Essays on Term Structure, Forward Premium Anomaly and Globalization

Term structure models have attracted tremendous amount of attention in the last two decades. My first paper specifies the dynamic and cross-sectional behavior of bonds in the framework of the Linear or general affine term structure model (ATSM). After revisiting the basic theory of ATSM under the physical probability measure, a linear version, LTSM is proposed. We find theoretical loads of ATSM and LTSM by solving Riccati equations, with parameters chosen for the solution to match that from the principal two-component models. This paper is the first which provides an empirical model, and numerically studies the state spaces that guarantee the Black-Scholes equation is uniquely solvable and that the yields are always positive, so it clarifies the condition of Duffie and Kan (1996).The second paper utilizes the LTSM to study the forward premium anomaly. This allows me to model the behavior of the risk premium theoretically and empirically. I test my model using data on the Canadian-U.S. exchange rate. The dynamic factors are captured by Composite Principal Component Analysis (CPCA) which supplies adifferent way to set up the global factors for both currencies. Different from previous work in this area, the LTSM and ATSM can account for the anomaly, and the theoretical interest rates are guaranteed to be positive.Freund and Weinhold (2000, 2002 & 2004) stated that the Internet stimulates international trade. However, the data provided by World Bank is certainly not an ideal measure. My third paper presents a new way to measure the diffusion ofthe Internet, downloaded from Cooperative Association for Internet Data Analysis(CAIDA). The major findings are that a significant and positive relationship existsbetween Internet distance and the bilateral international trade volumes across tencountries. Furthermore, the magnitude of elasticity is discussed and further supportthe conclusion of Freund and Weinhold.