University Finance Report: International Financial Management Analysis

This report presents an analysis of international financial management, specifically focusing on the factors that influence Foreign Direct Investment (FDI) acquisition. The study employs a systematic approach, utilizing regression analysis and a multinomial Logit model to examine the relationships between FDI acquisition and various independent variables, including institutional distance, uncertainty avoidance distance, and industry relatedness. Summary statistics and correlation matrices are provided to understand variable distributions and potential collinearity issues. The analysis includes control variables such as GDP, GDP growth rate, enterprise value, and transaction value. Variance Inflation Factor (VIF) is used to test for multicollinearity. The multinomial logit regression results are presented to analyze the relationship between ownership choice (full, majority, and minority acquisition) and the independent and control variables. The overall significance of the model is tested, and Nagelkerke R-squared is used to explain the proportion of the dependent variable explained by the independent and control variables. The report provides insights into the statistical significance of the coefficients and the direction of the relationships between the variables. The results indicate the impact of factors such as enterprise value, transaction value, and sectors on ownership choices.

Running head: INTERNATIONAL FINANCIAL MABAGEMENT
International Financial Management
Name of the Student
Name of the University
Author note

1INTERNATIONAL FINANCIAL MANAGEMENT
Analysis
Analysis is done following a systematic approach. The objective is to analyze factors
influencing FDI acquisition. Regression explains the cause and effect relations between
dependent and independent variables. The best way of analyzing the relation between FDI
acquisition and factors like institutional distance, uncertainty avoidance distance and industrial
relatedness is to conduct a regression analysis. Given the trichotomous nature of the dependent
variable that is ownership choice, a multinomial Logit model is formed. However, before
framing final model several things are considered.
First task is to observe the summary statistics of all the variables.
Variable Obs Mean Std. Dev. Min Max
UAD 1009 -2.16056 19.52557 -30 16
ID 1009 3.902991 0.56352 2.986649 4.618378
Ind_relate~s 1009 0.342914 0.474918 0 1
GDP 1009 1.20E+13 1.99E+12 7.59E+12 1.58E+13
GDP Growth 1009 -6.43295 3.633294 -12.4528 0.761495
Enterprise~e 1009 234.0319 2233.029 0 65380.65
Transactio~e 1009 45.04399 220.3831 0 3712.861
FD 1009 7.346978 3.913787 0.560856 15.58575
Sector 1009 0.282458 0.450418 0 1
Table 1: Summary Statistics of the chosen variable
The above table shows some centre measures of central tendency and dispersion. This
gives idea about distribution of the variables. For the regression model the chosen independent
variables are Institutional distance (ID), Uncertainty Avoidance Distance (UAD) and Industry
relatedness. All the variables that are GDP, GDP growth rate, Enterprise value, Transaction
value, Formal Distance (FD) and sectors are control variables. From the summary statistics mean

2INTERNATIONAL FINANCIAL MANAGEMENT
value of all the variables are positive except that is for UAD and GDP growth rate. Negative
GDP growth rate has adverse impact on the economic decision and hence on Ownership choice.
This is one factor behind keeping growth variable in the analysis as a control variable. From the
standard deviation measure, it is observed that Enterprise value and transaction value are highly
variable.
The next step in the research analysis is to examine the correlation among the choice variables.
For this purpose a correlation matrix is formed.
UAD ID
Ind_re~
s GDP
GDPGro~
h
Enterp~
e
Transa~
e FD Sector
UAD 1
ID 0.8992 1
Ind_relate~
s 0.0064 0.0132 1
GDP
-
0.5698 -0.727 -0.0227 1
GDPGrowt
h
-
0.7417
-
0.7517 -0.0497 0.5617 1
Enterprise~
e 0.0243 0.0008 -0.0423 -0.0005 -0.0125 1
Transactio
~e
-
0.0208
-
0.0214 0.005 -0.0262 0.0626 0.319 1
FD
-
0.2073
-
0.4438 -0.0238 0.6526 0.0601 0.0577 -0.0162 1
Sector 0.0261
-
0.0132 0.0801 0.0466 0.0556 -0.0357 -0.0489
-
0.0158 1
Table 2: Correlation matrix of all the independent and control variables

