CHAPTER 1

INTRODUCTION

Foreign Direct Investment (FDI) has become a major source of foreign capital
flow and a leading source of external finance for developing economies like India. It
comprises both inward foreign direct investment (IFDI) and outward direct
investment (ODI). IFDI is the investment made by non-resident investors in the
reporting country while ODI is the investment made by residents of the reporting
country to other economies. FDI is linked to globalisation which can be summarised
as opening-up of markets, transfer of technology, capital and people. Another main
aspect of globalisation is multilateralism. It would be noticeable that the former
cannot be effectual without later. One of the major objectives of international
economic reforms was to promote multilateralism. The economic base of
multilateralism lies in allocative efficiency. This means that economies are able to
export to the best destinations and import from most efficient sources. In this
respect, Foreign Direct Investment (in the form of multilateralism) implies
importing capital from a variety of sources as may be most efficient rather than
restricting to a bilateral basis. The facade of this fact would be to export capital
where it can be most efficiently utilised by combining capital with other resources
optimally. The base of this capital transfers is to have gains for both the home
economy as well as to host economy.

Over the last couple of decades, the importance of foreign direct investment
(both inward and outward) as a source of capital in the developing world has
increased significantly because it is non-debt creating and non-volatile source of
finance. There are dichotomous views on the effect of FDI on any economy. One of
the important findings reveals that impact of FDI totally depends upon the
absorptive capacity of host country (Haddad and Harrison, 1993). In this respect, it
is very important to understand, analyze and interpret whether FDI is boom or bane
for the overall growth and development of Indian economy.

To justify the need of the present study, we have discussed the review of the
related literature in the second chapter separately.

1
1.2 FDI SCENARIO IN INDIA SINCE INDEPENDENCE: IN BRIEF

India was exploited a lot before independence. Its trade was fraught, poverty
was a major challenge and there was no industrial development. But there was hope
for future projection that one day situation would improve. After independence,
Indians were very serious about foreign market trade and they did not want any
foreign firm to enter again in Indian subcontinent and exploit their resources. It was
all because of the fear of previous exploitation done by the Britishers. Government
of India made stringent laws and legislation for foreign investment and foreign
companies, so that they could not enter into Indian market easily. The Government
of India enacted strict legislation for regulating and controlling FDI inflows and
these regulations can be divided into following phases:

a. The phase of timid and discriminating attitude towards FDI (1948-1967).

b. The phase of selective and restrictive policy (1967-79).
c. The phase of semi-liberalization (1980-1990).
d. The phase of open door policy for FDI (1991 onwards).

After independence, FDI inflows in India were very low because of the strict
policies. Foreign investment was restricted in most of the sectors indicating the
aspiration of Indian Government to have less participation of foreign enterprises in
the Indian economy during the first phase. During the second phase (1967-79), the
big industrial houses and foreign enterprises were permitted to setup industries in the
core and heavy investment sectors. The foreign participation was restricted due to
Foreign Exchange Regulation Act (FERA), 1974. Only 40 per cent foreign
participation in the equity was allowed.

In the third phase, during the 1980-1990, the overall inflows of FDI fluctuated.
In 1982-83, Government of India liberalised facilities with regard to bank deposits
and equity shares of the corporate sector. RBI simplified the exchange control
procedural formalities to assist such investments. Government also provided various
relaxations to foreign investors, NRIs through exemption of taxes and removal of
quantitative ceilings. As per the International Financial Statistics Yearbook (2003),
almost all the countries except Germany were having positive FDI inflows in India
from 1980 to 1991. The top five countries which invested in India during 1981 were

2
USA, Germany, UK, Japan and Switzerland. Altogether they accounted for 86 per
cent of total FDI inflows in India. In 1990, top five countries who invested in India
were same except Japan that was out of the top five lists and Italy was among them
and these countries accounted for 57 per cent of total FDI inflows. At this time,
Government of India focused only on the internal part of the economy because of
large political upheavals in the form of Morarji Desai’s Government and Emergency
of 1984. India was more focussed on inward looking policy. There were high
fluctuations in inward FDI in India during the time period from 1981 to 1990 (see
factual data in chapter 4).

Fourth phase, 1991 onwards is known as the liberalised phase of Indian

economy. After mid 90’s Indian economy faced political disturbances and with other
economic issues that gave rise to severe financial crisis. At the same time, there was
a sudden breakout of Gulf war in January 1991, which led to Balance of Payment
crisis. Government of India just had two weeks reserves to pay for its imports. No
foreign companies were interested to invest in Indian economy leading to
downgrading of international credit. Consequently, India opened its doors for
foreign capital in 1991. Although, Government of India tried to boost the morale of
the domestic firms encourage them to manufacture more as well as explore new
market. In 1991-92, FDI inflows were US $129 million and it has increased to US
$61963 million in 2017-18. The credit of these increasing FDI inflows goes to
liberal policy of Government of India towards foreign capital (see factual data in
chapter 4).

According to International Financial Statistics Yearbook (2003), total 15,998

investment proposals were submitted by foreign firms, during the post-reform period
from 1992 to 2002. These proposals involved majorly for non-manufacturing sector,
bulk of the approvals were given to service sector. The one of the drawback of
Indian reform system was that they didn’t focus on the development of
manufacturing area. As an outcome, the economic liberalization process was taken
under Structural Adjustment Programme (SAP) with the support of IMF and World
Bank. There were series of reforms in 1991. New Industrial Policy (NIP) 1991
realized the role of FDI in industrial development in India by generating higher
competitiveness, efficiency and modernization. In NIP 1991, there were huge
3
changes like abolition of industrial licensing system except 18 industries which were
involved in manufacturing sector, risky (hazardous) chemicals and items of national
and social well-being; removal of restrictions under Monopolistic and Restrictive
Trade Practices (MRTP) Act, 1969. Foreign Investment Promotion Board (FIPB)
was established for providing single window clearance to companies. Dual approval
system was introduced which consists of automatic approval with 35 priority sectors
and secondly through formal Government approvals. Removal of restrictions on FDI
in low technology sectors and removal from necessary technology agreements
conditions etc. were the major steps of the reforms. The main objective behind these
reforms was to transforming India into an investor friendly nation for FDI. The FDI
limits and approval procedure in various sectors of the economy according to new
consolidated FDI policy 2017 is shown in following the table 1.1.

