We examine how residents of the United States allocate their stock portfolios internationally. We find that a large U. FDI establishes marginal differences in the endowments of information about different countries, which later translate into differences in stock portfolio holdings. We control for cross-country differences in capital controls, proximity along different dimensions, corporate governance, and economic and capital market development. Our results also hold for the G6 countries collectively. (JEL F21, F36, G11) How do investors allocate their stock portfolio internationally? The capital asset pricing model (CAPM) under purchasing power parity predicts that all investors hold the World Market Portfolio regardless of their nationality (Grauer, Litzenberger, and Stehle 1976). This prediction, however, is clearly at odds with the data (Karolyi and Stulz 2003), indicating that investors take into account factors other than the benefits of international diversification. Economists need alternative theories to describe how investors allocate their portfolios internationally.An important family of models asserts that cross-country differences in stock portfolio allocations arise because investors in different countries are endowed with different information sets (Gherig 1993;Brennan and Cao 1997;Kang and Stulz 1997). A long-standing conceptual difficulty with this argument is that information asymmetry would not be sustainable for long periWe are grateful to Geert Bekaert (the Editor) and two anonymous reviewers for excellent suggestions. For helpful discussions, we thank

We test the implications of a multi-asset equilibrium model in which a finite number of risk-averse liquidity providers accommodate non-informational trading imbalances. These imbalances generate predictable reversals in stock returns. An imbalance in one stock also affects the prices of other stocks. The magnitude of the cross-stock price pressure depends on the correlations of the stocks' underlying cash flows. The model implies that non-informational trading increases the volatility of stock returns. We confirm the model's implications using data from the Taiwan Stock Exchange. r 2008 Published by Elsevier B.V.JEL classification: G12; G15

No abstract

No abstract

No abstract