Continuous Assessment

Submitted By:

Name College Roll Number

Tanveesha Handa 2215147

Labhyashikha Anand 2215128

Bhargab Jyoti Deka 2215149

Hardik Arora 2215148

Raghav Verma 2215114

Naitik Jain 2215117

Submitted To: Dr. Pratibha Madan

Course: B.A.(Hons.) Economics
Subject: Introductory Econometrics
Introduction
An Econometric Analysis of Foreign Direct Investment (FDI): Key Influences and
Analytical Approach

Foreign Direct Investment (FDI) is a cornerstone of global economic growth, representing

cross-border business investments aimed at gaining influence or control. Through technology
transfer, skill development, and infrastructure improvement, FDI fosters economic
integration. To understand what attracts FDI, it's crucial to examine key macroeconomic
factors such as real GDP growth rate, net trade of goods and services, and exchange rate.

Impact of GDP Growth Rate on FDI

The real GDP growth rate reflects the health and stability of a nation's economy. High GDP
growth often indicates rising consumer demand, profitability, and robust infrastructure,
making it attractive to foreign investors seeking market expansion. With a growing economy,
businesses anticipate expanding markets and improved profitability while minimising
economic risks. Thus, the GDP growth rate serves as a key indicator for investors assessing
the long-term potential of their investments.

Significance of Net Trade in Attracting FDI

Net trade of goods and services, or the trade balance, impacts FDI by highlighting a nation's
competitive strengths. A surplus reflects efficient production and export markets that foreign
investors can access for market penetration or supply chain establishment. On the other hand,
a deficit may initially signal market challenges but could also indicate investment
opportunities to address unmet demands. Thus, foreign investors analyse trade balances to
gauge a nation's economic strengths and weaknesses.

Influence of Exchange Rate on FDI

Exchange rates significantly affect foreign investment decisions. Stable exchange rates
minimise currency fluctuation risks, fostering favourable investment conditions. A
depreciating currency often attracts investment by making assets relatively cheaper, while
prolonged volatility can deter investors due to financial risks. As such, exchange rates are a
crucial consideration for long-term projects.

Using Gretl for Econometric Analysis

Gretl (GNU Regression, Econometrics and Time-series Library) is an open-source software

ideal for analysing FDI determinants. It has an intuitive graphical interface and scripting
language to specify models incorporating GDP growth, trade balance, and exchange rates.
Gretl's diagnostic tools validate models by checking assumptions like multicollinearity and
heteroskedasticity. Its time-series features include unit root tests and cointegration analysis to
uncover long-term trends.

Moreover, Gretl offers data visualisation tools to create plots and summary statistics, making
it easier to convey how economic variables influence FDI.

Research Objectives
In this analysis, we aim to understand how FDI, as the dependent variable, is influenced by
key economic indicators such as GDP growth rate, Exchange rate, and oil BOP account
balance.. Using Gretl, we construct a multiple regression model with FDI as the dependent
variable and the chosen indicators as independent variables. We rigorously test the
assumptions of the regression model, including linearity, independence, homoscedasticity,
and normality of residuals, to ensure the reliability of our findings. Interpreting the results
involves examining regression coefficients, t-values, R-squared, Durbin-Watson, and White
tests to understand the significance and reliability of the independent variables in explaining
inflation. Finally, we calculate and interpret the Total Sum of Squares (TSS), Explained Sum
of Squares (ESS), and Residual Sum of Squares (RSS) to evaluate the overall fit of the
regression model and the proportion of variance explained by the independent variables.

