Analysis on Eviews

Objective of the Study

Methodology

The Analysis

Analysis on EVIEWS:

Step 1: Data series has been plotted to observe whether any missing points are present.

Graph:

[pic 1]

Checked the Graph for Infosys’s Sensex data named Close and obtained the above graph. It shows that data is non-stationary.

Hence there is need to test correlogram, the result is shown on next figure.

Step: 2: Examining correlogram of the series:        

[pic 2]

Observation:

From above figure it is clear that PACF shows AR(1) is present hence there is need to DE trend the data also ADF test is showing that there is unit root hence non stationary data.

Hence next is confirmatory test through unit root.

Step:3 Unit Root Test (Include in test equation: "Trend & Intercept")

[pic 3]

Observation: 

It is observed that H0 of unit root accepted at all levels. It confirms presence of non-stationarity through trend component.

So, Series is non-stationary and I(1) in nature.

Step - 4 - Stationerizing the Series by detrending (1st differentiation)

New "dd" has been created by following command.

Series dd=d(close,1,0)

[pic 4]

Now again taking correlogram test get white noise as shown in figure below. But we need to de seasonality it because we directly reached to white noise and also value of p is greater than .05

As shown in below figure-

[pic 5]

To de seasonalise data

Series dd =d(close,0,24)

And result is showing presence of AR(1)  and SAR(24) in PCF, probability is also 0

[pic 6]

Step 5: Estimation Stage

It is found presence of AR(1) and SAR(24) in PACF.

Hence to remove it -

ls d(close,0,24) ar(1) sar(24)

Estimation result is shown below-

[pic 7]

Observation: Coefficient of AR (1) is and SAR (24) are less than 1 and p-value is significant.

Now after this checking it with correlogram,

We reached White noise process