Since there is no cointegration relationship between the included variables, then there is no long run equilibrium relation between them but there may exist some significant relations in the short run that could be analyzed using the VAR model not the VECM which is irrelevant in your case
Thank you Dr. Doaa, however, i conducted another combined panel cointegration test ( Johansen fisher panel cointegration test), and the results indicates cointegration with six of the eight probabilities being lessthan 5%. i.e.
None 0.0000, 0.0000
At most 1 0.0000, 0.0000
At most 2 0.0003, 0.0108
At most 3 0.1400, 0.1400
What is your conclusion based on the above result?