Essays in financial econometrics

This thesis deals with two important areas in financial econometrics. The first is the estimation of the linear factor model and associated issues of misspecification and portfolio formation. These topics are investigated in chapters 1 and 2. The second area investigated is that of modelling the short term interest rate and in particular, the estimation via Gaussian approximation methods.; The first chapter examines the effects of misspecification in the linear factor model on the performance of multivariate F test and average F test under misspecification. In particular, two misspecification with autocorrelated factors and time-varying factor loadings are considered. The linear factor pricing literature uses a noncentral F distribution to evaluate the extent of misspecification. However, one central finding in this paper that under the two forms of misspecification considered, the test statistic does not follow a noncentral F distribution. A simulation study is conducted to gauge the extent of the misspecification. Another finding is that the average F test is more sensitive to misspecification than the multivariate F test.; The second chapter investigates the issues of portfolio formation and asset pricing tests. Work in empirical finance starts with grouping individual stocks into portfolios based on a particular attribute of the stocks. This paper examines the effect of this practice and whether using individual stocks solves the problem of grouping. Canadian stock return data is used. Three asset pricing tests, the multivariate F test, the average F test and a robust specification test by Hansen and Jagannathan (1997) are considered. It is found that (i) grouping of stocks based on different attributes can give different asset pricing inference using the same pool of stocks, (ii) using individual assets introduces survivorship problem and (iii) the three asset pricing tests can give different inference on the same model specification.; The third chapter compares the performance of three Gaussian approximation methods in estimating a nonlinear continuous time short-term interest rate model. The three approximation methods are: Nowman (1997), Shoji and Ozaki (1997) and Yu and Phillips (2001). The first finding is that the performance of the Nowman method and the Shoji and Ozaki method are similar. The second finding is that the Yu and Phillips method gives a less biased parameter estimates in the daily interval. The third finding is that the window width used in the Yu and Phillips method has a critical influence on parameter estimates with a large window width giving very poor model fit. An empirical study is implemented using Candadian and UK one-month interest rate data.