Item – Theses Canada

​
OCLC number
1006689265
Author
Hejazi, Walid,1963-
Title
Are term premia stationary?.
Degree
Ph. D. -- University of Toronto, 1994
Publisher
Ottawa : National Library of Canada = Bibliothèque nationale du Canada, 1995.
Description
3 microfiches
Notes
Includes bibliographical references.
Abstract
My thesis, entitled Are Term Premia Stationary?, consists of three essays that analyze the statistical properties of the term structure of interest rates. The first essay is a literature review. The second and third use recent advances in the analysis of non-stationary time series to study the unit root properties of interest rates and term premia. I explain the predictability of excess bond returns, and at the same time maintain in sample orthogonality of forecast errors and variables in the information set. I use monthly CRSP yield data on T-bills with maturities of up to 1 year, for the period 1964 to 1991. Are term premia stationary? I find evidence of non-stationary components in excess returns at short maturities, and in forward spreads for all maturities. This evidence is consistent with the presence of a highly persistent (unit root) component in term premia. The fact that these premia are not observed directly, but rather only in the presence of relatively large forecast errors, limits my ability to document their stochastic properties using traditional testing procedures. However, regressing these premia on non-stationary variables allows me to document evidence of unit roots in the term structure. Postulating a GARCH-M model for the term structure, I find the estimated premia on individual T-bills can be regarded as non-stationary processes. Although this estimated unit root component explains only a small fraction of total excess return variances, accounting for this component leaves no further evidence of unit roots in the term structure. Mean reversion in the term structure. I review the evidence on mean reversion in bond returns as well as its importance. I present new evidence on mean reversion in monthly returns to U.S. Treasury bills. I also show that standard tests for mean reversion can be misleading when the underlying time series contains a unit root. Using standard tests, I find mean reversion in excess returns, and forward spreads. I reject the null hypothesis of a unit root for these series. However, I find evidence of both mean reversion and non-stationarity in interest rates and forward rates.
ISBN
0315971991
9780315971998