Full text of DOI:10.1111/1467-7679.00165

Development Policy Review, 2002, 20 (2): 191-202
Does FDI Guarantee the Stability of
International Capital Flows? Evidence from
Malaysia
Graham Bird and Ramkishen S. Rajan∗
The conventional wisdom is that crises are largely due to swings in short-
term capital. Economies that finance their current account deficits mainly
via foreign direct investment (FDI) are therefore seen as being less
susceptible to a crisis. The analysis in this article, backed up by some
empirical evidence drawn from Malaysia, challenges the casual
presumption that the switch towards FDI alone will automatically imply
that extreme capital instability will become a thing of the past.
The early literature on currency crises focused on the question of the precise timing of a
speculative attack that was inevitable because of deteriorating macro fundamentals.
More recently it has concentrated on the issue of vulnerability as opposed to the
inevitability of a crisis. As such, while the initial focus was on the size and sustainability
of the current account deficit (i.e. what and how much is financed), the recent literature
emphasises the way in which the current account imbalance is financed.1
A key aspect of vulnerability that has been stressed in the literature is the
composition of international capital flows. The conventional wisdom is that crises are
largely due to swings in short-term capital (mainly bank loans in the case of East Asia).
Hence economies that finance their current account deficits mainly via foreign direct
investment (FDI) are seen as being less susceptible to a crisis. The spate of financial
crises in emerging economies in the 1990s, coinciding as they have done with increased
cross-border flows of capital, motivates our interest in examining the nexus between
crises and the composition of capital flows. In particular, we examine whether reliance
on FDI does in fact guarantee stability.
The layout of the article is as follows. The next section discusses the conventional
wisdom more fully and, with this in mind, examines trends in private capital flows to
developing countries in the 1990s. The following section examines external financing in
East Asia, focusing in particular on Malaysia, where a crisis occurred in spite of the fact
that there were low levels of short-term external debt, with FDI accounting for the bulk
of capital inflows (on average). The Malaysian case provides some reason to question
the conventional wisdom. Accordingly, the fourth section reconsiders it critically, and
concludes that the issue of crisis vulnerability is not simply a matter of the composition
∗
Surrey Centre for International Economic Studies, University of Surrey, UK, and School of Economics,
University of Adelaide, Australia and Institute of Southeast Asian Studies (ISEAS), Singapore,
respectively. They gratefully acknowledge research assistance by Senada Nukic and Regan Engelhardt.
1. More to the point, our focus is on the newer liquidity-based crisis models. While there is also a recent
strand of literature devoted to solvency issues (see Rajan, forthcoming) most of these insolvency models
draw inspiration from the original (first-generation) Krugman (1979) model.
 Overseas Development Institute, 2002.
Published by Blackwell Publishers, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.

192 Graham Bird and Ramkishen S. Rajan
of flows; care is needed before recommending policies to seek to bias capital flows
towards FDI. The final section offers a few concluding remarks.
Stability and the composition of capital flows: the
conventional wisdom
Received wisdom linking the composition of international capital flows to economic
instability and financial crisis is quite straightforward. It argues that short-term inflows
(or ‘hot money’) can be easily reversed, while longer-term flows (in the form of long-
maturity bonds and loans and especially FDI) cannot. Movements of hot money are
seen as being dominated by interest-rate differences and expected exchange-rate
changes which can alter rapidly, thus leading to capital volatility, while FDI is
determined by long-term fundamental economic characteristics which are more stable.
Indeed, FDI is often presented as being relatively irreversible in the short run. Since it
enhances the productive capacity of the host country, it produces the revenue stream
necessary to cover future capital outflows.2 But does the evidence confirm the greater
stability of FDI over other types of capital flows?
The empirical evidence: an initial look
Table 1 provides World Bank (1997) data on net private capital flows to developing
countries over the period 1990-98. Overall private flows increased about six-fold from
around US$43 billion in 1990 to about $304 billion in 1997 before declining sharply the
Table 1: Net long-term resource flows to developing countries,
1990-97 ($bn)
1990 1991 1992 1993 1994 1995 1996 1997 Average
Official development finance
Grants
Loans
56.4 62.7 53.8 53.6 45.5 54.0 34.7 44.2
29.2 35.1 30.5 28.4 32.7 32.6 29.2 25.1
50.6
30.4
20.3
6.6
27.2 27.6 23.3 25.1 12.6 21.4
11.6 13.3 11.1 10.0 2.5 10.0
5.4 19.2
-7.2 1.8
Bilateral
Multilateral
Private flows
Debt
15.6 14.4 12.2 15.2 10.4 11.3 12.6 17.4
41.9 53.6 90.1 54.6 160.6 189.1 246.9 256.0
15.0 13.5 33.8 44.0 41.1 55.1 82.2 103.2
13.6
149.1
48.5
17.1
24.8
6.6
Commercial banks
Bonds
Others
3.8
3.4 13.1
7.4 8.3 31.8 27.5 23.8 45.7 53.8
2.7 12.4 9.4 4.7 2.0 2.3 8.3
8.9 101.5 119.0 120.4
7.2 11.0 45.0 32.6 32.5 45.8 32.5
98.3 116.3 143.9 208.1 206.2 243.1 281.6 300.3
2.8
8.9 29.3 34.2
4.1
0.1
11.1
Foreign direct investment
Portfolio Equity
Total
23.7 32.9 45.3 65.6
3.2
74.4
26.2
199.7
Notes: Developing countries are defined as low- and middle-income countries with 1995 per capita incomes
of less than $765 (low) and $9835 (middle).
