厦门大学经济研究院计量经济学国际培训讲义,（2个PDF）英文,588k

文件一：LECTURE ON BASIC TIME SERIES MODELS
1 Introduction 1
2 BasicConcepts 1
2.1 Weak and Strict Stationarity . . . . . . . . . . . . . . . . . . . . . . . . . 1
2.2 Difference Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.3 Back-Shift Operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3 Stationary and Invertible ARMA Processes 6
3.1 Moving Average Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2 Autoregressive Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.3 Autoregressive Moving Average Processes . . . . . . . . . . . . . . . . . . 11
3.4 Invertibility of MA Processes . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.5 Vector AR processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.6 Impulse Responses and Error Variance Decomposition . . . . . . . . . . . 14
4 Box-Jenkins Approach 17
4.1 Differencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.2 Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

4.3 Model Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.4 Asymptotic Properties of the QMLE . . . . . . . . . . . . . . . . . . . . . 24
4.5 Diagnostic Checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5 Volatility Models 27
5.1 ARCHModels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
5.2 GARCHModels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.3 EGARCHModels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.4 GJR-GARCHModels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.5 Implementing GARCHModels . . . . . . . . . . . . . . . . . . . . . . . . 33
5.6 Stochastic Volatility Models . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.7 Realized Volatility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

文件二：Econometric Models
I would like to discuss some issues involving nonnormality, heteroscedasticity and endogeneity in the context
of three common econometric models: the linear regression model, the binary choice model and the sample
selection model. I will first present the benchmark cases when these issues are not present.