This project develops new theories and constructs new datasets to understand the determinants of net and gross foreign assets, the trade balance and the exchange rate. It also quantifies their respective roles in the dynamics of countries’ external deficits. In previous work, I constructed a database of US foreign assets and liabilities to analyse the sustainability of US current account deficits. I propose to build on this work along four related lines. First, I will generalize the analysis to other countries, contrasting the external balance sheets of large financially developed economies (US, UK) with those of small open economies (Canada, Australia). I will compare the historical role of the UK as a world banker to the current position of the US in the international monetary system. I will construct disaggregated databases of foreign assets at market value for these countries. Second, I will develop new theories of portfolio investment where international wealth transfers and predictable excess returns play a key role. These elements are rarely incorporated in open economy models but are essential for realism. I will develop and calibrate a new class of portfolio balance models compatible with the macroeconomic stylized facts on capital flows to study how countries’ capacity to accumulate foreign debt depends on changes in portfolio preferences (e.g. erosion of home bias). Third, I will use a disaggregated database of international investment positions of institutional investors to test for portfolio rebalancing at the microeconomic level. This exceptional database should also provide insights on the international propagation of financial crises. I will link the magnitude of price drops of given equities in crisis times to the institutional and geographical characteristics of their holders. Fourth, I will extend the methodology developed to analyze external adjustment to the issue of fiscal adjustment and twin deficits.