TABLE: 1.1
FDI LIMITS IN DIFFERENT SECTORS AND ENTRY ROUTES IN INDIA
S. FDI Limit Entry Route &
Sector/ Activity
No. (%) Remarks
1 Agriculture & Animal Husbandry 100 Automatic
2 Plantation Sector 100 Automatic
3 Mining 100 Automatic
4 Petroleum & Natural Gas (Exploring activities) 100 Automatic
Petroleum refining by the Public Sector
49
Undertakings (PSU), without any Automatic
disinvestment
5 Defence Manufacturing 100 Automatic up to
49%; Above 49%
under Government
route
6 Broadcasting 100 Automatic
7 Broadcasting Content Services 49 Government
Up-linking of Non-‘News & Current Affairs’ 100 Automatic
TV Channels/ Down-linking of TV Channels
8 Print Media (Newspaper, periodicals, of Indian
26
editions of foreign magazines dealing with Government
news and current affairs)
Publishing/printing of scientific and technical Government
100
magazines/specialty journals/ periodicals
9 Civil Aviation – Airports 100 Automatic
Air Transport Services Automatic upto
49% (Automatic up
100 to 100 % for NRIs)
Government route
beyond 49 %
10 Other Services under Civil Aviation Sector 100 Automatic

4
11 Construction Development: Townships,
100
Housing, Built-up Infrastructure, Hospitals Automatic
premises, education institutions
12 Industrial Parks (new & existing) 100 Automatic
Satellites- establishment and operation, subject
100
13 to the sectoral guidelines of Department of Government
Space/ISRO
14 Private Security Agencies 74 Automatic up to
49%; Government
route beyond 49 %
and up to 74%
15 Telecom Services 100 Automatic up to
49%; Above 49%
under Government
route
16 Cash & Carry Wholesale Trading 100 Automatic
17 E-commerce activities 100 Automatic
18 Single Brand retail trading 100 Automatic up to
49%; Government
route beyond 49 %
19 Multi Brand Retail Trading 51 Government
20 Duty Free Shops 100 Automatic
21 Railway Infrastructure 100 Automatic
22 Asset Reconstruction Companies 100 Automatic
23 Banking- Private Sector 74 Automatic up to
49%; Government
route beyond 49%
and up to 74%
24 Banking- Public Sector 20 Government
26 Credit Information Companies (CIC) 100 Automatic
Infrastructure Company in the Securities
49
27 Market Automatic
28 Insurance 49 Automatic
29 Pension Sector 49 Automatic
30 Power Exchanges 49 Automatic
31 White Label ATM Operations 100 Automatic
32 Financial services activities regulated by RBI, Automatic
100
SEBI, IRDA or any other regulator
33 Pharmaceuticals 100 Automatic
Food products manufactured or produced in
100
34 India Government
Prohibited Sectors:
35 Lottery, Betting and gambling, Nidhi company, Chit funds, Real estate business or
construction of farm houses, Manufacturing of cigars, cigarillos, cheroots and
cigarettes, of tobacco or of tobacco substitutes, Trading in Transferable
Development Rights (TDRs), Sectors/ Activities not open to private sector
investment e.g. (I) Atomic Energy and (II) Railway operations (other than permitted
activities).
Source: Consolidated FDI Policy 2017, DIPP
Note: Detailed table in Annexure I

5
1.3 NEED & OBJECTIVES OF THE STUDY

India is one of the fastest developing economies which has introduced

Liberalization, Privatisation and Globalisation (LPG) in 1991 and part of it, India
has relaxed the FDI regulatory framework on selective basis. The FDI plays an
important role in developing economies including Indian economy as it is a non-debt
creating long- term capital source and non-volatile in nature. It acts like stimulus to
competition, savings, innovation and capital formation leading to job creation,
industrial growth and economic development. There are dichotomous views
regarding the impact of FDI on the economic growth. Many researchers considered
that it generate many positive impacts, while some highlights its growth retarding
impact. But for developing economy like India, it can work as additional source of
finance. So there is a logic and rationale to study the growth, trends and pattern of
FDI in India on one hand and to analyze the performance of FDI in form of
relationship between FDI and economic growth of Indian economy on the other.

From the last few decades almost all the countries (developed as well as
developing nations) initiated to attract more and more FDI. It is because, some of the
Asian countries known as Asian Tigers like Singapore, China, Hong Kong and
South Korea has attain their status among the world economy by utilizing more
foreign investment. Under the globalization process, rapid development in various
parts of the world is taking place that has raise the question of attracting FDI for
economic growth and development in the emerging and developing economies.
India has also adopted liberal policy towards FDI since 1991 for economic growth.
This positive and open door policy followed by India towards foreign investment is
in great contrast to it’s an early restrictive approach. Due to the bad experience
during the British colonialism, some economic and social thinkers have been
criticizing the foreign investment. In this regard, it is very crucial to examine the
relationship between Foreign Direct Investment (FDI) and economic growth of
India, which will be helpful for the policy makers to evaluate and implement
suitable policies.

Many studies have been made from time to time all over the world to assess
the contribution of FDI in economic growth of various economies. Keeping this idea

6
in mind, present study is a humble attempt to analyse the growth and trend of FDI
inflows in India and its relationship with economic growth of India. In the light of
the above, we have set the following objectives for the present study:

1. To review the relevant literature in the field of FDI in general and its
relationship with economic growth and development in particular.

2. To study the pattern of concentration & dominance of FDI flows at

international level and the status of India in World’s FDI flows.

3. To analyze the growth, trend and pattern of Foreign Direct Investment Inflows
in Indian economy.

4. To access and analyse the determinants of Foreign Direct Investment inflows

in India that is to examine the most significant factors influencing FDI inflows
in Indian economy.

5. To examine the performance of Foreign Direct Investment in Indian economy

by analysing the relationship between FDI and economic growth of India.

6. To draw policy implications flowing from the results of the study for making
FDI as an engine of growth and development of Indian economy.

1.4 HYPOTHESES

The present study is confined to the time period during 1981-82 to 2017-18.
On the basis of existing researches in the field of FDI, the following null hypotheses
have been framed for the present study:

1. There is positive trend of FDI inflows in Indian economy over the period from
1981-82 to 2017-18.

2. There is positive impact of FDI inflows on economic growth of Indian

economy over the period 1991-92 to 2016-17.

1.5 SOURCES OF DATA

The present study is based on secondary data, which has been collected from
various sources such as newsletters of Secretariat of Industrial Assistance (SIA),

7
Department of Industrial Policy and Promotion (DIPP), Ministry of Commerce and
Industry, Government of India; Handbook of Statistics on Indian Economy by
Reserve Bank of India; various World Investment Reports (WIR) by United Nation
Conference on Trade and Development (UNCTAD) and from World Bank.

To analyze the status of FDI flows in Indian economy at the international level
from 1991-92 to 2017-18, data has been taken from various World Investment
Reports (WIR) by UNCTAD. To examine growth, trend and pattern of FDI inflows
in Indian economy, data from 1981-82 to 2017-18 is considered. The data has been
compiled from the various newsletters published by Secretariat of Industrial
Assistance (SIA), Department of Industrial Policy and Promotion (DIPP). To study
the performance of FDI inflows in Indian economy that is to explore the relationship
between FDI inflows and economic growth of India, data from 1991-92 to 2016-17
has been taken.

