Detailed explanation of methods and steps for cointegration analysis using eviews software

How to use eviews software for cointegration analysis

1. This equation is definitely a cointegration equation. OLS cannot be used because the error term is not white

noise. Moreover, all calculations related to cointegration basically use maximum likelihood estimator

(MLE). Including its sister fractional cointegration (FC).

2. Johansen test is used to test cointegration. In the original article, the intercept term is measured using the likelihood ratio (LR) test. This is normal. You have to use MLE to do it anyway. LR is easier than the other two tests.

3、cointegrating

equation

The cointegration equation is a statistical method. It cannot completely calculate all coefficients and must use one of them as the base point for normalization. Therefore, some coefficients may not be interpreted consistently in the calculation results.

The result is normalized with the first variable as the base unit. This is why the coefficient of HKE is 1 and there is no standard error.

4, repair

The positive coefficient is actually based on the first order difference of all original variables (HKE, HKM2, etc.)

(D(HKE),D(HKM2), etc.). Because if the original model is I(1) and cannot be calculated consistently, then first order

What follows the difference is I(0). Very easy to calculate. Even in some cases, OLS calculation can be used directly. standard

under each correction factor

error is also calculated in this way.

How to use eviews to test the correlation coefficient of two sets of data and draw a data change chart

The method of using OpenCV in Android is:

a. There are two folders under the OpenCV installation path "F:\OpenCV-2.3.1-android-bin". Copy the folder "OpenCV-2.3.1" to the directory where the Eclipse workspace is located, that is, in the upper-level directory of your project, then import it into the workspace, select the project in Package Explorer, and right-click on the standalone computer to pop up Select Properties in the menu, then select Android on the left side of the pop-up Properties window, then click the Add button on the lower right, select OpenCV-2.3.1 and click OK.

b. At this time, expand the project tree and you can see that a new OpenCV-2.3.1_src directory has been added, as shown below. Then the OpenCV Java API has been added correctly, otherwise the directory path where OpenCV-2.3.1 is placed is incorrect. correct.

c. Then you can import the OpenCV API package in the Java source file and use the OpenCV API. The format of the OpenCV API package is as follows:

Org.opencv.(OpenCV module name).(OpenCV class name)

For example:

Org.opencv.core.Mat

How to use eviews to get the cointegration equation in the cointegration J test

First judge the results of the cointegration test. According to the trace test (Figure 1 - previous table) and the maximum eigenvalue test (Figure 1 - next table), it can be judged that there are two cointegration equations, which are also output above. The result: indicate 2 ces. It can be judged mainly based on the p value behind the statistic. If p

You can write 2 cointegration equations, because the above is the conclusion:

Find the table corresponding to 2CES, 1.0000 corresponds to the explained variable, and then write two equations: [Pay attention to changing the positive and negative signs of the coefficients]

LOGGDP=-2.097LOGGSIP 2.2217LOGGTIP 10.3829

LOGGEC=-0.8943LOGGSIP 0.3717LOGGTIP 5.515

At the same time, you can also write a cointegration equation, which is the equation corresponding to 1CE, and prove that the residuals are stationary.

LOGGDP=-4.402LOGGEC-6.034LOGGSIP 3.858LOGGTIP 33.076

The above is the detailed content of Detailed explanation of methods and steps for cointegration analysis using eviews software. For more information, please follow other related articles on the PHP Chinese website!