Source: World Bank (1997).
2. The World Bank (1998, 1999b) has summarised the many benefits of FDI for host country economic
growth via technology transfer, crowding in of domestic investment, and the like. It notes that these
growth-inducing benefits are maximised when accompanied by sound domestic policies and greater
openness. These issues are not presented here as our focus is on capital account reversal (liquidity) as
opposed to issues of resource allocation.

Does FDI Guarantee the Stability of International Capital Flows?
193
next three years, owing to the East Asia-induced turmoil in the global market (World
Bank, 2002). In the remainder of this sub-section we concentrate on trends and patterns
during 1990-97, returning to the crisis period in the following section.
While the World Bank provides only annual data on capital flows, an indication of
the instability of the various forms of private capital flows may be derived by
computing coefficients of variation or CVs (Table 2). Interestingly, FDI flows have the
highest CVs, while portfolio flows have the lowest, suggesting superficially that FDI
flows are the most and portfolio flows the least variable. However, this conclusion is
misleading, as CVs do not take the trend into account. Thus, while the CV is larger for
FDI than for other types of capital flows, there is also a consistent (and predictable)
upward trend for FDI. Strictly speaking, what is needed is a measure of variation around
this trend rather than around the average. With insufficient data points to identify a
definite trend, a more useful indicator is perhaps the number of consecutive years over
which there are positive or negative changes, without a change in direction.
Table 2: Relative variability of various components of private flows
to developing countries, 1990-97
Private
flows
Debt
Commercial
bank lending
Bonds
FDI
Portfolio
equity
Vara
CVb
6692.6
44.9
972.8
20.1
238.5
14.0
360.0
14.5
1459.0
19.6
282.7
10.8
Notes: a) Var – variance; b) CV – coefficient of variation.
Source: Calculated from data in Table 1.
For commercial bank loans, there were consecutive annual directional changes in
each year between 1990 and 1993 inclusive. Although bank lending then increased
persistently between 1993 and 1997 with no further directional changes, the rate of
change varied, with rapid expansions in 1993-4 and 1994-5 being followed by much
more modest increases in 1995-6 and 1996-7 (in both percentage and absolute terms).
The pattern for bonds is a little different, with directional changes occurring only in
1992-3 and 1995-6. Bond flows increased sharply in 1992-3 when commercial bank
lending was declining, and fell in 1993-94 when bank lending increased. ‘Other debt
flows’ show four changes of direction, with these being fairly evenly spread over 1990-
97.
Portfolio equity flows exhibit three directional changes over 1990-97, although
these all occurred in the period between 1993 and 1997 when bank lending increased
persistently. Over 1990-93 portfolio investment persistently increased, although again
the rate of increase varied, with a particularly rapid increase occurring between 1992
and 1993. While not shown by the data, it should be noted that portfolio investment
comprises some relatively stable elements, such as investments by life insurance
companies and pension funds, along with highly unstable investments such as country
funds and mutual funds. Only FDI showed no directional change throughout the entire
1990-97 period. Moreover, as noted, the rise in FDI took place at a fairly persistent rate,
and on average constituted about half of all private flows. IMF data (IMF, 1998) also
show short-term flows exhibiting the greatest degree of variability.

194 Graham Bird and Ramkishen S. Rajan
‘Hot money’ and financial crises: theoretical background
At first sight, the foregoing data appear to be largely consistent with the conventional
wisdom; direct investment has been the most resilient form of external financing.3
Chuhan et al. (1996), Sarno and Taylor (1999) and the World Bank (1999a) reach a
similar conclusion. Empirical analysis suggests that emerging economies which are
most prone to currency crashes tend to have a relatively smaller share of FDI in total
capital inflows and a relatively higher share of short-term external debt. Using probit
analysis Frankel and Rose (1996) see the probability of a currency crisis as a function of
the stock of FDI and non-FDI liabilities. Based on a set of over 100 emerging
economies for the period 1971-92, they find that a low ratio of FDI to debt is linked to a
greater likelihood of a currency crisis. More specifically, a decline in FDI inflows by
1% of external debt is associated with an increase in the probability of crisis by 0.3%.4
While Frankel and Rose do not find overall indebtedness or the share of short-term debt
to have any statistical effects on the probability of crisis, other recent studies have
suggested that short-term indebtedness is a robust predictor of financial crises (Rodrik
and Velasco, 1999; World Bank, 1999a).