1.6 RESEARCH METHODOLOGY

The various econometric techniques are used to achieve the objectives of the
study. The present study aims and contemplates to use recent econometric
techniques, some of which are still emerging and have originated in the last few
years. Our study extensively depends upon time-series data for the purpose of
analysis. Generally, a time-series is a sequence of values of particular variable over
some period of time. Time series data is more challenging to analyze than cross-
sectional data in the fact that economic observations can rarely, if ever, be assumed
to be independent across time. An obvious characteristic of time-series data which
distinguishes it from cross-sectional data is that a time series data comes with a
temporal ordering. Second, economic time-series satisfies the intuitive requirements
for being outcomes of random variables. This current section aims to specify the
research techniques in order to achieve our research objectives.

To analyze the position of Indian economy at world’s FDI flows, the study
first presents the nature and extent of international flow of FDI during the time
period which is considered for the study. The available data have been processed
and presented in form of suitable tables and figures. Further, the study analyzed

8
growth, trend and pattern of FDI inflows in India. In the present study, to find out
growth rate following two methods are used:

 Annual Growth Rate (AGR)

 Compound Annual Growth Rate (CAGR)

1.6.1 Annual Growth Rate (AGR)

Annual growth rate is calculated by using following formula:

AGR = (Xt-Xt-1/Xt-1) × 100

Here Xt = Value of X variable in t time period
Xt-1 =Value of X variable in t-1 time period

1.6.2 Compound Annual Growth Rate (CAGR)

Ordinary Least Square (OLS) technique is used to compute compound annual

growth rate, by fitting the exponential function to the available data and exponential
trend equation is defined as

Y = ABt
Where B = 1+g and g is the compound growth rate.
The logarithmic transformation of this function is as:

Log Y = Log A + t Log B

Or Y* = b0 + b1t ...................................... (1.1)
Where, Y*= Log Y
b0= Log A
b1= Log B
Which is a log linear function.
The values of parameters b0 and b1 in equation (1.1) are estimated by using Ordinary
Least Square (OLS) method. The Compound Annual Growth rate (CAGR) is
computed by using following formula:

^
CAGR (g %) = [Antilog (b1) – 1] × 100

9
The analysis is based on interpretations of CAGR. To test the significance of
exponential model the test is applied as follows:


b1
t 
*

S.E. 
b1


Where b1 = estimate of the slope of the trend and

S.E. 
b1
= standard error of the slope estimate b1

Where the regression line is

Log Y = Log A + t Log B

After analyzing the growth rate of FDI flows at international level and for
India, an attempt is also made to examine the performance of FDI in Indian
economy by evaluating the relationship of FDI inflows in economic growth of
Indian economy for the time period from 1991-92 to 2016-17. To investigate the
relationship between FDI and economic growth of India, the data on Gross
Domestic Product (GDP) at constant price is used as a proxy for economic growth
and Foreign Direct Investment equity inflows is taken as proxy for Foreign Direct
Investment (FDI). The data of GDP and FDI equity is expressed in billions of
rupees. To investigate the relationship between two variables firstly, simple linear
regression model is used as follows:

1.6.3 Simple Linear Regression Model

In the regression model, GDP is taken as dependent variable and FDI is taken
as independent variable. We have used natural log transformation to determine the
degree of sensitivity of the dependent variable to change in the explanatory variable.
The simple linear regression model is as follows:

GDPt = α0 + α1 FDIt + Ut. ....................................... (1.2)

The logarithmic transformation of above model can be written as

In GDPt = α0 + α1 In FDIt + Ut ...................................... (1.3)

10
 The above equation (1.3) assumes no time lag between GDP and FDI.
However, to identify the time lag, lag regression model or regression model with
time lag is used. In this model, GDP is regressed on past values of FDI.

1.6.4 Lag Regression Model Approach/ Regression Model with Time Lag

The relationship between GDP and FDI can also be studied through a lag
regression model with time lag of the form as follows:

GDPt = α0 + α1 FDIt-1 + Ut, when time lag is 1

GDPt = α0 + α2 FDIt-2 + Ut, when time lag is 2
GDPt = α0 + α3 FDIt-3 + Ut, when time lag is 3 ..................................... (1.4)
..………………………………………...........
The above model explains that GDP of period‘t’ depends on past values of
FDI of period‘t-k’ where k goes from 1 to 24. In the above model, GDP is regressed
on each FDI individually through Ordinary Least Square.

The more scientific approach to examine the causal relationship between FDI
and GDP is the Granger Causality Test and Vector Error Correction Model
(VECM).

Now we will present the theory of stationarity, unit root testing procedure,
develop the concept of cointegration, Granger Causality test and VECM.

Most of time series data whether financial and economic are non-stationary,
which means that they have unit root and which will cause to spurious regression.
Moreover, the results based on classical regression analysis will not be valid and
reliable. In order to avoid violation of assumptions of classical regression model, the
time series, if it is trend stationary, is detrended, or if it is difference stationary; it is
differenced.

Multiple Unit Root test can be applied to confirm stationarity. The time series
are cointegrated if they are I (1) prior to differencing but become I (0) after
differencing and a linear combination of these series is stationary. Thus, the
economic forces try to restore equilibrium when these time series drift apart. It is
possible that they do not have long-run association ship but short-run association

11
ship exists and they may drift subsequently. How two time series are related and
how they influences each other is briefly described by Granger Causality and
VECM.

1.6.5 Stationarity of Time Series

A time series is severely stationary if the distribution of its values remains the
same as the time progresses. A covariance stationary or weakly stationary time
series satisfies the following conditions:

a. E (Yt) = μ (constant) for all t = 1, 2, 3……∞;

b. V (Yt) = E (Yt - μ) 2 = σ2 for all t;

c. Cov. (Yt, Yt+k) = constant for all t and k≠0,

This above equations implies that its mean, variance and auto covariance
remains constant over the specified time.

Stationarity is important for the persistent analysis of data, if the time series is
non-stationary then the results of classical regression analysis will not be valid. OLS
regression of non-stationary time series is meaningless and such phenomenon is
known as nonsense or spurious regression. The non-stationary time series can
provide a high R2 and very high value of t–ratios, even though the series are
unrelated. The standard assumptions for asymptotic analysis will not hold good after
running regression. The F-statistics will not follow F-distribution and t-ratio will not
follow t-distribution. So, it is essential to regress only a stationary time series over
another that is stationary, in order to have meaningful regression. In spurious
regression there may exist very strong first order auto correlation measured by very
small value of Durbin Watson‘d’ statistics. According to Granger & Newbold, if
R2> d or if R2 ≈ 1 then it indicates spurious regression.

