Distinguishing Across Models of International Capital Flows

International capital markets do not work perfectly. But are the imperfections severe enough that they need to be taken into account when modeling international capital flows? And if so, which of the many different possible forms of financial market imperfection best describe the data on international capital flows and, by extension, the data on international risk-sharing, consumption smoothing, and investment allocation. Moreover, given that different authors have emphasized different aspects of the international data (e.g., consumption and income, or alternatively, savings and investment, etc) to what extent are the conclusions that can be drawn as to the form of financial constraints operating across countries dependent on the data used? Towards an answer to these questions, in this paper we formulate a wide range of different theories of international financial market imperfections within a common modeling framework, and use a structural, simulated maximum likelihood approach to estimate and test across these theories. The theories we consider range from the benchmark of complete international financial markets, to theories of markets in which the countries' ability to trade is endogenously constrained by default risk or limited commitment, to models where the set of traded assets is limited by various forms of capital controls, all the way to international financial market autarky. We compute and estimate the corresponding dynamic models of international financial markets on subsets of our data that emphasize consumption smoothing, or alternatively investment sensitivity, as well as consumption, investment, capital and output jointly, focusing on the most recent decade of international capital flows, but also eventually planning to consider historical data.