A popular way to account for unobserved heterogeneity is to assume that the data are drawn from a finite mixture distribution. A barrier to using finite mixture models is that parameters that could previously be estimated in stages must now be estimated jointly: using mixture distributions destroys any additive separability of the log-likelihood function. We show, however, that an extension of the EM algorithm reintroduces additive separability, thus allowing one to estimate parameters sequentially during each maximization step. In establishing this result, we develop a broad class of estimators for mixture models. Returning to the likelihood problem, we show that, relative to full information maximum likelihood, our sequential estimator can generate large computational savings with little loss of efficiency.

We model two dimensions of bank globalization – bank nationality (a bank from the firm’s host nation, its home nation, or a third nation) and bank reach (a global, regional, or local bank) using a two-stage nested multinomial logit model. Our data set includes over 2000 foreign affiliates of multinational corporations operating in 20 European nations and over 250 banks that serve them. We find that these firms frequently use host nation banks for cash management services, and that bank reach may be strongly influenced by this choice of bank nationality. Our results suggest limits to the degree of future bank globalization.