1.6.6 Test for Stationarity

Stationarity of time series can be examined by using graphical analysis and

unit root tests.

12
1.6.6.1 Graphical Method

With the graphical method, stationarity of given variables can be examined at

level data, at first difference and at second difference.

1.6.6.2 Unit Root Test

Significant study on the existence of unit roots in macroeconomic time series

is done by Nelson and Plosser (1982). Their research is considered as an important
contribution to test stationarity and has lasting influence on future econometric
research and analysis. The presence or absence of unit roots is helpful in the
recognition of the hidden patterns of data. Non-existence of unit root implies
stationarity resulting in mean reversion in the sense that time series fluctuates
around a constant long-run mean. Absence of unit root implies that the series has
finite variance which is invariant of the time and the influence of transmitted shock
dissipates over the time horizon. A time series with unit root is characterized as non-
stationary process as it does not return to a long run deterministic path. The variance
of the time series tends to infinity as it is time dependent. Further, a non-stationary
time series suffers from permanent effect from random shock. A series with unit root
follows a random walk.

The procedure of the unit root test is to regress the equation (1.5)

Yt = ρYt-1 + Ut and -1 ≤ ρ ≤ 1 ..................................... (1.5)

If ρ = 1, then there is unit root problem that is Y is non-stationary. If │ρ│<1, then

there is no unit root problem that is Y is stationary. For theoretical reasons subtract
Yt-1 from both sides of the above equation and we obtain.

Yt – Yt-1 = ρYt-1 – Yt-1 + Ut

Yt – Yt-1 = (ρ – 1) Yt-1 + Ut

∆Yt = δYt-1 + Ut ...................................... (1.6)

Where, ∆Yt = Yt – Yt-1 and δ = ρ-1

13
 In practice, instead of estimating equation (1.5), the equation (1.6) is used to
test the null hypothesis that H0: δ=0 against an alternative hypothesis HA: δ0. If δ=
0, then ρ=1 that is there is unit root in the model which means time series under
consideration is non-stationary. But if δ <0, then time series is stationary. In order to
find out whether the estimated coefficient of Yt-1 in (1.6) is zero or not, two tests are
used:

 Dickey-Fuller (DF) Test

 Philips Perron (PP) Test

1.6.6.2. A. Dickey-Fuller (DF) Test

Dickey and Fuller have shown that under the null hypothesis H0: δ= 0, the
estimated t value of the coefficient of Yt-1 in (1.6) follows the τ statistic (Tau-
Statistic). The DF test is estimated in three different forms, that is, three different
null hypotheses.

Yt is random walk without drift: ∆Yt = δYt-1 + Ut ......................................(1.7)

Yt is random walk with drift: ∆Yt = β1 + δYt-1 + Ut ......................................(1.8)

Yt is random walk with drift around a stochastic trend:

∆Yt = β1 + β2 t+ δYt-1 + Ut ......................................(1.9)

In above three cases, the null hypothesis is that

H0: δ=0 (existence of unit root problem that is non-stationary) and

HA: δ0 (existence of no unit root problem that is stationary).

Ordinary Least Square (OLS) has been applied to estimate (1.7), (1.8) and (1.9) and
τ statistic is computed by using the following formula.



τ* =
S.E. 

14
 If the computed absolute value of τ statistic exceed the Dickey-Fuller critical
value, then reject null hypothesis and conclude that time series under consideration
is stationary, which is pre requisite condition for the application of Granger test.

And if the computed absolute value of τ statistic does not exceed the Dickey-
Fuller critical value then null hypothesis is accepted which implies the time series
under consideration is non-stationary. In this case Granger test cannot be applied.

The above same procedure is adopted to apply Philips Perron test to examine
the unit root problem.

1.6.6.2. B. Philips-Perron (PP) Test

To overcome the limitations of Dickey-Fuller test, which assumes that error

terms are serially uncorrelated and homogeneous, Phillips & Perron (1988) test is
conducted. It adjusts the DF test to take care of possible serial correlation in error
terms by adding the lag difference term of the dependent variable. The PP test, like
ADF, can also be performed with a constant, a constant with linear trend or neither.
The Phillips-Perron (1988) test for unit root uses non-parametric method to contain
the serial correlation in the error term without adding lagged terms. The PP test is
reliable and valid even though the disturbance terms are auto-correlated and
heterogeneous. The asymptotic distribution of PP’s t-statistics is the same as ADF’s
t-statistics; therefore, the MacKinnon (1991) critical value will be valid.

The null hypothesis of existence of unit root is same for both ADF and PP test.
Classical hypothesis testing suffers from limitation as it ensures that the null
hypothesis is not rejected unless there is evidence against it. These tests, because of
their low power, often indicate existence of unit root. Kwiatkowski et al. (1992),
therefore, suggested that tests based on stationarity as null can be used for
confirmation for unit root

1.6.7 Cointegration

In the literature, cointegration is the existence of long run equilibrium

relationship between X and Y. According to Granger (1988), only if X and Y are
cointegrated, then standard tests for causality are valid. Therefore, an important

15
precondition to causality testing is to check the co-integrating properties of the
variable under consideration. For this Yt is regressed on Xt as follows:

Yt = α0+ α1Xt + Ut .....................................(1.10)

The above regression is known as the cointegrating regression and slope parameter
(α1) is known as cointegrated parameter.

Engel and Granger (1987) first introduced the cointegration test and then it
was developed and modified by Stock and Watson (1988) and Johanson and Juselius
(1990). The test is very useful to examine the long run equilibrium relationships
between the variables. In the present context, we used following tests:-

 Engel-Granger test

 Co-integrating Regression Durbin-Watson (CRDW)

In Engel-Granger Test (1987), firstly residuals (Ut) are obtained by running
the following regression:

Yt= α0 + α1Xt + Ut &

Ut = Yt - α0- α1Xt

Secondly, Dickey-Fuller test for stationarity is applied on residuals (Ut)


U t   U t 1

Rejection of H0 suggests that Ut is stationary, in which case Yt and Xt are

cointegrated.
An alternative and quicker method for testing whether Y and X are cointegrated is
the CRDW, Sargan and Bhargava firstly provided their critical values.

In CRDW test, the Durbin-Watson‘d’ is obtained from the cointegrating regression.

At 1, 5 and 10 per cent level of significance, the critical values of ‘d’ are 0.511,
0.386 and 0.322 respectively.

If the calculated value of ‘d’ is more than the critical values then it can be concluded
that Y and X are cointegrated and Granger test can be applied.

