World Development Vol. 59, pp.

1628, 2014
2014 Elsevier Ltd. All rights reserved.
0305-750X/$ - see front matter
www.elsevier.com/locate/worlddev

http://dx.doi.org/10.1016/j.worlddev.2014.01.030

Foreign Portfolio Investment Flows to India:

Determinants and Analysis
REETIKA GARG and PAMI DUA *
University of Delhi, India
Summary. This paper analyzes the macroeconomic determinants of portfolio ows to India and nds that lower exchange rate volatility and greater risk diversication opportunities are conducive to portfolio ows. However, higher equity returns of other emerging
markets discourage these ows. Other conventional determinants of portfolio ows are domestic equity performance, exchange rate,
interest rate dierential and domestic output growth. An analysis of disaggregated portfolio ows shows that determinants of FIIs
are similar to aggregate portfolio ows, while ADR/GDRs are signicantly inuenced only by domestic equity returns, exchange rate,
domestic output growth, and foreign output growth.
2014 Elsevier Ltd. All rights reserved.
Key words FPI, FII, ADR/GDRs, ARDL model

1. INTRODUCTION

that these ows can expose the countries to new macroeconomic challenges.
India has not remained untouched by the developments in
the global nancial markets due to greater linkages of the Indian markets with the international markets. During the Asian
crisis as well as during the recent sub-prime crisis, overall balance of payments deteriorated due to massive reversal of portfolio ows. Table 1 indicates that in 2008, portfolio equity
ows to the Indian economy faced the sharpest reversal followed by the largest bounce back in 2009 compared to other
developing countries. In 2010, while ows to other economies
declined or increased marginally, portfolio equity ows to India increased by almost 90% compared to 2009, which indicates the condence of the investor in the Indian economy
that can be attributed to strong domestic fundamentals.
In the recent uncertain global scenario, it is important to
understand the factors that drive portfolio ows, to facilitate
ecient management of these ows in order to avoid any
imbalances arising out of large inows beyond the absorptive
capacity of the economy or sudden reversal of nancial capital
leading to a situation of capital crunch.
The literature that analyzes the determinants of portfolio
ows to developing countries has debated on the relative signicance of domestic and external factors. Studies by Calvo,
Leiderman, and Reinhart (1993, 1996), Taylor and Sarno
(1997), Fernaindez-Arias (1996) and Kim (2000), Byrne and
Fiess (2011) have emphasized that global factors like decline
in interest rates and slowdown in growth of industrialized
countries, have pushed capital to developing economies. On
the other hand Bohn and Tesar (1996), The World Bank
(1997), Mody, Taylor, and Kim (2001) and Felices and
Orskaug (2008), nd that domestic factors like equity index,
sucient availability of domestic reserves, and country
creditworthiness have attracted portfolio ows to developing

According to international nance theory, foreign portfolio

investment (FPI) ows are an inevitable outcome of investors
wanting to invest across countries in order to diversify the risk
of their portfolio and achieve higher returns. Some of the studies that have documented the benets of diversication across
countries include Grubel (1968), Levy and Sarnat (1970),
Solnik (1974a), Grauer and Hakansson (1987), Harvey
(1991), and De Santis and Gerard (1997). From the point of
view of the host country, especially the developing countries,
portfolio ows are considered to play a pivotal role in bridging
the savinginvestment gap and providing the much needed
foreign exchange to nance current account decit. The developing countries across the globe have been making conscious
eorts to attract foreign nancial capital which provides an
impetus to economic growth and nancial market development in the host country. DellAriccia et al. (2008) and
Obstfeld (2009) review the literature related to the benets
of nancial ows from the host country perspective.
The growing removal of restrictions on the trading of international nancial assets has led to a surge in the ow of nancial capital across the globe in the past two decades. The
investment by foreign investors is mainly geared toward the
fast growing developing economies, including India, which
provide protable investment opportunities. Table 1 indicates
that portfolio equity ows to developing countries increased
ve times from US $ 14 billion in 2000 to US $67 billion in
2005 and almost nine times to US$ 128 billion in 2010. Of
all the developing countries the major recipients of portfolio
equity ows have been China, Brazil, India, Russia, and South
Africa out of which China, India, and Brazil account for almost 70% of the total portfolio equity ows to all developing
countries (Figure 1).
Notwithstanding the benecial impacts of portfolio ows to
the investor as well as to the host country, these ows have
also been a source of concern. The foreign investor has to take
into consideration country and currency risk in addition to
other factors compared to investing in the home country.
From the host country perspective, portfolio ows are prone
to reversals and volatility. This is evident from the historical
and recent nancial crises that have brought into focus the fact

* We thank the anonymous referees for their valuable comments and suggestions. We are also grateful for the comments and suggestions received
on preliminary versions of this paper presented at the 48th Annual Conference of The Indian Econometric Society (TIES) and The Asian Meeting
of the Econometric Society (AMES). Final revision accepted: January 25,
2014.
16

FOREIGN PORTFOLIO INVESTMENT FLOWS TO INDIA: DETERMINANTS AND ANALYSIS

17

Table 1. Portfolio Equity Flows to Developing Countries (US $ Billion)

Years

All Developing Countries

China

Brazil

South Africa

Russian Federation

India

Mexico

Thailand

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010

13.95
5.60
5.80
24.30
39.90
67.47
107.70
132.95
53.36
108.75
128.41

6.91
0.80
2.20
7.70
10.90
20.35
42.86
18.51
8.72
28.16
31.36

3.07
2.50
2.00
3.00
2.10
6.45
7.72
26.22
7.56
37.07
37.68

4.17
1.00
0.40
0.70
6.70
7.23
14.96
8.67
4.71
9.36
5.83

0.15
0.50
2.60
0.40
0.20
0.10
6.48
18.67
15.00
3.37
4.80

2.48
2.90
1.00
8.20
9.10
12.15
9.51
32.86
15.03
21.11
39.97

0.45
0.20
0.10
0.10
2.50
3.35
2.80
0.48
3.50
4.17
0.64

0.90
0.40
0.50
1.80
1.30
5.12
5.24
4.27
3.80
1.33
3.43

Source: Global Development Finance, various issues.

30.69
30.48
31.84

China
21.37
20.44

India

Brazil

14.15
8.86
8.40

South Africa

Thailand

Russian Federation
0.00

3.52
3.17
2.14
1.75

26.44

20.68
21.87

2000-2010
2005-2010
2000-2004

11.36

5.47

4.30

5.00

10.00

15.00

20.00

25.00

30.00

35.00

Figure 1. Share of portfolio equity ows in total ows to all developing countries (percentages). Source: Calculations based on data from various issues of
Global Development Finance.

countries. Another set of studies in literature advocate complimentarity between domestic and external factors. These include Chuhan, Claessens, and Mamingi (1993), Montiel and
Reinhart (1999) and De Vita and Kyaw (2007) among others.
To explore further, this study examines the macroeconomic
determinants of portfolio ows to India. The determinants of
disaggregated components of portfolio ows i.e., Foreign
Institutional Investment ows (FII) and American/Global
Depository Receipts (ADR/GDR) are also analyzed, in order
to assess whether dierent components of portfolio ows are
driven by similar or dierent factors.
This paper makes an important contribution to the literature. Firstly, most of the literature that analyzes the determinants of portfolio ows has concentrated on the FII
component only. ADR/GDR ows have not received much
attention despite the fact that it has increasingly being used
to raise foreign capital. This study will examine the macroeconomic determinants of not only FII but also ADR/GDR ows
to India in order to fulll the existing gap in the literature.
Furthermore, this study examines a wider set of potential
determinants of FII ows to India compared to other studies
pertaining to the Indian economy such as Chakrabarti (2001),
Kaur and Dhillon (2010), Kumar (2011), Srinivasan and Kalaivani (2013). While the study by Gordon and Gupta (2003)
includes a wide range of determinants of portfolio ows, the
Ordinary Least Squares (OLS) methodology is used that
may yield biased and inconsistent estimates if the regressors

are endogenous. This study follows the Autoregressive

Distributed Lag (ARDL) approach to cointegration for estimating the long-run coecients which overcomes such problems. The long-run coecients are unbiased and the t-tests
are also valid, even if the regressors included in the specication are endogenous (Harris & Sollis, 2003).
The following section discusses the trends along with the
policy reforms related to portfolio ows in India. Section 3
describes the potential factors that drive portfolio ows. Section 4 discusses the empirical model, data, and methodology,
used for estimation. Section 5 discusses the econometric results
and Section 6 concludes.
2. TRENDS IN PORTFOLIO FLOWS AND POLICY
REFORMS
In 1992, the Indian government allowed FIIs to invest in the
nancial markets which required registration of FIIs with the
Securities and Exchange Board of India (SEBI). SEBI prescribes certain norms which are to be followed by the FIIs
for registration that includes compliance with Foreign Exchange Management Act 1999 for which they need permission
from Reserve Bank of India (RBI) and payment of registration fees. Overtime, the regulatory measures of SEBI have become liberal, which mainly included procedural relaxations
related to entry and exit of FIIs, raising the ceilings on

