2

I ran my data set in EViews and I got this:

enter image description here

My book says "[Clive] Granger suggested that to see if A granger-caused Y we would run (the equation) and test the 0 hypothesis that the coefficient of the lagged A jointly equal to zero. If we can reject the null hypothesis then we have evidence that A granger-causes Y."

So does my picture / output mean that

Real mortage rate(RMR) granger-causes GDP. RMR dosen't granger-cause GDP?

note that $d$ is first difference here, and KPI (consumer price index)

I'll post the whole picture in a link below if more information needed. https://i.stack.imgur.com/DqBP0.jpg

AdamO
  • 62,637
soet irl
  • 41

1 Answers1

1

In order to reject the null hypothesis that an explanatory variable does not Granger-cause the dependent variable, you need the p-value (Prob. column) to be lower than your significance level (standard values in econ are 1%, 2.5% or 5%).

In this case, you can affirm (at 2.5% significance level) that KPI and DPHOX Granger-cause DRMR. However you cannot reject the null hypothesis for GDP, therefore GDP does not Granger-cause DRMR.

However, as you're using the first difference for the dependent variable, it might be a good idea to use the first difference of the predictors.

Pierre Cattin
  • 492