We propose a Bayesian panel model for mixed frequency data, where parameters can change over time according to a Markov process. Our model allows for both structural instability and random effects. To ...