Error: the forecast error variance in the multivariate Kalman filter became singular

[jpfeifer]

Please read up on “stochastic singularity”. See e.g. Cannot have more observed variables than shocks?