Regression in Time Series

overview

The classical assumptions are insufficient to justify the use of a regression in time series data. This mostly due to the strict exogeneity sasumpitons which is violated when examing time series data. AS most real world data (certainly econometric data) is time series in nature, this limits the use of regression. However, the large sample properties of regressions are used to justify regressions in time series analysis when the data is assumed to be stationary. (see hayahsi for a full discussion)

Violation of Strict Exogeneity