ADB Economics

Working Paper Series

How Effective are Capital Controls in Asia?

Maria Socorro Gochoco-Bautista, Juthathip Jongwanich,
and Jong-Wha Lee
No. 224 | October 2010
ADB Economics Working Paper Series No. 224

How Effective are Capital Controls in Asia?

Maria Socorro Gochoco-Bautista, Juthathip Jongwanich,

and Jong-Wha Lee
October 2010
(Revised February 2011)

Maria Socorro Gochoco-Bautista is Senior Economic Advisor, Economics and Research Department, Asian
Development Bank; Juthathip Jongwanich is Assistant Professor, School of Management, Asian Institute
of Technology; and Jong-Wha Lee is Professor, Economics Department, Korea University. The authors
are grateful to Ruth Francisco, Anthony Baluga, and Lea Sumulong for excellent research assistance. The
views expressed in this paper are those of the authors and do not necessarily reflect the views or policies
of the Asian Development Bank or its Board of Governors or the governments they represent. The authors
accept responsibility for any errors in the paper.
Asian Development Bank
6 ADB Avenue, Mandaluyong City
1550 Metro Manila, Philippines
www.adb.org/economics

©2010 by Asian Development Bank

January 2011
ISSN 1655-5252
Publication Stock No. WPS113208

The views expressed in this paper

are those of the author(s) and do not
necessarily reflect the views or policies
of the Asian Development Bank.

The ADB Economics Working Paper Series is a forum for stimulating discussion and
eliciting feedback on ongoing and recently completed research and policy studies
undertaken by the Asian Development Bank (ADB) staff, consultants, or resource
persons. The series deals with key economic and development problems, particularly
those facing the Asia and Pacific region; as well as conceptual, analytical, or
methodological issues relating to project/program economic analysis, and statistical data
and measurement. The series aims to enhance the knowledge on Asia’s development
and policy challenges; strengthen analytical rigor and quality of ADB’s country partnership
strategies, and its subregional and country operations; and improve the quality and
availability of statistical data and development indicators for monitoring development
effectiveness.

The ADB Economics Working Paper Series is a quick-disseminating, informal publication

whose titles could subsequently be revised for publication as articles in professional
journals or chapters in books. The series is maintained by the Economics and Research
Department.
Contents

Abstract v

I. Introduction 1

II. Effectiveness of Capital Controls: A Literature Survey 4

III. Capital Controls in Emerging Asia 8

IV. Effects of Capital Controls: Evidence from Cross-Country Panel Data 12

A. Data and Methodology 12

B. Estimation Results 13
C. Asymmetric Effects of Capital Controls 17
D. Substitution Effects among Capital Controls 19
E. Endogeneity of Capital Controls 21

V. Conclusion 22

Appendix: Data Sources and Definitions 25

References 27
Abstract

This study examines the effects of capital account restrictions on capital flows in
nine Asian economies over the period 1995–2005 using panel regressions with
fixed effects. The results show that capital controls significantly affect capital
flows when such flows are disaggregated by asset type and direction of flow.
Tests for the presence of possible asymmetric effects, substitution effects, and
endogeneity of capital controls are conducted.
I. Introduction
In the last 3 decades, debates about the use of capital controls and their role in the
economy have periodically ebbed and waned. The recent revival of these debates is not
surprising in light of this history and in the context of continuing challenges in the global
economic environment spawned by the global financial crisis of 2008. The dilemma faced
by policy makers in striking the proper balance between keeping an economy open to
greater cross-border trade in financial assets while simultaneously attempting to insulate it
from the effects of swings in the international appetite for risk has not changed over time.

In the 1980s, the easing of restrictions on capital was pursued in the context of
economic liberalization pushed by multilateral institutions such as the World Bank. The
rationale underlying this was that welfare gains could be realized if capital were mobile
internationally. Capital-scarce countries could borrow from abroad at lower rates to invest
and consume more in the current period; and from greater income, pay back interest on
international borrowing. Capital-rich countries could realize the highest returns by lending
their savings internationally. Country-specific consumption risks thus could be efficiently
allocated and the marginal utility of consumption would equalize across countries
(Obstfeld 1994).

In the early 1990s, many Asian economies began to liberalize their capital accounts. The
focus then was on the proper sequencing of implementing economic reforms. Capital
restrictions were to be relaxed gradually and only after other markets had first been
liberalized and certain prerequisites fulfilled. It was recognized that unless an economy
had first undergone the necessary structural reforms to make it resilient, free capital
mobility could make such an economy vulnerable to external shocks. A gradual process
of liberalization and relaxation of capital controls was also seen as allowing policy makers
the ability to more effectively use the tools of macro policy to deal with shocks.

On the other hand, critics of capital account liberalization argued that that the gains from
unfettered trade in financial assets were not obvious (Bhagwati 1998). Liberalizing trade
in financial assets is different from liberalizing trade in goods. The contingent nature
of a financial contract meant that there is no guarantee that countries that borrowed
from abroad could repay their loans if, for example, the borrowed funds went primarily
into consumption rather than into investment, and no increase in future output was
forthcoming. Thus, critics of capital account liberalization advocated a limited degree
of financial openness to shield countries from external shocks, financial instability, and
extreme fluctuations in exchange rates (Krugman 1999, Stiglitz 2000 and 2002, Rodrik
2 | ADB Economics Working Paper Series No. 224

2006). They saw the judicious use of capital controls, either as a stand-alone tool or in
combination with other macro policy tools, as helping to insulate the domestic economy
from external shocks. These controls, for example, could drive a wedge between
domestic and global interest rates to reduce exchange rate pressures. Indeed, even
the International Monetary Fund (IMF) Articles of Agreement under Article VIII granted
member countries the right to maintain controls over capital transactions, but not over
current transactions.

While pursuing an overall strategy in favor of liberalization, policy makers in Asia

remained reluctant to completely do away with capital controls. They were afraid
that without adequate controls, capital inflows would cause the domestic currency to
appreciate in real terms and make their countries’ exports uncompetitive. With some
degree of capital controls, however, policy makers had an extra degree of freedom that
they could use to slow or prevent these large inflows, allow the use of expansionary
monetary and fiscal policy to boost employment and stock market values without
wreaking worsening external balance, and reduce the prospects of destabilizing capital
outflows especially during periods of political instability (Perkins and Woo 2000).1 Policy
makers also worried that capital inflows could suddenly stop and cause financial turmoil,
given shallow and underdeveloped domestic capital markets, while unabated and
large inflows could wreak havoc on financial stability by creating asset bubbles in the
nontradeables sector. Some felt that their economies were simply not ready for capital
account liberalization and could not yet safely intermediate capital flows.

These fears were validated when the Asian financial crisis hit in 1997. The apparent
success of Malaysia in using capital controls during the crisis resurrected interest in their
use along with debates about the wisdom of an open capital account. The IMF, which
had been on the verge of amending Article VIII to include capital account convertibility,
scrapped this plan all together in light of the crisis (Kawai and Takagi 2008). Since the
crisis, many countries in Asia have adopted more flexible exchange rate regimes, some
in the context of an inflation-targeting framework. By and large, economies successfully
recovered, and growth in Asia remained on track despite the bursting of the information
technology bubble in 2001. Capital controls were forgotten for a while.

However, the erstwhile relative calm in the global economic environment was disturbed
as external shocks once again impinged on macro stability and growth. A number
of emerging economies experienced large capital inflows that led to sharp currency
appreciation especially from mid-2006 until mid-2008, reawakening interest in capital
controls. Beginning in 2002, some countries in Asia had begun to re-employ capital
control measures as foreign bank flows into Asia turned to net inflows from outflows,
while both portfolio equity flows and carry trades accelerated and became more volatile
and sensitive to developments in global equity markets. Global commodity and fuel prices
1 In Malaysia, the fear was that the political struggle between Mahathir and Anwar would result in capital flight and
a collapse of the Malaysian ringgit and the stock market as had happened in Indonesia in May 1998 right before
Soeharto stepped down. Malaysia imposed capital controls the day before Anwar was fired by Mahathir.
 How Effective are Capital Controls in Asia? | 3

rose dramatically as well beginning in late 2006 until about the second quarter of 2008.
Policy makers became concerned with both adverse supply-side conditions and renewed
inflationary pressures.

This precarious state of affairs was in for an even larger shock with the collapse of
Lehman Brothers in September 2008. The ensuing global financial crisis provoked a
deleveraging process that saw large capital outflows from Asia as the United States (US)
economy went through a recession—one whose effects have been so severe that the
Great Depression became its comparator. The collapse of Asia’s export trade particularly
with the US dramatically reduced growth in the region in 2008 and 2009. Recovery and
the revival of growth prospects required large amounts of fiscal and monetary stimulus by
countries in the region to fill the loss of demand from the US.

