Essays on the Determinants of Foreign Exchange Rates

本論文研究的目的是利用市場微結構以及總體經濟理論來探討匯率的決定因子。本論文以三篇子研究來驗證（1）委託單流量以及流動性因子之整合模型在每日匯率變動中所扮演的角色，（2）跨國的消費成長差異在拋補的利率平價模型中是否能解釋每週匯率的變動。 第二章，我們著重在為何委託單流量模型應該加入流動性因子的原因。我們建立一個整合的模型去討論不同交易密度之外匯市場中，委託單流量以及實現的買賣價差如何影響匯率的變動。最後以一般化的動差模型來驗證該實證結果的穩固性以及一致性。我們的實證結果顯示委託單流量以及買賣價差在決定每日匯率變動上同時扮演重要的角色。有鑑於此，我們主張除了某些交易密度很高的貨幣外，委託單流量模型最好能考慮這些流動性因子在外匯市場微結構的影響性。 第三章，我們討論委託單流量在不同的訊息異質性與流動性下，對三種主要貨幣傳送訊息的能力。我們根據價格叢集性與交易量叢集性將樣本觀察值分成3×3的實驗設計矩陣來分析委託單流量傳送訊息的能力。該研究特別之處是我們分別以價格叢集性與交易量叢集性測度來代表不完美外匯市場的訊息異質性與流動性。我們發現市場參與者偏好在高訊息異質性以及與流動性時進行交易，這可能是因為市場參與者想要對其他交易者隱藏交易訊息並且為了避免交易的不確定性而對流動性有較高的需求。 第四章，我們探討當外匯市場出現未解的遠期溢酬現象時，拋補的利率平價模型對遠期匯率變動的估計會出現不合理的結果。此時，跨國間的消費成長差異是否具有改善利率平價模型估計偏誤的功能？實證結果指出我們的模型能夠降低利率平價模型對外匯遠期溢酬估計偏誤的情況。The purpose of this dissertation is to investigate the determinant of the foreign exchange rate according to the microstructure and macroeconomic theory. In particular, this dissertation proposes three essays in order to (1) investigate what roles of the order flow and liquidity factors play on the dynamics in the daily foreign exchange rates by estimating the unified model, (2) examine whether the consumption growth differential across the countries with the covered interest rate parity model can explain the change in the weekly foreign exchange rates. In Chapter 2, we aim to demonstrate why liquidity factors should be involved in the daily order flow model. We develop a unified model to show explicitly how the order flow and realized bid-ask spread affect the foreign exchange rate when trading density is different. Toward this end, robust estimation of the generalized method of moments is proposed. The proposed model is designed to reexamine the consistent estimates of our unified model with the order flow and the bid-ask spread. Empirically, our results provide consistent evidence that both order flow and liquidity factors play important roles in the determinant of the daily foreign exchange rate. Our findings suggest that the order flow model is better at incorporating these microstructure effects except for some currencies with a very high level of trading density. In Chapter 3, we discuss the ability of the information transmission of the order flow for major three currency pairs with the distinct heterogeneous information and liquidity. Our proposed model is designed to analyze the explanatory power of the order flow by using the price clustering and trade-size clustering measures to divide the total observations into the 3 by 3 matrix. In particular, heterogeneous information and liquidity are estimated by introducing measures under the setting of the price clustering and trade-size clustering in the imperfect foreign exchange rate markets. We find that traders prefer to trade when foreign exchange market is in the high heterogeneous information and the high liquidity. This result can be explained by the hidden information to reveal less information to other traders and the demand of liquidity to avoid the execution uncertainty. In Chapter 4, we investigate whether the consumption growth differential is involve in the covered interest rate parity when there is an invalid estimate of the interest rate differential caused by the forward premium puzzle. Empirically, we show that our proposed model is able to reduce the estimating bias caused by the forward premium puzzle