Long run variance estimation and robust regression testing using sharp origin kernels with no truncation

If ρ is passed to infinity with the sample size (T), the new kernels provide consistent LRV estimates. Within this new framework, untruncated kernel estimation can be regarded as a form of conventional kernel estimation in which the usual bandwidth parameter is replaced by a power parameter that serves to control the degree of downweighting. A data-driven method for selecting the power parameter is recommended for hypothesis testing. Simulations show that this method gives arise to a test with more accurate size than the conventional HAC t-test at the cost of a very small power loss.