There are models that conveniently explain the volatility of short-term capital
flows, covering both bank lending and portfolio flows (Chang and Velasco, 1998, 1999;
Goldfajn and Valder, 1997; Bikhchandani and Sharma, 2000). However, the essence of
these models is that a relatively small initial loss of confidence can translate quickly
into panic and a mass exodus of funds, especially when international reserves fall below
a threshold where they become insufficient to cover short-term liabilities. The
conventional wisdom is that it is these short-term flows that are highly liquid and
mobile and therefore make a country vulnerable to crisis.
It is easy to see how the above theory, combined with the empirical evidence for
developing countries, has resulted in the conventional wisdom that switching from
short-term to long-term capital flows may reduce the probability of currency crises. But
is the conventional wisdom unassailable? Is there any empirical evidence that runs
counter to it, and if there is, can this be explained? We now turn to examine in more
detail what happened to capital flows in East Asia during the 1990s.
Capital flows in East Asia in the 1990s
Boom and bust in East Asia
Data drawn from the IMF’s World Economic Outlook, for the early 1990s, show that net
private capital inflows to the Asia-5 economies were positive and exceeded the
corresponding current account deficit, resulting in a sustained accumulation of
international reserves. As revealed in Table 3, this accumulation was particularly high
in Thailand, which, along with Malaysia and Indonesia, was among the ten largest
3. Use of net data almost certainly serves to understate the de facto degree of capital volatility.
4. Hausmann and Fernández-Arias (2000) confirm the Frankel-Rose result but show that it is not robust when
extended to industrial countries. The authors note that these results may be because industrial countries
have a much larger stock of non-FDI liabilities than do developing countries and have a lower frequency
of crisis.

Does FDI Guarantee the Stability of International Capital Flows?
195
emerging market recipients of net private capital flows (Lopez-Mejia, 1999; World
Bank, 1997). On average, the ‘other net investment’, which includes short-term bank
flows, constituted a much higher share of overall capital flows in the Asia-5 economies
than the other Asian economies – about 75% in the case of Thailand, the ‘trigger’
country.
Table 3: Asia-5: net capital flows, 1990-97 (% of GDP)
1991 1992 1993 1994 1995 1996 Simple Averageb
1997c
Indonesia
Private capital flows
4.6
1.2
0.0
3.5
1.1
-2.4
2.5
1.2
0.0
1.4
1.1
-3.0
3.1
1.2
1.1
0.7
0.9
-1.3
3.9
1.4
.06
7.9
0.1
0.4
6.2
2.3
0.7
6.3
2.8
0.8
5.1
1.7
0.50
3.0
1.6
2.0
-0.4
0.1
1.0
1.8
Direct investment
Portfolio investment
Other investment
Official flows
3.1
2.7
-0.2
-0.7
-0.7
-2.3
0.7
-1.7
Change in reservesa
Malaysia
Private capital flows
Direct investment
Portfolio investment
Other investment
Official flows
11.2 15.1 17.4
1.5
5.7
0.0
-4.2
0.2
4.3
8.8
4.8
0.0
4.1
-0.1
2.0
9.6
5.1
0.0
10.2
7.2
0.0
4.7
5.3
0.0
-0.6
-0.1
3.6
8.3
0.0
2.9
0.4
8.9
0.0
6.2
7.8
0.0
9.7
4.5
2.9
-0.1
-0.6
-0.1
-2.5
0.0
-5.1
Change in reservesa
-2.6 -11.3 -17.7
Philippines
Private capital flows
Direct investment
Portfolio investment
Other investment
Official flows
1.6
2.0
0.3
0.2
3.3
-2.3
2.0
1.3
0.1
0.6
1.9
-1.5
2.6
1.6
-0.1
1.1
2.3
-1.1
5.0
2.0
0.4
2.5
0.8
-1.9
4.6
1.8
0.3
2.4
1.4
-0.9
9.8
1.6
-0.2
8.5
0.
4.1
1.8
0.2
2.1
2.0
-1.8
0.5
1.4
-5.3
4.5
0.8
2.1
Change in reservesa
-4.8
Thailand
Private capital flows
Direct investment
Portfolio investment
Other investment
Official flows
10.7
1.5
0.0
8.7
1.4
0.5
6.8
0.1
-2.8
8.4
1.1
3.2
4.1
0.2
-3.2
8.6 12.7
9.3
0.9
0.6
7.7
0.7
-1.2
11.5
1.6
1.4
-10.9
1.3
0.4
0.7
0.9
0.7
1.9
9.2
7.0 10.0
8.5
-12.6
1.1
-4.3
0.1
0.7
0.1
-4.3
4.9
9.7
Change in reservesa
-3.0
-4.4
Korea
Private capital flows
Direct investment
Portfolio investment
Other investment
Official flows
10.7
1.5
0.0
8.7
1.4
0.5
6.8
0.1
-2.8
8.4
1.1
3.2
4.1
0.2
-3.2
8.6 12.7
9.3
0.9
0.6
7.7
0.7
-1.2
11.5
1.6
1.4
-10.9
1.3
0.4
0.7
0.9
0.7
1.9
9.2
7.0 10.0
8.5
-12.6
1.1
-4.3
0.1
0.7
0.1
-4.3
4.9
9.7
Change in reservesa
-3.0
-4.4
Notes: a) Minus sign denotes a rise and vice versa; b) 1989 to 1996; c) estimates.
Source: IMF (1997).

196 Graham Bird and Ramkishen S. Rajan
There are by now many comprehensive discussions of the East Asian crisis, and we
do not intend going over well-travelled terrain.5 What is important for our present
purpose is that the collapse of the baht and other regional currencies was mainly due to
reversals of capital flows from the banking sector rather than portfolio equity
investments. Indeed, balance-of-payments data from the Institute of International
Finance (1999) reveal that the Asia-5 economies most afflicted by the regional crisis
saw a sharp reversal in net private capital flows of almost $130 billion between 1996
and 1998. This reversal primarily involved net (short-term) lending by foreign
commercial banks, which averaged about $60 billion in inflows between 1995 and
1996, but turned into a net outflow of about $30 billion over the following two years as
international banks became unwilling to roll over existing short-term debts to the
region. BIS data for 1999 also show that international bank lending to the Asia-5
economies remained buoyant at almost $50 billion in the first half of 1997, but swung to
-$40 billion in the third quarter of 1997, and then averaged close to -$100 billion for the
three following quarters.
This sudden reversal in bank lending is often portrayed as strong evidence of a
bank panic model (Chang and Velasco, 1998; Radelet and Sachs, 1998a, 1998b).6 A
much less highlighted aspect of the sharp contraction in private market financing is the
decline in portfolio flows during 1997-8, following the initial bank panic, as investors
too tried to scale down their regional financial exposures (‘flight to quality’), although
this is consistent with the Calvo-Mendoza crisis model (Calvo and Mendoza, 2000;
Rajan and Siregar, 2002).
In contrast, FDI flows have remained remarkably stable during the crisis period.7
The World Bank (1999b) has noted that the resilience in FDI was despite a fall in
market size (in foreign currency terms) and reduced immediate growth prospects, and
may be attributed to three factors: the sharp currency depreciations that reduced
production costs and asset values in foreign currencies; the decline in domestic asset
prices; and the greater potential for corporate restructuring in the crisis-hit economies.
Capital outflows from Malaysia, 1997-8
Following the devaluation of the Thai baht on 2 July 1997, Malaysia succumbed to the
crisis two weeks later on 14 July. The Malaysian monetary authorities allowed the
currency to depreciate despite having reserves of over US$ 20 billion.8 The ringgit
5. For detailed accounts of the East Asian crisis, see IMF (1997, 1998), Bird and Rajan (2001), Berg (1999),
Corsetti et al. (1999), Radelet and Sachs, (1999a,b), Rajan (1999) and the World Bank (1998, 1999a).
6. Of course, these ex-post swings in bank flows are only necessary and not sufficient evidence in support of
a bank panic model. Accordingly, at least in the case of Thailand, Rajan (2002) has provided data on the
foreign asset and liability positions in order to determine its ex-ante vulnerability to an external shock
(such as a devaluation), and then discusses the movements in capital withdrawals from the country
following the shock. Since the scenario of devaluation followed by collapse is closely intertwined with the
important issue of the illiquidity versus the insolvency of domestic financial institutions, this issue is also
examined, as are the consequences of the systemic liquidity crisis post-devaluation. The evidence
presented in its entirety strongly supports a bank panic view. Such a systematic exploration of the data
remains to be done for the other crisis-hit economies.
7. Indonesia was an important exception, FDI having collapsed because of ongoing socio-political
uncertainties (World Bank, 1999b; Rajan and Siregar, 2002). Latin America also shared this experience of
stable FDI flows during a boom and bust period (Hausmann and Fernández-Arias, 2000).
