Cross-correlation

[Emily89]

I am running two different variables in EViews using the Cross-Correlogram function.I want to know how this function can tell whether the first variable is a leading indicator of the second variable. Could any one explain the following result? Really appreciate the help!

Date: 07/25/13 Time: 11:18
Sample: 1990M01 2007M12
Included observations: 215
Correlations are asymptotically consistent approximations

X,Y(-i) X,Y(+i) i lag lead
.|. | .|. | 0 -0.0042 -0.0042
.|* | .|* | 1 0.1246 0.1244
.|. | **|. | 2 0.0423 -0.1493
*|. | .|* | 3 -0.0542 0.1253
*|. | .|. | 4 -0.0423 0.0185
.|* | .|. | 5 0.0936 -0.0157
*|. | .|* | 6 -0.0953 0.0823







I don't understand why the longest lag occur at i=2 is negative..what does it mean?

[Nalvarino]

Read the EViews User Guide I, chapter 12, pag. 422 (or if you have another version maybe it is in another page). In short, you would need to look at the positive lags to see if x is a leading indicator of y and the negative lags if you think that y is a leading indicator of x. At least that's what I get by watching the formula on the guide.

[trubador]

http://forums.eviews.com/viewtopic.php?f=18&t=789