16
1.6.8 Granger Causality Test

Granger Causality test (1969) has significant importance as it allows analyzing

which variable leads or precedes the other. Granger Causality provides information
for which a time series variable consistently and predictably changes before another
variable. Following Granger (1969), the Granger-casualty test has been developed to
ascertain whether or not the inclusion of past values of a variable X do or do not
help in the prediction of present values of another variable that is Y.

 If Y variable is better predicted by including past values of X variable than by

not including them, X variables said to Granger-cause Y variable.

 Similarly, if past values of Y variable can be used to predict X variable more

exactly than simply using the past values of X variable, then Y is said to
Granger cause X variable.

 There is bi-directional causality, if the analysis reveals that X Granger causes

Y and Y also Granger causes X,

To avoid spurious causality, both of the variables under consideration should

be stationary. However, the prime requirement of this technique is to test the order
of integration and unit root test has been used to ascertain the order of integration.
Therefore, we first spotlight the concept of unit root test and then cointegration.

1.6.8.1 Granger Causality Test Equations

The test involves measuring the following pair of regressions:

m m
Yt = α0 + ∑ αi Xt -i + ∑ βj Yt-j + Ut .................................(1.11)
i=1 j=1
m m
Xt = α0 + ∑ αi Xt-i + ∑ βj Yt-j + Ut ................................(1.12)
i=1 j=1
The number of lags ‘m’ in the above regressions is arbitrary and boils down to
a question of judgment. Generally, it is best to run the test for a few different values
of ‘m’.

17
 Equation (1.11) postulates that current Y is related to past values of itself as
well as that of X and (1.12) postulates a similar behavior for X.

The following Steps involved in implementing the Granger Causality test:

Step-I, regress current Y on all past values of Y but do not include the lagged X
terms. This is the restricted regression. From this regression, we obtain the restricted
residual sum of squares (RSSR).

Step-II, now run the regression including lagged X terms. This is the unrestricted
regression. From this regression, obtain the unrestricted residual sum of squares
(RSSUR).

Step-III, the null hypothesis is H0:∑ αi = 0. In other words, the lagged X terms do
not belong in the regression.

Step-IV, to test this hypothesis, F- test is applied as shown below:

RSSR- RSSUR /m
F= ----------------------- and F  (m, n-2m-1)
RSSUR/n-2m-1
Where ‘m’ is the number of lags; ‘n’ is the number of observation involved in
the model.

If the calculated F-value exceeds the critical F-value at the chosen level of
significance, the null hypothesis is rejected, in which case the lagged X variable
belongs in the regression. This is another way of saying X causes Y. The same steps
are used to test that whether Y causes X.

1.6.9 Vector Error-Correction Model (VECM)

According to Engel and Granger (1987), in case variables are cointegrated

then there exists a related error correction model wherein short term movements of
variables are affected by the deviation from the equilibrium. If the variables are
cointegrated, VECM is useful for both long-term and short-term (Ratanapakorn and
Sharma, 2007). The VAR is incapable of exploring long-term relations as well as it
is deficient in discovering short-term relations in presence of cointegration
(Mukherjee and Naka, 1995). VECM is more appropriate model for several macro-

18
economic variables as it distinguishes between stationary variables with transitory
(temporary) effects and non-stationary variables with permanent (persistent) effects
(Johansen & Juselius, 1991).

After applying VECM, Wald test have been used to know the short-run
relationship among variables. Breusch-Godfrey Serial Correlation LM Test,
Breusch-Pagan-Godfrey Heteroskedasticity test and Jarque-Bera Normality test has
used for residual analysis of VECM estimates.

Granger Causality tells ways to know correlation between the current value of
one variable and past value of others, it does not imply that movement of one
variable causes movement of other variables. The VECM can easily distinguish
between the short-run and long run causality as it can capture both short-run
dynamics between time series and their long-run relationships (Mashie & Mashie,
1969). However, to achieve the research objectives, applying test and for obtaining
results, research software SPSS version 16 and E-Views 10 student version is used.

1.6.10 Variance Decomposition

In econometrics, variance decomposition is other applications of

multivariate time series analysis that is used to aid in the interpretation of a vector
auto regression (VAR) model once it has been fitted. The variance decomposition
explores the size of information each variable contributes to the other variables in
the auto regression. The variance decomposition estimates the amount (how much)
of the forecast error variance of each of the variables can be explained by exogenous
shocks to the other variables. The VECM consist of two endogenous variables,
namely, GDP and FDI. Therefore, the model takes into account two types of shocks;
some shocks are transmitted through FDI channel while others are transmitted
through GDP channel. To examine the reaction of FDI and GDP to such shocks in
the present study, the Variance Decomposition under VAR environment is worked
out. The Variance Decomposition throw light on the proportion of forecast error
variance of a variable which is explained by an unanticipated change in itself as
opposed to that proposition attributable to the change in other variables. It allows the
detection of the percentage proportions in the variance of a variable that is driven by
the shocks that occur in the other variables. According to Enders (2003), the

19
variance decomposition shows that in what degree a variable changes under the
impact of the own shocks or the other variable’s shock.

1.6.11 Principal Component Analysis

Principle Component Analysis (PCA) is one of the important multivariate data

analysis techniques. This is also called ‘Hotteling transform’ or ‘Karhunen-leove
(KL) Method’. It transforms a number of (possibly) correlated variables into a
smaller number of uncorrelated variables that are called principal components (Paul
L.C. et. al, 2013). As in regression analysis technique, the larger the number of
explanatory variables allowed, the greater is the chance of over fitting the model
leading to produce such conclusions that fail to generalise to other datasets. While in
principal component analysis, strong correlated explanatory variables are reduced to
a few principal components and then regression is run against them. Thus, Principal
Component Analysis (PCA) is a statistical technique applied to reduce the
dimensionality of a larger set of possibly correlated variables into a smaller set of
linearly uncorrelated variables called principal components in a manner that the first
principal component account for the largest possible variance from the original data
set and each successive principal component accounts for a variance smaller than
that of the preceding principal component (I.T. Jolliffe, 2002).

There are different terms used in principal component analysis. The diagonal
of the correlation matrix have unities and the factor matrix shows the full variance.
The matrix that contains the factor loadings of all the variables on all the factors
extracted is termed as factor matrix. The simple correlations between the factors and
the variables are termed as ‘factor loadings’. The eigen values refer to the total
variance explained by each factor and the standard deviation measures the
variability of the data. The main task of principal component analysis is to
recognize the patterns in the data and to express the data by highlighting their
similarities and differences. Thus, principal component analysis is applicable when
main concern is to explore the minimum number of factors that will account for the
maximum variance in the data in the particular multivariate analysis.