Although Asia has recovered quickly while both Europe and the US still currently face
uncertain recovery prospects, there are fears that capital inflows will inundate Asia and
threaten macroeconomic and financial stability once again. The IMF itself now appears
to be taking a more nuanced position regarding capital controls, explaining that there
are conditions under which capital controls may legitimately be used by policy makers to
manage capital inflows in addition to both prudential and macroeconomic policy (Ostry
et al. 2010). More recently, policy makers in Asia have responded to capital inflows by
liberalizing capital outflows ostensibly to prevent sharp currency appreciation. However,
it is unclear how effective these policies will be as investors face little incentive to leave
Asia today in view of its strong recovery from the global financial crisis and continuing
pressures for domestic currencies to appreciate. It is also more difficult to measure capital
inflows today because investments into domestic bond markets in Asia are largely virtual
investments made using derivative instruments (McCauley 2008).

The recent re-imposition of capital restrictions in the Republic of Korea and Thailand
is evidence of attempts by Asian authorities to deal with large capital inflows. It is
reminiscent of the period beginning in 2002, when some economies in Asia employed
capital control measures as net foreign bank flows to Asia turned positive and portfolio
equity flows particularly sensitive to developments in global markets accelerated.
Authorities in several economies in Asia have also recently further liberalized capital
outflows in response to large inflows.

This study examines the effects of capital control measures on the volume and
composition of capital flows. It uses a methodology in Binici, Hutchison, and Schindler
(2010) and a panel of nine emerging Asian economies comprising the People’s Republic
of China (PRC); Hong Kong, China; India; Indonesia; the Republic of Korea, Malaysia;
the Philippines; Singapore; and Thailand, covering the period 1995–2005. In contrast
with previous studies, this study assesses the extent of de jure capital controls using
information on capital account restrictions by economy at a more disaggregated level,
4 | ADB Economics Working Paper Series No. 224

and adds to the few available multi-country studies, especially those focused on emerging
countries in Asia.

The remainder of this paper is organized as follows. Section II presents a brief survey
of the literature on the effectiveness of capital controls in various countries. Section III
shows the patterns of cross-border capital flows and the measures of legal restrictions
that have influenced them in emerging Asian economies. Section IV empirically
assesses the effects of capital control measures on total volume and asset composition
of international capital flows based on the cross-country panel data set. Section V
concludes.

II. Effectiveness of Capital Controls: A Literature Survey

The fears associated with capital account opening and an increasing degree of financial
integration between emerging economies and the rest of the world are not unfounded.
These have been collectively referred to as the “four fears”: fear of appreciation; fear
of “hot money”; fear of large inflows; and fear of loss of monetary autonomy (Magud
and Reinhart 2007, 647). Capital controls have thus been used to (i) reduce the volume
of capital inflows and alter the composition of flows in favor of longer maturities; (ii)
reduce the degree of nominal and real exchange rate volatility; (iii) reduce real currency
appreciation arising from capital inflows; (iv) give monetary authorities the ability to
implement more independent monetary policy; and (v) prevent the occurrence of financial
crisis and/or financial instability. Governments are understandably reluctant to completely
give up any degree of monetary policy autonomy and the ability to exercise control over
the volume and composition of capital flows. The task has become more daunting given
surges in capital flows, new financial instruments and greater sophistication of investors,
and limits to sterilization and other policy tools with which to respond to capital surges.

A wide variety of restrictions on cross-border capital flows have been used under different
conditions. These can be broadly classified as either (i) administrative or direct controls
or (ii) market-based or indirect controls (Ariyoshi et al. 2000, ADB 2010). Direct controls
involve the use of outright prohibitions on the transfer of funds and associated payments,
or explicit quantitative limits or approval procedures. Administrative controls are typically
intended to affect the volume of cross-border transactions by imposing administrative
obligations on the banking system. Indirect or market-based controls work on the price
or volume of a financial transaction or both of these and discourage these transactions
by making them more costly to undertake. An example of this is the explicit or implicit
taxation of cross-border flows (e.g., so-called Tobin taxes). Explicit taxes may be imposed
on financial transactions by type or maturity of asset. A form of implicit taxation is a
compulsory reserve or deposit requirement at the central bank by banks and nonbanks
engaged in cross-border financial transactions called unremunerated reserve requirement
 How Effective are Capital Controls in Asia? | 5

(URR). Likewise, dual or multiple exchange rate systems attempt to split the market
for domestic currency and raise the cost of domestic credit to speculators and prevent
them from establishing a net short position in the domestic currency. Other price and/or
quantity-based regulatory measures, such as those that discriminate between long and
short currency positions or between residents and nonresidents, are also examples of
indirect or market-based controls.

In the heady days of capital flows to emerging markets of the 1990s, Brazil (1993–1997),
Chile (1991–1998), Colombia (1995–1997), Malaysia (1994), and Thailand (1995–1997)
were among the countries that set limits on short-term capital inflows (Ariyoshi et al.
2000). Some form of market-based controls, largely through direct or indirect taxation of
inflows and other regulatory measures was used. Additionally, Brazil, Chile, and Malaysia
used administrative and direct controls (prohibition of nonresident purchases of money
market securities and nontrade-related swap transactions with nonresidents), an explicit
entrance tax on certain types of foreign exchange transactions (Brazil), as well as indirect
taxation of inflows through a URR (Chile and Colombia) (Ariyoshi et al. 2000).

The macro context in which such controls were imposed by these countries was not the
same. Thailand had a pegged exchange rate regime while the rest had heavily managed
exchange rate regimes. The ability of financial institutions to safely intermediate large
capital inflows was uncertain in some countries (Colombia, Malaysia, and Thailand); while
in other countries such as Chile, policy makers assessed that conventional policy tools
could not adequately deal with large inflows (Ariyoshi et al. 2000).

In the post-Asian financial crisis period, countries in Asia experienced large capital
inflows, which gave rise to unwelcome domestic currency appreciation. There were many
similarities in the responses to these inflows and currency appreciation, although the
macro context in which the controls were imposed may have differed across countries.
Countries used sterilization and direct foreign exchange market intervention, placed
some restrictions on capital inflows and a greater emphasis on the liberalization of capital
outflows (McCauley 2008), strengthened prudential regulations, and allowed greater
exchange rate flexibility. Countries seemed to have used a combination of some or all
of these measures rather than using only capital controls to deal primarily with currency
appreciation as well as asset bubbles.

Unfortunately, the effects of opening or restricting the capital account are not empirically
well established. Part of the problem with empirically verifying the advantages of
liberalizing the capital account is measuring the de facto openness of a country to
cross-border flows. Another problem is the lack of understanding regarding the channels
through which capital account restrictions or liberalization operates. This is due to several
factors including the inability to measure not only the degree of openness of the capital
account, but also the effect of attempting to raise or reduce the intensity of capital
restrictions on the degree of openness. In addition, the particular context in which controls
6 | ADB Economics Working Paper Series No. 224

are used, such as whether the exchange rate is being heavily managed, which may
lead banks to inadequately assess currency risk, as well as the presence of other macro
policy tools, make it difficult to measure and isolate the effects of capital controls on the
economy.

Hence, there is no single theoretical framework with which to analyze the effects of
capital controls, which have been implemented in different countries using different
methods or combinations, over different periods of time, with varying degrees of intensity
and length of time that the controls have been in place, and with differing timing of
adoption (such as whether the country is already in a crisis or not). Apart from difficulties
in measuring the openness of the capital account, when the effectiveness of capital
controls is assessed, it is important to keep in mind that ample historical evidence exists
to show that there are clear differences between de jure and de facto capital controls
(Edwards 2007).

Legal capital controls have been evaded in many ways. These include the simple
overinvoicing of imports and underinvoicing of exports, delay in repayments on trade
finance thereby getting a temporary loan, and other more sophisticated methods. The
intent of policy makers in imposing controls on capital may thus be undermined, and it
may be difficult to measure the extent to which capital controls actually bind.

The ability to evade controls may depend in part on the level of development of domestic
financial markets, where the trading of sophisticated instruments such as derivatives as
in Brazil and Chile; or the shift from one type of asset flow subject to restrictions to other
types of unrestricted flows, e.g., from debt-creating flows to foreign direct investment
(FDI), as in Colombia, facilitates evasion of the controls as experienced by these
countries in the 1990s (Ariyoshi et al. 2000). The incentive to evade controls increases
when the cost of circumventing them declines, for example, by maintaining large interest
rate differentials and pegging or heavily managing the exchange rate. Ironically, capital
controls being successful at driving a wedge between foreign and domestic interest rates
in the context of pegging or managing the exchange rate tends to invite more capital
inflows and complicates their management. The administrative burden imposed by new
capital control measures is significant as they have to be constantly revised to close
loopholes.

