Nonparametric pseudo-Lagrange multiplier stationarity testing

The framework of stationarity testing is extended to allow a generic smooth trend function estimated nonparametrically. The asymptotic behavior of the pseudo-Lagrange Multiplier test is analyzed in this setting. The proposed implementation delivers a consistent test whose limiting null distribution is standard normal. Theoretical analyses are complemented with simulation studies and some empirical applications.Time series, stationarity testing, limiting distribution, nonparametric regression, nonparametric hypothesis testing