18

WORLD DEVELOPMENT

investment in companies and greater coverage related to types

of funds that could invest as FIIs and the instruments in which
FIIs can invest. Initially only broad based funds were allowed
to invest in the Indian markets, but overtime foreign rms are
also allowed and even the denition of broad based funds has
also been widened. Presently, the insurance companies, banks,
hedge funds, mutual funds, asset management companies and
pension funds form the majority of FIIs investing in India. 1
Initially FIIs were allowed to invest only in the listed securities
of Indian companies, but overtime they have been allowed to
invest in unlisted securities, government securities, commercial
paper, and derivatives.
With regard to ADR/GDRs, Indian companies were allowed to raise capital by issuing depository receipts in the
world nancial markets in 1992. Initially in the 1990s Indian
rms refrained from listing in the US stock exchanges because
of the stringent accounting standard requirements of US Generally Accepted Accounting Principles (GAAP) and Securities
and Exchange Commission (SEC). The most attractive destination of the Indian rms was London Stock Exchange and
Luxemburg Stock Exchange. However, as observed by Pagano, Roell, and Zechner (2002), poor liquidity in the European
stock exchanges made Indian companies change their preferences and they started listing on US stock exchanges. This
was also because, overtime, Indian companies improved their
accounting standards, reecting greater transparency, which
allowed these rms to list in the US stock markets as well. Initially most of the Indian companies listing in the US, listed on
the National Association of Securities Dealers Automated
Quotation System (NASDAQ) and then later on New York
Stock Exchange (NYSE) as well because of the stricter norms
of NYSE.
In Figure 2, the annual trends in FII investment show that
after 199293, FII ows increased steadily till 199697. In
199798 and 199899, India experienced a massive reversal
in FII investment following the Asian crisis. Although Indian
economy was not directly involved in the crisis, and the intervention by RBI in the currency market as well as imposition of
capital controls was used to insulate the Indian economy from
the crisis (Dua & Sinha, 2007), still the foreign investment

ows to the Indian economy declined, specially the FIIs. However as compared to other East Asian countries, the magnitude
of adverse impact on India was small and short lived.
In 19992000 FII ows to the Indian economy bounced
back to their pre-crisis levels, which reected investor condence in the Indian markets. In 200203 FII ows dipped to
US$ 2.77 billion but in 200304, increased to US$ 10.9 billion.
In 2003, a working group that was set up for streamlining the
procedures of FIIs, which suggested a relaxation in the procedure of taking approval from both SEBI and RBI, to an approval only from SEBI. Moreover, in 2003, currency
hedging was allowed for FIIs without any restrictions.
From 200304 to 200607 FII ows followed a downward
trend. In 200607, in the wake of rising inationary pressures
and heightened liquidity, SEBI along with RBI tightened the
controls on inows and also liberalized outows. Interest rates
on Non-resident Indian (NRI) deposits were decreased, limits
were imposed on external commercial borrowings, restriction
on foreign investment by Indian companies and mutual funds
were eased. In 200708, FII ows surged to US$ 20 billion in
the backdrop of heightened global liquidity and bullish
domestic stock markets. In 200809, it was the huge reversal
in FII ows that created havoc in the Indian nancial markets
and led to a sharp decline in the total portfolio investment
ows to India. However, 200910 saw a massive revival of
FII ows to India.
The trends in ADR/GDRs show that in the initial years
after liberalization, ADR/GDR investment ows were greater
than FII ows received by the country. From 199596 to
200203, India received considerable amount of ADR/GDR
investment compared to FIIs. It was only after 200304 that
the magnitude of ADR/GDR investment seemed to be small
compared to FIIs but this was due to gigantic rise in FII
investment rather than a fall in ADR/GDR investment. The
ADR/GDR ows displayed an increasing trend from 2003
04 onward and reached their historical peak in 200708.
What is also noticeable is that for the period 200203 to
200607, while FII ows declined, ADR/GDR investment
ows increased. It may be possible that ADR/GDRs were
being used to circumvent restrictions on FII ows. The Report

34000

ADR/GDR

FII

Offshore Funds

FPI

US $ Million

24000

14000

4000

-6000

-16000

Figure 2. Trends in Components of FPI ows. Source: Reserve Bank of India, Handbook of Statistics 2013.

FOREIGN PORTFOLIO INVESTMENT FLOWS TO INDIA: DETERMINANTS AND ANALYSIS

of the Working Group on Foreign Investment by Ministry of

Finance, Government of India (2010), mentions that if there
are restrictions on entry and exit in the Indian nancial markets, then the foreign investor may direct its investment to
instruments of the Indian rms that are traded in nancial
markets outside India.
The composition of aggregate portfolio ows (Table 2)
shows that the share of FIIs in has always been higher than
the share of investment ows received through ADR/GDRs
since 199697, except in 200203 and in 200607. However,
ADR/GDRs comprise a signicant proportion of portfolio
ows. Oshore funds have always been a minor portion of
the total portfolio ows, however in the recent years since
200809, India has not been receiving any investment on account of oshore funds.
3. DETERMINANTS OF PORTFOLIO FLOWS
The discussion in the previous section indicates that Indian
government has followed a policy of gradual liberalization of
restrictions on portfolio ows to India. This has been done
mainly through liberalizing the investment limits, relaxing
the eligibility requirement of the rms that could invest in India and opening up greater number of instruments for foreign
investment. This implies a reduction in barriers to capital ows
and gradual increase in the degree of integration of the Indian
nancial markets with the global markets. In this context the
studies on asset pricing which are relevant for the Indian economy are those in the vein of Bekaert and Harvey (1995, 2003),
Stulz (1999), Carrieri, Errunza, and Hogan (2007) and Arouri
and Foulquier (2012) among others that discuss asset pricing,
assuming time varying integration of nancial markets, which
is a realistic assumption to make for emerging economies.
The early works on asset pricing have not taken into
account the fact that nancial markets may not be perfectly
integrated which could be due to the presence of capital ow
restrictions, taxes, informational asymmetry, or other risk factors that deter free movement of capital across borders. For
instance studies by Solnik (1974b), Stulz (1981a), Adler and
Table 2. Components of Total Portfolio Inows (Percentage in Total FPI)
Years

199697
199798
199899
199900
200001
200102
200203
200304
200405
200506
200607
200708
200809
200910
201011
201112
201213*

a. Global/American
depository
receipts

b. Foreign
institutional
investors

c. Oshore funds
and others

41.24
35.28
442.62
25.38
30.11
23.60
61.29
4.03
6.58
20.43
53.92
24.37
8.39
10.28
6.51
3.43
0.69

58.15
53.56
639.34
70.56
66.92
74.47
38.51
95.97
93.25
79.46
46.05
74.54
108.39
89.72
93.49
96.57
101.47

0.60
11.16
96.72
4.06
2.97
1.93
0.20
0.00
0.17
0.11
0.03
1.09
0.00
0.00
0.00
0.00
0.00

Source: Calculations based on data from Reserve Bank of India, Handbook of Statistics 2013.
*
Calculation based on provisional data.

