Changing the order of the equation changes the likelihood va

I am mesmerized by this;

It seems that changing the order of the equations changes the likelihood value (Bayesian DSGE estimation)???

Why? What should be the correct order of equation.

Many thanks

The equation order should not change the result.

Maybe the discrepancy comes from the fact that across each run, Dynare uses different random number series, which impact both the mode computation and the Metropolis. Note that we are currently changing this, so in the future the same result should be given across different runs.