On empirical likelihood test for predictability |
| |
Authors: | Kun Chen |
| |
Institution: | School of Statistics, Southwestern University of Finance and Economics, Chengdu, Sichuan, China |
| |
Abstract: | Predicting asset prices is a critical issue in statistics and finance. In this article, by incorporating the recent advances in nonparametric approaches, we propose the empirical likelihood test for the predictability for the direction of price changes. Under some regularity conditions, the test statistic has an asymptotic χ2 distribution under the null hypothesis that the direction of price change cannot be predicted. This test procedure is easy to implement and presents better finite sample performances than other popular causality tests, as reported in some Monte Carlo experiments. - Hightlights
We propose a non parametric likelihood test for predictability. The test involves no user-chosen parameter or estimation of covariance matrix. The test is simple to implement and has standard asymptotics. The test has significantly better sizes than several popular tests with satisfactory power.
|
| |
Keywords: | Empirical likelihood test predictability self-studentized size power |
|
|