19

Dumas (1983) and Lewis (1999) that have extended the asset
pricing models by Markowitz (1959), Sharpe (1964) and Lintner (1965) to an international setting, assume perfect capital
markets. Asset pricing models based on such assumptions
are not applicable to most of the emerging market economies,
where there are restrictions to ow of capital and the nancial
markets in these countries are only partially integrated with
the global markets (Harvey, 1995).
Some studies of international asset pricing have relaxed the
assumption of perfectly integrated capital markets and assume
partial integration of nancial markets such as Black (1974),
Stulz (1981b), Errunza and Losq (1985, 1989), Eun and Janakiraman (1986), Alexander, Eun, and Janakiramanan (1987),
Hietala (1989), Errunza, Losq, and Padmanabhan (1992),
Cooper and Kaplanis (2000). A major implication of these
studies is that assets to which investors have unrestricted access are priced lower than the assets to which investors have
restricted access. Although these studies assume partial integration of markets, they hold the degree of nancial market
integration to be constant overtime. But what is more appropriate is, to assume time varying market integration, which is
done in the studies by Bekaert and Harvey (1995, 2003), Stulz
(1999), Carrieri et al. (2007) and Arouri and Foulquier (2012)
among others, that is better suited in the context of reduction
in capital controls and gradual liberalization in emerging markets overtime.
The degree of market integration can have an important impact on expected returns, which depends on the risk of the
country portfolio. While in the case of complete integration
of nancial markets, risk of a country portfolio depends on
its covariance with the world market returns, in the case of
complete segregation of markets, it depends on the variance
in its own returns. In case of partially integrated emerging
economy, the investor expects to be compensated for the risk
of the country portfolio that depends on boththe covariance
with the world market returns (as in the case of complete integration of markets) as well as the variance in its own returns
(as in the case of complete segregation of markets) (Bekaert
& Harvey, 1995). Moreover, greater the degree of integration
of the countrys nancial markets with the world markets,
higher is the inuence of global factors in determining the foreign investment received by the country (Bekaert, 1995).
It is expected that as the market integration increases, expected returns from holding the countrys assets decrease
and correlation of countrys nancial markets with the world
markets increases, which reduces the possibilities of risk diversication (Bekaert, 1995; Bekaert & Harvey, 2000, 2002).
The aforesaid asset pricing models that assume partial integration of markets and time varying degree of integration of
markets indicate some of the factors such as correlation between domestic stock returns and global equity returns, variance in domestic equity returns and global factors i.e.,
global output growth and interest rates, that may inuence expected returns from assets in emerging economies and hence
the ow of capital to these countries. Based on these and other
factors that are considered to be important in the literature,
this study seeks to analyze the determinants of portfolio ows
to India. The potential variables used to explain portfolio
ows are described in detail.
Domestic Stock Market Performance: The domestic stock
market performance can have a positive or a negative
inuence on portfolio ows. An increase in portfolio ows
in response to bullish stock markets (Agarwal, 1997; Chakrabarti, 2001; Rai & Banumurthy, 2004; Richards, 2002) suggests
that foreign investors are chasing returns. On the other hand,
the relationship can turn negative if foreign investors buy (sell)

20

WORLD DEVELOPMENT

when equity index is falling (rising), with the expectation that

returns will rise (fall) in future (Gordon & Gupta, 2003; Grifn, Nardari, & Stulz, 2002).
Domestic Growth: Domestic output growth indicates the
soundness of macroeconomic and institutional fundamentals
of the host country, which are important in attracting capital
ows. A higher economic growth of the host country indicates
rapidly expanding economic activity which in turn would
mean greater protability from investing in the corporate sector.
Exchange Rate: In a world of exible exchange rate, capital
can earn a return not only through yields on assets, but also
through a change in exchange rate overtime. Appreciation of
currency of the host country is an additional avenue of gaining
returns for foreign investors.
Currency Risk: Volatility in exchange rate is expected to
have negative impact on portfolio ows because it represents
a higher degree of uncertainty in the returns received by foreign investor in terms of his home currency. Persson and
Svensson (1989) observe that increased exchange rate variability has negative impact on international trade and capital
ows.
Country Risk: Availability of sucient liquidity in the host
country indicates that in the event of withdrawal of funds by
the investors, the country does not default on the payments.
It is expected that countries with sucient stock of reservesto cover for imports and meet short run obligationsare perceived to be credit worthy and the probability
of their defaulting is less. Thus lower nancial risk of the country attracts greater portfolio ows.
Stock Return Risk: While investing in a developing country,
where stock markets are characterized by volatility, the foreign investor takes into account not only the returns from
investment in an asset, but also the variability associated with
the returns. This is important in determining the expected returns from investment. Unless the investor is suciently compensated for the variance in returns, a higher variability in
returns discourages foreign investment.
Risk Diversication: An investor invests in assets across
countries to diversify the risk of the portfolio. The objective
of the investor is to reduce the variance of the overall portfolio. If the addition of a countrys assets reduces the overall risk
of the portfolio (even if the variance of its own returns is high),
there are potential gains from diversifying internationally.
This depends on the degree of correlation between the domestic and the foreign markets. The lower the co-movement of
domestic equity returns with the global equity returns, the
greater the benets from diversication and hence higher the
portfolio ows received by a country.
Global Liquidity: Output growth in industrialized countries
represents the protability of investment in the corporate sector in these countries. A higher economic growth in industrialized countries could mean greater prots and hence greater
funds available for investment in developing countries. On
the other hand a slowdown in the economic growth could also
mean that the rms do not nd it protable to invest in their
home country and hence they chose to invest in developing
countries. In both the cases economic growth in industrialized
countries could lead to an increased global liquidity. The evidence in this regard is mixed. While Calvo et al. (1993, 1996),
Taylor and Sarno (1997) among others indicate a negative
relation between nancial ows to developing countries and
growth in industrial countries, Verma and Prakash (2011)
indicate a positive relation for the case of India.
Interest Rate Dierential: Interest rate dierential between
the host and the source country also determines portfolio

ows. The traditional open economy macroeconomic models

proposed by Mundell and Fleming suggest that in a world
where capital is mobile and exchange rates are xed, capital
ows occur so as to restore interest parity, i.e., capital moves
in or out of the country till the domestic and foreign interest
rates equalize. Investors invest their capital wherever the interest rates adjusted for risk are higher. While most of the studies
in the literature nd that portfolio ows are sensitive to
domestic and/or foreign interest rates, Verma and Prakash
(2011) nd that FII ows to India are not sensitive to interest
rate dierentials.
Returns in Other Emerging Markets: While diversifying
globally, foreign investors can either invest in emerging markets or they can invest in nancial markets of the industrialized countries. According to Buckberg (1996) investors
follow a two step process in deciding capital allocation.
Firstly, the total capital to be invested in emerging markets
is determined and then a part of that capital is allocated to
each emerging market depending on returns. This implies that
if total capital allocated to emerging markets is highwhich
will happen if equity returns in emerging markets is rising
then each emerging economy has a higher probability of
receiving greater amount of capital.
Alternately, it is also important to view dierent emerging
economies as competitors to each other, where each economy
is trying its best to receive a greater share of foreign investment. In this case, higher equity returns in emerging markets
implies a greater probability of foreign investment being received by competing economies.
Capital Controls: Although capital controls are used as a
tool to manage nancial ows, the evidence regarding their
eectiveness in the literature 2 is mixed. The broad view in
the literature is that capital controls may not be eective in
inuencing the overall magnitude of capital ows, but they
do aect the composition.
The variables described above capture dierent aspects of
the mechanism that drive portfolio ows which can be put together to obtain the empirical model that is discussed in the
next section.
4. EMPIRICAL MODEL, DATA AND METHODOLOGY
This study estimates the long-run macro econometric model
to analyze the determinants of portfolio ows to India. Based
on the factors discussed in the previous section, the following
empirical model is estimated:
F t a b S t c et r CRt d MSCI t h i  it f y t g y t
q Volet x alphat u betat W cc lt . . .

b > or < 0; c > 0; r > 0; d > or < 0; h > 0; f > 0; g > or

< 0; q < 0; x < 0; u < 0
Net portfolio inows, FII ows and ADR/GDR ows denominated in US $ billion are taken as the dependent variable (F).
Domestic stock market performance (S) is measured by the
monthly average of the BSE sensitivity index. Exchange rate
(e) is dened as the nominal eective exchange rate 36 country
(export based). 3 Currency risk proxied by volatility in
exchange rates (Vol(e)) is calculated using the Generalized
Autoregressive Conditional Heteroskedasticity (GARCH)
method. 4 Interest rate dierential (i  i*) is dened as the
dierence between 3 month Treasury bill rate for India
and 3 month London Inter Bank Oered Rate (LIBOR). 5

