The Bank has invited a guest, who gives a seminar on Bayesian Econo- metrics. Intrigued by the topic, your boss asks you about the possibility of estimating with Bayesian methods the same factor model that you have used to forecast GDP. From an inferential statistics point of view, what is the main dif- ference between the classical and the Bayesian estimation of the dynamic factor model? 8) When estimating the model with Bayesian methods, what are the implica- tions of changes in the starting values, and changes in the prior hype-parameters, on the final estimates of the model's parameters? And what are the implications of such changes on the final estimates of the common factor? 9) Assume that when writing down the state-space representation of the dy- namic factor model that you aim to estimate with Bayesian methods, you notice that some of the rows and columns of the covariance matrix of the transition equation's innovations turn to be full of zeroes. Why does this feature represent a problem for the implementation of the Carter and Kohn algorithm? How can we solve this problem?