A theoretical and empirical approach to the recurring carry trade

The carry trade and its unwind are often credited with the underlying cause of dramatic exchange rate movements. In this popular currency speculation, the speculators borrow the low-interest currencies, such as the Japanese yen, and lend in the high-interest countries, such as Australia and New Zealand. Speculators profit by the failure of uncovered interest parity. According to the parity, the low-interest currency would appreciate to offset the interest rate differential between the two countries involved. Though not fully accepted in the literature, the best candidate to explain the empirical anomaly is the foreign exchange risk premium. In essence, the excess returns in the carry trade are the reward for bearing the often dramatic unwind risk. In the first part of this dissertation, we build a theoretical model to support the risk-premium explanation, basing on the leverage cycle theory. We show that the risk associated with holding the foreign exchange may be driven by the credit spread between risk-free U.S. Treasuries and risky bonds. As the credit spread changes over time, pricing difference between the foreign currency suppliers and buyers also emerges. Suppliers ultimately adjust collateral demanded to rebalance the pricing difference. The exchange rate consequently moves with the collateral level. Through this channel, the exchange rate is linked with the credit spread. We also document, with comprehensive coverage, the co-integration relationship between the exchange rate and the credit spread as the empirical support of our modeling efforts. In the second part of this dissertation, we utilize the co-integration between the exchange rate and the credit spread to design a carry trade strategy. We compare the strategy to a few alternatives in prediction power and trading performance. The innovative strategy in general fares no worse than unsophisticated random walk. The superiority is not universal enough that we can refute the efficient market hypothesis, but the new strategy paves the way for future refinements. In the third part of this dissertation, we empirically show the connection between the carry trade and the information content of speculator positions in the currency futures. Specifically, we document that carry traders possess private information that drives the exchange rate changes. It is very likely that speculators in the currency futures market are also the carry traders.