20
1.6.11. A. Steps to Conduct the Principal Component Analysis

There are mainly three steps in Principal Component Analysis (PCA). These
are as follows:

1. To check the inter-correlations amongst the items, correlation are calculated

and Correlation matrix is yielded.

2. From the correlation matrix, the inter-correlated items or ‘factors’ are

extracted to yield ‘Principal Components’.

3. For the purposes of analysis and interpretation, these ‘factors’ are rotated.

There are two statistical calculations that help to make the decision: Eigen
values and Scree plots. An eigen value is ratio of the shared variance to the unique
variance accounted for in the construct of interest by each ‘factor’ yielded from the
extraction of principal components. An eigen value of 1.0 or greater is the criterion
accepted in the present literature for deciding if a factor should be further
interpreted. The logic behind the criterion of 1.0 is that the amount of shared
variance explained by a ‘factor’ should at least be equal to the unique variance the
‘factor’ accounts for in the overall construct.

While the Scree plots offers a visual aid in deciding how many factors should
be interpreted from the principal components extraction. In scree plot, the extracted
eigen values from the data are plotted against the order of ‘factors’. The ‘elbow’
or factor at which the screen plot has a significant reduction in eigen value and then
level's off is often taken as the criterion for selecting the number of ‘factors’ to
interpret.

On the basis of eigen values and scree plot, decision on how many ‘factors’
should be extracted is taken. These extracted ‘factors’ of inter-correlated items are
rotated. This rotation is done because it is possible for certain items to be highly
inter-correlated with items on different factors. The rotation forces these items in to
the factor with which it have the strongest association with the items of the factor.
This rotation enhances the interpretability of extracted factors; however, it cancels
out the ability to interpret the amount of shared variance associated with the factor.

21
For interpreting the factors, it should be keep in mind that any item that does not at
least have a correlation or ‘factor loading’ of 0.3 with the factor it has loaded on
should be discarded. We have analysed the relationship between FDI and its
determinants with the help of following equation:

FDIt = f (GDP, INF, MS, GDCF, TT, DSR, FOREX, REER, WPI) …..............(1.13)

Taking log both sides and adding error terms

Ln FDIt = α0+ β1LnGDPt + β2LnINFt + β3LnMSt + β4LnGDCFt + β5LnTTt + β6LnDSRt

+ β7LnFOREXt + β8LnREERt + β9LnWPI + µt ………………(1.14)

Here

α0 = Intercept

β = Coefficient of respective variables

t = 1991-92 to 2016-17

µt = Error term

GDP = Gross Domestic Product

INF = Infrastructure

MS = Money Supply

GDCF =Gross Domestic Capital Formation (Domestic Investment)

TT = Total Trade

DSR = Debt Service Ratio

FOREX= Foreign Exchange Reserves

REER = Real Effective Exchange Rate

WPI = Inflation Rate

We have applied Principal Component Analysis (PCA) on above variables and

find principal component. To find out the relationship between extracted principal

22
component and FDI, the study applied principal component regression on extracted
principal component named as ‘economic determinants’ and Foreign Direct
Investment.

1.6.11. B. Principal Component Regression

Principal component regression is a regression technique that is based on

principal component analysis (PCA). It considers regressing the outcome (extracted
principal component) on a set of explanatory or independent variables based on a
standard linear regression model. In Principal Component Regression, instead of
regressing the dependent variable on the explanatory variables directly, the principal
components of the explanatory variables are used as regressors. The main reason for
using principal component regression is to overcome the problem of
multicolinearity. The regression model for analysis in present study is as follows:

FDIt = f (EDt) ........................................(1.15)

FDIt = α0+α1EDt + µt ........................................(1.16)

FDI= Foreign Direct Investment

α0 = Intercept

α1 = coefficients

ED = Economic determinants

µt = Error Term

1.6.12 Inward FDI Performance Index

To judge the performance of FDI, the Inward FDI performance index is

calculated. This index is developed by the UNCTAD to measure a country’s relative
position in the world in terms of FDI performance. It is the ratio of a country’s share
in world FDI flows to its share in world GDP and can be calculated as follows:

Here: FDIi = Foreign Direct Investment Inflows in India

23
FDIw = Foreign Direct Investment Inflows in World

GDPi = Gross Domestic Product of India

GDPw = Gross Domestic Product of World

A Country’s Inward FDI Performance Index is equal to one, if the share of

country’s global inward FDI matches its relative shares in global GDP. Greater than
one value indicates a larger share of FDI relative to GDP, implying that one attract
more FDI than it can be expected on the basis of their relative GDP size. A value less
than one shows that a smaller share of FDI relative to GDP. A negative value shows
the disinvestment by the foreign investor in that time period.

1.7 CHAPTER SCHEME

To develop the study in an organized manner, the present research has been
divided into six chapters. We have not used the word ‘introduction’ in every chapter
separately, as first paragraph of each chapter depicts point one that is the
introduction of the chapter. The chapter scheme of the present study is as follows:

Chapter 1: Introduction

Chapter 1 explores the importance of FDI in general and for India in particular. It
highlights FDI scenario in Indian economy since independence, need & objectives
of the study, data sources, research methodology of the study, chapter scheme and
limitations of the study as well as the concepts & symbols used in present study.

Chapter 2: Literature Review

The relevant literature has been reviewed on different dimensions of FDI in general
and for India in particular and its relationships with growth and development.
Inferences have been drawn from this literature to justify the need of the present
study in the second chapter.

Chapter 3: Status of India in World’s Foreign Direct Investment(FDI) Flows

This chapter explores the trend, growth and pattern of FDI flows at world level as it
will be helpful to find out the status of India at world’s FDI flow. The chapter

24
analyses the growth and trend of FDI flows (inflows & outflows) in world, Asia and
BRICS countries. The chapter also measures the inward FDI performance index for
India.

Chapter 4: Growth, Trends and Determinants of FDI Inflows in India

It deals with FDI Inflows in Indian economy. Broadly, the chapter is divided into
two parts. First, FDI Inflows in pre reform period that is before 1991. Second, FDI
inflows in post reform period that is after 1991. The chapter depicts country wise,
sector wise, region wise growth and trend of FDI inflows in Indian economy. The
chapter also explores and analyses the determinants of FDI inflows in India.

Chapter 5: Performance of Foreign Direct Investment (FDI) in Indian

Economy

The performance of FDI in Indian economy by analysing the relationship between

FDI and economic growth of India is the subject matter of Chapter 5. It deals with
econometric techniques that have been applied to explore the relationship between
FDI inflows and economic growth of India. Results of different econometric
techniques and their analysis are also presented in this chapter.

