EViews.com

Yes it is possible but interpretation of coeffs might differ than what you'd expect. Say you have a model of y=0.5 log_x + e. What this means is that 1 unit increase of y explains 0.5% increase of x. But I might be wrong. Check J Woolridge (2002)

Yes there's three CE. Lucky you both results did not conflict. Anyway, what seems suspicious is your choice of lag length. Its bit too small relative to your sample size (3300). I recommend doing some diagnostics on VAR(5) to see any residual autocorrelation. Johansen Test is sensitive to underspeci...

VAR Residual Serial Correlation LM Tests Null Hypothesis: no serial correlation at lag order h Date: 04/26/12 Time: 15:57 Sample: 1961M01 2010M12 Included observations: 594 Lags LM-Stat Prob 1 33.80490 0.1121 2 31.13447 0.1846 3 53.37381 0.0008 4 51.58304 0.0014 5 33.08849 0.1289 6 33.04551 0.1299 7...

Its okay, I got it. For future reference for others facing the same problem, when this happens, it means that none of the variables have stochastic trend, in which case either you recheck stationarity of your variables or if your variables are confirmed I(1) i.e in theory in cases such as stock pric...

Hi all, I need help in figuring out the weird result I have gotten from Johansen Test. It seems that after running the test for 5 variables, the trace test indicated 5 Coint. Eq. which doesn't make sense. Is this a telltale sign that something is wrong eg model misspecification, lag selection etc......