Capital controls, of course, are only a single policy instrument, yet there may be multiple
policy objectives and macro tools that are simultaneously operative. This makes it difficult
to ascertain if the use of capital controls is successful—and if it is, how so. Practically
all countries that have used capital controls at any given time have faced large capital
inflows under a heavily managed exchange rate regime in an attempt to prevent real
currency appreciation because of its detrimental effects on a country’s exports. This is
a concern that particularly resonates among countries in Asia given their success with
export-led growth. The problem is that when policies to manage the exchange rate are in
 How Effective are Capital Controls in Asia? | 7

place with capital controls, it is difficult to assess the independent effects of such controls
on the exchange rate and on capital flows themselves.2

In general, the evidence is mixed as regards the effectiveness of capital controls in

achieving their intended objectives (Ariyoshi et al. 2000). Capital controls were only
temporarily able to drive a wedge between foreign and domestic interest rates and
to reduce pressures on the exchange rate in Brazil, Chile, Colombia, Malaysia, and
Thailand in the 1990s. In none of these countries were the interest rate differentials
between foreign and domestic interest rates and the reduction in exchange rate pressures
simultaneously achieved. In fact, all countries experienced significant real appreciations,
although less so in Malaysia and Thailand, where capital controls do not appear to have
reduced the amount of short-term inflows and lengthened maturities.

In the 1990s, some studies show that capital controls either did not have an independent
effect on total net private capital inflows (Cardenas 2007), or only had a temporary effect
on net private capital inflows without any significant effects on the real exchange rate
(Galindo 2007, Concha and Galindo 2008). However, there is no consensus on this as
some studies suggest that net private capital inflows did decline and in that sense, capital
controls were effective (Vargas and Varela 2008). Other studies find that capital controls
reduced the amount of external borrowing but did not have a significant impact on the
volume of non-FDI flows and significantly increased the volatility of the exchange rate
(Edwards and Rigobon 2009).

The adoption by many countries in Asia of inflation targeting as a monetary framework

and greater exchange rate flexibility in the post-Asian financial crisis period may have
obviated some of the inconsistencies in policy experienced in the past. Some of the
constraints to independent monetary policy, such as the desire to peg or manage the
exchange rate in the face of large capital inflows that doomed capital control measures in
the 1990s, may be relieved especially when taken alongside the liberalization of capital
outflows.

Were the various capital control measures in Asia adopted since 2003 effective
(McCauley 2008)? In general, they appear to have succeeded in driving a wedge
between foreign and domestic interest rates, thereby giving monetary authorities the
ability to quell pressures for the domestic currency to appreciate more so than was the
case for countries that used them in the 1990s. In some cases, the volume of certain
types of inflows was successfully reduced. The restriction on the ability of banks in the
PRC to borrow abroad seems to have been successful in reducing such bank inflows for

2 Edwards and Rigobon (2009) is an exception. They model the use of capital controls in the context of an exchange
rate band and the interactions between them. They find that capital controls were able to shield the nominal
exchange rate from vulnerability to external shocks. There is no consensus on their findings. Stiglitz (2002),
Eichengreen (2000), Eichengreen and Hausmann (1999), Stallings (2007), and Williamson (2003) generally support
Edwards and Rigobon’s findings while Calvo and Mendoza (1999), De Gregorio et al. (2000), and Larrain et al.
(2000) generally do not.
8 | ADB Economics Working Paper Series No. 224

6 months after the imposition of the controls, and have reduced interest rates on offshore
forward transactions in the renminbi–dollar exchange rate below onshore rates, even
as bank inflows subsequently grew. There was a sharp decline in fixed-income portfolio
flows into Thailand in response to the URR imposed against portfolio inflows in 2003. The
restriction on nonresident holdings of baht was also able to maintain the gap between
the onshore and offshore exchange rate of the baht to the US dollar. Restrictions on
foreign banks’ dollar borrowing from abroad appeared to have slowed down funding
by foreign banks of their branches in the Republic of Korea. The spread between the
offshore won yields from offshore cross-currency swaps and yields in the domestic market
widened significantly. Evidently, the Republic of Korea has enjoyed the greatest success
in promoting capital outflows, particularly of FDI to the PRC as well as portfolio outflows
particularly to the PRC and India (McCauley 2008). In India, the gap between onshore
Indian interest rates and offshore yields implied by nondeliverable forwards remained
modest.

In summary, the use of capital controls in Asia appears to be generally related to the
need to address undesirable currency appreciation due to large capital inflows. In this
regard, the evidence is mixed as to its effectiveness. There is also no consistent evidence
across time or regions to suggest that capital controls systematically worked to reduce
capital inflows. However, even in cases where capital controls were effective, the effects
were likely to be contemporaneous and temporary. Other policy tools used in conjunction
with capital controls, such as sterilization or exchange rate pegging or management, as
well as the monetary framework in place, e.g., inflation targeting, could complicate the
channels or undo the intended effects of capital controls.

III. Capital Controls in Emerging Asia

Capital accounts in Asia appear to be fairly open, as suggested by the high and generally
rising ratios of flows of foreign assets and liabilities to gross domestic product (GDP)
(Kawai and Lamberte 2008). Emerging Asia has been able to attract relatively consistent
amounts of direct investment (DI) inflows, although significant differences exist across
individual economies. The PRC has been the recipient of more than half of the region’s
DI flows in recent years. India is also a major recipient of DI flows. Portfolio investment
is highly volatile, with the risk of a sudden cut in capital inflows appearing to have
substantially increased (Kim and Ryou 2009).
 How Effective are Capital Controls in Asia? | 9

De jure capital account restrictions indexes constructed by Schindler (2009) for various
inflow and outflow asset categories (DI, portfolio equity, and portfolio debt) based on
data on capital controls obtained from the Annual Report on Exchange Arrangements
and Exchange Restrictions of the International Monetary Fund are used (IMF 2009).
Each index ranges from 0 to 1, with “0” being the least restrictive and “1” being the
most restrictive in terms of capital controls.3 As Binici, Hutchison, and Schindler (2010)
(hereafter “BHS”) point out, Schindler’s measure of capital control is more finely gradated
than the IMF’s binary capital controls dummy variable used in other studies. As such, the
indexes are helpful in detecting subtle differences across countries and over time in the
variety of controls employed.

Figure 1a shows that there is substantial heterogeneity in the evolution of de jure capital
controls among emerging Asian economies. Hong Kong, China; the Republic of Korea;
and Singapore have relatively lower levels of restrictions, while the PRC and India
have maintained a relatively high degree of capital account restrictions. The general
trend before the Asian financial crisis of 1997–1998 had been toward capital account
liberalization. Some degree of tightening can be observed right after the crisis especially
in severely affected economies such as Indonesia and Thailand.

There are differences in capital account restrictions for total flows across asset categories
as shown in Figure 1b. Financial credit restrictions rose dramatically after the 1997–1998
crisis and then declined beginning in 2003. To a lesser extent, the same is true of DI and
equity restrictions. Figure 1c shows an increase in the restrictiveness of capital controls
on inflows and outflows after the crisis, with control on outflows reaching levels higher
than they were prior to the crisis. Controls on inflows, in contrast, are at levels lower than
they were prior to the crisis.

3 See Schindler (2009) for a discussion of the methodology to construct Schindler’s capital control indexes.
10 | ADB Economics Working Paper Series No. 224

Figure 1: Capital Control Indexes, 1995–2005

(a) Average Total Capital Restrictions Index by Economy
1.0
Average Capital Restrictions Index

0.8

0.6

0.4

0.2

0
1995 96 97 98 99 2000 01 02 03 04 05
China, People’s Rep. of Hong Kong, China India
Indonesia Korea, Rep. of Malaysia
Philippines Singapore Thailand
(b) Total Capital Restrictions Indexes by Asset Type
1.0
Capital Restrictions Index

0.8

0.6

0.4

0.2
1995 96 97 98 99 2000 01 02 03 04 05
Direct Investment Equity Financial Credit
(c) Average Total Capital Restrictions Index by Direction of Flow
0.8
Capital Restrictions Index

0.7

0.6

0.5

0.4
1995 96 97 98 99 2000 01 02 03 04 05
Total Flow Inflow Outflow
Note: Average total capital restrictions index is the mean of direct investment, equity,
and financial credit restrictions indexes for total flows.
Source: Schindler (2009).
 How Effective are Capital Controls in Asia? | 11

This is shown also in Figure 2a. Except in the PRC and Indonesia, there is a generally
higher level of controls on outflows than inflows. Overall, the data suggest that economies
have followed different strategies of liberalizing (or, in some cases, tightening) capital
account restrictions. Financial credit restrictions have tended to be higher than
restrictions on equity in the Republic of Korea, Malaysia, the Philippines, Singapore, and
Thailand beginning in 2000 as shown in Figure 2b.