FOREIGN PORTFOLIO INVESTMENT FLOWS TO INDIA: DETERMINANTS AND ANALYSIS

Domestic output growth (y) is dened as the annual growth

rate of index of industrial production (IIP) for India. Global
liquidity proxied by foreign output growth (y*) is measured
by annual growth rate of IIP for Organization for Economic
Co-operation and Development (OECD) countries. Both the
growth series are calculated as the dierence between logarithm of index of a particular month and the twelfth lagged
month. This also helps to remove the seasonality in the IIP series if there exists any.
Country risk 6 is calculated using the methodology that is
similar to the one used by International Country Risk Guide
(ICRG) for calculating nancial risk. Based on the scale suggested by ICRG, risk points were assigned to foreign debt as
a percentage of Gross Domestic Product (GDP) (annual data),
net international liquidity as months of import cover (annual
data), foreign debt service as a percentage of exports of goods
and services (annual data) and current account as a percentage
of exports of goods and services (quarterly data). These points
were then aggregated to obtain credit risk indicator (CR)
(which took the same values for three months in each quarter),
where a higher risk point total suggests lower risk and vice versa. The components of the indicator are based on ratios and
not absolute values which ensures comparability overtime.
Emerging market MSCI Index 7 (MSCI) is used to capture
the stock performance in emerging markets. Stock return risk
(alpha) is captured by twelve month standard deviation in
MSCI-India. Risk Diversication (beta) is measured as the
twelve month correlation between MSCI-India and MSCIWorld. 8 To capture the liberal reform measures in 200304
(cc1) and restrictive capital controls (that restricted inows
and liberalized outows) in 200607 (cc2) dummy variables
are used. The impact of recent subprime crisis that originated
in US, but aected the real and nancial ows across the globe,
is captured through a dummy. When global liquidity is aected
due to any given exogenous reason, the ows of nancial
investments to emerging economies get hit. Thus it is necessary
to control for such eects. The main sources of the data used in
the analysis are Handbook of Statistics and Current Statistics
of RBI, International Financial Statistics of International
Monetary Fund (IMF), www.mscibarra.com, Federal Reserve
Bank and Global Development Finance reports.
(a) Methodology
The DickeyFuller Test with Generalised Least Squares
Detrending (DFGLS) proposed by Elliott, Rothenberg, and
Stock (1996) is used to test if a series contains a unit root.
To estimate the long-run coecients, the Autoregressive Distributed Lag (ARDL) approach to co-integration proposed
by Pesaran and Shin (1999) and Pesaran, Shin, and Smith
(2001) is used. This is because it is applicable irrespective of
the fact that the variables included in the analysis are I(0) or
I(1) or fractionally integrated.
The augmented ARDL (p, q1, q2, . . . qk) is given by the following equation (Pesaran & Pesaran, 1997; Pesaran et al., 2001)
uL; py t

k
X

bi L; qi xit k0 wt et

8t 1; . . . ; n

i1

where
uL; p 1  u1 L  u2 L2  . . .  up Lp

bi L; qi bi bi1 L bi2 L2 . . . biqi Lqi

8i 1; 2 . . . ; k

21

yt is an independent variable, L is the lag operator such that

Lp yt = ytp, wt is the s  1 vector of the deterministic variables such as the intercept term, time trends, dummies, or
exogenous variables with xed lags. xit represents ith independent variable where i = 1, 2, . . ., k where lag order of ith variable is qi.
The long-run slope coecients of the endogenous variables
are estimated as follows:
Ui

^i 1; ^
b
qi
^
/1; ^
p

8i 1; 2; . . . ; k

where ^
p and ^
qi , i = 1, 2, . . ., k are the selected (estimated) values of ^
p and ^
qi , i = 1, 2, . . ., k.
Ui gives the long-run impact of xit on yt.
The error correction model derived from the ARDL framework is used to test for Granger causality. This is done by testing the joint signicance of the error correction term and the
lags of each of the explanatory variables.
Before proceeding with the ARDL estimation it is important
to test whether there exists a long-run relation between the
variables, for which bounds testing approach is used.
To test whether there exists a long-run relation between the
variables, when all the variables are integrated of dierent order i.e., I(0) or I(1), bounds testing approach is used. This involves testing the null hypothesis of p = p1 = p2 = . . . =
pk = 0 against the alternative hypothesis of p1 p2
. . . pk 0 in the following equation
Dy t c0 py t1 p1 x1;t1 p2 x2;t1 . . . pk xk;t1
m
m
m
X
X
X

wDy ti
w1i Dx1;ti
w2i Dx2;ti
i1

...

i1
m
X

i1

wki Dxk;ti lt

i1

Pesaran and Pesaran (1997) have two sets of asymptotic critical values for the F-statistic. One is the lower bound critical
value which assumes that all the variables are integrate of order zero i.e., I(0). The other is the upper bound critical value
which assumes that all the variables are I(1). If the calculated
F-statistic is greater than the upper bound critical value then
the null hypothesis of no long-run relationship is rejected
and if it falls below the lower bound critical value, then the
null hypothesis of no long-run relationship is not rejected.
5. ECONOMETRIC RESULTS
(a) Results of the unit root test and the bounds test
The DFGLS unit root tests 9 suggest that Indian output
growth, foreign output growth, volatility in exchange rate
and beta of Indian market are stationary. All the other variables i.e., net FPI inows, net FII inows, net ADR/GDR
ows, domestic stock market index, exchange rate, emerging
market stock returns index, stock return risk, asset reserves
to imports ratio, and interest rate dierential are integrated
of order one. Since the variables under consideration are a
mix of I(0) and I(1), hence the ARDL method was found suitable for estimation.
The rst step is to check for cointegration between the
variables using the ARDL bounds test suggested by Pesaran
et al. (2001). A maximum lag value of four was used for
implementing the test. The dummy variable for crisis was also
included in the FPI and FII specications, since it was

22

WORLD DEVELOPMENT

observed that these ows were aected the most during the recent subprime crisis. For ADR/GDR a dummy variable to
control for three outlier observations was included. For all
the specications, the null hypothesis of no long-run relationship is rejected, as the calculated F-statistic lies above the
upper bound critical F value at 95% (Tables 35).
Once the bounds test ensures the presence of long-run cointegration between the variables, the next step is to estimate the
ARDL model and obtain the long-run coecients and then
perform Granger causality tests based on the short run error
correction model. The optimal lag length for the variables is
chosen such that there is no residual serial correlation in the
underlying model.
(b) Aggregate foreign portfolio investment inows and foreign
institutional investment inows
The decomposition of aggregate FPI ows shows that FII
ows are a dominant component of FPI ows and the peaks
and the troughs in total FPI ows closely matches with that
in FII ows (Figure 2), hence as expected, the estimation result
of aggregate FPI ows is quite similar to that of FII ows
(Tables 6 and 7).
For all the specications of FPI and FII (Tables 6 and 7), the
coecient on domestic stock market performance is found to
be positive and signicant indicating that well performing
domestic stock markets attract ows to India. The aggregate
FPI ows as well as the FII component respond in a positive
way to an appreciating exchange rate. This is because an appreciating exchange rate provides additional source of returns to
foreign investors which induces them to invest in India.
The improvement in performance of emerging market
stocks has a negative impact on FII as well as aggregate FPI
ows to India. This implies that when the overall returns from
investing in emerging market stocks increases, foreign investment received by India decreases. This can be analyzed in
terms of the income eect and the substitution eect. When
emerging markets as a whole perform better, then income effect would mean greater allocation of funds to emerging markets by foreign investor, because of which the probability that
greater funds are allotted to each emerging economy increases.
Once the allocation of funds to emerging markets is done, the
substitution eect comes into play i.e., each of the emerging
economies competes with the others for receiving these ows.
In case of India, substitution eect dominates the income eect
with regard to FII ows. Similar result holds for aggregate
FPI ows.

The dierential between domestic and foreign interest rates

has a positive association with portfolio and FII ows to
India. This can be attributed to increasing investment in debt
securities which are sensitive to interest rate dierentials. In
India, very recently during the crisis period, it has been
observed that FIIs have increasingly been investing in debt
instruments, which are less risky. 10
Domestic output growth is positive and signicant in all the
specications for both FPI as well as FII. An increasing
domestic output growth indicates higher levels of economic
activity and hence increasing corporate prots which is an
indicator of better investment opportunities in the country.
Thus to a foreign investor better growth performance of the
host country is a signal that demand for investment in the host
country will remain high as it is a rapidly expanding economy,
and hence the probability that returns from investment will
fall in future will be low. Moreover, rising domestic output
growth also reects strong economic fundamentals, which
encourages the investors to have greater faith in the Indian
economy.
The long-run coecient on volatility in exchange rate is negative and signicant in all the specications, indicating that
higher currency risk discourages FIIs and aggregate portfolio
ows to India. This is because increasing exchange rate volatility corresponds to increasing uncertainties in the returns
that will be received by the foreign investors in terms of their
own currency. Also, if the volatility in exchange rate is attributable to factors within the host country, then it also indicates
the presence of destabilizing forces within the economy.
The negative and signicant coecient on risk diversication measured by beta indicates that if beta increases i.e.,
comovement between Indian and global equity increases, then
aggregate FPI ows and FII ows decrease. On the other hand
declining beta indicates lower correlation between Indian and
global equity returns that presents the opportunities for reducing the risk of the portfolio by investing in Indian markets.
This conrms the fact that FIIs invest in India with the objective of diversifying risks.
The capital control dummy for restrictive policy reforms in
200607 is signicant for FII but insignicant for aggregate
FPI ows. This implies that the measures taken to restrict inows and liberalize outows were signicant in reducing FII
investment ows to India but aggregate portfolio ows remained unaected. The liberal measures taken in 200304
did not signicantly inuence both FPI as well as FII.
As expected, greater risk in asset returns discourages FII
ows as well as aggregate portfolio ows, however the eect