8. Athukorala (2001) notes that they lost US$ 1.5 billion trying to defend the currency.

Does FDI Guarantee the Stability of International Capital Flows?
197
promptly fell from about 2.5 ringgit per US dollar in mid-1997 to almost 5 to the dollar
by early 1998 before eventually being fixed at 3.8 per dollar in September 1998 when
the governemnt imposed capital controls. Growth of real output, which averaged 8.7%
in the first half of the 1990s, contracted by 7.5 percentage points the following year
(Athukorala, 2001).9
Since much of the discussion of the East Asian crisis has lumped the five crisis-hit
economies together, it has often gone unnoticed that Malaysia was an ‘outlier’ in terms
of capital inflows and outflows (Rajan and Siregar, 2002). In contrast to the other crisis-
hit East Asian economies, where the boom and bust was primarily related to reversals in
bank lending, the bulk of inflows to Malaysia prior to the crisis was in the form of FDI,
while outflows were primarily in the form of portfolio flows. Table 4 shows that the
current account, which averaged a deficit of about US$7 billion between 1995 and
1997, turned sharply into a surplus of more than US$ 9 billion in 1998. Since only about
a half of this turnaround was accounted for by a rise in international reserves, the
remainder must have taken the form of capital outflows. Malaysia’s capital account,
which had been in persistent surplus (of around US$5 billion on average) in the
previous three years, recorded a deficit of US$4.7 billion in 1998. While FDI to
Malaysia remained more or less stable during this crisis period, short-term flows, which
amounted to US$4.1 billion in 1996, registered declines of US$4 billion in 1997 and
US$5.5 billion in 1998.
Table 4: Malaysia: balance of payments, 1995-8
1995
1996
1997
1998
(US$bn)
Current account deficit
Capital account balance
Medium- and long-term flows, net
Private sector, net
-8.7
7.0
6.6
-4.9
7.4
5.4
-5.0
1.2
6.8
9.2
-4.7
3.6
4.2
5.1
5.1
2.2
Short-term flows, net
Errors and omissionsa
Overall balance
1.0
4.1
-2.1
2.5
-4.0
-1.6
-3.9
21.7
-5.5
-2.8
4.5
-0.7
-1.8
25.1
Official reserves (end period)
27.7
26.2
(% GDP)
Current account balance
Capital account balance
Net foreign direct investment
Net short-term flows
Overall balance
-10.0
7.2
3.8
1.2
-2.0
-4.9
5.3
3.6
4.1
2.5
-5.1
-0.4
4.0
-4.1
-4.0
-12.9
-10.5
3.0
-7.8
6.3
Note: a) includes portfolio capital.
Source: IMF, World Economic Outlook (various years).
9. For detailed discussions of the causes and consequences of the Thai and Malaysian crises, see Rajan
(2002) and Athukorala (2001), respectively. In the case of Malaysia, while Athukorala emphasises
growing monetary and financial weaknesses pre-crisis and the system’s inherent fragility, Kaminsky and
Reinhart (2000) and Dungey and Martin (2000) point to the importance of its extensive trade and financial
linkages or spillovers (direct and third-country) with Thailand.

198 Graham Bird and Ramkishen S. Rajan
An important detail regarding the dynamics of capital flows to and from Malaysia
during the crisis period needs to be noted. Net portfolio inflows, which averaged about
US$38 million between 1990 and 1996, turned around dramatically into a net outflow
of almost US$325 million in 1997; this was in excess of the cumulative inflows over the
entire period between 1980 and 1996.
Do we need to reconsider the conventional wisdom relating to FDI in the light of
the Malaysian experience?
Reconsidering the conventional view of FDI
Malaysia was affected by the crisis in Thailand despite the fact that the bulk of its
current account deficit had been financed through FDI. This suggests that, unless a
country’s current account deficits are financed almost entirely by FDI, the economy
remains vulnerable to capital reversals. Is it then wise for emerging economies
deliberately to bias capital flows towards FDI while curtailing other forms of capital
flow, particularly bank lending and portfolio investments?
This question may be answered in two ways. First, recent empirical investigations
into the causes of currency crises in emerging economies have raised doubts about the
existence of a direct link between FDI and the probability of currency crisis. For
instance, in a recent study involving 26 emerging economies during the crisis periods
1994 and 1997, Nitithanprapas and Willett (2000) find that low FDI is a robust indicator
of a country’s vulnerability to contagion only if combined with the current account
deficit and the real exchange rate. Thus they conclude that ‘the composite indicator of
current account, FDI, and real exchange rate is a useful indicator of external
vulnerability to financial contagion’ but FDI by itself may not be (Ibid.: 35).
