How to remedy the bad results of Bayesian estimation?

[jpfeifer]

See e.g.

Problem in Bayesian estimation

As I wrote above, your data treatment is wrong. Your variables have a trend. Please closely read “A Guide to Specifying Observation Equations for the Estimation of DSGE Models”. It clearly explains your problem.