Dynamic panel data analytics and a martingale approach to evaluation of econometric forecasts and its applications
Abstract/Contents
- Abstract
- Dynamic panel data models are widely studied and used in econometrics and biostatistics. We first introduce mixed models, which have many applications in risk analytics in finance and insurance, to analyze dynamic panel data. We next develop a martingale approach to the evaluation of forecasting methods. We apply this approach to the development of statistical inference procedures for a widely used measure, the accuracy ratio, for the evaluation of probability forecasts. A comprehensive asymptotic theory for the time-adjusted accuracy ratio is established, taking care of the time series aspects of the data. Our martingale approach does not require subjective modeling of the data, such as assuming whether time series are stationary, which is an obvious advantage in regulatory environments. We carry out an empirical study on default predictions of small and medium sized enterprises. We propose using the generalized linear mixed model (GLMM) to model the default probabilities, where as the logistic regressions are the most widely used method in the literature. We use the accuracy ratio to compare the forecasting methods, and the confidence intervals for the accuracy ratios show that the GLMM outperforms the logistic regression forecasts.
Description
Type of resource | text |
---|---|
Form | electronic; electronic resource; remote |
Extent | 1 online resource. |
Publication date | 2017 |
Issuance | monographic |
Language | English |
Creators/Contributors
Associated with | Wang, Zhiyu | |
---|---|---|
Associated with | Stanford University, Institute for Computational and Mathematical Engineering. | |
Primary advisor | Lai, T. L | |
Thesis advisor | Lai, T. L | |
Thesis advisor | Papanicolaou, George | |
Thesis advisor | Ying, Lexing | |
Advisor | Papanicolaou, George | |
Advisor | Ying, Lexing |
Subjects
Genre | Theses |
---|
Bibliographic information
Statement of responsibility | Zhiyu Wang. |
---|---|
Note | Submitted to the Institute for Computational and Mathematical Engineering. |
Thesis | Thesis (Ph.D.)--Stanford University, 2017. |
Location | electronic resource |
Access conditions
- Copyright
- © 2017 by Zhiyu Wang
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...