Rare Institutional Disruptions and Uphill Capital Flows

The puzzle of "uphill capital flows," where capital flows out of countries with relatively lower capital stocks and faster-growing TFP, has reattained prominence in the two decades preceding the recent financial crisis in the form of a large and persistent United States trade deficit with the rest of the world. Asymmetric investment risk has been shown in other studies to be a significant driver of capital flows between countries; how large does the risk have to be to drive capital uphill? This paper builds a model with two large open economies to assess the strength of asymmetric risk as an "uphill" force against the neoclassical "downhill" forces. I assess the model calibrating the foreign country as China, since its "downhill" forces are quite strong. The model shows that if risk arises from only the estimated Gaussian noise of TFP, relatively high coefficients of risk aversion are needed to rationalize the large and steadily growing holdings by foreigners of US assets over the decades preceding the crisis. However, TFP in middleincome countries is subject to other shocks besides Gaussian noise. Recent studies have documented the probability and magnitude of drops in output attributable to rare disruptions to political institutions. This paper shows that this additional risk, even at moderate levels of risk aversion, is enough to drive Foreign investors to engage in decades-long flows of precautionary savings into the United States and quantitatively dominate the effect of Foreign TFP catchup.