Hi everyone,
I am currently trying to identify the relationship between a sentiment index I created and the market I am observing. I am basically performing a basic time series analysis and a panel analysis and I am stuck in both in performing Johansen Cointegration tests.
1. When I do the analysis regarding my sentiment index and the market index, I observe that both are I(1) and the residuals are I(0). Accordingly, I use vec to perform a Johansen Cointegration and thus to determine the degree of the relation. Is it possible to retrieve the residuals from the Johansen Cointegration Equation to calculate the R² of the equation? If not, is there some other indicator I could use to describe the effectiveness of my sentiment?
2. I could not find any way to perform a Johansen Fisher Panel Cointegration test and retrieve the statistics from that test, e.g. the coefficient, the z-statistics etc. I could only find xtfisher, but that only checks for unit-roots. Is it pssibel to perform a fisher type johansen cointegration test with panel data in stata?
Any help is much appreciated.
Best
Micha
I am currently trying to identify the relationship between a sentiment index I created and the market I am observing. I am basically performing a basic time series analysis and a panel analysis and I am stuck in both in performing Johansen Cointegration tests.
1. When I do the analysis regarding my sentiment index and the market index, I observe that both are I(1) and the residuals are I(0). Accordingly, I use vec to perform a Johansen Cointegration and thus to determine the degree of the relation. Is it possible to retrieve the residuals from the Johansen Cointegration Equation to calculate the R² of the equation? If not, is there some other indicator I could use to describe the effectiveness of my sentiment?
2. I could not find any way to perform a Johansen Fisher Panel Cointegration test and retrieve the statistics from that test, e.g. the coefficient, the z-statistics etc. I could only find xtfisher, but that only checks for unit-roots. Is it pssibel to perform a fisher type johansen cointegration test with panel data in stata?
Any help is much appreciated.
Best
Micha
