Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/220103 
Year of Publication: 
2015
Series/Report no.: 
Discussion Paper No. 14
Publisher: 
Institute for Applied Economic Research (ipea), Brasília
Abstract: 
This paper shows that the LM statistic for testing first order serial correlation in regression models can be computed using the Kalman Filter. It is shown tha.t when there are missing observations, the LM statistic for this tesi is equivalent to the tesi statistic derived by Robinson (1985) using the likelihood conditional on the observation times. The Kalman Filter approach is preferable because the test statistic for first order serial correlation in t.emporally aggregated regression models can be obta.ined as an extension of the previous case.
Subjects: 
First Order Serial Correlation
LM test
Regression Models
Missing Observations
Temporal Aggregation
Kalman Filter
Document Type: 
Working Paper

Files in This Item:
File
Size
729.33 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.