Chapter 6: Conclusions and Policy Implications

The last chapter provides the conclusions and draws policy implications. To
consolidate the answer of the research questions and objectives, this chapter
synthesizes the overall findings, which follows the research implications for
researchers and policymakers. The future scope of research also falls in the domain
of this chapter.

1.8 LIMITATIONS OF THE STUDY

The success of any study depends upon the availability of relevant information
and its proper use in terms of processing it, presenting it and exposing it to statistical
techniques. There are various limitations faced by all economic / scientific studies
and the present study is not an exception. The limitations of the present study are as
follows:

25
 The present study is based on secondary data; there are always some problems
with the data. The major problems are data gaps, change in the concepts and
coverage of data. In this context, very important lacuna is that data on some
variables are not available.

 The assumption is debatable that FDI is the only cause for growth of Indian
economy in the post liberalised period. There are no proper methods to support
the validity of this assumption available to segregate the effect of FDI.

 To examine the performance of FDI, FDI equity inflow is taken as proxy for
FDI and Gross Domestic Product is taken as proxy for economic growth.

 The study is open for all limitations of the Ordinary Least Square (OLS)
method, which are discussed in any standard book of econometrics.

 There has been a problem of sufficient and homogenous data from different
sources on outward direct investment (ODI) from Indian economy due to
which, it is not covered in the present study.

1.9 CONCEPTS & SYMBOLS USED IN THE STUDY

 Bi-directional: Existence of two way /directional relationship among

variables.

 Brown Field Investment: Brown-field investment is that investment in which

a foreign entity purchases an existing facility to start new production.

 Cointegration: Cointegration means existence of linear and long term

relationship among given variables.

 Equity Capital: Equity capital includes equity in branches, all shares in

subsidiaries and associates (except the non-participating and preferred shares
that are treated as debt securities and included under other direct investment
capital) and other capital contributions like provisions of machinery etc.

 FDI: “The numerical guideline of ownership of 10 per cent of ordinary shares

or voting stock determines the existence of a direct investment relationship.

26
 An effective voice in the management, as evidenced by an ownership of at
least 10 per cent, implies that the direct investor is able to influence or
participate in the management of an enterprise; it does not require absolute
control by the foreign investor.” (OECD Benchmark Definition).

 Green Field Investment: A green field investment is that investment in which

a foreign investor builds its operations in another country from the ground up.

 Lag Regression Model: Lag regression model is a model in which a

regression equation is used to predict current values of a dependent variable
based on the lagged (past period) values of this explanatory variable.

 Non-Stationary: A non-stationary time series is one whose statistical

properties such as mean, variance, autocorrelation, etc. are all not constant
over time.

 Other Direct Investment Capital (or Inter-Company Debt Transactions):

It consists the borrowing and lending of funds, including debt securities and
trade credits, between direct investors and direct investment enterprises as well
as between two direct investment enterprises which share the same direct
investor. It has been mentioned earlier that deposits and loans between
affiliated deposit institutions are recorded under other investment rather than
direct investment.

 Reinvested Earnings: Reinvested earning covers the direct investor’s share

(in proportion to direct equity participation) of earnings that is not distributed,
as dividends by associates or subsidiaries and income of branches not remitted
to the direct investor. If these earnings are not identified, all branches’
earnings are considered, to be distributed, by convention.

 Simple Linear Regression Model: Simple linear regression is a statistical

method that allows us to summarize and study relationships between two
quantitative variables; one is dependent and other is independent.

 Stationary: A stationary time series is that whose statistical properties like

mean, autocorrelation, variance, etc. are all constant over the period of time.

27
 Uni-directional: Existence of one way/directional relationship among
variables.

 Unit Root Test: In statistics, it is used to examine whether a time series

variable is non-stationary and possesses a unit root. The null hypothesis is
generally defined as the presence of a unit root and the alternative hypothesis
is stationarity.

 ADF: Augmented Dickey Fuller test

 BRICS: Group of Brazil, Russian Federation, India, China and South Africa

 CAGR: Compound Annual Growth Rate

 CIS: Commonwealth of Independent States

 CRDW: Cointegrating Regression Durbin Watson

 D.F.: Degrees of Freedom

 D.W.: Durbin Watson

 DF Test: Dickey Fuller test

 DIPP: Department of Industrial Policy and Promotion

 FDI: Foreign Direct Investment

 FERA: Foreign Exchange Regulation Act

 FIPB: Foreign Investment Promotion Board

 FPI: Foreign Portfolio Investment

 GDP:Gross Domestic Product

 Null Hypothesis

 Alternative Hypothesis

 FDI: Inward Foreign Direct Investment

28
 F: International Monetary Fund

 LPG: Liberalisation. Privatisation and Globalisation

 M&A: Merger and Acquisition

 MNCs:ulti-National Companies

 MNEs: :ulti-National Enterprises

 MRTP: Monopolistic and Restrictive Trade Practices

 NIP:New Industrial Policy

 ODI: Outward Direct Investment

 OECD:Organisation of EconomicCooperation and Development 

 OLS: Ordinary Least Square

 PCA: Principal Component Analysis

 PP Test: Phillips-Perron Test

 PPP: Purchasing Power Parity

 R2& Adjusted R2: Coefficient of determination 

 RSSR: Restricted residual sum of squares 

 RSSUR: Unrestricted residual sum of squares. 

 SAP: Structural Adjustment Programme

 S.E.: Standard Error 

 SIA: Secretariat of Industrial Assistance

 τ: Tau Statistic

 Ut: Random Disturbance Term

29
 UAE: United Arab Emirates

 UNCTAD: United Nations Conference on Trade and Development

 VAR: Vector Auto Regressive Model

 VECM: Vector Error Correction Model

 WIR: World Investment Report

1.10 CONCLUSIONS
An overview of Foreign Direct Investment in the Indian economy along with the
need, objectives of the study and sources of data falls in the domain of this chapter. The
framework to conduct the research in details quantitative and econometric
methodologies employed is presented here. Variety of econometrics tools and
techniques will be employed such as Augmented Dickey-Fuller (ADF) test, Philips-
Perron (PP) test and Johansen Cointegration technique, Granger Causality test, Vector
Error Correction Model (VECM) and Variance Decomposition method used to examine
long-run and short-run causal relationships between the variables. Principal Component
Analysis method will be used to explore the determinants of FDI inflows in India. It
also includes the chapter scheme of the thesis with limitations of the study, concepts and
symbols used.

30
REFERENCES:

 Constantin, C. (2014), “Principal Component Analysis- A Powerful Tool in

Computing Marketing Information”, Bulletin of the Transilvania University of
Brasov Series V: Economic Sciences, Vol. 7 (56) No. 2 – 2014.

