人民大学高级计量经济学讲义13.ppt

Econometric Theory Lecturer: Dr.Jingtao Yi Room 715 Business School, RUC Lecture 13: k-Variable Linear Regression XI 1. Autocorrelation; 2. Testing for Autocorrelated Disturbances; 3. Estimation under Autocorrelated Disturbances; 4. Autoregressive Conditional Heteroscedasticity (ARCH). Autocorrelation Forms of Autocorrelation Forms of Autocorrelation Forms of Autocorrelation OLS and Autocorrelation The consequences of applying OLS to the case of autocorrelation are the same as those for the case of heteroscedasticity. Under autocorrelation, OLS estimator is unbiased, consistent, but inefficient and the inference procedures are invalid. Testing for Autocorrelation Testing for Autocorrelation Testing for Autocorrelation Estimation under Autocorrelation Estimation under Autocorrelation Autoregressive Conditional Heteroscedasticity (ARCH) Cross-section analysis: heteroscedasticity but assume that no autocorrelation Time-series analysis: autocorrelation but assume that no heteroscedasticity or assume that homoscedasticity Time-series analysis with heteroscedasticity: autoregressive conditional heteroscedasticity (ARCH) ARCH Testing for ARCH * *