The term structure of interest rates shows the relationship between yields of zero-coupon bonds and their maturities. The empirical performance of the single-factor model of the affine term structure models, such as Vasicek (1977) and Cox, Ingersoll, and Ross (1985), has not been entirely satisfactory. The curve fitting methods, and particularly the spline method, used in practice to estimate the term structure are ad hoc and thus subject to arbitrage opportunities. Guo (1998) used the fundamental Partial Differential Equation (PDE) for bond pricing to derive a linear discount function, which is consistent with no-arbitrage. He showed that this is the unique linear solution to the PDE. This solution, the exponential-polynomial model or EP model for short, has n unobserved state factors that drive a stochastic discount process for pricing bonds so as to rule out arbitrage opportunities. In this thesis, we conduct an extensive cross-sectional analysis of the EP model on two different data sets: prices for daily Treasury bills, notes and bonds from the New York Federal Reserve Bank quotation sheets from July 1989 to October 1996, and daily Canadian bills, notes and bonds prices for the time period from June 1992 to May 1995. We estimate the model by applying a minimization criterion. The cross-sectional analysis shows that the EP model is able to describe adequately the term structure of interest rates. For the US data, we find that every term structure from the sampling period can be fully represented by either nine or ten state factors. Eigenvalue analysis indicates that the first three principal components are underlying the term structure movements. We conduct a time series analysis on the three principal components. They are found to be best described by ARMA/GARCH processes. We form two types of GARCH forecasts of the three principal components and test their out-of-sample performance. We conclude that the three principal components are predictable in a statis
| Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.37778 |
| Date | January 2000 |
| Creators | Mazigh, Monia. |
| Contributors | Whitmore, G. A. (advisor) |
| Publisher | McGill University |
| Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
| Language | English |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Format | application/pdf |
| Coverage | Doctor of Philosophy (Faculty of Management.) |
| Rights | All items in eScholarship@McGill are protected by copyright with all rights reserved unless otherwise indicated. |
| Relation | alephsysno: 001809687, proquestno: NQ70095, Theses scanned by UMI/ProQuest. |
Page generated in 0.0022 seconds