I would like to find the long-run relationship between domestic and international prices. I have run the Johansen Cointegration test on level with 2 and 6 lags, but I have got mixed results. With 2 lags the trace-test confirm one or more cointegrating vectors but trace-test does not confirm cointegration. However, with 6 lags suggested by AIC and HQ information criteria, their is no cointegration at all. How should I proceed further?
Thanks Dear Sharam,
I did selected the right lag length as you mentioned. My question is whether to run the Johansen Cointegration test on level of price series or on their first differences?
Thanks Mr. Sharma for your prompt reply,
Sorry for the miss spelling of your name.
Your comments were useful.
I run cointegration on level series, but the results of trace test and max test does not support each other. The results are also different for 2 lags suggest by BIC and 6 lags suggested by AIC and HQ. With 2 lags trace test rejects the null of no cointegration at 5% level, but max test does not accept null at 5% but accept the alternative of two or more cointegrating vectors at 10% level. However, with 6 lags both tests fail to reject the null of no cointegration. What should I conclude? Is it possible to consider the trace test results with 2 lags and continue by estimating the VECM.
Waiting for your reply.
Dear Mr. Sharma,
How would we decide the optimal lag length while running the Johansen Test? The test seems to be very sensitive to lag lengths.
Yes this problem has already been solved with me. Generally speaking we convert all prices into log form and then conduct unit root and cointegration tests. However, it may not be wrong if one use the level form as well. I am sharing two of my papers below, if they are of any help to you:
http://www.foodandagriculturejournal.com/vol3.no4.pp27.pdf
https://ideas.repec.org/a/ebl/ecbull/eb-15-00640.html