Table 3. ARDL Bounds Testing Approach for Cointegration Dependant Variable: Net Foreign Portfolio Investment ows (US$ Billion) 1995M102011M10
F-Statistic [p-value]

Critical F value (95%)

Lower Bound

Critical F value (90%)

Upper Bound

Result

Lower Bound

Upper Bound

1.956

3.085

Reject H0

2.141

3.250

Reject H0

2.035

3.153

Reject H0

Model IV: FPI = f (S, e, MSCI, i  i , y, Vol (e), beta, Dummy)

F(8,151)=4.9321[.000]
2.365
3.553

2.035

3.153

Reject H0

Model V: FPI = f (S, e, MSCI, i  i*, y, Vol (e), cc1, cc2, Dummy)
F(7,154)=5.1915[.000]
2.476
3.646

2.141

3.250

Reject H0

Model I: FPI = f (S, e, CR,MSCI, i  i*, y, y*, Vol (e), Dummy)

F(9,146)=3.8679[.000]
2.272
3.447
*

Model II: FPI = f (S, e, MSCI, i  i , y, Vol (e), Dummy)

F(7,156)=5.2065[.000]
2.476
3.646
*

Model III: FPI = f (S, e, MSCI, i  i , y, Vol (e),alpha, Dummy)

F(8,151)=4.2707[.000]
2.365
3.553
*

H0: No long-run relationship.

H1: There exists long-run relationship.

FOREIGN PORTFOLIO INVESTMENT FLOWS TO INDIA: DETERMINANTS AND ANALYSIS

23

Table 4. ARDL Bounds Testing Approach for Cointegration Dependant Variable: Net Foreign Institutional Investment ows (US$ Billion) 1995M10
2011M10
F-Statistic [p-value]

Critical F value (95%)

Lower Bound

Critical F value (90%)

Upper Bound

Result

Lower Bound

Upper Bound

Model I : FII = f (S, e, CR,MSCI, i  i*, y, y*, Vol (e), Dummy)

F(9,146)=4.0513[.000]
2.272
3.447

1.956

3.085

Reject H0

Model II: FII = f (S, e, MSCI, i  i*, y, Vol (e), Dummy)

F(7,156)=5.3720[.000]
2.476
3.646

2.141

3.250

Reject H0

2.035

3.153

Reject H0

2.035

3.153

Reject H0

2.141

3.250

Reject H0

Model III: FII = f (S, e, MSCI, i  i , y, Vol (e),alpha, Dummy)

F(8,151)=4.3675[.000]
2.365
3.553
*

Model IV: FII = f (S, e, MSCI, i  i , y, Vol (e), beta, Dummy)

F(8,151)=5.2981[.000]
2.365
3.553
*

Model V: FII = f (S, e, MSCI, i  i , y, Vol (e), cc1, cc2, Dummy)

F(7,154)=5.3587[.000]
2.476
3.646
H0: No long-run relationship.
H1: There exists long-run relationship.

Table 5. ARDL Bounds Testing Approach for Cointegration Dependant Variable: Net American/Global Depository Receipt Inows (US$ Billion) 1995M10
2011M10
F-Statistic [p-value]

Critical F value (95%)

Lower Bound

Critical F value (90%)

Upper Bound

Result

Lower Bound

Upper Bound

1.956

3.085

Reject H0

Model II: ADR = f (S, e, MSCI, i  i , y, y , Vol (e), Dummy)

F(8,151)=6.1459[.000]
2.365
3.553

2.035

3.153

Reject H0

Model III: ADR = f (S, e, i  i*, y, y*, Dummy)

F(6,161)=9.9744[.000]
2.649

3.805

2.262

3.367

Reject H0

4.049

2.425

3.574

Reject H0

3.805

2.262

3.367

Reject H0

Model VI: ADR = f (S, e, y, y , beta, Dummy)

F(6,161)= 9.6733[.000]
2.649

3.805

2.262

3.367

Reject H0

Model VII: ADR = f (S, e, y, y*, cc1, cc2, Dummy)

F(5,164)=11.0678[.000]
2.850

4.049

2.425

3.574

Reject H0

Model I : ADR = f (S, e, CR,MSCI, i  i*, y, y*, Vol (e), Dummy)

F(9,146)=5.2562[.000]
2.272
3.447
*

Model IV: ADR = f (S, e, y, y , Dummy)

F(5,166)=12.8455[.000]
2.850
*

Model V: ADR = f (S, e, y, y , alpha, Dummy)

F(6,161)=10.1512[.000]
2.649
*

H0: No long-run relationship.

H1: There exists long-run relationship.

is statistically insignicant. Country risk is not signicant in

inuencing portfolio ows as well as FII ows indicating higher investor condence in the Indian economy. Portfolio ows
and FIIs are driven more by domestic output growth compared to foreign output growth.
The Granger causality tests based on the error correction
model of the ARDL specication shows that all the variables
that are signicant determinants also signicantly Granger
cause FPI and FII ows to India. This means that they help
to improve the predictive performance of these ows (Tables
8 and 9).
(c) American/global depository receipts
ADR/GDRs present a mechanism of indirect investment by
foreign investor in the stocks of Indian companies. This prevents them from the complication of conversion of foreign currency into domestic currency (rupees) and does not involve any
additional entry costs that would otherwise be incurred if the

foreign investor decided to invest directly in the Indian nancial

markets. In case, the investment by foreign investor is done directly in the domestic (Indian) companies through the nancial
markets, then the foreign investor has to monitor not only the
changes in the price of assets in which he decides to invest, but
also the changes in exchange rate as the process involves conversion of foreign currency into domestic (Indian) currency.
But this does not imply that investment through ADR/
GDRs is independent of exchange rate changes. It is only
the case that the burden of currency risk is not explicitly borne
by the foreign investor, who purchases the receipts. For instance, when ADR/GDRs are issued, then each receipt is
backed by a certain number of shares of the domestic (Indian)
company. Once this ratio between ADR/GDR shares and
domestic shares is decided, the price of each receipt (ADR/
GDR) is decided. The pricing decision is based on the price
of the shares of the company in the domestic (Indian) market,
and the corresponding exchange rate between the two economies.

24

WORLD DEVELOPMENT
Table 6. Long-Run Coecients Dependant Variable: Net Foreign Portfolio Inows (US $ Billion) 1995M10 to 2011M10
Variable

S
e
MSCI
CR
i  i*
y
y*
Vol(e)
alpha
beta
CC1
CC2
Crisis dummy
Intercept

Coecient [p-value]
Model I ARDL
(4,1,2,1,1,1,2,1,1)

Model II ARDL
(4,1,2,1,1,2,1)

Model III ARDL

(4,1,2,1,1,2,1, 1)

Model IV ARDL
(4,1,2,1,1,2,2,1)

Model V ARDL
(4,1,2,1,1,2,2)

.000401[.002]
.175880[.013]
.00556[.021]
.06365[.377]
.120720[.178]
.116040[.032]
.005151[.910]
.37593[.046]

.000368[.002]
.183200[.004]
.00509[.021]

.000404[.001]
.150470[.023]
.00597[.011]

.000472[.000]
.208990[.001]
.0065[.008]

.000457[.000]
.138100[.029]
.004690[.023]

.136060[.110]
.119040[.010]

.10335[.267]
.14336[.004]

.05270[.579]
.124090[.006]

.067200[.439]
.125890[.004]

.378680[.036]

.31934[.086]
.00408[.712]

.56554[.015]

.40540[.017]

.73823[.065]
.48530[.211]
1.0970[.158]
Yes
Yes

Yes
Yes

Yes
Yes

Yes
Yes

Yes
Yes

Table 7. Long-Run Coecients Dependant Variable: Net Foreign Institutional Investment Inows (US $ Billion) 1995M10 to 2011M10
Variable

S
e
MSCI
CR
i  i*
y
y*
Vol(e)
alpha
beta
CC1
CC2
Crisis dummy
Intercept

Coecient [p-value]
Model I ARDL
(4,1,2,1,1,1,2,1,1)

Model II ARDL
(4,1,2,1,1,2,2)

Model III ARDL

(4,1,2,1,1,2,2,2)

Model IV ARDL
(4,1,2,1,1,2,2,1)

