This paper consists three chapters that the first two chapters study the flight-to-quality sovereign debt crises while the third chapter studies the persistent market exclusion. In the first two chapters, I characterize risk averse lenders' optimal bond holdings under which flight-to-quality debt crises can arise when there are large differences in borrowing countries' future default risks, and do this within a dynamic, stochastic general equilibrium model. In this paper, there is a substitution effect between different bonds because borrowing countries are competing with each other on international borrowing. The relative differences in countries' fundamentals, rather than the absolute fundamentals, determine the magnitude of the substitution effect, and thus the direction of lenders' cross-border capital movements. Specifically, when the differences in countries' fundamentals are large enough, international lenders would like to move toward countries with relatively low future default risks, which improves these countries' borrowing conditions and deteriorates other countries'. Furthermore, safe countries accommodate lenders' capital movements by issuing more debt, which reduces the borrowing resources available to other countries, further intensifies the difficulties faced by countries with deteriorated borrowing conditions, and may finally force them to default. Such forces were quantitatively important in explaining the empirical evidence from the recent European Debt Crisis: European peripheries had difficulty raising funds in international markets, while in countries such as Germany, and the United States, the yields declined and the debt positions rose since 2010. In the third chapter, I characterize the lender's optimal recovery plan under which debt recovery after default is decreasing in duration of market exclusion. In this paper, the borrower's endowment realization is private information that is persistent, and she trades with the lender repeatedly. The lender has to choose a recovery plan which specifies the amount of debt recovery the borrower should repay for market reentry after default, while the borrower retains the right to decide on whether to repay the debt recovery. In equilibrium, a borrower with a high endowment realization would like to repay a higher debt recovery in order to regain market access earlier and to avoid the high output cost. In contrast, a borrower with a low endowment realization would prefer to stay with the market exclusion in order to get a higher haircut on her debt. As a consequence, the equilibrium in this paper features decreasing debt recovery and increasing probability of market reentry over time after default, both are consistent with the empirical findings in recent studies. I show that the lender choosing such a plan to separate the borrower's states by using time. In particular, given the lender has to maintain market exclusion after default in order to support the ex ante equilibrium borrowing in the market, she would prefer to allocate the ex post inefficiency in an efficient way.