Figure 2: Capital Control Indexes by Economy, 1995–2005

(a) Average Inflow Restrictions versus Average Outflow Restrictions
China, People’s Rep. of Hong Kong, China India
1.0
0.8
0.6
0.4
0.2
0.0
Capital Restrictions Index

Indonesia Korea, Rep. of Malaysia

1.0
0.8
0.6
0.4
0.2
0.0

Philippines Singapore Thailand

1.0
0.8
0.6
0.4
0.2
0.0
1995 96 97 98 99 200001 02 03 04 05 1995 96 97 98 99 2000 01 02 03 04 05 1995 96 97 98 99 2000 01 02 03 04 05

Average Inflow Restrictions Average Outflow Restrictions

(b) Financial Credit Restrictions versus Equity Restrictions

China, People’s Rep. of Hong Kong, China India
1.0
0.8
0.6
0.4
0.2
0.0
Capital Restrictions Index

Indonesia Korea, Rep. of Malaysia

1.0
0.8
0.6
0.4
0.2
0.0

Philippines Singapore Thailand

1.0
0.8
0.6
0.4
0.2
0.0
1995 96 97 98 99 2000 01 02 03 04 05 1995 96 97 98 99 2000 01 02 03 04 05 1995 96 97 98 99 2000 01 02 03 04 05

Financial Credit Restrictions Equity Restrictions

Note: Average inflow (outflow) restrictions index is the mean of direct investment, equity, and financial credit restrictions
indexes for inflows (outflows).
Source: Schindler (2009).
12 | ADB Economics Working Paper Series No. 224

IV. Effects of Capital Controls: Evidence from Cross-

Country Panel Data
A. Data and Methodology

In order to examine the effect of de jure restrictions on aggregate flows (volume of capital
flows); on particular asset categories (DI, equity, and debt flows); and on capital inflows
and outflows in these categories (composition of capital flows), the following baseline
regression equation is estimated:

 KF 
ln   = α 0it + α1KCit + X itθ + ε it
 N it (1)

(KF/N)it is capital flows per capita; KCit is the capital controls index; and Xit is a vector of
control variables. The latter includes the growth of real GDP per capita; the real interest
rate differential between country i and the US; the ratio of stock market capitalization
to GDP and the ratio of domestic credit to the private sector to GDP as proxies for an
economy’s level of financial development; the ratio of merchandise trade to GDP as a
measure of a economy’s openness to trade; the ratio of natural resource exports as a
percentage of total merchandise exports as a proxy for the natural resource endowment
of a country, which is assumed to make it more attractive as an investment destination
the more resource-abundant it is; and an overall business rating index from the
Economist Intelligence Unit (EIU) as an indicator of an economy’s institutional quality and
governance. For more details on the definition and sources of data used in this paper,
see the Appendix.

Annual capital flow figures in terms of aggregate flows, direction of flow and type of
asset from 1995 to 2005 are derived using the methodology in BHS from an updated and
extended version of the database developed by Lane and Milesi-Ferretti (2007) (hereafter
“LM”). Using data on capital flows from the IMF’s Balance of Payments Statistics (BOP)
data including calculations for capital gains and losses, LM generate estimates for stock
positions of countries using estimates of their international investment position as a
benchmark. These stock data are then converted into flows by taking first differences.

BOP data measure net capital inflows and outflows during a recording period while
international investment position data measure the stocks of external assets and liabilities
at the end of the period. Capital inflows measure net purchases or sales by nonresidents
of domestic assets while outflows measure net purchases or sales of foreign assets by
residents. Hence, both capital inflows and outflows can also take on negative values
(LM). Negative values for inflows (outflows) can be considered as outflows (inflows). As in
 How Effective are Capital Controls in Asia? | 13

BHS, to construct the best counterpart of inflow and outflow data from the derived flows
obtained by taking first differences of the LM stock data, the following formulation is used:

Inflows = − min(derived flow assets, 0) + max(derived flow liabilities, 0) (2)

Outflows = max(derived flow assets, 0) − min(derived flow liabilities, 0) (3)

As BHS point out, the way the data are constructed implies that changes in the stocks
can arise from both sales/purchases and from valuation changes. Since capital account
restrictions only affect actual transactions and not valuation changes, the effects of capital
controls may be underestimated using the LM measure. The estimates obtained can thus
be interpreted as lower bounds.

Schindler’s capital control indexes are differentiated by asset type (i.e., financial credit,
equity, collective investment, money market, DI). These capital control indexes are
matched with capital flows in the different asset categories in the following way: equity
restrictions index for equity flows, the financial credit restrictions index for debt flows, and
DI restrictions index for FDI flows.

The data set is a panel composed of nine emerging Asian economies: the PRC;
Hong Kong, China; India; Indonesia; the Republic of Korea; Malaysia; the Philippines;
Singapore; and Thailand for the period 1995–2005. The model is estimated by panel
regression with fixed country and time effects. Standard errors have been corrected for
general forms of heteroskedasticity and uses White’s robust standard errors.

B. Estimation Results
1. Total Flows

Total flows are the sum of inflows and outflows. Table 1a presents the results of tests of
the effectiveness of capital controls on total flows aggregated by the type of asset shown
in each column of the table. The types of assets include FDI, equity, the sum of FDI and
equity, debt, and the sum of all three. The rows of each table indicate which type(s) of
capital the capital restrictions is (are) considered and the control variables used. The
different regression models can be identified by their respective numbers found in the
columns under the asset type names in each table.

The coefficient on the variable for capital controls in Table 1a shows that equity
restrictions significantly reduce the sum of FDI and equity flows as well as the sum of
FDI, equity, and debt flows. Similarly, restrictions on financial credit significantly reduce
the sum of FDI, equity, and debt flows. Restrictions on DI are not found to have any
statistically significant effect on the different types of total flows.
14 | ADB Economics Working Paper Series No. 224

Table 1: Panel Regression Results with Country and Time Fixed Effects, 1995–2005
(dependent variable: ln(capital flow/capita))

Total/By Direction of Flow Type/Composition of Asset

Explanatory Variables FDI Equity FDI + Debt FDI + Equity +
Equity Debt
(a) Total Flows (Inflow + Outflow) (1) (2) (3) (4) (5)
Direct investment restrictions -0.093 -0.322 -0.324
[0.15] [0.80] [0.91]
Equity restrictions -0.518 -1.338 -1.596
[0.80] [2.81]*** [3.78]***
Financial credit restrictions -1.035 0.600
[1.43] [1.80]*
Real interest rate differential 0.049 0.018 0.012 -0.118 0.015
[1.28] [0.69] [0.81] [1.47] [1.03]
ln(real GDP per capita) 0.193 3.763 0.308 -6.722 -0.340
[0.13] [2.20]** [0.31] [0.46] [0.38]
Stock market capitalization/GDP 0.157 0.452 0.286 -1.931 0.278
[0.60] [1.58] [1.59] [1.03] [1.48]
Domestic credit to private sector (percent of GDP) 0.009 0.012 0.013 -0.049 0.012
[0.91] [1.50] [2.43]** [0.80] [2.62]**
Merchandise trade (percent of GDP) -0.006 0.005 -0.006 0.006 -0.010
[1.21] [0.68] [1.54] [0.44] [2.28]**
Natural resources 0.103 0.040 0.063 0.263 0.029
(percent of merchandise exports) [2.06]** [0.52] [2.06]** [1.41] [0.96]
Overall business rating, EIU -0.482 -0.156 -0.626 0.442 -0.555
[1.57] [0.42] [3.04]*** [0.19] [2.75]***
Observations 99 99 99 40 99
Number of countries 9 9 9 9 9
R-squared 0.39 0.52 0.58 0.54 0.61
(b) Inflow (6) (7) (8) (9) (10)
Direct investment restrictions -1.842 -1.053 -0.508
[4.83]*** [2.44]** [0.96]
Equity restrictions -0.187 -0.333 -0.929
[0.32] [0.45] [1.09]
Financial credit restrictions -3.221 0.288
[1.19] [0.57]
Real interest rate differential 0.062 0.114 0.066 -0.115 0.070
[2.19]** [1.65] [2.28]** [0.28] [2.05]**
ln(real GDP per capita) -0.016 4.439 -1.022 32.403 -1.498
[0.01] [1.94]* [0.58] [0.45] [0.83]
Stock market capitalization/GDP 0.115 0.603 0.215 -17.879 0.275
[0.39] [1.05] [0.62] [2.15] [0.76]
Domestic credit to private sector (percent of GDP) 0.006 -0.004 0.012 0.017 0.011
[0.50] [0.24] [1.01] [0.09] [0.94]
Merchandise trade (percent of GDP) -0.003 -0.013 -0.005 0.210 -0.010
[0.49] [1.31] [0.83] [2.38]* [1.32]
Natural resources 0.075 0.006 0.063 0.337 0.028
(percent of merchandise exports) [1.23] [0.08] [1.16] [0.34] [0.46]
Overall business rating, EIU -0.209 0.189 0.238 5.555 0.225
[0.53] [0.40] [0.54] [0.88] [0.42]
Observations 92 79 94 26 94
Number of countries 9 9 9 8 9
R-squared 0.47 0.62 0.56 0.94 0.54
continued.
 How Effective are Capital Controls in Asia? | 15