Similarly Bussiere and Mulder (1999) test for the significance of FDI (to GDP
ratio) in the crises in emerging economies in 1997 and 1998. They find that the variable
was not statistically significant at the 10% level, although it had the correct sign,
suggesting to them ‘only a limited reduction in vulnerability as a result of FDI financing
of the deficit’ (Ibid.: 17).
Second, a potential criticism of the conventional view regarding differing degrees
of stability of various capital flows is that it fails to take account of the complex
interactions between FDI and other flows. Examining each flow individually,
particularly during short periods of time (such as year-to-year variations), may at best
be an unreliable indicator of the degree of risk of various classes of flow, and at worst
could be highly misleading.10 Capital that flows in as FDI may well flow out under
another guise.
Hausmann and Fernández-Arias (2000) have recently found that the standard
deviation of FDI is not very different from that of total net flows, especially in the case
of Latin America, and that the volatility of FDI itself has been on the rise. Furthermore,
while the overall share of FDI in capital flows had been rising in many developing
countries during the 1990s, this failed to make the overall capital account more stable.
10. Claessens et al. (1995) computed statistical measures of volatility for a group of 10 developed and
developing countries (France, Germany, Japan, Great Britain, and the United States; Argentina, Brazil,
Indonesia, Korea, and Mexico) and failed to unearth any systematic pattern in the volatilities of the various
types of capital flows.

Does FDI Guarantee the Stability of International Capital Flows?
199
Even though FDI has become the single largest component of capital flows for
developing countries, this has not been discernibly matched by declining volatility in
the international capital market or a reduced incidence of financial crises. This is
consistent with Dooley et al. (1994) who find that a high level of FDI is associated with
greater and not lower variability in capital flows.
What could explain this? Contrary to popular belief, FDI is not ‘bolted down’,
although the physical assets it finances are. Foreign investors can use the physical assets
as collateral to obtain a loan from banks and can then place the funds abroad. In other
words, the foreign investor may hedge the firm’s FDI exposure by borrowing
domestically and taking short-term capital out of the country. Hence, a firm may be
doing one thing with its assets and a completely different thing with the manner in
which it finances them. This appears consistent with the Malaysian capital flows
reported above, where portfolio outflows in 1997 outweighed the cumulative inflows
between 1980 and 1996. Apparently the portfolio outflows must have entered via some
other account (such as FDI or bank loans).
The World Bank (1999b: 54) has also recently cautioned against the presumption
that FDI necessarily implies greater financial stability by pointing out that:
(d)uring a crisis, ‘direct investors’ may contribute…..to capital withdrawals by
accelerating profit remittances or reducing the liabilities of affiliates toward their
mother companies. While these are non-FDI flows, they result from decisions by
foreign investors. It is difficult to determine the extent to which foreigners involved in
direct investment took out capital through non-FDI flows during the financial crisis
because the data are available only with considerable delay. In addition to long-term
determinants, FDI is affected by many short-run factors….., such as movements in host
countries’ exchange rates and asset prices and growth prospects, as well as the
economic environment in FDI source countries.
The IMF (1998) has similarly drawn attention to the fact that the distinction
between portfolio and FDI flows in the balance of payments can be somewhat arbitrary
and that the proportion of FDI flows in aggregate capital flows may be overstated.11
Small differences in equity ownership, which may serve to reclassify financial flows,
are unlikely to represent substantially different investment horizons.
The preceding analysis suggests that FDI may be negatively correlated with other
capital flows, particularly portfolio flows. A negative correlation is also consistent with
the famous Modigliani-Miller theorem which argues that various forms of capital flow
are just alternative ways of financing a particular activity, and that, with perfect capital
markets, the manner in which projects are financed is irrelevant. Thus, if one
component of capital flow increases, ceteris paribus, another must fall.12 However, Das
Gupta and Ratha (2000) claim that ‘FDI adds to liquidity in the short term and improves
the medium-term outlook of a particular sector or the economy as a whole’. On these
grounds, we should expect FDI and portfolio flows to be positively related.
11. Conversely FDI may actually be understated in some cases. For instance, some part of the recorded short-
term borrowing by Thailand was actually FDI and intra-banking transfers (Ostry, 1997).
12. Needless to say, the ‘real world’ does not fit the perfect capital markets assumption, with taxation,
information asymmetry and other frictions being prevalent, as discussed by Hausmann and Fernández-
Arias (2000).

