Global ownership patterns

To analyze multinationals’ market entry, I assemble a new dataset of cross-border mergers and acquisitions (M&A) for 1985–2019, with sales for the parent and target firms. Seven main facts emerge: (1) the number of acquisitions per parent has a Pareto distribution; (2) the number of different industries and countries entered per parent has a similarly shaped Pareto distribution; (3) parents entering unpopular industries or countries are large, as measured by global sales; (4) at the parent level, the number of industries and the number of countries entered is highly correlated; (5) the percent of zero parent-level M&A flows across destination industry-countries is 99%, and obeys a gravity relation; (6) single-industry and single-country parents are very prevalent (around 50%), but their sales share is low (around 10%); (7) there is positive assortative matching between parent and target sales. A balls-and-bins model (à la Armenter and Koren, 2014) quantitatively matches many of the facts due to the sparsity of M&A data. Three unmatched facts provide information for future theories: parent entry by market popularity, single-market multinational parents’ prevalence and sales share, and assortative matching between parent and target sales.