Table 1: continued.
Total/By Direction of Flow Type/Composition of Asset
Explanatory Variables FDI Equity FDI + Debt FDI + Equity +
Equity Debt
(c) Outflow (11) (12) (13) (14) (15)
Direct investment restrictions 1.170 0.813 0.953
[2.40]** [1.73]* [1.88]*
Equity restrictions 2.499 0.657 0.402
[1.93]* [0.90] [0.50]
Financial credit restrictions -5.018 -0.492
[0.70] [1.28]
Real interest rate differential -0.021 -0.037 -0.010 -0.187 -0.011
[0.31] [1.05] [0.41] [0.30] [0.49]
ln(real GDP per capita) 2.383 4.583 3.680 50.130 3.809
[1.14] [1.40] [1.92]* [0.40] [1.71]*
Stock market capitalization/GDP 0.288 0.129 0.182 -17.855 0.086
[0.71] [0.22] [0.68] [1.30] [0.32]
Domestic credit to private sector (percent of GDP) 0.012 -0.003 0.009 -0.005 0.001
[1.00] [0.13] [0.73] [0.02] [0.06]
Merchandise trade (percent of GDP) -0.007 0.029 0.002 0.183 -0.001
[0.83] [2.19]** [0.29] [1.76] [0.13]
Natural resources 0.150 0.063 0.112 0.352 0.075
(percent of merchandise exports) [2.43]** [0.68] [2.00]** [0.30] [1.13]
Overall business rating, EIU -1.341 -0.514 -1.137 3.342 -0.964
[3.15]*** [0.64] [2.98]*** [0.46] [2.66]***
Observations 95 91 98 26 98
Number of countries 9 9 9 8 9
R-squared 0.26 0.35 0.34 0.93 0.32
* significant at 10%; **significant at 5%; *** significant at 1%.
EIU = Economist Intelligence Unit, FDI = foreign direct investment, GDP = gross domestic product.
Note: Robust t-statistics in brackets. All specifications include time dummies and a constant but not reported. Data on debt flows
are unavailable for earlier years.
Source: Authors‘ estimates.

Among the control variables, the ratio of domestic credit to the private sector to
GDP, natural resource exports as a percentage of merchandise exports, and the EIU
overall business rating index are those that are generally statistically significant in the
regressions. A higher ratio of domestic credit to the private sector as a proportion of
GDP gives rise to larger total aggregate flows of FDI and equity, as well as of total
aggregate FDI, equity, and debt flows. As expected, a larger ratio of natural resources
to merchandise trade increases FDI, the sum of FDI and debt, and the sum of all three
types of capital flows, but not equity flows alone nor debt flows. The coefficient on the
EIU overall business rating index, a proxy for institutional quality and governance, is
significantly positive in the regressions for the sum of FDI and equity flows and the sum
of all three types of flows. This suggests that good governance and institutional quality
tend to increase aggregate FDI and equity flows as well as the aggregate of all three
types of flows, another intuitively appealing result. A higher level of real GDP per capita
also significantly increases total equity flows. Similarly, a higher ratio of merchandise
16 | ADB Economics Working Paper Series No. 224

trade as a percentage of GDP or greater trade openness also significantly increases total
aggregate flows of FDI, equity, and debt.

Since the available data on debt flows are limited as there are many observations with
missing values, the effectiveness of capital controls on debt flows is tested using a panel
Tobit regression model with country and time dummies. As in the earlier panel regression
results shown in Table 1a, the panel Tobit regression results show that financial credit
restrictions have no statistically significant effect on total debt flows.4

2. Inflows

Table 1b presents the results of tests of the effectiveness of capital controls on total
inflows aggregated by type. As seen from the table, capital controls on DI inflows
significantly reduce FDI inflows as well as the sum of FDI and equity inflows. Financial
credit restrictions also reduce debt inflows but this effect is significant only when panel
Tobit regression is used.5

In general, it appears that FDI is the main type of inflow affected by capital controls on
FDI inflows alone. In contrast with the earlier results, the sum of the three types of flows
is not affected by capital restrictions on capital inflows.

Among the control variables, it is the real interest rate differential that is significant in
most regression models and affects capital inflows of different types, except debt and
equity inflows considered separately. An increase in the real interest rate differential in
favor of the domestic country significantly increases capital inflows to the country. This
seems to imply that real interest rates are higher in the recipient country because the
marginal productivity of capital is likewise relatively higher there and is thus attractive to
investment. In one regression model, a higher level of real GDP per capita is found to
significantly increase equity inflows while in another, the proportion of merchandise trade
in GDP is found to increase total debt inflows.

3. Outflows

Table 1c shows that capital controls have a statistically significant effect on almost types
of total outflows except total debt outflows.6 Direct investment restrictions significantly
affect FDI outflows, the sum of FDI and equity outflows, and the sum of FDI, equity, and
debt outflows. Restrictions on equity outflows also significantly affect equity outflows.
However, the sign on the coefficients of the capital control restrictions is positive in all
cases in which it is statistically significant. The findings imply that capital outflows actually
increase when countries impose restrictions to try and prevent capital outflows. This
4 Data for other total flows are also fitted using the same panel Tobit regression model. Results using panel Tobit
regression and panel regression model for other total flows are similar.
5 The coefficient on financial credit restrictions is –3.22 with a p-value of <0.001.
6 The coefficient of capital restrictions is significant for the sum of FDI and equity flows when using least squares
panel regression model only but not when the panel Tobit model is used.
 How Effective are Capital Controls in Asia? | 17

may be so because a country that restricts capital outflows sends a signal to market
participants that it is worried about loss of confidence and the possibility of capital flight
and financial instability, and may signal difficulties in repatriating profits, any of which
could precipitate capital outflows. The results suggest that capital controls may not be the
best instrument to use to reduce the volume of capital outflows or engineer a shift in the
type of capital outflows.

The idea that restrictions on capital outflows may affect perceptions about confidence
in the recipient country are further bolstered by the finding that the frequently significant
control variable for capital outflows is evidently the EIU overall business rating. Real GDP
per capita is also significant in the regressions as are the fraction of natural resources in
merchandise trade and merchandise trade as a proportion of GDP.

C. Asymmetric Effects of Capital Controls

The possibility of asymmetric effects, in which the effect of tightening capital controls may
be different from relaxing controls, is tested by estimating the following equation:

 KF 
ln 
 N it
( ) +
( ) −
 = α 0it + α1KCit + α 2 KCit ⋅ D + α 3 KCit ⋅ D + X itθ + ε it (4)

In equation (4), D+ is a dummy variable that takes on a value of 1 when capital controls
are tightened from the previous period and 0 otherwise, while D− is the dummy used to
represent loosening of capital controls from the previous period. The results of estimating
equation (4) for the presence of asymmetric effects are presented in Tables 2a, 2b, and 2c.

Table 2a shows that there are generally no statistically significant asymmetric effects on total
flows with one exception. In Model (2), a reduction in restrictions on equity flows leads to a
statistically significant increase the amount of equity flows. However, tightening restrictions
on equity flows does not have a statistically significant effect on such flows.
Table 2b shows that there are no significant asymmetric effects on capital inflows.

Table 2c shows that there are generally no statistically significant asymmetric effects on
capital outflows except in the case of tightening equity restrictions that reduce the sum
of FDI, equity, and debt outflows. However, loosening equity restrictions do not have a
statistically significant effect on these outflows.
18 | ADB Economics Working Paper Series No. 224

Table 2: Panel Regression Results with Asymmetric Effects, 1994–2005

(dependent variable: ln(capital flow/capita))
Total/By Direction of Flow Type/Composition of Asset
Explanatory Variables FDI Equity FDI + Equity Debt FDI + Equity
+ Debt
(a) Total Flow (Inflow + Outflow) (1) (2) (3) (4) (5)
DI restrictions -0.216 -0.548 -0.298
[0.24] [0.91] [0.58]
DI restrictions x Reduced restriction dummy -0.493 0.092 0.318
[1.03] [0.26] [0.84]
DI restrictions x Increased restriction dummy -0.050 0.145 0.217
[0.09] [0.44] [0.85]
Equity restrictions -0.539 -1.499 -1.599
[0.57] [2.14]** [2.26]**
Equity restrictions x Reduced restriction dummy -0.628 0.080 -0.019
[1.82]* [0.36] [0.10]
Equity restrictions x Increased restriction dummy -0.102 -0.093 -0.077
[0.28] [0.43] [0.35]
FC restrictions -2.016 0.611
[0.85] [1.32]
FC restrictions x Reduced restriction dummy -0.747 -0.071
[0.50] [0.18]
FC restrictions x Increased restriction dummy 0.091 0.052
[0.12] [0.18]
Observations 90 90 90 40 90
Number of countries 9 9 9 9 9
R-squared 0.39 0.49 0.57 0.54 0.60
(b) Inflow (6) (7) (8) (9) (10)
DI restrictions -2.053 -1.448 -1.475
[2.52]** [1.11] [1.09]
DI restrictions x Reduced restriction dummy -0.112 -0.142 -0.498
[0.20] [0.27] [0.77]
DI restrictions x Increased restriction dummy 0.119 0.722 1.402
[0.11] [0.49] [0.90]
Equity restrictions 0.000 0.000 0.000
[.] [.] [.]
Equity restrictions x Reduced restriction dummy 1.773 1.628 1.498
[1.55] [1.38] [1.25]
Equity restrictions x Increased restriction dummy 0.482 0.434 0.575
[0.81] [0.64] [0.66]
FC restrictions -5.819 -0.024
[1.23] [0.03]
FC restrictions x Reduced restriction dummy 3.457 0.061
[1.18] [0.09]
FC restrictions x Increased restriction dummy 5.106 0.719
[0.98] [0.98]
Observations 83 71 85 26 85
Number of countries 9 9 9 8 9
R-squared 0.45 0.63 0.59 0.98 0.58
continued.
 How Effective are Capital Controls in Asia? | 19

Table 2: continued.

