Dear Professor Pfeifer,

I have a model with 6 shocks. Accordingly I have 6 data sets to estimate the parameters. the variables (of the data sets) are not linearly related. But I still have stochastic singularity. Then I tried to start from just one shock and one data set (and so on), and I figure out that the model even works with 6 shocks and one data set! Could I have your advice on this matter?

an additional question: in the latest version of stable Dynare, stoch_simul after Estimation works with “mean” of estimated variables or still with “mode”?

I appreciate your time.

Kind regards,

Leo

# 6 shocks- 5 data sets but no stochastic singularity

- Having more shocks than observables is not a problem, only the other way round.
- If you get stochastic singularity, then the observables are linearly related.
- After MCMC it is the mean, after ML it is the mode