xtserialpm: A portmanteau test for serial correlation in a linear panel model

Vincenzo Verardi, Koen Jochmans

Research output: Contribution to journalArticle

Abstract

We introduce the commandxtserialpmto perform the portmanteautest developed in Jochmans (2019). The procedure tests for serial correlation ofarbitrary form in the errors of a linear panel model after estimation of the regressioncoefficients by the within-group estimator. The test is designed for short panelsand can deal with general missing-data patterns. The test is different from therelated portmanteau test of Inoue and Solon (2006) that is performed byxtistest(Wursten 2018) in that it allows for heteroskedasticity. In simulations documentedbelow,xtserialpmis found to provide a much more powerful test thanxthrtest(Wursten 2018), which performs the test for first-order autocorrelation of Bornand Breitung (2016). Comparisons withxtistestandxtserial(Drukker 2003)are also provided. These tests perform well under stationarity but break downeven under mild forms of heteroskedasticity
Original languageEnglish
Number of pages13
JournalStata Journal
Publication statusAccepted/In press - 2019

Keywords

  • xtserialpm
  • heteroskedasticity
  • fixed-effect model
  • portmanteau test
  • serial correlation
  • short panel data unbalanced panel

Fingerprint Dive into the research topics of 'xtserialpm: A portmanteau test for serial correlation in a linear panel model'. Together they form a unique fingerprint.

  • Cite this