3INTERNATIONAL FINANCIAL MANAGEMENT
The next issue to be addressed is whether collinearity is present or not. Correlation matrix
shows degree of association among the variables. Here, one gests pair wise correlation. In a
multivariate regression analysis, presence of multicollinearity can create a major problem and
even falsification of the result obtained. The problem of multicolinearity is said to exit when one
or more independent variables are related. For a statistically acceptable result the independent
variables should be independent of each other. The presence of any kind of relation between
them gives a biased result where the coefficients are either over estimated or under estimated
based on the nature of the relation. Therefore, test for multicollinearity should be done before
estimating the regression model. There are different testing methods for multicollinearity. These
include computation of Variance Inflation factor (VIF), estimation of Condition Index (CI) or
such others. In the paper variance inflation factor (VIF) is used for this purpose. A value of VIF
less than 10 implies multicollinearity is zero or negligible. If the estimated value of VIF is
greater than or equal to 10 then multiocollinearity can be a severe problem and in that case
dropping the correlated with other variables should be dropped.
Source SS df MS
Number of obs
= 1009
F( 8, 1000) 746.08
Model 329151.33 8 41143.917 Prob > F 0.000

4INTERNATIONAL FINANCIAL MANAGEMENT
Residual 55146.656 1000 55.146656 R-squared 0.8565
Adj R-squared 0.8554
Total 384297.99 1008 381.24801 Root MSE 7.4261
Coef. Std. Err T P > t [95% Conf. Interval]
ID 34.74279 .8098321 42.9 0.000 33.15362 36.33195
Ind_relatedness -0.1983576 0.4965931 -0.4 0.690 -1.172842 0.7761265
GDP 1.51E-13 2.2E-13 0.68 0.495 -2.82E-13 5.83E-13
GDPGrowth -7.03E-02 1.20E-01 -0.59 0.558 -3.06E-01 1.65E-01
EnterpriseValue 0.0000883 0.000111 0.79 0.427 -0.0001296 0.0003061
TransactionValue 0.0003974 0.0011283 0.35 0.725 -0.0018168 0.0026116
FD 1.140175 0.0966761 11.79 0.000 0.9504636 1.329886
Sector 1.906563 0.5241122 3.64 0.000 0.8780773 2.935049
_cons -148.9093 4.286158 -34.74 0.000 -157.3202 -140.4984
Table 3: Regression for independent variables
Variable VIF 1/VIF
ID 3.81 0.262694
GDP 3.51 0.285198
GDPGrowth 3.47 0.288159
FD 2.62 0.382143
Transactio~e 1.13 0.884738
Enterprise~e 1.12 0.890147
Sector 1.02 0.981699
Ind_relate~s 1.02 0.983602

5INTERNATIONAL FINANCIAL MANAGEMENT
Mean VIF 2.21
Table 4: Variance inflation Factor
VIF for all the variables are less than 10 making the mean VIF 10. This indicates that the
chosen variables behave independently and hence can be used for the regression model.
Once the choice of variables is made then the regression analysis can be done. The list of
dependent and independent variables taken for analysis is given below
Variables Definition source
Dependent Variable
Ownership Full acquisition is coded by 1, when the acquirers buy 100% equity
of the target firms; Majority acquisition is coded by 2, when the
bidders buy 50%-99% shares of targets; minority acquisition is
coded by 3, when the acquirers buy less than 50% of the targets.
Thomson
One
Banker
Independent Variables
Uncertainty Avoidance
Distance
The differences in one of cultural dimension index(Uncertainty
Avoidance Distance) among the home (acquirer) and host (acquired)
country
The
website of
Hofstede
Institutional Distance The institutional differences between countries of acquirer and
acquires through calculating the six dimensions from Kaufmann et
al. (2009)
World
Bank
Industry Relatedness The similarity of industry among the home and host country, which
is regarded as if the SIC code of industry in countries matched.
Code 1 means the industry is related, Code 0 is means unrelated.
Thomson
One
Banker
Control Variables
GDP difference The difference in gross domestic product between home and host World

6INTERNATIONAL FINANCIAL MANAGEMENT
country Bank
GDP Growth difference The difference in growth rate of gross domestic product between
home and host country
World
Bank
Enterprise Value The value of the companies acquired Thomson
One
Banker
Transaction Value The value of a specific transaction Thomson
One
Banker
Financial Distance The difference in financial development among host and home
country
Cross-
National
Distance
Database
Sector
(Manufacturing/Service)
To test if the deal is implemented in manufacturing or service sector.
Code 1 is represented the transaction of manufactures, otherwise 0.
Thomson
One
Banker
Main analysis of the present paper is based on the result of multinomial logistic model.
A typical logit model is used to make regression analysis when the dependent variable contains
binary outcome. Multinomial logit model is an extension of simple form of logit. It helps to
examine the relation for variables that have outcome in three or more level. Here there are 3
choices coded as 1,2 and 3 representing full acquisition, majority acquisition with share between
50 to 99% and minority acquisition with share less than 50% respectively. A choice has to made
about the base or reference outcome. The regression result shows two sets of relation with the
independent variables in relation with the base or reference outcome. Base outcome is the one
that appear with the highest frequency. A frequency table is constructed to choose the base.

