Stochastic singularity


#1

Hi team, my code has more external shocks than observable variables, and my observable variables shouldn’t be linear related. But still I get the error message of stochastic singularity, unless I delete the last observable (lbl). Is there any way I can keep my last observable and avoid the stochastic singularity problem? Thanks. forq.rar (15.6 KB)


#2
  1. You should try to find out why that variable is linearly related to the others
  2. You can always add measurement error to that or another variable to break the linear dependence.