200 Graham Bird and Ramkishen S. Rajan
The nature of the relationship between FDI and other capital flows is therefore an
empirical issue. In one of the few direct tests of the nexus between the various forms of
capital flow, Bosworth and Collins (1999) examine the degree of correlation between
FDI, portfolio investment, and loans. They use a data set of 58 emerging economies
over the period 1978-95. The correlations (total, cross-country and cross-period) turn
out to be all slightly positive but mostly insignificant.13 When they focus specifically on
a subset of 18 emerging economies (eliminating those with no portfolio capital inflows),
they find a slight increase in the correlation between portfolio capital and other inflows,
including FDI, but none of the coefficients are statistically significant. This may be
taken either as evidence that positive and negative effects offset one another, or that
portfolio and FDI flows react to different factors (Reinhart, 2000).
As further indicative evidence, a simple correlation test between portfolio and FDI
using Malaysian capital flows data from the IMF (annual between 1982 and 1998)
reveals a negative correlation between FDI and portfolio investment (-0.47), which is
significant at the 10% critical level. While these results are certainly not conclusive
(particularly in light of the very limited number of observations), when taken with the
above discussion, they suggest caution in terms of accepting the conventional wisdom
about FDI.
Concluding remarks
Short-term capital volatility has been seen as lying at the heart of recent financial crises.
The policy debate has focused on reducing the instability of short-term capital flows by
controls or by taxation and regulation, and on switching the composition of capital
flows to the longer-term end, particularly in the form of FDI. That a country can reduce
its vulnerability to crisis by increasing the share of FDI in capital inflows has become
the conventional wisdom.
Although at a highly aggregated level there appears to be some empirical
justification for this view, a more detailed examination of the evidence and of the
underlying analytics counsels a degree of caution. Drawing in particular on the case of
Malaysia, which did have a high proportion of FDI relative to capital flows, it may be
seen that changing the composition of capital inflows to the long end provides no
guarantee of financial stability. Indeed, increasing FDI may itself be associated with and
causally connected to increased instability in portfolio flows, implying that the apparent
stability of FDI may be of spurious importance. A potential danger is then that policy
measures designed to encourage FDI may involve not only a distortionary cost but also
little gain in terms of enhanced financial stability. Such a shift could increase rather than
reduce the need for other measures to stabilise short-term capital movements.14
13. The only statistically significant one was that between FDI and loans in the time dimension, which was
just 0.09. This is consistent with the fact that FDI tends to be accompanied by an increase in bank loans.
14. However, Jeanne (2000) argues that it is not clear that short-term debt contracts ought to be discouraged as
they may play a socially advantageous function in reducing agency problems. The World Bank (1999b)
surveys recent literature on short-term debt. Hausmann and Fernández-Arias (2000) find that countries
which are riskier and financially under-developed tend to have lower aggregate private capital inflows but
a higher share of FDI in overall capital inflows.

Does FDI Guarantee the Stability of International Capital Flows?
201
References
Athukorala, P. (2001) Crisis and Recovery in Malaysia: The Role of Capital Control.
Cheltenham: Edward Elgar.
Berg, A. (1999) The Asia Crisis: Causes, Policy Responses, and Outcomes. Working
Paper No.99/138. Washington, DC: IMF.
Bikhchandani, S. and Sharma, S. (2000) Herd Behaviour in Financial Markets: A
Review. Working Paper No.00/48. Washington, DC: IMF.
Bird, G. and Rajan, R (2001) ‘Banks Financial Liberalisation and Financial Crises in
Emerging Markets’, The World Economy 24 (7): 889-910.
Bosworth, B. and Collins, S. (2000) ‘Capital Flows to Developing Economies:
Implications for Saving and Investment’, Brookings Papers on Economic Activity
1: 143-69.
Bussiere, M and Mulder, C. (1999) External Vulnerability in Emerging Market
Economies: How High Liquidity Can Offset Weak Fundamentals and the Effects of
Contagion? Working Paper No.99/88. Washington, DC: IMF.
Calvo, G. A. and Mendoza, E. (2000) ‘Rational Contagion and the Globalization of
Securities Markets’, Journal Of International Economics 51: 79-113.
Chang, R. and Velasco, A. (1998) The Asian Liquidity Crisis. Working Paper No.6796.
Cambridge, MA: National Bureau of Economic Research.
Chang, R. and Velasco, A. (1999) ‘Liquidity Crises in Emerging Markets: Theory and
Policy’, in B. Bernanke and J. Rotemberg (eds), NBER Macroeconomics Annual
1999. Cambridge, MA: MIT Press.
Chuhan, P., Perez-Quiros, G. and Popper, H. (1996) International Capital Flows: Do
Short-term Investment and Direct Investment Differ? Policy Research Working
Paper No.1507. Washington, DC: World Bank.
Claessens, S., Dooley, M. and Warner, A. (1995) ‘Portfolio Capital Flows: Hot or
Cold’, World Bank Economic Review 9: 153-74.
