[经济学]多元回归分析.ppt

10.1 The Nature Of Time Series Data 10.1 时间序列数据的性质 10.2 Examples Of Time Series Regression Models 10.2 时间序列回归模型的例子 10.3 Finite Sample Properties Of Ols Under Classical Assumptions 10.3 经典假定下的OLS有限样本性质 引言：OLS方差与序列相关 10.1 The Nature Of Time Series Data 10.1 时间序列数据的性质 时间序列数据的性质：时间序列与截面数据 The Nature Of Time Series Data: time Series vs. Cross Sectional 时间序列数据有一个时间上的顺序，而截面数据则没有 Time series data has a temporal ordering, unlike cross-section data 时间序列数据是随机过程的一个实现值 A time series data set is one realization of a stochastic (i.e. random) process 由于我们面对的不再是个人的随机样本，须要对原有假设做出一些更改 Will need to alter some of our assumptions to take into account that we no longer have a random sample of individuals 说明：时间序列数据是随机过程的一个实现值 时间序列数据是随机过程的一个实现值 a time series data set is one possible outcome, or realization, of the stochastic process. 按照时间顺序排列的一个随机变量序列，称为随机过程，时间序列过程。 Formally, a sequence of random variables indexed by time is called a stochastic process or a time series process 一个时间序列所有的可能实现值构成的集合，相当于截面分析中的总体。 The set of all possible realizations of a time series process plays the role of the population in cross-sectional analysis. 10.2 Examples Of Time Series Regression Models10.2 时间序列回归模型的例子 Examples of Time Series Models静态模型:static model 一个静态模型表达了同一时期各个变量之间的关系： yt = b0 + b1zt + ut A static model relates contemporaneous variables: yt = b0 + b1zt + ut Examples of Time Series Models有限分布滞后模型: Finite Distributed Lag Models 有限分布滞后模型（finite distributed lag，FDL）则允许一个或多个变量的滞后值对当期的被解释变量产生影响 ： yt = a0 + d0zt + d1zt-1 + d2zt-2 + ut A finite distributed lag (FDL) model allows one or more variables to affect y with a lag: yt = a0 + d0zt + d1zt-1 + d2zt-2 + ut 一般， q阶有限分布滞后模型包括z的q阶滞后项 More generally, a finite distributed lag model of order q will include q lags of z yt = a0 + d0zt + d1zt-1 +…+ d2zt-q + ut Examples of Time Series Models有限分布滞后模型: Finite Distributed Lag Models 称d0称为即期/冲击倾向，即期/冲击乘数(impact propensity，impact multiplier)，反映z变化1单位，导致的y的即时变化。 即期/冲击倾向dj随着时间j的变化，称为滞后分布(lag dist