 Consolidated FDI Policy (2017), Department of Industrial Policy and

Promotion, Ministry of Commerce and Industry, Government of India, August
2017.

 Dunning, John H. (1977), “Trade, Location of Economic Activity, and the

Multinational Enterprise: A Search for an Eclectic Approach”, in the
International Allocation of Economic Activity, ed. B. Ohlin, P. O. Hesselborn,
P. M. Wijkman. New York: Holmes and Meier.

 Dunning, John H. (1981), “International Production and the Multinational

Enterprise”, London George Allen & Unwin.

 Dunning, John H. (1998), “Location and the Multinational Enterprise: A

Neglected Factor?” Journal of International Business Studies, 29, 1 (First
Quarter 1998): 45-66.

 Enders, W. (2004), “Applied Econometric Time Series”, 2nd, ed. Hoboken,

NJ: John Wiley Sons, Inc., 2004.

 Engel, R.F. & Granger, C.W.J. (1987), “Co-integration and Error

Correction: Representation, Estimation and Testing”, Econometrica, 55, pp.
251-276.

 Granger, C.W.J. (1969), “Investigating Causal Relations by Econometric

Models and Cross-Spectral Methods”, Econometrica, 37, pp. 424-438.

 Granger, C.W.J. (1981), “Some Properties of Time Series Data & Their Use
in Econometrics Model Specification”, Journal of Econometrics. Vol. 16, 1,
pp. 121-130.

31
 Granger, C.W.J. (1988), “Some Recent Development in a Concept of
Causality”, Journal of Econometrics, 39, pp. 199-211.

 Granger & Newbold (1974), “Spurious Regression in Econometrics”,

Journal of Econometrics, Vol. 2, pp. 11-120.

 Haddad, Mona & Harrison, Ann (1993),“Are There Positive Spillovers

from Direct Foreign Investment? Evidence from Panel Data for Morocco”,
Journal of Development Economics, 42 (1), 51-74.

 Harbison, F. H. (1973), “Human Resources as the Wealth of Nations” Oxford

University Press, New York.

 Hymer, S. H. (1976), “The Multinational Operations of National Firms: A

Study of FDI” MIT Press, Cambridge, MA.

 International Financial Statistics Yearbook (2003), International Monetary

Fund. Statistics Dept, Aug. 29, 2003, Vol. 56, ISBN 9781589062665/0252-3027.

 Johansen and Juselius, K. (1990), “Maximum Likelihood Estimation and

Inference on Cointegration– with Applications to the Demand for Money”,
Oxford Bulletin of Economics and Statistics, Vol. 52, No. 2, pp. 169–210.

 Johansen & Juselius (1991), “Estimation and Hypothesis Testing of Co-

Integrating Vectors in Gaussian Vector Auto Regression Models”,
Econometrica, 59, pp. 1551-1580.

 Johansen, S. (1991), “Estimation and Hypothesis Testing of Cointegration

Vectors in Gaussian Vector Autoregressive Models,” Econometrica, Vol. 59,
No. 6, pp. 1551–1580.

 Jolliffe, Ian T. (1982),"A Note on the Use of Principal Components in

Regression", Journal of the Royal Statistical Society, Series C. 31 (3): 300–
303. doi:10.2307/2348005. JSTOR 2348005

 Kwiatkowski, Denis & Phillips, Peter C. B. & Schmidt, Peter & Shin,
Yongcheol, (1992), “Testing the Null Hypothesis of Stationarity Against the

32
 Alternative of a Unit Root: How Sure are We that Economic Time Series have
a Unit Root?," Journal of Econometrics, Elsevier, Vol. 54(1-3), pages 159-
178.

 MacKinnon, J. G. (1991), “Critical Values for Cointegration Tests,” Chapter

13 in Long-Run Economic Relationships: Readings in Cointegration, ed. R. F.
Engle and C. W. J. Granger. Oxford, Oxford University Press

 Mashi, A. & Mashi, R. (1969), “Empirical Tests to Discern the Dynamic

Causal Chain in Macroeconomic Activity: New Evidence from Thailand and
Malaysia Based on a Multivariate Co-Integration/Vector Error Correction
Model”, Journal of Policy Modeling, 18, pp. 531-60.

 Mukherjee, T.K. and Naka, Atsuyuki, N. (1995), “Dynamic Relations

between Macroeconomic Variables and the Japanese Stock Market: An
Application of a Vector Error Correction Model”, The Journal of Financial
Research, Vol. 28, no 2, pp. 223-237.

 Nelson, C. R. and Plosser, C. I. (1982), “Trends and Random Walks in

Macroeconomics Time Series Some Evidence and Implications”, Journal of
Monetary Economics 10(1982), pp- 139-162, North-Holland Publishing
Company.

 Organisation for Economic Co-operation and Development (1999), OECD

Benchmark Definition of Foreign Direct Investment, 3rd Edition, Paris.

 Paul, L.C., Suman, A.A., Sultan, N. (2013), “Methodological Analysis of

Principal Component Analysis (PCA) Method”, IJCEM International Journal
of Computational Engineering & Management, Vol. 16 Issue 2, March 2013,
ISSN (Online): 2230-7893 www.IJCEM.org

 Phillips, P. & Perron (1988), “Testing for a Unit Root in Times Series
Regression”, Biometrica, 75, pp. 335-346.

33
 Ratanapakorna, O. and Sharma, S.C. (2007), “Dynamic Analysis between
the US Stock Returns and the Macroeconomic Variables”, Applied Financial
Economics, 2007, 17, pp. 369–377.

 Romer, P.M. (1986), “Increasing Returns and Long-Run Growth”, Journal of

Political Economy, 94 (5), October, 1002-1037.

 Romer, P.M. (1990), “Endogenous Technological Change”, Journal of

Political Economy, 98 (5), Oct. (Part 2), S71-S102.

 Sargan, J.D. and Bhargava, A.S.(1983), “Testing Residuals from Least

Squares Regression for Being Generated by the Gaussian Random Walk”,
Econometrica, 51, 153-174.

 Seethapathi, K. (2006), “FDI Issues in Emerging Economics”, ICFAI

University Press,

 Stock, J.H. and Watson, M.W. (1988), “Testing for Common Trends”,
Journal of the American Statistical Association, April 1988.

 Tharwat, Alaa (2016), “Principal Component Analysis-A Tutorial”, DOI:

10.1504/IJAPR.2016.079733,https://www.researchgate.net/publication/30916
5405.
 https://statistics.laerd.com/spss-tutorials/principal-components-analysis-
pcausing-spss- statistics.php
 UNCTAD “World Investment Report” various issues, Geneva
 www.dipp.nic.in
 www.rbi.nic.in
 www.worldbank.org/
 www.unctad.org/en

34