2009 Fiscal Year Final Research Report
On the inference of stochastic regression models
| Project/Area Number |
19530184
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Economic statistics
|
|---|
| Research Institution | Nagoya City University |
|---|
Principal Investigator |
HODOSHIMA Jiro Nagoya City University, 大学院・経済学研究科, 教授 (30181514)
|
|---|
| Project Period (FY) |
2007 – 2009
|
| Keywords | bootstrap / 非正規性 / 漸近分布 / シミュレーション / 弱外生性 |
|---|
| Research Abstract |
I have mainly considered estimation and testing of the asymptotic covariance matrix of the QMLE or GMME in stochastic regression models where explanatory variables are stochastic. In these models, I have obtained asymptotic results and also simulation results using some bootstrap methods when explanatory variables are given (conditional models) as well as not given (unconditional models). In addition, I have obtained an empirical result using foreign exchange rate data in the conditional theteroskedastic VAR model derived from multivariatet distribution, an analysis of cross-correlations using high-frequency future data of energy prices, a new weak exogeneity result, and a note on a bootstrap procedure in White's test of heteroskedasticity.
|