Literature review
Impact of exchange rate on FDI: A comparative study of India and China

Dr VB Khandare

Source:
https://www.researchgate.net/profile/Vilas-Khandare/publication/330080676_Impact_of_exc
hange_rate_on_FDI_A_comparative_study_of_India_and_China/links/5c2c55cba6fdccfc707
70189/Impact-of-exchange-rate-on-FDI-A-comparative-study-of-India-and-China.pdf

This study was conducted to examine the impact of exchange rates on foreign direct
investment in India and China. For analysing the impact of the exchange rate on FDI
correlation and regression analysis techniques have been used. It observed that from 1991 to
2014 foreign direct investment in India increased by 458.89 times in absolute terms whereas,
the FDI in China increased by 29.43 times in absolute terms during the same period. The
exchange rate shows a 2.68 times decrease in the value of the Indian rupee in terms of the US
dollar and a 1.15 times decrease in China's Yuan in terms of the US dollar during the study
period. It is found that there is a positive correlation between FDI and the exchange rate in
India. For China the correlation between FDI and exchange rate is negative.
On the Causal Links Between FDI and Growth in Developing Countries
Hanrick Hansen & John Rand

Source : https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1467-9701.2006.00756.x

Many recent studies analysing foreign direct investment and growth have shown a positive
association between FDI and GDP. But there seems to be ambiguity with respect to the
direction of causality.This study analyses the causal relationship between these variables in a
sample of 31 developing countries covering three continents over the time period 1970–2000.

Variable Interpretation

A) Real GDP growth (Annual per cent change)

Description: Real GDP growth represents the annual percentage change in the
inflation-adjusted value of all goods and services produced by an economy, providing a
measure of economic growth.
Source: https://www.imf.org/external/datamapper/NGDP_RPCH@WEO/IND?year=2024
International Monetary Fund (IMF)
Unit: Percent (%)
Range: -5.8% - 8.8%

B) Foreign Direct Investment (FDI) (Annual net inflows)

Description: FDI represents the amount invested by residents of foreign countries in the
domestic country in the form of assets.
Source: World Bank
(https://data.worldbank.org/indicator/BX.KLT.DINV.WD.GD.ZS?locations=IN)
Units: Number
Range: 2143628110 -64362364994

C) Exchange Rate
Average Annual 1 dollar into the Rupee exchange rate
Source:https://cimsdbie.rbi.org.in/DBIE/#/dbie/reports/Statistics/Financial%20Market/Forex
%20Market
(RBI)
Range: 33.4498-74.225

D )Net Trades in Goods and Services (BOP) (Annual Net inflows)

Description: It represents India’s Balance of Payment account.
Source: https://data.worldbank.org/indicator/BN.GSR.GNFS.CD?locations=IN
World Bank
Unit: Number
Range: -136063399026 – -5122207192

Estimation and interpretation

Multiple Regression

Regression Output
Covariance Output Matrix
The covariance output of regression coefficients indicates the relationship between the
coefficients of independent variables.
Positive covariance: A positive covariance between two coefficients shows a direct relation
between the variables.
Negative covariance: A negative covariance between two coefficients shows a direct relation
between the variables.
The magnitude of the covariance coefficient: The magnitude of the covariance coefficient
indicates the strength of the relationship between the coefficients. Larger values indicate a
stronger relationship.
Statistical significance: If the covariance is statistically significant (usually determined by the
p-value), it suggests that the relationship between the coefficients is unlikely to be due to
random chance.

ANOVA
Actual vs Fitted Line for FDI (For each year)

Predicted vs Actual FDI

Individual Hypothesis Testing
Objective: To determine whether individual explanatory variables, Real GDP Growth Rate,
net trade of goods and services and exchange rate have a statistically significant effect on the
dependent variable, FDI.

Confidence Interval:
The confidence interval represents the upper and the lower range in which the explanatory
variables will have a significant impact on the dependent variable.

Interpretation
95% confidence intervals for two-tail test of all the explanatory variables are given at 22
degrees of freedom.

t-tests, Critical values and p-values

Interpretation:

Real GDP Growth Rate:

H0: Real GDP Growth Rate is statistically insignificant.

H1: Real GDP Growth Rate is statistically significant
P-value(0.0839766) > Level of Significance(0.05), therefore, Ho is accepted i.e.Real GDP
Growth Rate is statically insignificant.

