Professional Documents
Culture Documents
Zubair's Eviews Project
Zubair's Eviews Project
Analyzing Econometrics
Data by Using EViews
Interest rate effect on Investment
Thanks to Almighty Allah-The Creator, The Supreme, The True Judge, The Perfect
and Purest for all. Definitely, He is the Creator and Commandant, and gave us the
strength to accomplish our assigned management project.
History of all great work is to witness that no great work was ever done without the
support of your delegating hardworking team efforts.
I am submitting our generous gratitude to our instructor Sir Khair Bux Mangiro
lecturer at Sukkur IBA, University Kandhkot Campus, for his sincere and great
efforts to guide and make us able to complete this mega assignment report.
Research Methodology:
As I discussed previous the following research is calculated to see the impact of interest rate on
the investment. As an extension, an additional variable of foreign exchange rate is also used to
reduce the omitted variable bias and to check the multiple regression relationship among
variables. The panel data is used in this study that includes both, different intervals 1974 to 2000
as well as different variables (Interest, Fore Exchange Rates, and Interest) in Pakistani
background. Just the data used in this study is the secondary data that is resulting from a number
of sources like; SBP (State Bank of Pakistan), WB (World Bank), BS (Bureau of Statistics).
Subsequently, the data is evaluated in the EViews 12 where the data is processed under required
tests. Finally the results are interpreted.
Variables:
As mentioned earlier, in the
beginning I would be working on Interet
two important variables interest and
14
investment. Inflation is considered
12
as the independent explanatory
10
variable and imports are used as
8
dependent variable. Interest’s data is
6
derived from SBP (State Bank of
4
Pakistan’s) web which is
2
denominated in percentage and
0
similarly the investment data is 1970 1975 1980 1985 1990 1995 2000 2005
derived from WB’s (World Bank) portal which is denominated in the US billion dollars. Hence
to make the data into a single demonization the logarithm is taken first and then they are
interpreted. The data is time series data that starts from 1974and to 2020—having total 25 years
of data. Following is the data graphically that helps to understand and observe data more clearly.
Graphically representation:
Investment is time series data which starts from 1974 to 2000 (X-axis) and on Y-axis the
percentage of inflation is given. The graph is not trending neither have disruptions and thus from
the picturing we can say that may the data is stationary. In 1995the data has the highest peak that
shows the highest percent of inflation in the given history of Pakistan.
Equation
Ln Imports = B0 + B1 (ln interest)
Where B0 is a constant of the equation and B1 is a coefficient of Interest.
Ln Imports = 1.000075- 0.063605 (ln Interest)
Interpretation
B0: As mentioned B0 is nothing but an intercept. It has no meaning but just it shows the
value of independent variable when it touches Y-axis (X=0). The value of intercept is
1.000075 that shows investment would 1.000075 billion when there is no inflation.
12
10
8
B1: B1 is a slope of the equation
inerest
6
and coefficient of the ln
Investment
inflation. That simply one 4 Linear (Investment)
percent increase in inflation
2
would cause 0.0635% increase
in the imports as it has a 0
positive slope. Is B1 significant? 5 6 7 8 9 10 11 12 13 14
Let’s check it. From checking investmen
B1‘s significance level we make
hypothesis. Null Hypothesis would be H 0 B1 = 0 and an alternative hypothesis would be,
H1 = B1 ≠ 0. This can be tested using t-statistics if the t-statistic is greater than 1.96 then
we will not reject our null hypotheses at 5% significance level otherwise reject. So our t-
statistics is 0.638935 which is lower than 1.96 (5% significance level). Hence we are not
rejecting our null hypothesis and can conclude that B1 is insignificant. That simply means
inflation cannot represent import in the given data set.
R-Squared: R-squared is also called coefficient of determination. R-squared shows us
the percentage of the dependable variable explained by independent variable. At the same
time it also suggests how strong and reliable model is? It is also called the goodness of fit
indicator. In this case the R-squared is 0.06317 or simply 6.3 % of changes in imports are
caused by the inflation. This is too small number because the remaining 98% of the
imports are represented by other variables which are not incorporated in this model.
SER (Standard Error of Regression): SER of regression shows the difference between
the actual points and the estimated points built be regression equation simply how wrong
your estimated model is from your actual model. SER is also a measure of goodness of
fit. In this model the standard error is 0.089821that shows 8.98% error and it’s a huge
difference that is an indicator that our regression model is not fit and need revisions.
Akaike Information: Akaike Information is a statistics test that is tells the goodness of
fit of model like R2 does and among other similar tests Akaike Information is considered
as the most used and accurate. The near the Akaike Information to 0 the stronger the
model is. But Akaike Information in this model is -1.9082 which is bit far from 0. Simply
it gives us an indication that model we are using is not fir.
Schwarz criterion: Schwarz criterion is similar to Akaike Information which is also a
fitness of good criteria. Similar to Akaike Information if the results are near 0 then model
said to be strong otherwise not. So our Schwarz criterion is -1.81142 which is relatively
far from 0 that also gives us an indication that our model is not fit.
Hannan Quinin: Hannan Quinin is similar to the Schwarz criterion and Akaike
Information criterion. Simply it tells how fit the model is which you are using. If the
Hanin Quinin score are near to 0 that indicates that the model is fit otherwise not. Here in
this case the results are 1.88033which is relatively far from 0. Hence, we can conclude
that the model is not fit.
Durbin Watson: Durbin Watson test a statistical test used to know the auto-correlation
in the data set. If the results of Durbin Watson test is 2 that means there is no auto-
correlation if it’s greater than two means there is negative auto-correlation. Similarly if
it’s less than means there is auto-correlation but that is positive. Here in this case the
Durbin Watson is 0.575617. That simply means there is a positive auto-correlation
between the Log of Imports and Log of inflation.