The global financial crisis of 2008/09 has reminded both policymakers and academics of the powerful effect of sudden changes in the direction of capital flows. A tightening of borrowing constraints was an important contributor to these sudden changes and forced many borrowers into rapid deleveraging. Based on their experience in the 1990s, a number of emerging market economies had prepared for such shocks by accumulating foreign reserves. This dissertation analyzes the effects of such credit shocks and the optimal precautionary response in emerging economies.

Chapter 1 is a brief introduction that motivates the topic and overviews main results of the subsequent chapters.

Chapter 2 takes the view of a small open economy. It develops a formal model of why emerging markets simultaneously hold external debt and external reserves. Reserves may be held simultaneously with debt even when their return is lower because they are valuable for self-insurance. Two key assumptions generate this finding. First, the economy may experience a sudden stop in its access to new foreign debt issuance. Second, debt has longer maturity than reserves. When a sudden stop occurs, the maturity difference allows the agent to repay the debt gradually, giving a liquidity advantage to reserves. I numerically show that the model economy optimally chooses simultaneous holding for most periods. The model also generates contrasting responses of reserves to the sudden stop shock and the endowment shock, consistent with the data.

Chapter 3 takes the view of a firm in an emerging economy. It investigates the relationship between credit shocks and firm financing patterns. After empirically establishing that banking crises are followed by stagnation in credit and that investment is financed less by debt and more by internal fund or equity at the time of banking crises, I develop a dynamic model of the firm consistent with this finding. In the model, the firm increases its reliance on retained earnings or equity issuance in response to a negative credit shock. In the long-run distribution, the introduction of a credit shock leads to a lower average debt and higher volatility in equity payout, debt, and capital. An extended period of negative credit shocks leads to a creditless recovery where investment is financed not by debt but by retained earnings or equity issuance.