Model V ARDL
(4,1,2,1,1,2,2)

.000387[.002]
.118960[.079]
.00590[.011]
.065881[.342]
.127350[.138]
.092972[.070]
.034617[.429]
.37134[.040]

.000329[.004]
.145900[.016]
.00483[.026]

.000374[.003]
.110860[.088]
.00562[.014]

.000438[.001]
.16392[.007]
.0064[.007]

.000432[.000]
.098750[.093]
.00439[.026]

.155880[.060]
.114600[.010]

.11274[.219]
.13871[.004]

.0708[.442]
.11628[.008]

.083400[.308]
.120420[.003]

.463880[.034]

.43546[.059]
.00899[.444]

.5432[.015]

.4963[.013]

.66831[.085]
.4370[.231]
1.3340[.065]
Yes
Yes

Yes
Yes

Yes
Yes

Yes
Yes

Yes
Yes

Table 8. Granger Causality Test from ECM: Net Foreign Portfolio Inows
Null hypothesis

S does not Granger cause FPI

e does not Granger cause FPI
MSCI does not Granger cause FPI
i  i* does not Granger cause FPI
y does not Granger cause FPI
Vol(e) does not Granger cause FPI
beta does not Granger cause FPI

Chi-square statistic [p-value]

Model II

Model IV

106.5756[.000]
68.7153[.000]
56.7942[.000]
56.7825[.000]
55.7905[.000]
77.4479[.000]

108.7294[.000]
68.6112[.000]
58.0089[.000]
58.8540[.000]
57.1757[.000]
80.4300[.000]
56.9355[.000]

Now if the price of the shares of the company in the domestic (Indian) market increases, then the price of ADR/GDR
will also increase so that the zero arbitrage condition holds.
In case the domestic currency (rupee) is getting devalued, the
price of ADR/GDR being issued will be higher.
Exchange rate also inuences investment in ADR/GDRs in
the case when dividends are issued by the domestic (Indian)
company. The domestic company will issue dividends in terms
of domestic currency (rupees) but this dividend will be received

Conclusion

Reject
Reject
Reject
Reject
Reject
Reject
Reject

H0
H0
H0
H0
H0
H0
H0

by the foreign investor, who has purchased ADR/GDR, only

after conversion into foreign currency. In this case again, a
devaluation in the domestic currency vis-a`-vis the foreign currency would mean a decline in the amount of dividend received by the foreign investor in terms of foreign currency.
The empirical results for the investment ows received by
India on account of ADR/GDRs support both the above
arguments. The estimation of the long-run coecients
(Table 10) show that the coecient on domestic stock market

FOREIGN PORTFOLIO INVESTMENT FLOWS TO INDIA: DETERMINANTS AND ANALYSIS

25

Table 9. Granger Causality Test from ECM: Net Foreign Institutional Investment Inows
Null hypothesis

Chi-square statistic [p-value]

S does not Granger cause FII

e does not Granger cause FII
MSCI does not Granger cause FII
i  i* does not Granger cause FII
y does not Granger cause FII
Vol(e) does not Granger cause FII
beta does not Granger cause FII

Conclusion

Model II

Model IV

106.153[.000]
61.1412[.000]
49.9603[.000]
48.6887[.000]
48.9052[.000]
75.9569[.000]

102.909[.000]
64.2641[.000]
54.5394[.000]
54.0778[.000]
53.2539[.000]
79.6101[.000]
54.0243[.000]

Reject
Reject
Reject
Reject
Reject
Reject
Reject

H0
H0
H0
H0
H0
H0
H0

Table 10. Long-Run Coecients Dependant Variable: Net American/Global Depository Receipt Inows (US $ Billion) 1995M10 to 2011M10
Variable

S
e
MSCI
CR
i  i*
y
y*
Vol(e)
alpha
beta
CC1
CC2
Dummy
Intercept

Coecient [p-value]
Model I ARDL
(2,1,4,0,4,4,2,1,0)

Model II ARDL
(2,2,4,0,4,2,2,0)

Model III ARDL

(2,2,4,4,2,2)

Model IV ARDL
(2,2,4,2,2)

Model V ARDL
(2,2,4,2,2,3)

Model VI ARDL
(2,2,4,2,2,1)

Model VII
ARDL (2,2,4,2,2)

.000034[.111]
.041209[.002]
.00026[.505]
.003525[.787]
.002684[.870]
.011093[.216]
.01672[.013]
.00741[.736]

.000034[.100]
.041924[.001]
.00025[.502]

.000020[.000]
.036956[.001]

.000018[.000]
.035407[.000]

.000023[.007]
.03357[.001]

.000019[.000]
.036502[.000]

.000032[.013]
.033577[.001]

.003399[.832]
.011318[.205]
.01646[.016]
.00635[.772]

.000432[.977]
.011000[.207]
.01652[.012]

.010896[.198]
.01688[.007]

.013041[.141]
.019522[.033]

.011415[.177]
.01821[.005]

.014838[.110]
.02131[.004]

.00076[.640]
.04042[.435]
.04861[.485]
.16323[.276]
Yes
Yes

Yes
Yes

Yes
Yes

returns and exchange rate is positive and signicant in all the

specications. This indicates that well performing domestic
stock markets and an appreciating domestic currency encourages ADR/GDR ows received by India.
This is because improvements in the performance of shares
of the company in the domestic stock market (which back the
ADR/GDRs) lead to improved performance of ADR/GDRs,
through arbitrage. Similarly for the reasons stated above an
appreciating rupee would lead to better returns from investing
in ADR/GDR and a higher value of dividends issued on
ADR/GDR.
The estimated long-run coecients also show that increase
in foreign growth rate reduces the amount of ADR/GDR
investment ows received by India. This is because a higher
foreign growth rate implies a higher economic activity which
in turn means higher corporate prots and hence better investment opportunities in the developed countries. This could lead
to diversion of investment by foreign investors toward the
nancial instruments of developed countries.
On the other hand increase in foreign growth rate which reects higher corporate prots, also means greater availability
of funds with foreign investors for investment, and thus a
higher probability of increased investment in Indian securities.
However, a negative sign on the coecient of foreign output
growth means that these funds are not invested in ADR/
GDR of Indian companies and the former eect dominates.
The coecient on domestic output growth is positive which
means that an increase in the Indian growth increases the
investment ows through ADR/GDR mechanism. This is because on one hand the foreign investor perceives an improvement in growth performance of India (i.e., the country to
which the rm oating ADR/GDR belongs) as an indicator

Yes
Yes

Yes
Yes

Yes
Yes

Yes
Yes

that the returns on the instrument will be promising in future

because of the expansion in the economic activity of the country. On the other hand, more importantly, a higher growth
rate also means an increase in the requirement of capital by Indian rms for expansion, which is met by greater issuance of
ADR/GDRs by Indian rms. Thus not only the demand for
foreign investment by Indian rms increases, but foreign
investors also nd it protable to invest in Indian ADR/
GDRs when growth of the Indian economy increases. In this
case both the demand as well as supply side factors facilitate
increased ow of foreign capital to India through ADR/
GDRs.
Unlike FII ows the coecient on risk diversication which
is measured by beta is insignicant for ADR/GDR ows.
This indicates that investment through ADR/GDRs does
not respond to risk diversication opportunities in India.
The emerging market equity performance, interest rate
dierential, country risk, currency risk, risk related to asset
returns and capital control measures are also found to be
insignicant in driving ADR/GDR investment. The Granger
causality tests indicate that domestic stock market performance, exchange rate, domestic and foreign output growth
Table 11. Granger Causality Test from ECM: Net American/Global
Depository Receipt Inows
Null hypothesis
S does not Granger cause ADR
e does not Granger cause ADR
y does not Granger cause ADR
y* does not Granger cause ADR

Chi-square statistic
[p-value] Model IV
87.1700[.000]
108.0586[.000]
88.8691[.000]
89.3956[.000]

Conclusion
Reject
Reject
Reject
Reject

H0
H0
H0
H0

26

WORLD DEVELOPMENT

82000

FPI

US $ Million

62000

FPI+FDI
Current Account Deficit

42000

22000

2000

-18000

Figure 3. Current Account Decit, Nondebt Creating Capital Flows (FPI + FDI) and FPI. Source: Reserve Bank of India, Handbook of Statistics 2013.

