Loading...
Please wait, while we are loading the content...
Similar Documents
Exact Simulation-Based Tests in Multivariate Regressions: Applications to Asset Pricing Models 1
| Content Provider | Semantic Scholar |
|---|---|
| Author | Beaulieu, Marie-Claude Dufour, Jean-Marie Khalaf, Lynda |
| Copyright Year | 2000 |
| Abstract | Multivariate regressions (MR) are among the simplest empirical models of financial econometrics. It is well known however that despite their simple statistical structure, standard asymptotically justified MR-based tests are unreliable. Exact tests have been proposed for a few specific hypotheses [e.g. Gibbons, Ross and Shanken (Econometrica 1989), Shanken (Journal of Finance 1986), Velu and Zhou (Journal of Empirical Finance 1999), Stewart (Econometric Reviews 1997)], most of which depend on normality. In this paper, we propose likelihood based exact market-model tests for possibly non-linear hypotheses, allowing for a wide class of error distributions which include normality as a special case. The proposed test procedures are computationally attractive and may be easily obtained by simulation. For the Gaussian model, our test results serve to unify existing results on efficiency tests. In non-Gaussian contexts, we re-consider efficiency tests allowing for multivariate student-t errors and unknown zero-beta rate. In this case, we propose a set estimate for the intervening degrees-of-freedom parameter, which serves to devise a confidence-set based exact Monte Carlo test. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.fields.utoronto.ca/programs/cim/financial_math/finance_seminar/00-01/dufour2-lecture.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
