important in the comparison with other Asian economies. Capital inflows its own shock only account 25% share of importance. On the other hand, foreign shocks seem to have less effect in motivating capital inflows to Korea, Malaysia, Philippines, Singapore and Taiwan. Interestingly, capital own shocks account at least 50% of motivation in capital flows. The results are somewhat different from previous studies such as Calvo et al. (1993) and Ying and Kim (2001). The plausible explanations are different sample periods. The previous works do not include data since 1997, the year of Asian currency crisis. Thus, the motivation of capital movements is relatively difficult to be understood by aggregate variables. We suggest further research in using disaggregate data, such as bonds or equity data to decompose the capital inflows would be helpful application related to this issue.

4. Conclusion

It is interesting to understand that capital inflows into Asian-7 countries have substantially changed. The structural difference in motivating capital inflows into Asian economies makes empirical results somewhat different from previous studies. In fact, the inclusion of data since 1997 results in foreign shocks and domestic shocks play relatively less roles in explaining the motivation of capital flows. We tried to eliminate data after 1997, and results were similar to literatures. This project shed some light in further research in following perspectives. First, the sluggish economic growth and low interest rate are the major explanations for the capital inflows to Asian economies, but not the whole story for capital flows to Asia after 1997. Second, the disaggregate data of capitals might provide solution to answer the motivation of capital movements.

5．References

Ahmed, Shaghil, Barry, W., Ickes, Ping Wang, Byung Sam Yoo, 1993. “International business cycles”,

American Economic Review 83(3), 335-359.

Alejandro, Lopez-Mejja, 1999. “A Survey of Causes, Consequences, and Policy Responses”, International Monetary Fund, IMF Working Paper 17: 44-80.

Blanchard, J., Olivier, & Danny Quah. 1989. “The dynamic effects of aggregate supply and demand disturbances”, American Economic Review 79(4), 655-673.

Calvo, Guillermo A., Leonardo Leiderman, and Carmen M. Reinhart, 1993. “Capital inflows and real exchange rate appreciation in Latin America: the role of external factors”, IMF Staff Papers 40(1): 108-151.

Calvo, Guillermo A., Leonardo Leiderman, and Carmen M. Reinhart, 1993. “Inflows of Capital to Developing Countries in the 1990s”, Journal of

Economic Perspectives (10)2:

Chang, Koyin, Larry Filer and Yung-Hsiang Ying, 2002. “A Structural Decomposition of the Taiwanese Business Cycles”, 13(1): 53-64.

Chuhan, Punam, Stijin Claessens and Nlandu Mamingi, 1998. “Equity and Bond Flows to Latin America and Asia: the Role of Global and Country factors”, Journal of Development Economics 55: 439-463.

Clarida, R., & Jodi Gali, 1994. “Sources of real exchange rate fluctuations: how important are nominal shocks”, Carnegie-Rochester Conference

Series on Public Policy 41, 1-56.

Faust, Jon, and Eric Leeper, 1997. “When do Long-run Identifying Restrictions Give a Reliable Results?” Journal of Business and Statistics 48(2): 389-418.

Goldstein, Morris. 1998. “The Asian Financial Crisis: Causes, Cures, and Systemic Implications”, Institute

for International Economics, Washington. DC.

Hussein, K. and L. De-Mello, 1999. “The International Capital Mobility in Developing Countries; Theory and Evidence”, Journal of

International Money and Finance 18(3): 367-81.

Lee, Jaewoo & Menzie, D., Chinn, 1998. “The current account and exchange rate: a structural VAR analysis of major currencies”, NBER working paper 6495. Rogers, H., John, 1998. “Monetary Shocks and Real

Exchange Rates”, Board of Governors on the Federal Reserve System. International Finance Discussion

Papers No. 612 .

Shapiro, D., Matthew, & Mark Watson. 1988. “Sources of Business Cycle Fluctuations”, in Stanley Fischer, ed., NBER Macroeconomic Annual, MIT

Press.

Ying, Yung-Hsiang, & Yoonbai Kim, 2001. The Empirical Analysis of Capital Flows: the Case of Korea and Mexico. Southern Economic Journal 67(4),954-68.