Exchange Rate:

H0: Exchange Rate is statistically insignificant.

H1: Exchange Rate is statistically significant
P-value(0.00937787) < Level of Significance(0.05), therefore, Ho is rejected i.e. Exchange
Rate is statistically significant.

Net Trade of goods and services :

H0: Net Trade of goods and services is statistically insignificant.

H1: Net Trade of goods and services is statistically significant
P-value(0.000151953) < Level of Significance(0.05), therefore, Ho is rejected i.e.Net Trade
of goods and services is statistically significant.

Summary Statistics

Restricted sum of squares

H0 : β1+ β2 Exchange rate+ei
H1: β1+ β2 Exchange rate + β3 Gdp growth rate + β4 Net Trades in goods and services

Test Statistic F(2,22) = 4.49705

Critical Values at 10% Level of Significance = 2.5613

Test Statistic > Critical Value . Therefore Null Hypothesis is rejected at 10% Level of
Significance

Critical Value at 5% Level of Significance = 3.4434

Test Statistic > Critical Value . Therefore Null Hypothesis is rejected at 5% Level of
Significance

Critical Value at 2.5% Level of Significance = 4.3828

Test Statistic > Critical Value . Therefore Null Hypothesis is rejected at 2.5% Level of
Significance

Interpretation -
Null Hypothesis H0 is rejected at 10%,5%,2.5% level of significance.This shows that the
restricted model is not better. Unrestricted model provides better results and there are no
improvements in choosing a model with only the exchange rate as an independent variable.
Model Specification test results

Tests for assumptions violation

Multicollinearity
Multicollinearity refers to the correlation of independent variables within a regression model,
making it challenging to isolate the unique effects of each predictor. This analysis focuses on
three economic indicators—Real GDP Growth, Net Trade in Goods and Services, and the
Exchange Rate—to understand their relationships and the potential issues ofmulticollinearity.
Using Variance Inflation Factors (VIF) and Belsley-Kuh-Welsch (BKW) collinearity
diagnostics, the results and implications are carefully evaluated.

Variance Inflation Factors (VIF) Analysis

The Variance Inflation Factor (VIF) assesses the degree of multicollinearity by measuring the
inflation of an estimated regression coefficient's variance due to correlations among the
predictors. In this analysis, the VIF values are 1.275 for Real GDP Growth, 1.371 for Net
Trade, and 1.545 for the Exchange Rate. Since all values are well below the threshold of
10, which typically signals problematic multicollinearity, these results suggest that none of
the predictors are highly correlated with one another. This finding indicates that each variable
provides unique information to the regression model.
Belsley-Kuh-Welsch Collinearity Diagnostics

Belsley-Kuh-Welsch diagnostics provide additional insights into multicollinearity by

analysing the eigenvalues and condition indices of the covariance matrix. The eigenvalues
reveal the spread of data information, and their disparities hint at potential collinearity
problems. The condition indices quantify the eigenvalue dispersion, with indices above 30
indicating strong linear dependencies. The analysis produced condition indices of 1.000,
3.445, 4.712, and 16.116, all well below 30. Thus, no strong collinearity issues are present,
although one index above 10 suggests a moderate correlation. Furthermore, the variance
proportions, which show the contribution of each predictor's variance to the condition
indices, revealed that the constant and exchange rate variables have proportions of
0.975 and 0.936, respectively, associated with condition indices above 10. This suggests
some multicollinearity involving these two variables, yet the correlation is not severe enough
to threaten the interpretability of the regression results.

Interpretation

The comprehensive analysis of VIF and Belsley-Kuh-Welsch diagnostics reveals that the
dataset does not suffer from significant multicollinearity. All VIF values are well below 10,
and no condition indices exceed 30, indicating that the regression coefficients are not
substantially inflated. The slight multicollinearity between the constant and the exchange rate
does not pose a major challenge for the model's reliability.

