| The rise in international capital flows is one of the most noteworthy economic facts of the last two decades. Paralleling these changes, open economy theory now emphasizes that individuals forecast future events when making present decisions such as accumulating foreign assets. In these models, individuals are supposed to know how long current changes in the economy will last.; Chapter 2 moves away from this unrealistic assumption. Economic agents in the model have limited information and learn progressively about the true size and persistence of changes. The predictions of the model under learning vary substantially from those under full information in ways consistent with the data: (i) investment reacts sluggishly to changes in productivity despite reasonable costs to building capital, (ii) agents may finance investment through domestic saving even when it is optimal to finance it entirely through foreign borrowing, and (iii) saving and investment will be highly correlated despite perfect capital mobility. This correlation arises because of learning and does not depend on specific assumptions about the persistence of changes.; Some authors have argued that direct testing of open economy models yields more reliable evidence about capital mobility than saving-investment correlations. The finding in this literature is that net capital flows are substantially more volatile in the data than in the models. Chapter 3 questions the validity of these findings. Using simulation methods, I show that the econometric tests commonly used are highly biased towards finding excess volatility as well as rejecting the model.; If the true extent of international capital mobility is an open debate, there is no denying the challenges posed by speculative flows. In particular, recent financial crises have prompted a reconsideration of capital controls, hampered by the lack of reliable measures of restrictions. Chapter 4 builds new disaggregated measures that improve upon standard indices used in the literature. Chapter 5 uses these disaggregated measures to test whether controls insulate countries from foreign monetary disturbances. This important feature of capital controls has not been tested across countries before. Controls are found to have negligible effects, suggesting their adoption may be of limited use to strengthen the international financial architecture. |
