Four essays on international reserves

Essay one investigates the Korean reserve demand and its structural change after the 1997 Asian financial crisis. The reserve dynamics for the pre- and post-crisis periods are reasonably specified by the error correction model. Our structural break tests provide mixed evidences for the change in the reserve demand after the crisis. In general, our empirical results may provide the evidence that the Korean reserve demand became more sensitive to the adjustment cost and the openness, but less sensitive to the opportunity cost after the crisis, which is consistent with the rapid reserve accumulation of Korea after the crisis.;In essay two, we investigate the effects of financial market volatility on the reserve holding behavior of four Asian countries that experienced the financial crisis in 1997---Korea, Indonesia, the Philippines, and Thailand. We model the financial market volatility and estimate its effects on reserve dynamics, reserve accumulation, and reserve volatility, respectively. Our empirical results support the evidence for the structural change in the effects on reserve holding behavior for Korea after the crisis in 1997. This would be one of the evidences of the precautionary motive for reserve holdings after the crisis. However, except for Korea, our results demonstrate mixed results for the other three countries.;Essay three investigates the possible links between the reserve accumulation by developing countries and an endogenous decrease in the magnitude of the opportunity cost effect. An increase in the reserves decreases the risk of liquidity problems imposed by a sudden capital outflow, which results in a decrease in the opportunity cost of holding reserves. Thus, a high level of reserves may make the negative opportunity cost effect smaller. Generally, our empirical results support this interpretation.;In essay four, we develop a simultaneous supply/demand model of reserve accumulation to examine the opportunity cost effect. If the opportunity cost is also determined simultaneously by the supply side, the coefficient of the marginal opportunity cost effect obtained from OLS estimation would be biased upward. Our empirical results identify an upward bias, giving us a stronger, or reinforced, estimated opportunity cost effect when we incorporate the supply side.