Heteroscedasticity test
As it relates to statistics, heteroskedasticity refers to the error variance, or dependence of
scattering, within a minimum of one independent variable within a particular sample.
To assess the presence of heteroskedasticity, we perform White's test, which is a general test
for heteroskedasticity. The null hypothesis of this test is that there is no heteroskedasticity in
the regression model.

Test Results:
Test Statistic: TR^2 = 9.476906
p-value: 0.3945

Interpretation:
The p-value of White's test for heteroskedasticity is 0.3945, which is greater than the
conventional significance level of 0.05. Therefore, we fail to reject the null hypothesis that
there is no heteroskedasticity in the regression model.

Conclusion:
Based on White's test, there is no evidence of heteroskedasticity in the regression model. The
assumption of homoscedasticity is not violated, and the OLS estimators remain efficient.

Autocorrelation

Let The equation to be estimated is ln (Xt) = β1 + β2 ln (Yt) + et (1)

t = 1, 2, . . . , The error term contains all of the economic factors other than x that affect y. If
changes in any of these other factors (shocks) affect y for more than one year, then the errors
of the model will not be mutually independent of one another. The errors are said to be
serially correlated or autocorrelated.
DURBIN-WATSON d TEST: DECISION RULES

Null Hypothesis Decision if

No positive autocorrelation Reject 0 < d< dL

No positive autocorrelation No decision dl ≤ d≤ du

No negative autocorrelation Reject 4 - dL<d<4

No negative autocorrelation No decision 4- du≤ d≤ 4-dL

No positive or negative Do not reject du < d < 4 - du

autocorrelation

Critical Value (with 3 explanatory variables at the specified number of data point):
dL=0.906 dU=1.408

Since the value of d is 0.724 from the result, we can conclude that there is a positive
autocorrelation (from decision rule table).
Remedial Measure for Auto-Correlation:

Data collected by us was cross-sectional, giving units in the data a time (time series
approach) and then calculating AR1 regression model.
Now calculating the Durbin-Watson result shows there is no autocorrelation.
Conclusion

This econometric analysis aimed to understand the determinants of Foreign Direct Investment
(FDI) by examining key macroeconomic indicators such as Real GDP Growth Rate,
Exchange Rate, and Net Trade of goods and services. The analysis was conducted using
Gretl, a powerful tool for regression analysis, diagnostics, and interpretation.

Estimation and Interpretation:

The multiple regression analysis revealed the following:

1. Real GDP Growth Rate: The Real GDP Growth Rate was found to be statistically
insignificant (p-value = 0.0839766), suggesting that it does not have a significant effect on
FDI.

2. Exchange Rate: The Exchange Rate was found to be statistically significant (p-value =
0.00937787), indicating that it has a significant impact on FDI.

3. Net Trade of goods and services: Net Trade of goods and services was found to be
statistically significant (p-value = 0.000151953), suggesting that it also significantly
influences FDI.

The restricted sum of squares test indicated that the unrestricted model, including Exchange
Rate and Net Trade of goods and services as independent variables, provides better results
compared to a model with only Exchange Rate.

Assumptions Violation:

1. Multicollinearity: VIF and Belsley-Kuh-Welsch diagnostics revealed that the dataset does
not suffer from significant multicollinearity.

2. Heteroscedasticity: White's test for heteroskedasticity showed no evidence of

heteroskedasticity in the regression model.

3. Autocorrelation: While the initial calculation suggested positive autocorrelation, the

Durbin-Watson test indicated no autocorrelation.

In conclusion, Exchange Rate and Net Trade of goods and services are significant
determinants of Foreign Direct Investment, while Real GDP Growth Rate does not
significantly influence FDI. The model constructed with Gretl provides a reliable framework
for understanding the relationship between these economic indicators and FDI. However,
further research and refinement of the model may be necessary to capture additional factors
influencing FDI inflows.