7INTERNATIONAL FINANCIAL MANAGEMENT
Ownership Choice Freq. Percent Cum.
1 543 53.82 53.82
2 135 13.38 67.2
3 331 32.8 100
Total 1,009 100
Table 5: Frequency of different ownership choice
From the frequency table it is seen that the frequency for fully acquisition is 543, for
majority acquisition the corresponding frequency is 135 and that for minority acquisition is 331.
It is very clear that ownership choice of full acquisition appears maximum number of times.
Hence, it is chosen as the base outcome.
Multinomial logistic regression
Number of obs
= 1009
LR chi2(14) = 218.39
Prob > chi2 = 0.000
Log likelihood = -867.72559 Pseudo R2 = 0.1118
Ownership
Choice Coef
Std.
Err Exp (b) Z P > |z|
[95% Confidence
Interval]
1 Base Outcome
2
UAD
.014384
7 0.01315 1.014489 1.09 0.274 -.0113945 0.040164
ID -.638667 .552063 0.527996 -1.16 0.247 -1.72069 0.443357
Ind_relatedness -.459455 .206027 0.631628 -2.23 0.026 -0.86326 -0.05565
GDP 5.33e-14 9.16e-14 1 0.58 0.561 -1.26E-13 2.33E-13
GDPGrowth -.015965 .050176 0.984162 -0.32 0.750 -0.11431 0.082379

8INTERNATIONAL FINANCIAL MANAGEMENT
EnterpriseValue
.000673
8 .000523 1.000674 1.29 0.197 -0.00035 0.001699
TransactionValue
-2.22e-
06 .000578 0.999998 -0.00 0.997 -0.00113 0.00113
FD -.012467 .04256 0.987611 -0.29 0.770 -0.09588 0.07095
Sector -.208419 .216713 0.811867 -0.96 0.336 -0.63317 0.21633
_cons
.688393
4 2.58104 1.990515 0.27 0.790 -4.37036 5.747143
3
UAD .0479622 .010975 1.049131 4.37 0.000 0.026452 0.069473
ID -1.27451 .467394 0.279567 2.73 0.006 -2.19059 -0.35844
Ind_relatedness -.992010 .174949 0.370831 5.67 0.000 -1.3349 -0.64912
GDP 1.60e-14 7.36e-14 1 0.22 0.828 -1.28E-13 1.60E-13
GDPGrowth .0682524 .039657 1.070635 1.72 0.085 -0.00947 0.145979
EnterpriseValue .0041456 .000798 1.004154 5.20 0.000 0.002582 0.005709
TransactionValue -.012377 .00272 0.987700 4.55 0.000 -0.01771 -0.00705
FD .0237726 .034846 1.024057 0.68 0.495 -0.04452 0.092069
Sector -.757366 .182056 .468899 -4.16 0.000 -1.11420 -.400543
_cons 5.072575 2.17158 159.585 2.34 0.019 .8163657 9.328785
Log-Lik Intercept Only: -976.922 Log-Lik Full Model: -867.726
D(982): 1735.451 LR(14): 218.392
Prob > LR: 0.000
McFadden's R2: 0.112 McFadden's Adj R2: 0.081
Maximum Likelihood R2: 0.195 Cragg & Uhler's R2: 0.227
Count R2: 0.538 Adj Count R2: 0.000
AIC: 1.779 AIC*n: 1795.451
BIC: -5036.013 BIC': -107.725
Cox and Snell: Maximum Likelihood R2: 0.195
Nagelkerke : Cragg & Uhler's R2: 0.227
Table 5: Result of Multinomial Logit Regression

9INTERNATIONAL FINANCIAL MANAGEMENT
Before entering into the detail, analysis of the model the overall significant of the model
needs to be tested first. The null hypothesis here is that the model is not a good model. Hence,
lower the probability of accepting the null hypothesis greater is the overall significance of the
model. The statistics used for this test is chi square. Lower the P value for the statistics higher is
the chance for rejecting the null hypothesis and obtains a significant model. For the chosen
model it is observed Prob>chi -=0.000. This implies a overall significant model. Correlation
coefficient gives idea about part of the dependent variable explained by the independent
variables. However, for a logistic regression standard R square used for linear regression model
cannot be used. Nagelkerke R square is a more appropriate measure here. The output of
multinomial logit regression gives nagelkerke R square value as 0.227. This has the implication
for indicating how much proportion of the dependent variable is explained by the independent
and control variables together. It can be said that the independent variables and control variables
able to explain 22.7% variation of the dependent variable.The above table shows the relation of
majority acquisition with independent and control variables over the full acquisition and the
same for minority acquisition. Sign of the co efficient indicates the direction of the relationship.
The co efficient should be statically significant to make any prediction about the relationship.
The null hypothesis used for testing significance of the coefficient assumes the co efficient equal
to zero. Corresponding alternative hypothesis is co efficient is different from zero. Rejection of
the null hypothesis indicates the significance of the estimated co efficient. The simplest way to
take the decision is to observe the p value for each estimate, which is the probability of accepting
the null hypothesis. Greater the p value lower is chance for rejecting the null hypothesis. When p
value is less than 0.05 than the probability of accepting the null hypothesis is very low and hence
can be rejected making the coefficient significant.

10INTERNATIONAL FINANCIAL MANAGEMENT
Looking at the control variable for majority acquisition, all the variables are statistically
significant as P values are greater than 0.05. Coming to minority acquisition slightly different
result is obtained as some of the control variables are statistically significant. The significant
variables are enterprise value, transaction value and sectors. P values of all these three variables
are 0.00. This implies a complete rejection of null hypothesis of having the coefficient zeros.
Therefore, enterprise value, transaction value and sectors have significant impact on ownership
choice. Enterprise value represents the value of the acquired firm. The coefficient is positive and
significant and therefore has the implication high enterprise value benefit the acquire. The
transaction value captures the value of a certain transaction. The co efficient is negative and
hence implies and adverse effect on the ownership choice. Sector is a binary variable assuming
the value 0 and 1. 1 stand for manufacturing sector and 0 stand for service sector. The significant
of this variable indicates that when acquisition is made in a manufacturing industry then it has an
effect on the dependent variable.
Constant term is also significant for minority acquisition is significant unlike for minority
acquisition case where all the control variables including as well as the constant term appears as
non-significant.
The variable Uncertainty avoidance distance is positively related with majority
acquisition. Value of the concerned co efficient is 0.014. When uncertainty avoidance distance
increases by one unit then majority acquisition can be increased by 1% over the full acquisition
given significance of the variable. The value for UAD is 0.274 making acceptance of null
hypothesis in favor of non-significance. Hence, the relation remains in determinant. The
Corresponding co efficient of UAD in case of minority acquisition is also positive. The value
here is 0.048. This indicates UAD has a positive influence for minority acquisition over full

11INTERNATIONAL FINANCIAL MANAGEMENT
acquisition. Unit change in uncertainty avoidance causes a nearly 5% increase in ownership
choice to wards minority acquisition. For minority acquisition, the variable UAD is statistically
significant with a p value = 0.00. The result supports the hypothesis that higher value of
uncertainty avoidance is associated with higher ownership choice for minority acquisition in
contrast to full or majority acquisition. Uncertainty avoidance distance measures the associated
risk on part of the acquirer in doing investment in doing business or investment in the emerging
market. Therefore, greater uncertainty avoidance increases the likely hood of minority
acquisition. This statement is supported by the collected data. A statistically significant and
positive co efficient for this variable indicates that uncertainty avoidance, and minority
acquisition moves in the same direction significantly. With higher uncertainty avoidance,
acquirer retains a minority share reducing the possibility of majority and full acquisition.
For institutional distance, there seems a negative relation between ID and majority
acquisition. This means increase in institutional distance has an adverse impact on majority
acquisition. Value of the estimated co efficient is -0.638. Therefore, on account of unit increase
for this variable there might be a 63% increase in majority acquisition. This means institutional
distance is likely to have a much greater impact on ownership choice than uncertainty avoidance
distance. However, this conclusion can only be drawn if the variable turns out to be statistically
significant. Too test the significance p value is considered. The noted p value is 0.247 which is
greater than the significance level 0.05. This makes the variable statistically insignificant. As the
co efficient turns out to be statistically insignificant its effect on majority acquisition should not
be considered. For another ownership category “minority acquisition” institutional distance has
similar kind of inverse relation. Corresponding co efficient value here is -1.27. The p value is
0.006 that is within the significance. The p value suggests significance of the co efficient and