signicantly Granger cause ADR/GDR ows to India

(Table 11).
6. CONCLUSIONS AND IMPLICATIONS
This paper analyzes the macroeconomic determinants of
portfolio ows to India for the period 19952011 where portfolio ows to India have been increasing since 2002 and accelerated in 2006 but slowed down in 2008 due to the global
nancial crisis. The slowdown in portfolio ows occurred in
the backdrop of rising current account decit.
In the past decade, nondebt creating capital ows (foreign
direct investment + portfolio ows) of which portfolio ows
is one of the components, have been sucient to nance current account decit as shown in Figure 3. However, more recently, the current account decit has far exceeded these
ows implying that debt creating ows 11 are required to nance the balance. This is substantiated by the recent rise in
deposits of the nonresident Indians and external commercial
borrowings.
In the current context identifying the determinants of portfolio ows is not only important to attract greater ows that
can be used to nance the current account decit but also
for predicting these ows, so that policy reactions can be formulated in advance to manage these ows.
The results indicate that higher currency risk discourages
portfolio ows as it increases the uncertainty in returns received by the foreign investor in terms of its home currency.
It is also observed that higher equity returns in emerging markets decrease portfolio ows to India indicating that India
faces competition from other emerging economies in terms
of attracting portfolio ows. Evidence is also found in favor
of the fact that investors invest in India with the objective of
risk diversication. The conventional determinants i.e., well

performing domestic stock market, interest rate dierential

and an appreciating exchange rate also encourage portfolio
ows to India. Greater risk related to asset returns discouraged portfolio ows though the eect is not signicant. The results also indicate that portfolio ows are determined more by
strong domestic output growth rather than foreign output
growth. Country risk does not signicantly inuence portfolio
ows, which indicates greater investor condence in the Indian
economy.
An analysis of the disaggregated components of portfolio
ows i.e., FII and ADR/GDR ows shows lower currency
risk, lower emerging market equity returns and greater risk
diversication opportunities in addition to strong domestic
equity performance, appreciating exchange rate, higher interest rate dierential and rising domestic growth are conducive
to FII ows. Higher asset returns risk discourages FIIs though
the eect is not signicant. As observed for aggregate portfolio
ows, country risk and foreign output growth are statistically
insignicant in inuencing FIIs. For ADR/GDRs, domestic
stock market performance, exchange rate, domestic output
growth, and foreign output growth, are observed to be important determinants. The results also suggest that the policy
reforms that restricted inows and liberalized outows in
200607 reduced FII ows to India but did not inuence
investment through ADR/GDRs.
The results suggest that India may be able to attract portfolio ows by maintaining strong domestic growth, lower exchange rate volatility through intervention in currency
markets and making the domestic nancial market performance less vulnerable to global shocks by expanding the
domestic investor base in nancial markets. To deal with the
extreme episodes of turmoil or crisis, outows of foreign capital can be controlled using policy measures that include imposition of restrictions on capital outows.

NOTES
1. Currently there are 1506 FIIs registered with SEBI out of which 56%
belong to Mauritius (The Times of India, 2012). This may be because the
funds from other countries are also mobilized to India via Mauritius to
ensure tax benets which accrue due to the Double Tax Avoidance
Agreement (DTAA) between India and Mauritius.

2. See Habermeier, Kokenyne, and Baba (2011) for a review of literature.

3. The results obtained broadly remain similar if nominal rupee-dollar
exchange or 36 country trade based nominal eective exchange rate is
used.

FOREIGN PORTFOLIO INVESTMENT FLOWS TO INDIA: DETERMINANTS AND ANALYSIS

4. To check for robustness, the 3 month moving average standard
deviation in nominal eective exchange rate was alternatively used to
measure volatility in nominal eective exchange rate, and the results are
qualitatively similar.
5. Alternately it can also be dened as the dierence between 3 month
Treasury bill rate for India and 3 month Treasury bill rate for US. The
results are robust to this denition.
6. Country Risk is also calculated as the ratio of total reserves and
imports, both denominated in US million dollars. The FPI, FII and
ADR/GDR specications were estimated with reserves to import ratio
and the results obtained were broadly similar. Aggregate portfolio
ows as well as disaggregated FII and ADR/GDR ows to India
overtime were found to be insensitive to the credit worthiness of the
country.
7. The Morgan Stanley Capital International (MSCI) Emerging Markets
Index is a free oat-adjusted market capitalization weighted index which
measures equity market performance of emerging markets. The MSCI
Emerging Markets Index consists of 21 emerging market country indices.
These include 5 ve American countries (Brazil, Chile, Columbia, Mexico,
and Peru); 8 eight European, Middle East and African countries (Czech
Republic, Egypt, Hungary, Morocco, Poland, Russia, South Africa, and
Turkey); and 8 eight Asian countries (China, India, Indonesia, Korea,
Malaysia, Philippines, Taiwan, and Thailand)

27

8. Risk diversication is also captured by covariance between the Indian

equity returns and global equity returns as a ratio of variance in global
equity returns. Lower the beta, greater the opportunity for diversication
and hence higher the portfolio ows received by India. This was calculated
as the ratio of twelve month covariance between MSCI-India and MSCIWorld and twelve 12 month variance of MSCI-World. The FPI, FII, and
ADR/GDR specications were estimated using this measure instead of
correlation and the results obtained were broadly similar.
9. Augmented Dickey-Fuller Test (Dicky & Fuller, 1979; Dicky &
Fuller, 1981), Phillips-Perron Test (Phillips Peter & Perron, 1988), and
the Kwiatkowski, Phillips, Schmidt, and Shin Test (Kwiatkowski, Phillips,
Schmidt, & Shin 1992) were also conducted and all of them give the same
result as the DFGLS Test
10. The sensitivity of FII ows to interest rate dierential indicates that
debt securities are seen as a safer alternative to equity, especially during
global crisis period. This recent phenomenon of foreign investors investing
in less risky debt securities, has been true for most of the developing
countries (Global Development Finance, 2012). For all the developing
economies, debt ows increased almost 3 times from US$166 billion in
2009 to US$ 495 billion in 2010. On the other hand equity ows increased
by about 25% from US$ 508 billion in 2009 to US$ 634 billion in 2010.
11. Non-debt creating ows include external assistance, external commercial borrowings, short-term trade credit, banking capital, non-resident
Indians deposits.

REFERENCES
Adler, M., & Dumas, B. (1983). International portfolio choice and
corporation nance: A synthesis. The Journal of Finance, 38(3),
925984.
Agarwal, R. N. (1997). Foreign portfolio investment in some developing
countries: A study of determinants and macro economic impact. The
Indian Economic Review, 32(2), 217229.
Alexander, G., Eun, C. S., & Janakiramanan, S. (1987). International
listing and stock returns: Some empirical evidence. Journal of Financial
and Quantitative Analysis, 23, 135152.
Arouri, M. E. H., & Foulquier, P. (2012). Financial market integration:
Theory and empirical results. Economic Modeling, 29, 382394.
Bekaert, G. (1995). Market integration and investment barriers in
emerging equity Markets. World Bank Economic Review, 9, 75107.
Bekaert, G., & Harvey, C. R. (1995). Time varying world market
integration. Journal of Finance, 50(2), 403444.
Bekaert, G., & Harvey, C. R. (2000). Foreign speculators and emerging
equity markets. Journal of Finance, 55, 565614.
Bekaert, G., & Harvey, C. R. (2002). Research in emerging market nance:
Looking to the future. Emerging Markets Review, 3(4), 429448.
Bekaert, G., & Harvey, C. R. (2003). Emerging markets nance. Journal of
Empirical Finance, 10, 355.
Black, F. (1974). International capital market equilibrium with investment
barriers. Journal of Financial Economics, 1, 337352.
Bohn, H., & Tesar, L. (1996). US equity investment in foreign markets:
Portfolio rebalancing or return chasing?. The American Economic
Review, 86(2), 7781.
Buckberg, E. (1996). Institutional investors and asset pricing in emerging
markets. IMF Working paper 96/2.
Byrne, J. P., & Fiess, N. (2011). International capital ows to emerging and
developing countries : National and global determinants. Working
papers 2011_01. Business School Economics, University of Glasgow.
Calvo, G. A., Leiderman, L., & Reinhart, C. M. (1993). Capital inows
and real exchange rate appreciation in Latin America: The role of
external factors. IMF Sta Papers, 40(1), 108151.
Calvo, G. A., Leiderman, L., & Reinhart, C. M. (1996). Inows of capital
to developing countries in the 1990s. Journal of Economic Perspectives,
10(2), 123139.
Carrieri, F., Errunza, V., & Hogan, K. (2007). Characterizing world
market integration overtime. Journal of Financial and Quantitative
Analysis, 42, 915940.