Total/By Direction of Flow Type/Composition of Asset

Explanatory Variables FDI Equity FDI + Equity Debt FDI + Equity
+ Debt
(c) Outflow (11) (12) (13) (14) (15)
DI restrictions 0.249 1.169 1.411
[0.27] [1.21] [1.48]
DI restrictions x Reduced restriction dummy -0.973 -0.195 -0.620
[1.45] [0.21] [0.61]
DI restrictions x Increased restriction dummy 0.407 -0.596 -0.730
[0.52] [0.75] [0.85]
Equity restrictions 3.971 1.223 1.271
[2.02]** [1.02] [1.16]
Equity restrictions x Reduced restriction dummy -0.016 -0.176 -0.330
[0.01] [0.25] [0.46]
Equity restrictions x Increased restriction dummy -0.113 -0.711 -1.133
[0.15] [1.19] [1.89]*
FC restrictions -5.999 -0.245
[0.67] [0.51]
FC restrictions x Reduced restriction dummy 0.000 0.932
[.] [1.07]
FC restrictions x Increased restriction dummy 1.245 -1.025
[0.33] [0.99]
Observations 86 82 89 26 89
Number of countries 9 9 9 8 9
R-squared 0.25 0.38 0.34 0.93 0.36
* significant at 10%; **significant at 5%; *** significant at 1%.
DI = direct investment, FC = financial credit, FDI = foreign direct investment.
Note: The asymmetric effects are with country and time fixed effects. Robust t-statistics in brackets. All specifications include time
dummies and a constant but not reported. Data on debt flows are unavailable for earlier years. All specifications include all
control variables, namely, real interest rate differential, ln(real GDP per capita), stock market capitalization/GDP, domestic
credit to private sector, merchandise trade, natural resources, and EIU’s overall business rating; time dummies; and a
constant but not reported.
Source: Authors‘ estimates.

D. Substitution Effects among Capital Controls

The possibility of substitution effects—whether a particular type of capital flow is

significantly affected by capital controls imposed on other types of assets—is tested. To
test for the presence of substitution effects, a particular type of capital flow is regressed
against its own capital restriction as well as all the other different types of capital
restrictions entered as separate explanatory variables individually per regression model:

 KFj 
ln   = α 0it + α1KC jt + α 2KCkt + α 3 KCit + X itθ + ε it (5)
 N it
where KFj is j-type of capital flow and KCk and KCl are the capital controls for other types
of capital flows (i.e., type k and type l).
20 | ADB Economics Working Paper Series No. 224

The results are presented in Table 3. Table 3a shows that equity restrictions and financial
credit restrictions have a statistically significant effect on total FDI flows. When equity
restrictions are imposed, FDI flows decline. Surprisingly, equity restrictions seem to
have a greater effect on FDI flows than on equity flows. In the case of financial credit
restrictions, however, there is a statistically significant positive effect on total FDI flows.
This latter finding seems to suggest that FDI flows and debt flows, the latter being the
type of flows for which financial credit restrictions are designed to affect, are substitutes
of each other. Note also that DI restrictions have a positive but statistically insignificant
effect on debt flows. This may be because restrictions on financial credit may be
seen as a prudential measure to prevent asset bubbles from being created. Financial
credit restrictions may not be harmful to FDI as FDI may not generally depend on the
availability of local credit nor be financed by debt.

Table 3: Panel Regression Results with Substitution Effects,1995-2005

(dependent variable: ln(capital control/capita)
FDI Equity Debt
(a) Total (Inflow + Outflow) (1) (2) (3)
Total Total Total
Direct investment restrictions 0.045 -0.643 0.430
[0.08] [1.15] [0.62]
Equity restrictions -2.656 -0.263 2.070
[4.32]*** [0.38] [0.79]
Financial credit restrictions 1.444 -0.154 -1.130
[4.31]*** [0.27] [1.22]
Observations 99 99 40
R-squared number of countries 0.52 0.53 0.56
(b) Inflow (7) (8) (9)
FDI FDI Equity
Inflow Inflow Inflow
Direct investment restrictions -2.221 -0.488 -1.860
[4.51]*** [1.37] [0.25]
Equity restrictions 0.698 0.008 -2.125
[1.14] [0.01] [0.09]
Financial credit restrictions 0.251 1.590 -3.235
[0.48] [1.99]* [0.69]
Observations 92 79 26
R-squared number of countries 0.48 0.67 0.95
(c) Outflow (13) (14) (15)
FDI FDI Equity
Outflow Outflow Outflow
Direct investment restrictions 1.063 0.404 0.912
[1.98]* [0.52] [0.67]
Equity restrictions -0.848 2.854 -1.196
[0.92] [2.28]** [0.44]
Financial credit restrictions 0.194 -1.604 -0.442
[0.34] [3.24]*** [0.25]
Observations 95 91 36
R-squared 0.27 0.41 0.74
* significant at 10%; **significant at 5%; *** significant at 1%.
FDI = foreign direct investment.
Note: The substitution effects are with country and time fixed effects. Robust t-statistics in brackets. All specifications include time
dummies and a constant but not reported. Data on debt flows are unavailable for earlier years. All specifications include all
control variables, namely, real interest rate differential, ln(real GDP per capita), stock market capitalization/GDP, domestic
credit to private sector, merchandise trade, natural resources, and EIU’s overall business rating; time dummies; and a
constant but not reported.
Source: Authors‘ estimates.
 How Effective are Capital Controls in Asia? | 21

Table 3b shows that financial credit restrictions tend to increase equity inflows, suggesting
the presence of a substitution effect as well between equity inflows and debt inflows.

In terms of outflows, however, Table 3c shows that restrictions on financial credit reduce
equity outflows. This suggests that debt and equity outflows are more complements of
each other than substitutes.

In general, therefore, financial credit restrictions appear to be effective in affecting FDI

flows, and equity inflows and outflows. Total FDI and debt flows as well as equity and
debt inflows appear to be substitutes of each other, while equity and debt outflows appear
to be more complements of each other.

E. Endogeneity of Capital Controls

It is possible that capital controls may be endogenous if policy makers react to the
effects of capital flows, such as currency appreciation or the accumulation of reserves,
by imposing or withdrawing capital controls, or by changing the degree of restrictiveness
of capital controls. To test for endogeneity, a set of panel instrumental variable (IV)
regressions by direction of capital flow and by asset category is estimated, all with
country and time fixed effects.

Tests are performed to determine whether capital flows influence the use of capital
controls or their degree of restrictiveness. Cardoso and Goldfajn (1998) use expected
returns as an instrument for capital flows. In their paper, total private capital flows are
assumed to depend on expected returns and changes in capital controls.7 Changes in
capital controls, in turn, depend on total net capital flows. By substitution, the reduced
form equation has changes in capital controls depending on both expected returns
and capital flows as shown in equation (6) below. In this study, the real interest rate
differential between a country’s and that for the US is used as a proxy for expected
returns. Specifically, the following reduced form functional relationship is used:

 KF 
∆KCit = f  ln
N
(
g ( i − iUS )it ) (6)
 IT 
where (i − iUS) is the differential in real interest rates between country i and the US. The
results are shown in Table 4a. None of the coefficients on capital flows are statistically
significant. Hence, no evidence of an endogeneity problem is found.

