Panel data unit root common vs individual

[EconometricAarhus]

Hello.

I am doing a random effects model on bonds. N=373 and T=60. In the preliminary tests I am about to do unit root testing. I am in dought of which unit root test to rely my data on. What are the pros and cons of doing common unit root testing vs individual unit root?

And what about co-integration? Should I do that also?

NB I have attached an example.

Thanks and best regards.