Johansen Test (Web Application)

Johansen Test

The Johansen test enables the user to reveal existence of stationary linear combinations of time series, being integrated of the first order, and it is one of the system estimation methods, using the maximum likelihood method relating to vector autoregression models. It is a part of the group of cointegration tests.

Make sure that the results panel is displayed.
Select two or more variables in the data view area.
Go to the Descriptive Statistics tab on the results panel.
Select the Johansen Test item in the Cointegration Tests button menu.

Test parameters:

Endogenous Variables. Create a list of variables, tested for cointegration. The list contains all selected variables by default. At least two endogenous variables are required to execute the test.

Exogenous Variables. Create a list of variables, influencing tested variables. Exogenous variables are absent by default.

Managing the contents of the variables used in the test. Each variable has the corresponding checkbox. If the checkbox is selected, the variable is used in the test; if it is deselected, the test is executed without variable.
Moving of variable between lists. Select one or more variables in the list and use the and the buttons.
Executing additional operations with variables. Click the button. Select the required commands in the menu that opens.

Calculation Start. Specify calculation start point.

Calculation End. Specify calculation end point.

Type of Error Correction Model. Select the error correction model in use:

Without Trend in Autoregression, without Constant in Cointegration Equation.
Model without Trend or Constant in Cointegration Equation and Autoregression.
Model with Constant in Cointegration Equation and in Autoregression.
Model with Linear Trend and Constant in Cointegration Equation, without Trend in Autoregression, with Linear Trend in Data.
Model with Trend and Constant in Autoregression and Cointegration Equation, with Square Trend in Data.

Autoregression Order. Specify the order of variables autoregression.

Missing Data Treatment. Select missing data treatment method. The Casewise method is used by default, that is, empty values are excluded. Calculations are executed without considering them. For details about missing data treatment methods see the Missing Data Treatment section.

The test is executed simultaneously for all selected variables. To find cointegration relations, the maximum likelihood relation values are compared with the N percent (N = 1%, 5%, 10%) significance level. If the maximum likelihood relation value is greater than the significance level, the hypothesis about cointegration relations existence is rejected.

For example: