|
MAXIMUM EMPIRICAL LIKELIHOOD AS AN ALTERNATIVE TO GENERALIZED METHOD OF MOMENTS FOR A FINITE SAMPLE CASE |
|---|---|
| รหัสดีโอไอ | |
| Creator | Gulfam Haider, Abdul Jabbar, Mudassar Rashid |
| Title | MAXIMUM EMPIRICAL LIKELIHOOD AS AN ALTERNATIVE TO GENERALIZED METHOD OF MOMENTS FOR A FINITE SAMPLE CASE |
| Contributor | - |
| Publisher | TuEngr Group |
| Publication Year | 2563 |
| Journal Title | International Transaction Journal of Engineering, Management, & Applied Sciences & Technologies |
| Journal Vol. | 11 |
| Journal No. | 7 |
| Page no. | 11A07J: 1-11 |
| Keyword | GMM, MEL, Entropy Maximization, Biasness, Endogeneity, Non-parametric approach, Information Theoretic Approach. |
| URL Website | http://TuEngr.com/Vol11_7.html |
| Website title | ITJEMAST V11(7) 2020 @ TuEngr.com |
| ISSN | 2228-9860 |
| Abstract | The study explored the implications of Maximum empirical likelihood (MEL) as an alternative approach to the Generalized Method of Moments (GMM) on a real economic model using finite samples. The literature frequently discussed that the MEL has good finite samples properties in contrast to GMM however, it has not been tested empirically. This study used the MEL approach to an economic model that has finite samples and endogeneity problems. For this purpose, the most famous economic model Keynesian consumption function is used using data of several countries with a small sample of thirty observations. The analysis found that elasticity estimated by MEL and GMM is significantly different and in some of the cases differences are drastically high. It is believed that such differences are due to poor finite sample properties of GMM. Thus, the study suggests that GMM provides biased and less efficient estimates in finite samples. Therefore, the researchers are recommended to use MEL rather than GMM to estimate economic models having a problem of endogeneity in a finite sample. |
