This paper has two major objectives. First, we develop and implement a Bayesian generalized factor model that allows for non-orthogonality of the idiosyncratic factors and the flexibility of cross-sectional and time series dimensions. Second, we evaluate the significance of the orthogonality...