Chakrabarti, R. (2001). FII ows to India: Nature and causes. Money and
Finance, 2(1), 6181.
Chuhan, P., Claessens, C. A., & Mamingi, N. (1993). Equity and bond
ows to Asia and Latin America: The role of global and country factors.
Policy research working paper series 1160. The World Bank.
Cooper, I., & Kaplanis, E. (2000). Partially segmented international
capital market integration budgeting. Journal of International Money
and Finance, 43, 287307.
De Santis, G., & Gerard, B. (1997). International asset pricing and
portfolio diversication with time-varying risk. Journal of Finance, 52,
18811912.
De Vita, G., & Kyaw, K. S. (2007). Growth eects of FDI and portfolio
investment ows to developing countries: A disaggregated analysis by
income levels. Applied Economics Letters, 16(3), 277283.
DellAriccia, G., di Giovanni, J., Faria, A., Kose, A., Mauro, P., Ostry, J.
D., et al. (2008). Reaping the benets of nancial globalization. IMF
occasional paper no. 264.
Dicky, D. A., & Fuller, W. A. (1979). Distributions of the estimators for
autoregressive time series with a unit root. Journal of the American
Statistical Association, 74(366), 427431.
Dicky, D. A., & Fuller, W. A. (1981). Likelihood ratio statistics for
autoregressive time series with a unit root. Econometrica, 49(4), 10571072.
Dua, P., & Sinha, A. (2007). Insulation of India from the East Asian crisis:
An analysis. The Singapore Economic Review, 52(3), 419443.
Elliott, G., Rothenberg, T. J., & Stock, J. H. (1996). Ecient tests for an
autoregressive unit root. Econometrica, 64(4), 813836.
Errunza, V., & Losq, E. (1985). International asset pricing under mild
segmentation: Theory and test. Journal of Finance, 40, 105124.
Errunza, V., & Losq, E. (1989). Capital ow controls, international asset
pricing, and investors welfare: A multi-country framework. Journal of
Finance, 44, 10251037.
Errunza, V., Losq, E., & Padmanabhan, P. (1992). Tests of integration,
mild segmentation and segmentation hypotheses. Journal of Banking
and Finance, 16, 949972.
Eun, C., & Janakiraman, S. (1986). A model of international asset pricing
with a constraint on the foreign equity ownership. Journal of Finance,
41, 897914.
Felices, G., & Orskaug, B. (2008). Estimating the determinants of capital
ows to emerging market economies: A maximum likelihood disequilibrium approach. Working paper no. 354. Bank of England.

28

WORLD DEVELOPMENT

Fernaindez-Arias, E. (1996). The new wave of private capital inows: Push

or pull?. Journal of Development Economics, 48(2), 389418.
Global Development Finance (2012). External debt of developing countries.
The World Bank.
Gordon, J. P., & Gupta, P. (2003). Portfolio ows into India: Do domestic
fundamentals matter? IMF working paper no. 03/20.
Government of India (2010). Report of working group on foreign
investment. Ministry of Finance.
Grauer, R. R., & Hakansson, N. H. (1987). Gains from international
diversication: 19681985 returns on portfolios of stocks and bonds.
Journal of Finance, 42, 721741.
Grin, J. M., Nardari, F., & Stulz, R. M. (2002). Daily cross-border equity
ows: Pushed or pulled? NBER working paper no. 9000.
Grubel, H. G. (1968). Internationally diversied portfolios. American
Economic Review, 58, 12951314.
Habermeier, K., Kokenyne, A., & Baba, C. (2011). The eectiveness of
capital controls and prudential policies in managing large inows. IMF
sta discussion note SDN/11/14.
Harris, R., & Sollis, R. (2003). Applied time series modelling and
forecasting. Chichester: John Wiley and Sons.
Harvey, C. R. (1991). The world price of covariance risk. Journal of
Finance, 46(1), 111157.
Harvey, C. R. (1995). Predictable risk and returns in emerging markets.
Review of Financial Studies, 8, 773816.
Hietala, P. T. (1989). Asset pricing in partially segmented markets
evidence from the Finnish market. Journal of Finance, 44, 697718.
Kaur, M., & Dhillon, S. S. (2010). Determinants of foreign institutional
investors investment in India. Eurasian Journal of Business and
Economics, 3(6), 5770.
Kim, Y. (2000). Causes of capital ows in developing countries. Journal of
International Money and Finance, 19(2), 235253.
Kumar, R. (2011). Determinants of FIIs in India: Evidence from Granger
causality test. South Asian Journal of Marketing and Management
Research, 1(1), 6168.
Kwiatkowski, D., Phillips, P. C. B., Schmidt, P., & Shin, Y. (1992).
Testing the null hypothesis of stationarity against the alternative of a
unit root: How sure are we that economic time series have a unit root?.
Journal of Econometrics, 54(13), 159178.
Levy, H., & Sarnat, M. (1970). International diversication of investment
portfolios. American Economic Review, 60, 668675.
Lewis, K. (1999). Trying to explain home bias in equities and consumption. Journal of Economic Literature, 37(2), 571608.
Lintner, J. (1965). The valuation of risk assets and the selection of risky
investments in stock portfolios and capital budgets. The Review of
Economics and Statistics, 47(1), 1337.
Markowitz, H. M. (1959). Portfolio selection: Ecient diversication of
investments. Cowles foundation monograph 16. New York: John Wiley
and Sons.
Mody, A., Taylor, M. P., & Kim, J. Y. (2001). Modeling economic
fundamentals for forecasting capital ows to emerging markets.
International Journal of Finance and Economics, 6(3), 201216.
Montiel, P., & Reinhart, C. M. (1999). Do capital controls and
macroeconomic policies inuence the volume and composition of
capital ows? Evidence from the 1990s. Journal of International Money
and Finance, 18(4), 619635.

Obstfeld, M. (2009). International nance and growth in developing

countries: What have we learned? IMF sta papers, Vol. 56, No. 1.
Pagano, M., Roell, A. A., & Zechner, J. (2002). The geography of equity
listing: Why do companies list abroad?. The Journal of Finance, 57(6),
26512694.
Persson, T., & Svensson, L. E. O. (1989). Exchange rate variability and
asset trade. Journal of Monetary Economic, 23(3), 485509.
Pesaran, M. H., & Pesaran, B. (1997). Working with Microt 4.0: An
interactive econometric software package. Oxford: Oxford University
Press.
Pesaran, M. H., Shin, Y., & Smith, R. (2001). Bounds testing approaches
to the analysis of level relationships. Journal of Applied Econometrics,
16(3), 289326.
Pesaran, M. H., & Shin, Y. (1999). An autoregressive distributed lag
modelling approach to cointegration analysis. In S. Strom (Ed.),
Econometrics and economic Theory in the 20th century: The Ragnar
Frisch centennial symposium. Cambridge: Cambridge University Press.
Phillips Peter, C. B., & Perron, P. (1988). Testing for a unit root in time
series regression. Biometrika, 75(2), 335346.
Rai, K., & Banumurthy, N. R. (2004). Determinants of foreign institutional investment in India: The role of return, risk, and ination. The
Developing Economies, 42(4), 479493.
Richards, A. (2002). Big sh in small ponds: The trading behaviour and
price impact of foreign investors in Asian emerging equity markets.
Research discussion paper 200405. Reserve Bank of Australia,
Sydney.
Sharpe, W. F. (1964). Capital asset prices: A theory of market equilibrium
under conditions of risk. The Journal of Finance, 19(3), 425442.
Solnik, B. H. (1974a). Why not diversify internationally rather than
domestically?. Financial Analyst Journal, 30, 4854.
Solnik, B. H. (1974b). An equilibrium model of the international capital
market. Journal of Economic Theory, 8(4), 500524.
Srinivasan, P., & Kalaivani, M. (2013). Determinants of Foreign institutional investment in India: An empirical analysis. MPRA paper no.
43778.
Stulz, R. M. (1981a). A model of international asset pricing. Journal of
Financial Economics, 9, 383406.
Stulz, R. M. (1981b). On the eects of barriers to international investment.
Journal of Finance, 36, 923934.
Stulz, R. (1999). International portfolio ows and security markets. In M.
Feldstein (Ed.), International capital ows. University of Chicago
Press.
Taylor, M. P., & Sarno, L. (1997). Capital ows to developing countries:
Long- and short-term determinants. World Bank Economic Review,
11(3), 451470.
The Times of India, May 12, 2012.
The World Bank (1997). Private capital ows to developing countries: The
road to nancial integration. World Bank Policy Research Report,
USA: Oxford University Press.
Verma, R., & Prakash, A. (2011). Sensitivity of capital ows to interest rate
dierentials: An empirical assessment for India. RBI working paper 7/
2011.

Available online at www.sciencedirect.com

ScienceDirect