An alternative test of endogeneity is also implemented. In Table 4b, the two-period lag of
capital flows is used as an alternative instrument. As can be seen from the table, none
of the coefficients on the capital control variables are statistically significant. Again, no
evidence of an endogeneity problem is found.
7 In this study, the log of capital flows per capita, by direction of flow and asset type, is used in lieu of net capital
flows in Cardoso and Goldfajn (1998).
22 | ADB Economics Working Paper Series No. 224

Table 4: Test for Endogeneity, 1996–2005

Type/Composition of Asset
FDI Equity FDI + Equity Debt FDI + Equity
+ Debt
(a) Dependent Variable: Change in Restrictions (1) (2) (3) (4) (5)

ln(total capital flow) 0.023 -0.003 0.024 0.269 -0.022

[0.35] [0.02] [0.31] [0.90] [0.28]
Observations 90 90 90 40 90

(6) (7) (8) (9) (10)

ln(Capital inflow) -0.016 -0.044 -0.033 -0.033 -0.057
[0.27] [0.86] [1.04] [1.04] [1.53]
Observations 83 71 85 26 85

(11) (12) (13) (14) (15)

ln(capital outflow) 0.199 -0.086 0.824 -0.177 -0.177
[0.25] [0.76] [0.20] [1.30] [1.30]
Observations 86 82 89 36 89

(b) Dependent Variable: Change in Restrictions (16) (17) (18) (19)

ln(total capital flow) 0.040 0.010 0.000 — -0.079

[0.19] [0.14] [0.00] — [0.36]

Observations 81 81 81 81

(20) (21) (22) (23)

ln(capital inflow) 0.084 0.030 -0.262 — -0.097
[0.43] [0.48] [0.37] — [0.33]
Observations 67 46 71 71

(24) (25) (26) (27)

ln(capital outflow) -0.006 0.099 -6.072 — -0.246
[0.05] [1.38] [0.01] — [0.27]
Observations 73 66 79 79
9 9 9 9
* significant at 10%; ** significant at 5%; *** significant at 1%.
“—“ denotes insufficient observations.
FDI = foreign direct investment.
Note: With country and time fixed effects. Absolute value of z-statistics in brackets. Instrument for ln(capital flow) in (a) is real
interest rate differential between country i and the US; 1-year lagged ln(capital flow) for (b).
Source: Authors‘ estimates.

V. Conclusion
This study examines the effects of capital account restrictions on capital flows in nine
emerging Asian economies over the period 1995–2005 using panel regressions with fixed
effects. The type of capital restriction imposed matters for capital flows by both type of
asset and direction of capital flows.
 How Effective are Capital Controls in Asia? | 23

In terms of total flows, the results show that capital restrictions on equity flows reduce the
sum of total volume of FDI and equity flows as well as the sum of FDI, equity, and debt
flows. Financial credit restrictions reduce the sum of FDI, equity, and debt flows. Total
flows are significantly affected by the openness of the economy, the natural resource
endowment of an economy, and by perceptions about the ease doing business in an
economy, which in this study is regarded as a proxy for the quality of an economy’s
institutions and governance.

In terms of capital inflows, restrictions on DI inflows reduce FDI inflows as well as the
sum of FDI and equity inflows. Financial credit restrictions likewise reduce debt inflows,
but this effect is only detected when panel Tobit regression is used. In general, the results
suggest that controls on capital inflows largely impact FDI inflows. The real interest rate
differential was the most important control variable found, consistent with the idea that
the relative marginal efficiency of investment and profitability of investment matter for FDI
inflows. A general policy prescription, therefore, is for policy makers to be encouraged to
find ways to improve the profitability of investment, rather than resort to capital controls to
affect the type of asset flows their economies receive.

In terms of capital outflows, capital restrictions affect most asset types (with the exception
of total debt outflows) but in a perverse way, as the use of controls on capital outflows
actually increases the amount of these outflows. This finding suggests that it is not
possible for policy makers to prevent capital outflows, reduce their volume, or target
particular types of capital outflows through the use of capital controls on outflows. Hence,
it may be best for economies to liberalize rather than constrain capital outflows if they
want to prevent such outflows.

The results here stand in contrast to those in BHS in which both debt and equity controls
reduce capital outflows with little effect on capital inflows. The findings in this study also
suggest that debt outflows are not generally affected by capital controls on debt while the
effect on debt inflows is only detected when a panel Tobit regression is used. Financial
credit restrictions seem to be the most effective type of capital control.

Tests for the possible endogeneity of capital controls are conducted using panel IV
regression with fixed effects and the real interest rate differential or the second lag of
capital flows as alternative instruments. No statistically significant endogeneity problem
is detected. The presence of asymmetric effects is also tested by using dummy variables
that distinguish between episodes of tightening versus relaxing capital controls. Loosening
equity restrictions increases the amount of total equity flows but tightening such
restrictions has no statistically significant effect on equity flows. Also, tightening equity
restrictions reduces the sum of FDI, equity, and debt outflows, while loosening such
restrictions do not have a statistically significant effect on these outflows.
24 | ADB Economics Working Paper Series No. 224

The possibility of substitution effects in which imposing capital controls on one type of
asset flows leads to statistically significant effects on the volume of other types of asset
flows is likewise tested. The results suggest that total FDI flows and total debt flows are
substitutes of each other as are debt inflows and equity inflows. However, debt outflows
and equity outflows appear to be more complements of each other than substitutes.
Financial credit restrictions affect FDI flows, equity inflows, and equity outflows. They
reduce equity outflows and do not reduce total FDI flows and equity inflows.

Why capital controls work, which types of capital flows they work on, under what
circumstances they work in, and for what purpose they are employed in a particular
country are questions not addressed by this study. These questions need to be examined
in a country-specific context and are the subject of future research.
 How Effective are Capital Controls in Asia? | 25

Appendix: Data Sources and Definitions

Variables/Frequency/ Notes and Definitions Sources of Data
Period Covered
Dependent Variables
Investment flows (total, Following Binici et al. (2010), flows data are derived by taking Lane and Milesi-
inflow, outflow) per capita the first difference of Lane and Milesi-Feretti (2007) stock Feretti (2007),
($M/‘000 population)/ data. Flows data are divided by working age population downloaded 25
annual/1995–2005 (‘000). November 2010.
Equity Includes equity securities issued by monetary authorities, Population data
general government, banks, and other sectors. from the World
Debt Includes bonds and notes; and money market instruments Bank’s World
issued by monetary authorities, general government, banks Development
and other sectors. Indicators,
downloaded 2
December 2010

Foreign direct investment Includes equity capital, reinvested earnings, and other capital.
Explanatory Variables
Capital restriction Index of financial openness (Range: 0 to 1, from least to most Schindler (2009),
indexes (total/average, regulated). Total/average restriction index is a simple average downloaded 25
inflow, outflow)/annual/ of the inflow and outflow capital restriction index by asset November 2010
1995–2005 category.
Financial credit Covers restrictions on credits other than commercial credits
granted by all residents, including banks, to nonresidents or
vice versa.
Equity Refers to shares and other securities of a participating nature
and bonds, securities, and other securities with an original
maturity of more than 1 year.
Direct investment Covers restrictions on investments for the purpose of
establishing lasting investment economic relations both
abroad by residents and domestically by nonresidents. These
investments are essentially for the purpose of producing
goods and services, and, in particular, in order to allow
investor participation in the management of an enterprise.
Growth in real GDP per Derived using GDP per capita data (constant 2000 US$). World Bank World
capita (percent) Development
Real interest rate Domestic real interest rate minus US real interest rate. Indicators
differential (percent) Database,
Merchandise trade (percent Total trade (exports + imports)/GDP. downloaded 25
of GDP) November 2010
Natural resources Exports of natural resources (including agricultural raw
(percent of merchandise materials, fuel, and metals and ores) as a proportion of total
exports) merchandise exports.
Domestic credit to private Refers to financial resources provided to the private sector,
sector (percent of GDP) such as through loans, purchases of nonequity securities, and
trade credits and other accounts receivable, which establish a
claim for repayment.
26 | ADB Economics Working Paper Series No. 224

Variables/Frequency/ Notes and Definitions Sources of Data

Period Covered
Stock market capitalization “Value of listed shares to GDP, calculated using the following Beck and
(percent of GDP)/ deflation method: {(0.5)*[Ft/P_et + Ft-1/P_et-1]}/[GDPt/P_at] Demirgüç-Kunt
annual/1995–2005 where F is stock market capitalization, P_e is end-of-period (2009) revised as
CPI, and P_a is average annual CPI.” of March 2010,
downloaded 25
November 2010
Overall business rating, EIU’s overall business environment rating (Range: 1 to 10 from Economist
EIU/annual/1995–2005 low to high or most favorable environment) is an average Intelligence Unit
of EIU’s rating for the following: market opportunities, World Investment
macroeconomic environment, political environment, Services Database,
infrastructure, policy toward private enterprise, labor market, downloaded 2
tax regime, financing, foreign trade and exchange, and policy December 2010
environment for foreign investment.
CPI = consumer price index, EIU = Economist Intelligence Unit, GDP = gross domestic product.
Note: Investment flows per capita are available only for countries and years for which national authorities report their
international investment position. Missing observations: People's Republic of China (1995–2004); Hong Kong, China and
the Republic of Korea (1995–2000); India (1995–1996); Indonesia; Malaysia; the Philippines; and Singapore (1995–2001);
Thailand (inflow for 1995, outflow and total for 1995–2000).
 How Effective are Capital Controls in Asia? | 27

