Abstract

Complex ownership structures as of MultiNational Enterprises (MNE) can lead to geographical misallocations of the outward foreign direct investment (FDI) statistics, thus making FDI statistics by immediate counterpart economy not enough to understand the ultimate sources of the investments and their ultimate destinations, also called Ultimate Host Economies (UHE). In this regard, a specific guidance and methodology is being developed in the scope of the international statistical manuals’, and experimental methods are investigated within the FDI statistical community.
In this paper, we conduct a network analysis on the reconstructed global outward FDI network in terms of connectivity and centrality measures, which let us highlight the biggest global players as starting point/final destination of the investment chains, as well as the biggest intermediaries acting as passing-though jurisdictions.
Moreover, we refine a model proposed in previous work for estimating FDI statistics by UHE by means of maximum entropy techniques. The model was based on absorbing Markov chains over the reconstructed global outward FDI network whose limiting distribution provides the FDI distribution by UHE over the world’s countries. The refinement takes into account also the economy from where of the investment is made when computing the passing-through component of a given country, which was not the case in the preliminary model set up.