How to Test Unit Root and remove Unit Root from data and how to make data Stationary Using E-views

[Music] Somaly cool my name is Imran Khan and this is Esau selling in this video I will show you how to test you know truth in the data using a view but before doing that let's talk about what is stationary data and what is unit root the data will be shatori when the mean and the variance are constant over time but if any of them changes then we have unit root in the data we want the data to be stationary because if there is a unit root in the data and if we apply any tests on such such data then we will have forth and incorrect results now look at the graph in this graph this is a upward turning graph and in this graph the mean and the variance both are changing with the time so it means that this graph has a unit root the data is not stationary while in the next graph you can see that the mean is constant over time and the variance as well so this data has no unit root and the data is stationary now let's go to eviews and test the unit root now I have my Ezio's file open here in the image I have for one variable with a name exchange and there are 252 observation now I'm going to apply the unit to test on this data so for doing that click on click then see these statistics and then unit root test enter the name of the variable which is exchange and click OK now we have another window and this is the window for augmented dickey-fuller text editor test because document 8 dickey-fuller test test the unit group in the data and first of all we'll check the level it means that the data we are going to test the Euler tool on the raw data on the data which is in its original form so first of all I will check level here now in leg length this is the built-in function and I will keep it as it is and in the third I select first of all intercept and remember that I will test the unit tube all of these three options and if there is same results in all these these options then we will decide whether the data has unit truth or not so first of all I will check it check the level and intercept here and click OK and the maximum length in your leg is the body force expenses is okay click on OK now we have another window this is the result of augmented dickey-fuller test and you know the null hypothesis then I have our null hypothesis that exchange has unit food and here the probability value is more than 5 percent which is eighty nine point zero two portion which means that the data has unit fruit now with that we have a test statistic T statistic value which is negative point four eight and there are two checks to check whether the data has unit root or not first of all is this interval it is it is not significant it means the null hypothesis is not rejected and which which is obvious that we have unit ood but we can also judge the unit root from here if the value of this piece test statistic is less than these three values which is critical values test critical values if the T statistic value is less than this then you come to know that the data has unit rude so now let's check the other option which is intercept plus channel intercept and keep it on level and click OK now you can see that we have a probability very which is sixty five point zero eight person and here my nail hypothesis is not rejected it means that data has you know true here and it is also clear from this result that the T statistics is less than the test critical values so it means that when I check chan and n constant both then I will come to know that my data has unit root now let's move to the third option which is none it means that I'm neither going to include intercept not both so click OK now you can see that the probability value is 99.85% which is more than 5% it means that that I has you know truth but when you look at this the T statistic value is more than the test critical value so what does it mean it means that we have to either include only constant or intercept are both the trend and intercept and we cannot do any tests on this data by excluding both of them.i or any one of them there must be others interpret or both of them so now let me test the unit truth on the first difference and check intercept here and click OK now you see that the probability value is less than 0.5 percent of 0.05 and the T statistic value is more than lettuce critical values that mean - it means that we have no unit using the data when we get the first difference now if I select the second option which is trend and intercept then also my data has become stationary on this because the probability value is less than 5% and the T value is more than the test clip critical values so it means that when I take the first difference of this data my data will become stationary but remember that you have to include either intercept or trend energies of both because if you include none of them then the result will not be true now let's move to excel file and take the false difference of the data and also see the result in graphical form and this is my data and if I insert the graph with this data you can see that there is a unit root in the data because the mean and the variance are changing constantly now if I take the first difference of this data second lu- first one and drag it to the end then this Lu is the first difference of my data and now if I insert the graph of this data you can see that the mean is constant and the meeting is caused and so it means that the unit tooth has gone and the data has become stationary so that's all for now thanks for watching see you in the next video

Channel: Ilmi Dunya

Views: 36,233

Rating: 4.6571426 out of 5

Keywords: unit root test, eviews, stationary data, first differene, esoft learning, research, Unit, Root, Testing, Augmented, Dickey, Fuller, Test, P-value, ADF, fixed effect regression, random effectt regression, multiple regression, tutorials, regression, spss, econometrics, random effect, fixed effect, education business, tutoring business, financial estimation, banking estimation, eviews test, time series data, how to apply adf test in eviews, how to apply unit root test in eviews, eviews tutorial

Id: IzJZWOW9iG0

Length: 9min 32sec (572 seconds)

Published: Thu Jan 26 2017