Reason
of applying johansson co-integration
1. The main reason of applying johnson co-integration is all
variables dependent and independent are stationary at first difference and its
oder of integration is I(1).
2. Johensen co-integration model is applying on time series
data.
3. johnson co-integration use two types of statistics,
number one is trace-value and 2nd one is Maximum-Eigen-value-statistics
Hypothesis testing
H0= there is no co-integration
H1= co-integration exist
If trace value is greater than its critical value it
means H0 reject which shows co-integration exist in this model.
If maximum Eigen value is greater than its critical value
it means H0 reject which shows co-integration exist in this model.
johansson co-integration
Method of applying
Select all variable, first variable should be dependent
variable then put figure on ctrl butn and select independent variables.
Open these variables as group͞͞- go to view and select
johansen co-integration
The main
advantage of johansen cointegration over Engle granger co-integration (EG
cointegration) that Engle granger
cointegration identify only one co-integrating relation on the other hand,
johansen cointegration can identify more than one co-integrating equation
If you are not
sure about the condition of the data so you should use the option of Summarize
all 5 set of assumption, it is better for appropriate lag length.
According to Akaike information criteria
option number four and lag one is appropriate. 3 and 4 option is appropriate
for time series data so we select option 4.
Results shows two
tables of maximum eigen value these are the value of trace statistica and its
critical value and these are the value of maximum Eigen value statistics and
here it’s a critical value.
Hypothesis
None = there
is one co-integrating vector which is your equation in which your dependent
variable is GDP if this variable is significant like this the trace statistics
is greater than the critical value, it means reject H0. If none is significant
its mean only one co-integrating relation is exist
At most 1= one
co-integrating relation exist
At most 2= two co-integrating relation exist
If trace and Egien
value shows different results in that case trace is more appropriate.
Normalized co-integrating coefficient
The variable
having value below the one which means the dependent variable same as we have
selected and these are the independent variables like capital and FDI what we
have to interpret that the sign would be opposite, let’s suppose there is a
minus 95.78 is the Normalized coefficient of capital, it is not a -95.78, it
would be a 95.78 which means we have to take the opposite sign if capital shows
the negative value it means it has positive effect on GDP. Standard errors are
in parenthesis.