Corsetti, G., Pesenti, P. and Roubini, N. (1999a) ‘What Caused the Asian Currency and
Financial Crisis? Part I: Macroeconomic Overview’, Japan And The World
Economy 11: 305-73.
Dasgupta, D. and Ratha, D. (2000) What Factors Appear to Drive Private Capital
Flows to Developing Countries? And How Does Official Lending Respond? Policy
Research Working Papers No.2392. Washington, DC: World Bank.
Dooley, M., Fernández-Arias, E. and Kletzer, K. (1994) Recent Private Capital Inflows
to Developing Countries: Is the Debt Crisis History? Working Paper No.4792.
Cambridge, MA: National Bureau of Economic Research.
Dungey, M. and Martin, V. (2000) ‘Measuring Contagion in the East Asian Currency
Crisis’. (October) (mimeo).
Frankel, J. and Rose, A. (1996) ‘Currency Crisis in Emerging Markets: Empirical
Indicators’, Journal of International Economics 41: 351-68.
Goldfajn, I. and Valdes, R. (1997) Capital Flows and the Twin Crises: The Role of
Liquidity. Working Paper No.97/87. Washington, DC: IMF.
Hausmann, R. and Fernández-Arias, E. (2000) Is FDI a Safer Form of Financing?
Working Paper No.416. Washington, DC: Inter-American Development Bank.
Institute of International Finance (1999) ‘Capital Flows to Emerging Market
Economies’. Washington, DC: IIF (April).

202 Graham Bird and Ramkishen S. Rajan
IMF (1997) World Economic Outlook: Interim Assessment. Washington, DC: IMF.
IMF (1998) World Economic Outlook 1998. Washington, DC: IMF.
IMF (1999) International Capital Markets: Developments, Prospects, and Key Policy
Issues. Washington, DC: IMF.
Jeanne, O. (2000) ‘Foreign Currency Debt and the Global Financial Architecture’,
European Economic Review 44: 719-27.
Kaminisky, G. and Reinhart, C. (2000) ‘On Crises, Contagion, and Confusion’, Journal
of International Economics 51: 145-68.
Krugman, P. (1979) ‘A Model of Balance of Payments Crises’, Journal of Money
Credit and Banking 11: 311-28.
Lopez-Mejia, A. (1999) Large Capital Flows: A Survey of the Causes, Consequences,
and Policy Responses. Working Paper No.99/17. Washington, DC: IMF.
Nitithanprapas, E and Willett, T. (2000) A Currency Crises Model That Works: A
Payments Disequilibrium Approach. Claremont Working Papers No.2000-25.
Claremont, CA: Claremont Colleges.
Ostry, J. (1997) Current Account Imbalances in ASEAN Countries – Are They a
Problem? Working Paper 97/51. Washington, DC: IMF.
Radelet, S. and Sachs, J. (1998a) The Onset of the East Asian Financial Crisis. Working
Paper No.6680. Cambridge, MA: National Bureau of Economic Research.
Radelet, S. and Sachs, J. (1998b) ‘The East Asian Financial Crisis: Diagnosis,
Remedies, Prospects’, Brookings Papers on Economic Activity 1: 1-74.
Rajan, R. (1999) ‘Economic Collapse in Southeast Asia’, Policy Study. Claremont, CA:
The Lowe Institute of Political Economy (July).
Rajan, R. (2002) (Ir)relevance of Currency Crises Theory to the Devaluation and
Collapse of the Thai Baht. Princeton Studies in International Economics No.88.
Princeton, NJ: International Economics Section, Princeton University.
Rajan, R. and R. Siregar (2002) ‘Private Capital Flows in East Asia: Boom, Bust and
Beyond’, in G. de Brouwer (ed.), Financial Markets and Policies in East Asia.
London: Routledge.
Reinhart, C. (2000) ‘Comments on Bosworth and Collins’, Brookings Papers on
Economic Activity 1: 170-79.
Rodrik, D. and Velasco, A. (1999) Short-term Capital Flows. Working Paper No.7364.
Cambridge, MA: National Bureau of Economic Research.
Sarno, L. and Taylor, M. (1999) ‘Hot Money, Accounting Labels and the Permanence
of Capital Flows to Developing Countries: An Empirical Investigation’, Journal of
Development Economics 59: 337-64.
World Bank (1997) Private Capital Flows to Developing Countries: The Road to
Financial Integration. New York: Oxford University Press.
World Bank (1998) Global Development Finance 1998. New York: Oxford University
Press.
World Bank (1999a) Global Economic Prospects and the Developing Countries. New
York: Oxford University Press.
World Bank (1999b) Global Development Finance 1999. New York: Oxford University
Press.
World Bank (2002) Global Development Finance 1999. New York: Oxford University
Press.