Journal of Statistical and Econometric Methods

Bootstrap of Kernel Smoothing in Quantile Autoregression Process

• [ Download ]
• Times downloaded: 10616

• Abstract

  The paper considers the problem of bootstrapping kernel estimator of conditional quantiles for time series, under independent and identically distributed errors, by mimicking the kernel smoothing in nonparametric autoregressive scheme. A quantile autoregression bootstrap generating process is constructed and the estimator given. Under appropriate assumptions, the bootstrap estimator is shown to be consistent.