References
ADB. 2010. Asian Development Outlook 2010. Asian Development Bank, Manila.
Ariyoshi, A., K. Haberneier, B. Laurens, I. Otker-Robe, J. Canales-Kriljenko, and A. Kirilenko. 2000.
Capital Controls: Country Experiences with Their Use and Liberalization. IMF Occasional
Paper No. 190, International Monetary Fund, Washington, DC.
Beck, T., and A. Demirgüç-Kunt. 2009. A New Database on Financial Development and Structure
(Revised as of March 2010). Available: econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/
EXTRESEARCH/0,,contentMDK:20696167~pagePK:64214825~piPK:64214943~theSite
PK:469382,00.html. Downloaded 25 November 2010.
Bhagwati, J. 1998. “The Capital Myth: The Difference between Trade in Widgets and Dollars.”
Foreign Affairs 7(May/June):7–12.
Binici, M., M. Hutchison, and M. Schindler. 2010. “Controlling Capital? Legal Restrictions and
the Asset Composition of International Financial Flows.” Journal of International Money and
Finance 29:666–84.
Calvo, G., and E. Mendoza. 1999. “Empirical Puzzles of Chilean Stabilization Policy.” In G. Perry
and D. Leipziger, eds., Chile: Recent Policy Lessons and Emerging Challenges. The World
Bank, Washington DC.
Cardenas, M. 2007. “Controles de Capitales en Colombia: Funcionano O No?” In Debate de
Coyuntura Economica. Bogota: Fedesarrollo.
Cardoso, E., and I. Goldfajn. 1998. “Capital Flows to Brazil: The Endogeneity of Capital Controls.”
IMF Staff Papers 45(1):161–202.
Concha, A., and A. Galindo. 2008. “An Assessment of Another Decade of Capital Controls
in Colombia: 1998–2008.” Paper presented at the XIII LACEA Meeting, Rio de Janeiro.
Unpublished.
De Gregorio, J., S. Edwards, and R. Valdes. 2000. “Controls on Capital Inflows: Do They Work?”
Journal of Development Economics 63:59–83.
EIU. 2010. World Investment Services Database. Economist Intelligence Unit. Available: https://
portal.eiu.com/sso/cas/login.asp?service=http%3A%2F%2Fwww.eiu.com%2Fsso%2Fcas%2Fc
lient&brand=&renew=true&auth=ip. Downloaded 2 December 2010.
Edwards, S. 2007. “Capital Controls, Sudden Stops, and Current Account Reversals.” In S.
Edwards, ed., Capital Controls and Capital Flows in Emerging Economies: Policies, Practices
and Consequences. Chicago: University of Chicago Press.
Edwards, S., and R. Rigobon. 2009. “Capital Controls on Inflows, Exchange Rate Volatility and
External Vulnerability.” Journal of International Economics 78:256–67.
Eichengreen, B. 2000. “Taming Capital Flows.” World Development 28(6):1105–16.
Eichengreen, B., and R. Hausmann. 1999. Exchange Rates and Financial Fragility. NBER Working
Paper No. 7418, National Bureau of Economic Research, Massachusetts.
Galindo, A. 2007. “Controles de Capitales en Colombia: Funcionan o No?” In Debate de Coyuntura
Economica. Bogota: Fedesarrollo.
International Monetary Fund. 2009. Annual Report of Exchange Arrangements and Exchange
Restrictions (AREAER). Washington, DC.
Kawai, M., and M. Lamberte. 2008. Managing Capital Flows in Asia: Policy Issues and Challenges.
ADB Institute Research Policy Brief No. 26, Asian Development Bank Institute, Tokyo.
Kawai, M., and S. Takagi. 2008. A Survey of the Literature on Managing Capital Inflows. ADB
Institute Discussion Paper No. 100, Asian Development Bank Institute, Tokyo.
Kim, T.-J., and J.-W. Ryou. 2009. “Determinants of Cross-border Financial Capital Flows in
East Asia: The Case of Korea.” Paper presented at the Oxford Business and Economics
Conference, 24–26 June, Oxford.
28 | ADB Economics Working Paper Series No. 224

Krugman, P. 1999. “Currency Crises.” In M. Feldstein, ed., International Capital Flows. Chicago:
National Bureau of Economic Research and University of Chicago Press.
Lane, P. R., and G. M. Milesi-Ferretti. 2007. “The External Wealth of Nations Mark II: Revised and
Extended Estimates of Foreign Assets and Liabilities.” Journal of International Economics
73(2): 223–50.
Larrain, F., R. Labtan, and R. Chumacero. 2000. “Chapter 4: What Determines Capital Inflows?
An Empirical Analysis for Chile.” In F. B. Larrain, ed., Capital Flows, Capital Controls, and
Currency Crises: Latin America in the 1990s. Michigan: University of Michigan Press.
Magud, N., and C. Reinhart. 2007. “Capital Controls: An Evaluation.” In S. Edwards, ed., Capital
Controls and Capital Flows in Emerging Economies: Policies, Practices and Consequences.
Chicago: University of Chicago Press.
McCauley, R. 2008. Managing Recent Hot Money Flows in Asia. ADB Institute Discussion Paper
No. 99, Asian Development Bank Institute, Tokyo.
Obstfeld, M. 1994. International Capital Mobility in the 1990s. CEPR Discussion Paper No. 902,
Centre for Economic Policy Research, London.
Ostry, J., A. Ghosh, K. Haberneier, M. Chamon, M. Qureshi, and D. Reinhardt. 2010. Capital
Inflows: The Role of Controls. IMF Staff Position Note SPN/10/04, International Monetary
Fund, Washington, DC.
Perkins, D., and W. T. Woo. 2000. “Malaysia: Adjusting to Deep Integration with the World
Economy.” In W. T. Woo, J. Sachs, and K. Schwab, eds., The Asian Financial Crisis: Lessons
for a Resilient Asia. Cambridge: MIT Press.
Rodrik, D. 2006. “Goodbye Washington Consensus, Hello Washington Confusion? A Review of the
World Bank’s Economic Growth in the 1990s: Learning from a Decade of Reforms.” Journal
of Economic Literature 44(4):619–35.
Schindler, M. 2009. Measuring Financial Integration: A New Data Set. IMF Staff Papers No. 56,
International Monetary Fund, Washington, DC.
Stallings, B. 2007. “The Globalization of Capital Flows: Who Benefits?” The Annals of the
American Academy of Political and Social Science 3(610):201–16.
Stiglitz, J. 2000. “Capital Market Liberalization, Economic Growth, and Instability.” World
Development 28(6):1075–86.
———. 2002. Globalization and its Discontents. New York: W.W. Norton.
Vargas, H., and C. Varela. 2008. Capital Flows and Financial Assets in Colombia: Recent
Behavior, Consequences and Challenges for the Central Bank. Borradores de Economia No.
502, Banco de lad Republica de Colombia, Bogota.
Williamson, J. 2003. “An Agenda for Restarting Growth and Reform.” In P. Kuczynski, ed., After the
Washington Consensus. Institute for International Economics, Washington, DC.
World Bank. 2010. World Development Indicators Database. Available: data.worldbank.org/data-
catalog/world-development-indicators/wdi-2010. Downloaded 25 November 2010.
About the Paper
Maria Socorro Gochoco-Bautista, Juthathip Jongwanich, and Jong-Wha Lee find that capital
controls significantly affect total capital flows, as well as inflows and outflows. Restrictions
on financial credit appear to be more effective than other forms of controls. Results also
show that liberalizing capital outflows may be a better policy than imposing controls on
them, because the latter results to even larger capital outflows.

About the Asian Development Bank

ADB’s vision is an Asia and Pacific region free of poverty. Its mission is to help its developing
member countries substantially reduce poverty and improve the quality of life of their
people. Despite the region’s many successes, it remains home to two-thirds of the world’s
poor: 1.8 billion people who live on less than $2 a day, with 903 million struggling on
less than $1.25 a day. ADB is committed to reducing poverty through inclusive economic
growth, environmentally sustainable growth, and regional integration.
Based in Manila, ADB is owned by 67 members, including 48 from the region. Its
main instruments for helping its developing member countries are policy dialogue, loans,
equity investments, guarantees, grants, and technical assistance.

Asian Development Bank

6 ADB Avenue, Mandaluyong City
1550 Metro Manila, Philippines
www.adb.org/economics < 0113 2083 >
ISSN: 1655-5252
Publication Stock No. WPS113208 Printed in the Philippines