A Question about Bayesian Estimation

Dear all,

It is well known that if the number of structural shocks in the DSGE model is less than the number of observable variables, then the problem of stochastic singularity arises. What if there are more shocks than the observables? Stochastic singularity will not arise in this case. Is there any other potential problems, like identification problem, associated with this?



Generally, that is not a problem. See for example Schmitt-Grohé/Uribe (2012): What’s news in business cycles. If you don’t introduce a collinearity problem due to the additional shocks entering linearly in the same way in the same equation, there should be no identification problem. Still, you might want to check identification.

Dear Johannes, thank you as always!